**New Trends in Sport and Exercise Medicine**

Editor **Daniela Galli**

MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin

*Editor* Daniela Galli University of Parma Italy

*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal *Applied Sciences* (ISSN 2076-3417) (available at: https://www.mdpi.com/journal/applsci/special issues/trends in sport exercise medicine).

For citation purposes, cite each article independently as indicated on the article page online and as indicated below:

LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Volume Number*, Page Range.

**ISBN 978-3-0365-2502-0 (Hbk) ISBN 978-3-0365-2503-7 (PDF)**

© 2021 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.

The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.

## **Contents**



## **About the Editor**

**Daniela Galli** In 1997, she obtained a single-cycle master's degree in Biological Sciences cum laude from "La Sapienza" University (Rome, IT). In December 2000, she obtained a diploma of specialization in Applied Genetics, and in 2004, she obtained a PhD in Morphogenetic and Cytological Sciences from "La Sapienza" University. From 2005 to 2007, she was awarded a post-doc fellow at Pasteur Institute (Paris, France). From 2007 to 2013, she spent time in Pavia and Parma as a post-doc fellow. From 2013–2017, she worked as a researcher at the University of Parma (Parma, IT). Since 2018, she has been an Associate Professor at the University of Parma. Her research themes cover subjects including stem cell application studies for the prevention and treatment of heart attacks, mechanisms of cardiogenesis, and tissue engineering studies for bone regeneration. Her most recent publications also concern anthropometric evaluations of children in relation to diet and sport and postural and metabolic studies in elite athletes.

### *Editorial* **Special Issue "New Trends in Sport and Exercise Medicine"**

**Daniela Galli**

Department of Medicine and Surgery, University of Parma, 43125 Parma, Italy; daniela.galli@unipr.it

### **1. Introduction**

The practice of regular physical activity has been proposed as a determinant in many disciplines, from wellness to physiotherapy; in fact, it reduces the risks of cardiovascular diseases and diabetes. Moreover, physical exercise decreases the incidence of some types of cancer, such as breast and colon cancer. Finally, rehabilitation protocols need correct exercise training to reach the complete "return to play" of the patients. Unfortunately, the mechanisms associated with the beneficial effects of physical activity are still under study. Therefore, advances in all aspects of sport and exercise medicine will be relevant for physicians, recreational sport practitioners and elite athletes.

This was the aim of this Special Issue "New trends in sport and exercise medicine", that achieved great success. Sixteen papers have been published, which are briefly described below. They range from mobile applications in physiotherapy to changes in bioactive lipids in half-marathoners [1–16].

However, sport and exercise medicine are wide subjects and require more papers to clarify their different aspects. Therefore, we proposed a new Special Issue (https: //www.mdpi.com/journal/applsci/special\_issues/Sport\_Exercise\_Medicine\_II, accessed on 8 September 2021) to continue on this path and receive new insights in sport and exercise medicine.

### **2. Sana: A Gamified Rehabilitation Management System for Anterior Cruciate Ligament Reconstruction Recovery**

The Anterior Cruciate Ligament (ACL) surgery reconstruction is necessary in approximately 76.6% of ACL ruptures. The rehabilitation following ACL reconstruction includes long-term programs that are often performed at home by patients, without supervision. The failure of rehabilitation programs increases the onset of secondary pathologies in the years after surgery. In this paper, Kungwengwe and Evans [1] studied a mobile application, called Sana, based on gamification theory, to improve the results of post-operative physiotherapy after ACL reconstruction. The program content is based on the Royal National Orthopedic Hospital guidelines for ACL rupture reconstruction. The data show that the patients are satisfied with the application, find it convenient, engaging and user-friendly, and that physiotherapists state that the system improves therapy administration and interactions with the patients, suggesting that the Sana system can make a positive contribution to ACL rupture post-surgery rehabilitation.

### **3. Effect of Physical Exercise on the Release of Microparticles with Angiogenic Potential**

Microparticles (MP) are extracellular vesicles involved in cell-to-cell communications. They are present in physiological conditions, but their levels change in response to oxidative stress, inflammation, hypoxia and shear stress. Interestingly, these situations are also caused by physical activity. Notably, microparticles induced by physical activity are involved in angiogenesis. Di Credico et al. [2] reviewed the effects of different types of exercise on microparticle release in correlation with new vessel formation. Although training status, drugs and disease can affect microparticle formation, future studies will be necessary to clarify the effect of acute and chronic exercise on vascular adaptations.

**Citation:** Galli, D. Special Issue "New Trends in Sport and Exercise Medicine". *Appl. Sci.* **2021**, *11*, 8353. https://doi.org/10.3390/app11188353

Received: 30 August 2021 Accepted: 6 September 2021 Published: 9 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

### **4. The Effects of Wild Ginseng Extract on Psychomotor and Neuromuscular Performance Recovery Following Acute Eccentric Exercise: A Preliminary Study**

Ginseng is a widely diffused anti-inflammatory agent with positive actions on the nervous system and cognitive function. Thus, it is a reasonable alternative to decrease exerciseinduced inflammation. Jung et al. [3] studied the effect of ginseng extract (700 mg/day) on psychomotor and muscular performance after acute eccentric exercise. Although they only found a significant decrease in exercise-induced inflammatory markers such as Interleukin-6 (IL-6), this preliminary study opens new issues, which a future project, with a large sample, will elucidate.

### **5. Explosive Strength Modeling in Children: Trends According to Growth and Prediction Equation**

Strength training is usually only performed in adults. Recently, the relevance of muscular strength in youth and adolescents has been shown to prevent cardiovascular diseases, metabolic disorders and obesity. However, it is very important to monitor developmental aspects in parallel with training and increases in performance. Looking at this scenario, Carnevale Pellino et al. [4] elaborated a model of predictive equations to correlate anthropometric parameters, fitness levels and the maturation of children.

This will provide practical suggestions for coaches who should consider age, body shape and body mass for strength training and, in particular, lower limb performance.

### **6. Contributions of Anthropometric and Strength Determinants to Estimate 2000 m Ergometer Performance in Traditional Rowing**

Traditional rowing is a widely diffused sport, and many scientists are studying the importance of physiological characteristics to optimize the rowers' performance. Although, to date, the papers were focused on Olympic rowing, the paper by Sebastia-Amat et al. [5] analyzes the contribution of physical determinants such as height, body mass and strength of the lower/upper body on row ergometer performance in traditional rowing. The authors found that higher anthropometric determinants were positively correlated with higher performance, but upper body strength factors (determined by bench pull training) showed the best correlation with positive performance, evidencing the role of upper body strength in rowing execution.

### **7. Effects of Achilles Tendon Moment Arm Length on Insertional Achilles Tendinopathy**

The Achilles tendon is the largest tendon in the body, and it is very important in the gait cycle because it links the ankle joint to the triceps surae. The insertional Achilles tendinopathy (IAT) is a painful chronic disorder that requires suture bridge surgical therapy. This technique is still under investigation and its level of effectiveness is unclear. Miyamoto et al. [6] performed a retrospective study to analyze the moment arm length in both healthy subjects and in post-operative patients. They observed a significant increase in the force of the IAT group after surgery, suggesting that a long moment arm is one of the causes of IAT and that suture bridge surgery is an effective technique that reduces the Achilles tendon moment arm.

### **8. Immediate Effects of an Inverted Body Position on Energy Expenditure and Blood Lactate Removal after Intense Running**

Overtraining and overreaching are challenges for all athletes. Thus, physiological and psychological recoveries after training are important for health and optimal performance.

Kim and Park [7] studied the effect of a cool-down strategy with an inverted body position (IBP) on fatigue perception, blood lactate concentration and heart rate in 22 athletes. While the fatigue perception was not significantly modified by this recovery technique, blood lactate level and energy expenditure were significantly lower than in controls, suggesting that this technique could be useful in many disciplines that have short breaks during matches, such as martial arts and wrestling.

### **9. Ultrasound-Guided Percutaneous Needle Electrolysis and Rehab and Reconditioning Program for Rectus Femoris Muscle Injuries: A Cohort Study with Professional Soccer Players and a 20-Week Follow-Up**

One of the most common sports injuries is rectus femoris muscle strain. Therefore, it is very important to elaborate rehabilitation programs that can induce a fast "return to play" in the athletes. Valera-Garrido et al. [8] studied the effects of a combination of percutaneous needle electrolysis and a specific rehabilitation program in professional soccer players with Rectus femoris muscle strain. The authors demonstrated that the combination of the two techniques permitted a safe "return to play", with no re-injuries seen in a long-term check.

### **10. Effectiveness of Training Prescription Guided by Heart Rate Variability Versus Predefined Training for Physiological and Aerobic Performance Improvements: A Systematic Review and Meta-Analysis**

During physical activity, the balance between the training load and the physiological responses to training, including neuromuscular and cardiovascular activity, is very important. Mendellin Ruiz et al. [9] conducted a systematic review comparing the effects of heart-rate variability-guided training versus a predefined training designed to optimize the aerobic performance of athletes.

Although they did not find significant differences between the two training methods, this review evidences the potential of the heart-rate variability methodology, suggesting that additional studies are necessary to specify the physiological effects of this method with respect to the generally used strategies.

### **11. Exercise Training of Secreted Protein Acidic and Rich in Cysteine (Sparc) KO Mice Suggests That Exercise-Induced Muscle Phenotype Changes Are SPARC-Dependent**

The beneficial effects of physical exercise have been well documented, but the molecular basis of these positive outcomes is still unclear. One of the proteins induced by exercise is Secreted Protein Acidic and Rich in Cysteine (SPARC). Ghanemi et al. [10] performed functional studies in *Sparc null* mice, showing that muscle strength and performance are SPARC-dependent. Future studies on the roles of exercise-related proteins in muscle physiology and diseases will be important to clarify the benefits of physical activity for human health. In fact, all these mediator proteins represent potential molecular targets for pharmacological therapy.

### **12. Effects of High-Impact Weight-Bearing Exercise on Bone Mineral Density and Bone Metabolism in Middle-Aged Premenopausal Women: A Randomized Controlled Trial**

Osteoporosis is a serious problem in western countries. High-impact exercise seems to be important for bone mass increase, but the majority of studies were performed in post-menopausal women. Prevention of the onset of osteoporosis would also be important to characterize the effects of high-impact exercise on middle-aged, pre-menopausal women. Kim et al. [11] performed this type of study and showed that a physical activity program (intensity set at 60–80% of maximal heart rate) reduces age-associated changes in bone markers, although it does not increase the bone mineral density.

### **13. Physical Activity and Redox Balance in the Elderly: Signal Transduction Mechanisms**

Reactive Oxygen Species and cellular antioxidant machinery are very important in studies of age-associated diseases. Physical exercise stimulates both these components, creating contrasting effects. The intensity and volume of the exercise, together with the physiological characteristics of the people, seem to be the most important parameters. To better characterize the molecular pathways correlated with the oxidant/antioxidant effects of exercise, Galli et al. [12] reviewed the existent literature on this subject. Although more studies will be necessary to reach a comprehensive view of all the involved pathways, the

authors evidenced that, in the future, precision medicine will be able to administer physical activity in a personalized way, transforming wellness in targeted prevention.

### **14. Pulmonary Effects Due to Physical Exercise in Polluted Air: Evidence from Studies Conducted on Healthy Humans**

Air pollution is associated with many chronic diseases (cancer, pulmonary disorders, etc.). However, outdoor physical activity is widely diffused in countries with high levels of pollution. Araneda et al. [13] reviewed the studies conducted on healthy people who performed physical activity in environments with low air quality. The authors concluded that exercise should be performed in safe environmental conditions by reducing pollution, increasing installation of correct filters in gyms and increasing mobile stations for urban races to monitor the load of pollutants.

### **15. Augmentation Index Is Inversely Associated with Skeletal Muscle Mass, Muscle Strength, and Anaerobic Power in Young Male Adults: A Preliminary Study**

Arterial stiffness is correlated with performance parameters such as cardiorespiratory endurance and muscle strength. In fact, the increase in cardiorespiratory endurance reduces body mass index, low-density lipoprotein cholesterol and systolic blood pressure, which represent significant risk factors for cardiovascular disease. In young adults, this correlation is still unclear; thus, Lee et al. [14] investigated the relationship between muscle strength parameters and arterial stiffness in young males. They showed an inverse significant correlation with one of the parameters used to measure arterial stiffness, such as the Augmentation index, suggesting that strength training in young adults contributes to changes in arterial stiffness.

### **16. The Role of Cholinesterases in Post-Exercise HRV Recovery in University Volleyball Players**

Exercise recovery is a complex process that depends on the equilibrium between the sympathetic and the parasympathetic branches of the autonomous nervous system on the heart rate. Notably, the parasympathetic branch slows the heart rate through the increase in acelylcholine release and, therefore, the activation of the Acetylcholinesterase enzyme that is abundant in the sino-atrial node. The levels of Acetylcholinesterase are generally indirectly measured by Cholinesterase in the blood. Hoyos-Flores et al. [15] found a positive correlation between heart rate variability and Cholinesterase levels after long-term and intermittent high-intensity training in volleyball players, suggesting that Cholinesterase and heart rate could be considered as the internal load of the recovery phases after training.

### **17. Changes in Plasma Bioactive Lipids and Inflammatory Markers during a Half-Marathon in Trained Athletes**

After a competitive marathon cytokines, inflammation and oxidative stress increase in athletes. However, the relationship between bioactive lipids and inflammatory mediators is still uncharacterized. Gaggini et al. [16] studied the levels of Ceramides, Diacylglycerols and Sphingomyelin, together with inflammatory markers in athletes, after a half-marathon. Although they did not find a significant correlation between bioactive lipids and inflammatory markers, they observed a significant decrease in lipids after a race, suggesting that these molecules could represent new biomarkers to characterize exercise adaptation and evaluate specific training for different subject categories.

**Funding:** This research received no external funding.

**Acknowledgments:** I wish to thank all the authors and the reviewers who have contributed to this Special Issue. Additionally, I wish to thank the Editor in Chief and the editorial office of the journal. Finally, special thanks go to Sara Zhan, my Assistant Editor, who helped me in all the different phases of preparation of this Special Issue.

**Conflicts of Interest:** The author declares no conflict of interest.

### **References**


## *Article* **Sana: A Gamified Rehabilitation Management System for Anterior Cruciate Ligament Reconstruction Recovery**

### **Tanaka Kungwengwe \* and Richard Evans**

College of Engineering, Design and Physical Sciences, Brunel University London, Uxbridge UB8 3PH, UK; richard.evans@brunel.ac.uk

**\*** Correspondence: kungwengwe@outlook.com

Received: 29 June 2020; Accepted: 14 July 2020; Published: 16 July 2020

**Abstract:** The anterior cruciate ligament (ACL) provides stabilization support for the back and forth motion of the knee joint. ACL ruptures account for 50% of all sports-related knee injuries with approximately 76.6% of them requiring reconstructive surgery, necessitating long-term patient rehabilitation. Compliance with rehabilitation management programs, following ACL reconstruction, is fundamental for the successful restoration of the knee's kinematics and reducing the risk of secondary osteoarthritis. Existing recovery programs are often paper-based and require patients to perform exercises at home, unsupervised, resulting in a low level of self-efficacy; by promoting self-efficacy in home-based settings, rehabilitation outcomes can improve. This paper reports the design development of the Sana system, a mobile and wearable application that adopts behavioral design principles and gamification theory to improve long-term post-operative outcomes for ACL reconstruction recovery. A feasibility study was conducted from 15 October 2019–13 May 2020, employing the double diamond framework and a human-centered design approach (BS EN ISO 9241-210: 2019). Eighteen participants were recruited, including eight domain experts (in fields such as user experience design, human factors, and physiotherapy), and ten representative users who had undergone long-term rehabilitation for musculoskeletal injuries.

**Keywords:** rehabilitation; compliance; anterior cruciate ligament reconstruction (ACLR); range of motion (ROM); self-efficacy; behavioral design; gamification

### **1. Introduction**

Anterior cruciate ligament (ACL) ruptures account for 50% of all sports-related knee injuries. Approximately 76.6% of them result in reconstructive surgery, requiring long-term patient rehabilitation [1]. ACL tears often occur in activities that involve sudden stops or changes in movement and have profound effects on knee kinematics with recurrent knee instability as the main cause [2]. Anterior cruciate ligament reconstruction (ACLR) surgery is an integral aspect of treatment, but the economic repercussions are significant, with an estimated 15,000 primary ACL reconstructive surgeries performed in the UK annually, at a cost of £63 million [3]. To improve the success rate of recovery treatments, patients must remain compliant and perform exercises correctly. Failure to do so increases the chances of secondary knee osteoarthritis within 10–15 years post-surgery [4].

The digital health device market for ACLR rehabilitation has grown exponentially in recent years. With the emergence of Healthcare 4.0, patient-oriented systems, using the automated capture of patient data and transparent sharing with healthcare providers, can lead to improved healthcare delivery [5]. In the context of home-based physiotherapy (HBPT), following ACL reconstruction, a technological shift from traditional paper-based methods of treatment to automated interventions has occurred,

significantly improving physiotherapy services for both patients and caregivers [6]; however, there is an identified need for assistive interventions for post-ACLR recovery patients to ensure optimal recovery.

Non-compliance to HBPT is a significant problem prevalent among patients recovering from musculoskeletal (MSK) injuries, with yearly rates as high as 50–65% [7]. High rates of non-compliance are attributed to the fact that caregivers in public and private healthcare settings continue to administer paper-based treatment procedures. Paper-based procedures fail to provide personalized care and user engagement and, as a result, present barriers to rehabilitation compliance and self-efficacy. Numerous digital rehabilitation interventions exist on the market (e.g., Tracpatch and Re.Flex) that assist in knee surgery rehabilitation, but these are predominantly targeted towards the recovery of total knee replacement surgeries and not ACL injuries. Existing rehabilitation systems typically use augmented reality (AR) and interactive smartphone dashboards to provide patients with real-time feedback on their performance during rehabilitation exercises. Further, they enable caregivers to monitor patients' progress and advise on exercises remotely. Data can also be captured by patients' self-reporting recovery outcomes.

Currently available products and services combine smart mobile applications with wearable hardware devices to allow exercises to be performed and remotely monitored in at-home environments. Through Internet of Things, wireless sensors and biometrics, physiotherapists and doctors can monitor the kinetic and kinematic movements of patients post-surgery. For example, ROMBOT (https://rombot.com/) connects patients with licensed physiotherapists in the United States to create personalized movement programs to assist in pain management, recovery, and the prevention of injuries. Similarly, CyMedica Orthopedics developed e-vive which delivers tele-rehabilitation services connecting clinicians with patients that are undergoing partial and total knee replacement. Gait Up (https://www.gaitup.com/), developed as a spin-off of the University Hospital of Lausanne and the Swiss Institute of Technology of Lausanne, employs sensors, algorithms, and biometrics to convert inertial sensor signals into meaningful motion insights. Finally, physiotherapists at Macquarie University in Australia developed the REPS Recovery Exercises system, a recovery exercise application that allows stroke patients to complete exercises at home as part of a self-maintenance rehabilitation program.

As digital technologies and connectivity improves, rehabilitation management service providers are embedding gamified elements into their core functionality, providing more appealing and personalized experiences to wider audiences. During rehabilitation, gamification provides intrinsic motivation to keep patients on track and allows for provision of immediate feedback, making the process more fun, enjoyable, and interactive. Although numerous products are available for post-surgery rehabilitation management, limitations do exist, including: interfaces tend to be overtly technical and system orientated; they often fail to consider self-efficacy principles mentioned by Magee et al. [8] and Henderson and Cole [9], based on motivation and positive re-enforcement in their interface design. Future ACL recovery systems must (1) measure the quality of patients' adherence to exercise instructions, (2) measure performance through monitoring the quality of pre-determined exercises, and (3) share patient performance data periodically with caregivers to assist in the creation of well-informed, timely treatment decisions. This paper reports on the design development of a HBPT system that aims to improve the long-term post-operative outcomes of ACLR surgery. The platform focuses on addressing phase 1 (initial rehabilitation), where patients are primarily expected to regain their baseline range of motion.

### *1.1. Post-Operative ACLR Surgery*

The time taken to reach full recovery from ACLR surgery varies between patients but can take up to 12 months. During this time, patients are normally provided with a HBPT program, with exercises that are specific to their needs, considering the severity of knee damage and the level of activity a patient is hoping to reobtain [10]. Table 1 provides a breakdown of each phase of rehabilitation, as established by the Royal National Orthopaedic Hospital (National Health Service (NHS) Trust) [11].


**Table 1.** Rehabilitation guidelines for patients undergoing ACL reconstruction rehabilitation.

### *1.2. Related Work*

Gamification is defined as the use of game design features and elements in non-game contexts [12]. Extant research has supported the use of gamification in the rehabilitation process for musculoskeletal injuries [13,14] and other health-related conditions [12,15], especially to encourage engagement and increase patients' adherence to exercise programs. Qiu et al. [16] proposed Fun-Knee, a two-dimensional (2D) cartoon-based fishing game designed using UNITY engine. The authors identified that gamification can achieve significant levels of pain distraction, thereby increasing rehabilitation compliance. In their study, patients performed a series of cognitively demanding tasks that diverted their attention away from their injury. Patients could also set the exercise difficulty based on their needs. Egan et al. [17] established that pain and physical discomfort are significant barriers to exercise compliance and determined that pain distraction through gamification is one method that can be adopted to address this. Findings from Qiu et al. [16] showed that gamification methods are useful for enhancing rehabilitation compliance due to their motivating, progress-tracking, and goal-oriented nature. Further, Brown et al. [12] applied motivational principles and game elements to their study, including visualization of progress and automated goal setting activities to enhance engagement and participation. By incorporating these features, adherence and completion rates rose to 97%. Several studies have also shown that the inclusion of concept-based goal setting increases patients' motivation [18,19]. Gunaydin et al. [18] considered three simple game scenarios where the patient accumulated stars according to the correctness of their movement. This feedback was observed to improve the motivation of patients; however, the clinical benefits of the system require further validation using biostatistical analysis before conclusive results can be drawn.

Clausen et al. [19] proposed the GenuSport application, a muscle training program that consists of a strength monitoring unit with three integrated sensors and two primary game modes: maximal strength and flight control. Direct performance feedback was provided after each game. Their findings show that 80% of patients had high levels of compliance, meaning that they trained at least once per day. During the study, none of the patients required face-to-face contact and were able to use the device autonomously after adequate instruction. However, the researchers described seeing an overall decrease in frequency of use during the study and attributed this to the two game modes not developing over time in difficulty. From this, it can be assumed that a limitation of gamification is its need to be continuously evaluated and further developed to maintain long-term engagement.

Qui et al. also expressed similar challenges with regards to sustaining patients' long-term engagement and the fear of exercises becoming routine after initial interest. Deacon et al. [20] proposed a system that utilized the existing architecture of Nintendo Wii Balance with the aim of improving balance controls to reduce the risk of falls in older patients. Feedback provided to the user was focused on positive reinforcement which occurred once the patient had completed an exercise. The benefits of this approach were that it allowed patients to understand when they had completed a task to the best of their ability. The key extrinsic motivational factor was the hope of improved quality of life from using the system. Their results showed a need to implement audible feedback; users stated that they would find this more stimulating and engaging. However, since basic ACLR rehabilitation exercises require the patient to remain in a specific position to ensure accurate knee angle detection, Wii platforms are not ideal for ACLR rehabilitation.

Overall, gamification, like other persuasive architectures has merit, if implemented in the correct manner [12,21]. However, current research has criticized gamification in the healthcare sector. For example, Deloitte [22] described gamification as 'just another over-glorified tech trend', while Bogost [23] considered it a quick fix adopted by innovators to increase and promote engagement. Brown et al. [12] stated that these criticisms likely derive from the belief that "features such as points and leader boards miss the essence and power of games as motivational techniques which have the potential to encourage behavior change or adherence to treatment programs positively". Their study outlined the importance of encouraging engagement to treatment rather than technology, and that designers should consider this when applying gamification methods to contexts where the intention is to alleviate suffering and improve well-being [24]. To date, few studies have implemented gamification strategies to address rehabilitation requirements, specific to ACL reconstruction recovery. Therefore, the aim of this paper is to report the development of mobile system intervention capable of improving long-term post-operative outcomes for patients following ACLR surgery.

### **2. Methods**

To design the proposed system, a feasibility study was conducted over an eight-month period from 15 October 2019–13 May 2020. The double diamond framework was employed to model the investigation, as shown in Figure 1. This consisted of four phases: discover, define, develop, and deliver. Referring to the guidelines outlined in 'BS EN ISO 9241-210: 2019–Part 210 Human-centred design for interactive systems' allowed the authors to ensure that the project adhered to the strict design protocols outlined by the NHS for developing digital health systems [25].

**Figure 1.** Study design based on the double diamond framework.

### *2.1. Discover and Define Phases*

First, two semi-structured interviews were conducted with one physiotherapist specializing in MSK rehabilitation at multiple stages of the design process. The first interview was conducted during the discovery phase and provided an initial understanding of ACL injuries and a basis for mapping out an initial patient-user journey called pre-Sana. The second follow-up interview was conducted during the define phase and helped validate design assumptions while addressing gaps in the authors' knowledge. The inclusion criteria for the expert interviews were a minimum of 7 years of professional or academic experience in MSK rehabilitation. Second, four representative patients were recruited for semi-structured interviews. This helped identify a range of pain points that patients experienced during HBPT, following total knee reconstruction procedures. The main inclusion criteria were users aged between 21 and 40 who had experienced MSK injury of the knee in the last year and had undergone long-term rehabilitation lasting between 2 and 6 weeks. The primary method used for data collection was voice recording, while insights were transcribed manually. NVivo 12 was used for data analysis. A best-fit approach was employed which ensured that insights were grouped based on their relation to the following themes: education, engagement with caregiver, self-efficacy, and support/education. This criterion was based on prior research conducted in the behavioral design and strong predictors of rehabilitation compliance [8].

### *2.2. Develop Phase*

First, five open-moderated card sorting activities were conducted with representative patients. This user-centered design technique helped uncover the target users' expectations, thereby, allowing the authors to design a system infrastructure that was fit for purpose and delivered an optimal user experience. In the card sorting sessions, participants were tasked with organizing a set of 30 cards that corresponded with predetermined application features and functionalities into groups. The software used for data capture was OptimalSort, which provided qualitative insights that were presented using a dendrogram. Since only five participants were recruited for the study, the best merge method was used for data analysis; this method draws assumptions about more prominent clusters based on individual pair relationships. To obtain qualitative insights, participants were encouraged to think aloud and provide verbal justifications for their pairings. These insights were transcribed and analyzed manually.

Second, a design review with Cambridge Consultants, a technology consultancy with specialisms in medical device development, was conducted to validate the technical and design feasibility of the proposed system. Five participants were recruited who aligned with the specialisms required for the project delivery, including user experience (UX) research, UX design, human factors engineering, and software development. The design review was structured using the approach outlined by Harris [26] with insights categorized as challenges, ideas, likes, and questions.

Finally, online pilot usability tests were conducted over Zoom with two principle UX designers to fine-tune the study design and prototype before user testing. The inclusion criteria for these participants were a minimum of 7 years of professional experience in UX design for medical applications. To ensure that the results were appropriate to use in the final study, participants were recruited who also matched the criteria for the target user demographic of the system. However, since participants were not undergoing rehabilitation at the time of the study, context-setting and role-play instructions were provided during pre-study briefing. Participants were recruited from Cambridge Consultants who had participated in the design review study. The System Usability Scale method, proposed by Lewis and Sauro [27], was used for data analysis. Using the Fibonacci sequence (1–5), 'task criticality' and 'impact' scores were assigned to each task performed in the test. Issue frequency (%) was calculated by dividing the number of occurrences by total participants. The last stage involved calculating the 'severity' of each issue by multiplying the three variables—criticality score, impact score, and issue frequency [28]. The results obtained helped inform decisions regarding usability issues during design and development.

### *2.3. Deliver Phase*

To conclude, one summative evaluation was conducted to assess the overall user experience of the proposed system against design specification and usability heuristics: visibility of system status, provide information feedback, recognition rather than control, aesthetics and minimalist design, and consistency and standards. The inclusion criteria were a minimum of 7 years of professional or academic experience in musculoskeletal rehabilitation. The interview was recorded, with insights being transcribed and analyzed manually.

### **3. Results**

### *3.1. Discover Phase*

The goal of the discovery phase was to establish potential bottlenecks in the patient-user journey. This required interviewing physiotherapists who specialized in MSK rehabilitation and users who had undergone total knee reconstruction surgery followed by long-term rehabilitation. In order to design a solution that effectively addressed users' primary needs, it was first essential to understand the challenges users faced from their point of view and empathize with their experiences. Four user interviews were conducted to gain insights into these challenges and pain points. Through analysis of interview findings, we categorized responses into three core challenges: (1) self-efficacy, (2) education, and (3) engagement with the primary caregiver. Further details are given below.


### *3.2. Define Phase*

The goal of the define phase was to consolidate the data collected during the discovery phase. A patient-user journey, called pre-Sana, depicting the fundamental pain points and design opportunities of the existing paper-based rehabilitation process was developed, as shown in Figure 2.

**Figure 2.** Current paper based anterior cruciate ligament reconstruction (ACLR) rehabilitation procedure: pre-Sana.

### *3.3. Develop Phase*

The development phase focused on creating and refining the iteration flows and digital interface for the proposed system. Engagement with potential users and technical experts was maintained throughout the phase through rapid prototyping, iteration, and testing loops. This helped to ensure that assumptions were being continuously validated against the developed requirements specifications. The first stage in development involved the creation of a Unified Modeling Language (UML) diagram (Figure 3), to visually represent the system's information architecture along with its main actors, roles, and actions. The UML diagram was crucial in defining the initial high-level categories required for conducting the affinity mapping activity. Categories included view profile, therapy and exercises, dashboard, contact physiotherapist, and smart sleeve. These represented the features that patients would primarily interact with when navigating through the system. Considering that caregivers would have an external data collection dashboard, their engagement with the Sana app would primarily be when communicating remotely with the patient via the 'Contact' navigation item.

Once the system architecture was determined, an affinity mapping activity was conducted. The key research findings from the discovery phase were converted into a range of design ideas and potential system functions and features. These were then organized into the following established high-level categories: profile, therapy and exercises, dashboard, and contact. This activity was conducted to identify gaps in the authors' domain knowledge. A list of assumptions and questions was documented (Table 2) which was used to structure a follow-up interview with a physiotherapist (see Table 2).

• Profile: What information, if any, is a patient required to know regarding their injury? How often is a patient called up for follow-up treatment? What kind of advice or medication will a patient be given regarding pain and swelling management?


The follow-up interview helped answer the key questions and assumptions identified during the affinity mapping activity. Adjustments to the affinity map were made using insights gathered from the interview. This included disregarding features such as patient appointment scheduling, diet monitoring, and the ability for patients to choose their exercises from a list of approved exercises. From our findings, it was clear that exercise quality, exercise compliance, and pain levels were the most critical metrics caregivers use when completing assessment of whether a patient is successfully recovering.

**Figure 3.** Unified Modeling Language (UML) diagram for the Sana system.


**Table 2.** Results of the follow-up interview with the physiotherapist.

To determine the most appropriate content to be displayed on the system, we first acknowledged that content is often structured based on what makes sense to the designer, not to the end user. Sherwin [29] highlighted this as the number one usability problem in system design. For the proposed system, usability issues were mitigated by conducting a series of open-moderated card sorting activities. These assisted with the identification of target user demographics mental models and the establishing of the system infrastructure that was representative of these mental models. The data presented in Table 3 shows the key insights extracted from the dendrogram analysis in Figure 4. It highlights the features/cards that participants commonly grouped together and the labels they assigned each grouping during the card sorting activity.


**Table 3.** Card sort groupings.

**Figure 4.** Dendrogram analysis.

Patient User Journey: Post-Sana

To design a new user journey that illustrates how users interact with the proposed system, a design review was conducted with Cambridge Consultants. This critique was crucial for validating the technical and design feasibility of the proposed system. It helped establish which areas required prioritization and refinement with subsequent development. Participants that specialized in UX design, UX research, human factors engineering, and software development participated in the critique. Insights from the affinity mapping and card sorting activities were instrumental for drafting the user interfaces for each stage of the user journey. The final user journey for the proposed system is shown in Figure 5.

(**b**)

**Figure 5.** (**a**) User journey for interaction with the proposed system (steps 1–4). (**b**) User journey for interaction with the proposed system (steps 5–7).

From the feedback collected during the design critique, shown in Table 4, it was clear that the focus areas for further development focused on the dashboard and exercise user workflows. As a result, the following actions were established, as shown in Table 4.



**Table 4.** Feedback collected from the design critique.

Following the design critique meeting, the decision was made to prioritize exercise reports when developing the system's dashboard. The justification for this was that the feature directly aligns with the exercise workflow, which is the primary workflow used by patients. A series of sketch concepts were produced to illustrate the process of developing the workflow, as shown in Figure 6.

Similarly, a series of sketch concepts were produced to refine the exercise workflow and address some of the issues raised during the design review, relating to AR visualization, gamification, and positive feedback; these are illustrated in Figure 7.

### **Figure 6.** Dashboard workflow.

**Figure 7.** Exercise workflow.

### **4. The Proposed System: Sana**

This chapter presents the developed system that incorporates the three key components of the Sana system, as presented in Figure 8. The system allows users to: (1) assess quality (i.e., measure the quality of adherence to exercise instructions at a specific time), (2) monitor performance (i.e., measure performance by monitoring the quality of a pre-determined set of exercises over time), and (3) share data (i.e., share performance data periodically with caregivers, helping them make well informed and timely treatment decisions). Figure 9 maps the user journey post-Sana development and illustrates the emotions experienced by users during system use.

**Figure 9.** User journey map post-Sana development.

In Figure 10, screen prints of the developed system are shown.

**Figure 10.** Screen prints of the developed Sana system.

### *4.1. Exercise Calibration*

Setting realistic targets was deemed essential for promoting rehabilitation compliance. To ensure that each rehabilitation plan is optimally personalized to patients' specific needs and that realistic goals are set, patients calibrate their exercises using their uninjured knee during the service onboarding phase. The calibration activity is a crucial aspect of the overall user workflow as it helps physiotherapists to establish a patient's baseline. The baseline is then used as a metric measurement for treatment progression while providing patients with realistic rehabilitation goals. Figure 11 shows the calibration process.

**Figure 11.** Calibration process.

### *4.2. Integration of AR and Exercise Gamification*

The Sana system builds on the fundamental principles of gamification through goal setting, allowing patients to be more engaged with their exercises. Testament to the validity of implementing this approach, Qiu et al. [16] highlighted how gamification can be an effective method for enhancing rehabilitation compliance due to its motivating, progressive, and goal-oriented nature. The Sana system adopts a simplified gamification approach to minimize the risk of patients pushing themselves too hard and worsening their injury. The patient has one goal: to reach the green circle, as shown in Figure 12, which is set based on their baseline maximum range of motion. Nielsen [30] stated that "sustaining a match between system and the real-world" is a critical usability heuristic; therefore, the Sana system is designed to mimic key aspects of real-life physiotherapy sessions through the use of audio feedback, as the patient progresses through their exercises. The benefits of this could be either positive reinforcement or reminding the patient to stop if they are experiencing pain.

**Figure 12.** Screen print of the gamification element of the Sana system.

### *4.3. Progress and Performance Feedback*

During the development of the Sana system, the progress dashboard, as shown in Figure 13, had to be thoughtfully designed. While it may be easy to present progress and performance data to a patient through a digital interface only, this had to be achieved in such a way that would continuously keep patients motivated rather than overwhelmed. This is especially important considering that ACLR rehabilitation can take up to 12 months. Progress data are presented simply and unobtrusively via progress bars and colloquial syntax (e.g., "You're 30% there!"). Through this approach, patients perceive the system as friendly and human-oriented, thereby increasing user engagement and long-term adherence to the rehabilitation program. Behavioral design and nudge theory are two fundamental pillars of the dashboard design. Positive reinforcement is used throughout to elicit feelings of personal accomplishment and to boost self-efficacy. Magee et al. [8] highlighted that positive reinforcement methods lead to higher levels of rehabilitation compliance; however, feedback provided by the primary caregiver is irreplaceable. The Sana system facilitates direct patient–caregiver communication by incorporating a chat facility. Magee et al. [8] also stated that patients who are aware of their clinicians' feelings towards their performance and progress are more likely to strive for positive rehabilitation outcomes.

**Figure 13.** Screen print of the progress and feedback element of the Sana system.

Finally, a summative evaluation was conducted with an occupational therapist to assess the proposed system against the design specification and relevant usability heuristics. Overall, the concept was well-received. The participant stated that being able to review performance remotely was extremely beneficial for rehabilitation purposes. Regarding visual design, the therapist highlighted the simplicity, unobtrusive and friendly nature of the system as favorable characteristics which would lead to high adoption rates. Similarly, the participant did not experience any issues with navigation or performing set tasks, reflecting a good overall user experience. The participant agreed with the idea of using the uninjured knee to calibrate exercises to determine patients' baseline, describing it as "genius", and went on to suggest that they could see applications of this technology in other aspects of rehabilitation, such as shoulder and hand therapy.

### **5. Conclusions**

This paper reports the development of a mobile-based system that improves post-operative outcomes for patients that have undergone long-term rehabilitation for ACLR surgery. The authors developed Sana, a gamified rehabilitation management system with performance monitoring, data sharing, and quality assessment capabilities. Its content is based on rehabilitation guidelines established by the Royal National Orthopaedic Hospital, which incorporates rehabilitation concepts and treatment modalities that are an integral aspect of ACLR rehabilitation. The proposed system builds on the fundamentals of behavioral design and gamification theory to improve patient self-efficacy and engagement during rehabilitation. Findings show that target users found the proposed system to be convenient and unobtrusive, engaging, informative, easy to use, and user-friendly. Additionally, physiotherapists and primary caregivers stated that the proposed system can improve treatment administration and decision-making processes, provide a personalized rehabilitation experience for patients, and facilitate communication between patients and caregivers. These findings suggest that the proposed system is feasible and would make a positive contribution to ACLR rehabilitation and clinical practice.

### *5.1. Study Limitations*

Of course, this study has several limitations. First, time represented a significant limitation. Due to the eight-month constraint of our project, it meant that only one deliverable was feasible, and a longitudinal clinical study could not be conducted to test the efficacy of the system in its intended context and quantify the error of using the system on the opposite leg; this will be conducted as future work. Second, due to limited human resource availability, the authors could not concentrate on all system elements during the time period of the project. Third, although most of the technology required to deliver the proposed system successfully was readily available, there were aspects of the process that would have benefited from additional technical resources, notably, eye-tracking software for user testing; this type of software is widely used in most software development projects and would have provided additional quantitative metrics on visual hierarchy, object weight, and problematic areas within the system's user interface. Using the opposite leg for calibration is a feature that was favored by therapists who participated in the study. Justification being, it aligns with current factors they consider when establishing patient specific rehabilitation goals. However, this feature could present challenges for patients with two injured limbs or those whose opposite leg is significantly weaker than the injured leg, pre-injury. To confirm its efficacy, the platform would need to undergo extensive user testing incorporating different user cases and scenarios. Likewise, it would need to be validated with a wider range of healthcare professionals.

### *5.2. Future Developments*

Overall, the platform was positively received by target users, healthcare specialists, and technologists specializing in medical device development and design alike, which justifies the need to develop research and development further. Throughout our study, questions were raised regarding whether the simplicity of the implemented gamification concept would be enough to provide users with actionable visual feedback, informing them of whether they are performing their exercises correctly and accurately; determining this would require back-end software development. The next phase of the project will involve working with software engineers to develop the platform using Unity engine and testing a fully functional prototype; this will be reported in future research. Development efforts were primarily placed on the exercise, calibration, and progress review workflows, as these were seen as critical components of the proposed system. In the future, other key functionalities of the system, such as chat and my profile will need to be refined to complete the application's service. An integral part of assessing the overall efficacy of the proposed system will involve developing the Sana sleeve and clinician's dashboard and conducting a clinically controlled longitudinal study

using the entire system (Sana app, sleeve, and dashboard). Developing the Sana sleeve will require establishing a team consisting of mechanical and material engineers who are specialists in this area. However, academic groups [16,31,32] and commercial organizations, such as Loomia, have developed similar smart textile technologies that could be integrated, reporting high success rates. There is strong evidence to suggest that a smart textile approach is a potential future direction.

**Author Contributions:** Conceptualization, T.K.; methodology, T.K. and R.E.; validation, T.K.; investigation, T.K.; data curation, T.K.; writing—original draft preparation, T.K.; writing—review and editing, T.K. and R.E.; supervision, R.E.; project administration, T.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors would like to thank Cambridge Consultants, the workshop technicians at Brunel University London for technical support provided during the research, and all those involved in the data collection exercises.

**Conflicts of Interest:** The authors declare no conflicts of interest. All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Brunel University London (Refs: 19079-LR-Nov/2019- 21331-1 and 19079-LR-Nov/2019- 21331-2).

### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review* **E**ff**ect of Physical Exercise on the Release of Microparticles with Angiogenic Potential**

### **Andrea Di Credico, Pascal Izzicupo, Giulia Gaggi, Angela Di Baldassarre \* and Barbara Ghinassi**

Department of Medicine and Aging Sciences, University "G. D'Annunzio" of Chieti-Pescara, 66100 Chieti, Italy; andrea.dicredico@unich.it (A.D.C.); izzicupo@unich.it (P.I.); giulia.gaggi@unich.it (G.G.); b.ghinassi@unich.it (B.G.)

**\*** Correspondence: angela.dibaldassarre@unich.it

Received: 18 June 2020; Accepted: 14 July 2020; Published: 16 July 2020

**Abstract:** Cellular communication has a fundamental role in both human physiological and pathological states and various mechanisms are involved in the crosstalk between organs. Among these, microparticles (MPs) have an important involvement. MPs are a subtype of extracellular vesicles produced by a variety of cells following activation or apoptosis. They are normally present in physiological conditions, but their concentration varies in pathological states such as cardiovascular disease, diabetes mellitus, or cancer. Acute and chronic physical exercise are able to modify MPs amounts as well. Among various actions, exercise-responsive MPs affect angiogenesis, the process through which new blood vessels grow from pre-existing vessels. Usually, the neo vascular growth has functional role; but an aberrant neovascularization accompanies several oncogenic, ischemic, or inflammatory diseases. In addition, angiogenesis is one of the key adaptations to physical exercise and training. In the present review, we report evidence regarding the effect of various typologies of exercise on circulating MPs that are able to affect angiogenesis.

**Keywords:** microparticles; microvesicles; extracellular vesicles; molecular markers; cell–cell communication; physical exercise; physical activity; angiogenesis; secretome; paracrine signaling; cellular crosstalk

### **1. Introduction**

The capacity to communicate between cells, tissues, and organs is a fundamental requisite to ensure appropriate physiological functions [1]. One of the most important mechanisms governed by cell communication and signaling is represented by the formation of vasculature. It is well known that vascular formation and growth are processes strictly regulated by growth factors and bioactive molecules. Hypoxia represents an important stimulus for angiogenesis: it activates hypoxia-inducible factors (HIFs) that upregulate various angiogenic genes, including vascular endothelial growth factor (VEGF) [2]. On the other hand, other different factors (e.g., thrombospondin-1, statins) are involved in the inhibition of angiogenesis [3].

Neovascular growth can be a physiological (e.g., in injured tissues [4]) or a pathological process, as it occurs in diabetic retinopathy, maculopathy, in tumors, and in many other disorders. Physiological angiogenesis represents the results of an accurate balance between pro-angiogenic and anti-angiogenic signals [5]. In pathological conditions, this balance is compromised and the new vessels do not provide a functional vascular perfusion [6].

Small extracellular vesicles, named microparticles (MPs), can be involved in angiogenic processes [7]. MPs are defined as heterogeneous small membranous vesicular structures derived from different cell types [8]. They are lacking nucleus and ranging in size from 0.1 to 1 μm. The most abundant population of MPs derives from platelets, but also other different cell types release these vesicles into the circulation such as endothelial cells [9], leucocytes [10] erythrocytes [11], epithelial cells [12], muscle cells, and various tumor cell lines [12]. Therefore, MPs express surface protein and antigens deriving by their parental cell as well as cytoplasmic and nuclear content (e.g., proteins, mRNAs, microRNAs, small-interfering RNAs, long non-coding RNAs), that are signature of their cellular origin [13]. MPs are released in response to cellular activation or apoptosis, elicited by a variety of stimuli, such as inflammation, oxidative stress, or mechanical/hemodynamic fluctuations depending on the involved parental cell [14].

Physical exercise can induce modifications in MP concentration [15]. In fact, MP concentration is modified by stimuli such hypoxia and shear stress [16], conditions elicited to a different extent by physical exercise. It has been demonstrated that MPs produced in response to physical exercise have a stimulating effect on new vessels growth [17].

The purpose of this review is to summarize the effects of different patterns of physical exercise on the concentration of circulating MPs with a known angiogenic potential.

### **2. Vasculogenesis, Arteriogenesis, and Angiogenesis**

Vessels can be formed mainly by three processes: vasculogenesis, arteriogenesis, and angiogenesis. Vasculogenesis is the formation of new vessels that occurs in the embryo as well in the fetal annexes involved in the transportation of maternal nutrients: the placental villi and the yolk sac [18]; it derives from the differentiation of mesodermal cells into angioblasts, that proliferate, migrate, and couple to produce a primitive tube-like vessel. Consequently, vessels continue to develop as angioblasts differentiate in endothelial cells, form a vascular lumen, and deposit the basal lamina [19]. Various growth factors are involved in the molecular signaling of vasculogenesis: the fibroblast growth factor (FGF)-2, involved in mesodermal induction as well as in the induction of the angioblasts from the mesoderm [20], VEGFs family, that plays a pivotal role for its stimulating effects, [21], and finally the transforming growth factor-β family (TGF-βs) that contributes to vasculogenesis, with a dose dependent effect (low doses stimulate, while high doses inhibit endothelial cell growth [22]).

Arteriogenesis is the process whereby small pre-existing arterioles are transformed in functional arteries. One of the major stimuli inducing arteriogenesis is represented by shear stress, a mechanical signal able to activate transcription factors such as early growth response protein-1 (erg-1), that in turn lead to gene expression of chemokines like monocyte chemoattractant protein-1 (MCP-1), and adhesion molecules like intercellular adhesion molecule-1 (ICAM-1). These molecules are important for the docking of monocytes on endothelial cells, a fundamental step in arteriogenesis. Moreover, also growth factors such as VEGF, FGF, and TGF-β are involved in this process [23,24].

Lastly, angiogenesis is the process in which new blood vessels rising from pre-existing vessels, and it can occur mainly through mechanisms such as sprouting and intussusception (also known as vessel splitting or non-sprouting angiogenesis) [25]. Angiogenesis is a fundamental process during embryonic, fetal, and adult life and plays a pivotal role in many physiological process (e.g., as in skeletal growth, in wound healing, in pregnancy); however, angiogenesis represents also an important hallmark of pathological processes such as cancer and several noncommunicable diseases. Angiogenesis is governed by a precise balance between angiogenic growth factors and inhibitors. VEGF-A and its tyrosine kinase receptors represent the main and best characterized signaling pathway involved in developmental angiogenesis [26]. However, other molecules play a key role in endothelial cell proliferation and migration. They are represented by FGFs, platelet-derived growth factor (PDGF), and epidermal growth factor (EGF) [27]. Therefore, angiogenesis is a complex process controlled by several molecules and can have useful as well as detrimental effects on human health, and for this reason the various underlying mechanisms are object of great interest.

### **3. Extracellular Vesicles Classification**

Human cells are able to release various types of membrane vesicles in response to different stimuli, generally called extracellular vesicles (EVs). EVs include exosomes, MPs (also called microvesicles), and apoptotic bodies [28] (Table 1).


**Table 1.** General characteristics of extracellular vesicles (EVs).

\* PS = phosphatidylserine.

Exosomes are vesicles surrounded by a double layer of phospholipids and their size ranges between 50 and 100 nm in diameter. They are released in both physiological condition and upon activation by exocytosis of multivesicular bodies (MVBs) [29]. Exosomes can communicate with other cells either through a direct contact between surface proteins or by vesicles–cell membrane fusion or by endocytosis. They expose phosphatidylserine (PS) on the external membrane and various markers such as LAMP1, TSG101, CD63, CD81, and CD9 [30]. The main functions of exosomes include antigen presentation and immunostimulatory activity. In addition, exosomes are able to horizontally transfer mRNA and miRNA as well as oncogenic receptors [31].

MPs, formerly described as "platelet dust", are now well known cell–cell communicators and they act as signaling molecules in physiological homeostatic processes or as a consequence of pathological condition including diabetes, chronic kidney disease, rheumatoid arthritis, multiple sclerosis, and cardiovascular diseases [32]. MPs are found in most body fluids such as plasma, saliva, urine, and cerebrospinal fluid. They represent a heterogeneous population of vesicular-like structures composed by a phospholipid bilayer and their sizes range from 100 nm to 1 μm in diameter; however precise cut-off values remain to be established. MPs found in bloodstream may originate from various vascular cells (e.g., platelets, monocytes, erythrocytes, granulocytes) [33]. However, among all circulating MPs, platelet-derived MPs (Plt-MPs) are the most abundant population, representing 70–90% of the total [34]. They are produced by budding of the plasma membrane, in a process called ectocytosis. After their release, MPs carry proteins that are signature of their cell of origin, and transport various enzymes, RNA, and miRNA [35]. MPs generally expose PS on their outer leaflet (Annexin V+), however some evidence shows that MPs can also be PS negative [36]. In addition, a subpopulation of various phenotypes of MPs expose tissue factor [37]. The release of MPs is induced upon cell activation or apoptosis.

Apoptotic bodies are EVs larger than MPs and exosomes. Their sizes range from 500 nm to 3 μm and they expose PS; however, their outer membrane differs from MPs due to its permeability [38]. Apoptotic bodies are released by membrane blebbing of cells undergoing apoptosis and they may horizontally transfer their content, such as cell organelles, fragmented DNA, and oncogenes.

### **4. Methods of Detection of Microparticles**

Flow cytometry (FC) is one of the most common methods of detection, quantification, and size evaluation of MPs [39]. In addition, FC allows the phenotypic characterization of MPs using fluorescent antibodies against the diverse antigens expressed by the cell of origin [40]. However, FC has some limitations. For example, MPs with a size ranging from 0.1 to 0.4 μm are too undetected by most cytometers [41]. In addition, the type of instrument, settings, and sample preparation procedures can cause a high level of variability of the results [42].

The nanoparticle tracking analysis (NTA), on the other hand, is able to detect vesicles smaller than those recognized by FC. Such type of analysis correlates the speed of movements, tracked using a microscope, with the size of MPs [43]. Nevertheless, NTA is unable to discriminate vesicles derived from cells from other particles of small dimensions such as high-density lipoprotein (HDL) [44]. Fluorescent nanoparticle tracking analysis (F-NTA), diversely from NTA, is based on the fluorescence of the particles [45].

Other techniques to analyze the EVs are based on image analyses obtained by optical and electron microscopy. Optical microscopy permits to assess the size and morphological features of MPs and of apoptotic bodies, while cannot provide information on their phenotype. This kind of information can, anyway, be obtained by fluorescent microscopy labeling MPs with fluorescent probes that recognize specific surface markers [41]. However, this approach does not permit to specify the size of the EVs [41]. An accurate morphological analysis of MPs is obtained by transmission electron microscopy (TEM), that represents the most used method to assess the composition, morphology, size, and membrane structure of MPs [34]. However, TEM requires a considerable amount of time for both preparation and analysis of the sample.

Western blotting (WB) is a commonly used technique that allows to identify the cellular origin of MPs [46]; anyway, this methodological approach presents some important limitations: indeed, it requires a large amount of MPs to be performed and the analysis needs to be completed with complementary methods such as NTA or TEM [47,48].

Finally, another method for MPs detection is the enzyme-linked immunosorbent assay (ELISA). This is also a quantitative analysis, that measures MPs indirectly: indeed, such method relies on the binding of MPs to specific conjugate-antibodies adsorbed to a well plate [49]. ELISA permits the specification of MPs subtypes, but fresh plasma is indispensable, because the freezing-thawing process leads to an increase of annexin-binding MPs [50].

However, is important to note that most of the studies analyzing MPs in exercise settings utilize the flow cytometric approach [17,51–56].

### **5. Production Mechanisms of MPs**

It is well known that MPs can be released upon activation or apoptotic stimuli; however, most cells constitutively shed MPs by ectocytosis and significant concentrations of these vesicles can be detected in the plasma [57]. The common mechanism for the production of MPs is represented by a change in the asymmetry of the phospholipids composing the plasma membrane. The lipid bilayer of normal cells presents an asymmetrical composition of phospholipids [58]. PS and phosphatidylethanolamine (PE) are located in the internal layer of the plasma membrane whereas the external layer is composed of phosphatidylinositol (PI), phosphatidylcoline (PC), sphingomyelin (SM), and other glycolipids. In order to allow MPs formation, this precise distribution is disrupted. Such changes in membrane architecture are controlled by three enzymes: flippase, floppase, and scramblase [59]. Flippase is involved in the maintenance of the phospholipid asymmetry between the membrane layers. Floppase controls the ATP-dependent translocation of PS to the external layer in response to cellular activation. Finally, scramblase is responsible for the membrane randomization allowing phospholipids to flow down their concentration gradients. In resting cells, only flippase is active, contributing to the internalization of negatively charged phospholipids (e.g., PS and PE) and to the maintenance of the physiological membrane asymmetry. Nevertheless, various events providing cellular activation are able to increase intracellular calcium concentration, causing the flippase inhibition and the activation of floppase, scramblase, and other calcium-sensitive enzymes such as calpain and gelsolin, resulting in MPs release [60].

An important mediator of apoptotic MPs production is Rho-associated kinase (ROCK I), a protein kinase activated via caspase-mediated cleavage. ROCK I may allow the phosphorylation of myosin light-chains, triggering cytoskeletal rearrangement and formation of apoptotic MPs through the coupling of actin-myosin filaments to the plasma membrane [61].

### **6. MPs and Angiogenesis**

Vesicle-like structures have recently emerged as important mediators for cellular communication in both physiological and pathological processes [30]; in particular, MPs released by various cell types may play an important role in angiogenesiss [62,63], by modulating the production of proangiogenic factors through changes in the secretome of endothelial cells [64,65]. Like exosomes, MPs can communicate with other cells through ligands on their membrane that interact with specific receptors of the target cells; alternatively, MPs can fuse with the plasma membrane of the target cells, delivering its cytoplasmic content (e.g., proteins, miRNAs, RNAs). In this way, MPs may elicit the synthesis of proangiogenic or antiangiogenic factors and affect the main processes involved in formation of new vessels, such as endothelial cells adhesion, migration, and proliferation [66].

Endothelial cells-derived MPs (E-MPs) are clearly involved in the formation of blood vessels. A pioneer in vitro study highlighted that vesicles released by endothelial cells can transport integrin-β1 and active matrix metalloproteinases-2 and 9 (MMP-2 and MMP-9), thus inducing endothelial cells invasion and tube-like structure [67]. In addition, it has been demonstrated that upon stimulation with the proinflammatory interleukin-3, endothelial cells release vesicles containing pro-angiogenic factors such as miR-126-3p and Stat5 [68]. Moreover, another study shows that endothelial MPs are able to transfer miR-126 from endothelial cells to vascular smooth muscle cells, corroborating the angiogenic properties of endothelial MPs [69].

Additionally, Plt-MPs are involved in angiogenesis. These vesicles can promote angiogenic effects by transferring various cytokines, such as PDGF, VEGF, FGF-2, and activating key proteins such as PI3k, ERK, and src kinase [70]. Various studies found that the angiogenic potential of the platelet-derived MPs can have both functional positive and negative effects: when Plt-MPs were infused in a rat model of myocardial ischemia, indeed, they increased the quantity of functional capillaries [70]; on the other hand, MPs shed by platelets can favor angiogenesis involved in tumor progression, upregulating important angiogenic factors, such as VEGF, HGF, and interleukin-8 [71].

### **7. Angiogenesis Induced by Physical Exercise**

Physical activity has beneficial effects on human health [72–75], and regular physical exercise improves health in both physiological and pathological conditions [76–79]. Numerous evidence demonstrates the inverse relationship between regular exercise and cardiovascular diseases, hypertension, stroke, type 2 diabetes mellitus, metabolic syndrome, obesity, osteoporosis, and various types of cancer [76,80–83]. Depending on the type of physical exercise, we can observe different effects on the organism: endurance (or aerobic) exercise is characterized by a high number of muscular contractions, relative low load, and an important involvement of the cardiorespiratory system, whereas resistance exercise is represented by low number of muscular contraction but with relative higher loads and thus it relies mainly on the neuromuscular system [84]. These two types of exercise can be considered two extremes in a line, and between these two, numerous types of exercise exist, based on the modulation of the exercise variables (e.g., intensity, volume, density, rest). Each type of exercise can increase different responses and adaptions, triggering diverse cellular pathways.

Cardiovascular endurance training is the strongest stimulus for exercise-induced angiogenesis; this results in an improved capillary to fiber ratio in the skeletal muscle to provide an enhanced blood flow and oxygen delivery to the exercising muscles [85–89]. Endurance exercise stimulates the angiogenesis activating several pathways (Figure 1). It induces an important reduction of the partial pressure of oxygen that leads to a local tissue hypoxia. This modification in oxygen levels decreases the hydroxylation of HIF-1alpha trough the inhibition of prolyl hydroxylase domain enzymes (PHD); in this way, HIF-1alpha become more stable and translocate into the nucleus, where forms an active complex with HIF-1beta [90] and activates genes involved in angiogenesis such as VEGF [84]. Another key regulator of VEGF synthesis is the peroxisome proliferator-activated receptor gamma co-activator 1-alpha (PGC-1alpha), which is modulated by aerobic exercise in several ways: intense endurance exercise produces an energy deficit and an increase of the AMP/ATP ratio that activate the AMP-activated protein kinase (AMPK) responsible for the PGC-1alpha phosphorylation. Moreover, an acute bout of exercise increases the NAD+/NADH ratio that promotes the sirtuin (specifically SIRT1) activity and PGC-1alpha deacetylation [91]. Finally, aerobic exercise affects also the estrogen-related receptor gamma (ERR gamma) that controls the skeletal muscle vascularization by modifying the VEGF expression [92].

**Figure 1.** Representation of the main pathways activated in skeletal muscle in response to exercise. In addition to the canonical molecular pathways, aerobic exercise may affect the release of microparticles (MPs) that carry soluble factors involved in angiogenesis. AMP = adenosine monophosphate; ATP = adenosine triphosphate; NAD = nicotinamide adenine dinucleotide; NADH = reduced nicotinamide adenine dinucleotide; PGC-1α = peroxisome proliferator-activated receptor gamma coactivator 1-alpha; HIF = hypoxia-inducible factor; VEGF = vascular endothelial growth factor; PDGF = platelet-derived growth factor.

In the last years it was evidenced that physical exercise induces modification of the MPs release into the bloodstream [93]. These EVs carry various molecules important for the tissues' crosstalk during and post exercise [94,95] and can be also a cargo of soluble factors involved in endothelial cell proliferation, migration, and adhesion. This results in the generation of new vessels in response to exercise.

### **8. MPs and Physical Exercise**

Physical exercise, particularly endurance exercise, is known to stimulate shear stress due to its hemodynamic effects [96]. Intraluminal shear stress has a beneficial effect on vasculature because it is able to ameliorate vascular dilation through mechanisms mediated by nitric oxide [97]. In addition, exercise-induced shear stress can modulate the production of both E-MPs and Plt-MPs. For example, it was seen that high shear stress can reduce the concentration of E-MP while low shear stress is prone to increase such concentration [98]. Moreover, both high and low shear stress can elicit the production of MPs from platelets [99] (Figure 2).

**Figure 2.** This figure shows the possible mechanisms involved in the production of platelet-derived MPs and endothelial-derived MPs and their role in stimulating angiogenesis. Both platelets and endothelial cells are activated in response to mechanical and chemical stimuli provided by exercise, that in turn elicit the MPs production by ectocytosis. The branching point refers to the site in which MPs stimulate neo-vascularization.

Acute exercise leads to an augmentation of the sympathetic activity, increasing both epinephrine and norepinephrine concentrations [100]. In this regard, studies report that norepinephrine is able to stimulate Plt-MPs release [101]. Along with hormonal response, acute exercise stimulates the secretion of many cytokines from skeletal muscle, named myokines [102]. Among these myokines, IL-6 is released in the bloodstream and elicits multiorgan effects [103]. It has been reported that IL-6 is also able to stimulate Plt-MPs formation [104], thus exercise can be a potent stimulus to enhance such mechanism.

The following sections provide an overview of various studies (summarized in Table 2) investigating the effects of physical exercise and training on MPs concentration, with an emphasis on endothelial and platelet-derived MPs, due to their specific involvement in angiogenesis [17,105–107].




**Table 2.** *Cont.*

Schwarz Whal P et al.

Twelve men triathletes/cyclists

± 3.4)

 (24.7

4 × 4 min at 95% PPO vs. 4 × 30 s 1 h of moderate (46 ± 2% VO2max) vs.

heavy (67 ± 2% VO2max) intensity

semirecumbent

 cycling

all-out

CD31+/CD42b− (apoptotic E-MPs)

CD62E+ (apoptotic E-MPs)

CD41+ (Plt-MPs)

Before exercise, and 0, 30,

↓ three exercise protocols

↔ CD62E+ for both MI and HI↑ CD41+ at 30 and 60 min

of HI exercise

CD31+/CD42b− after all

> 60, 180 min post-exercise

At baseline, 30, 60, 80, 100, 120, 150, 180, and 240 min

of the protocol

(2014) [117] Wilhelm EN et

Nine healthy young

men (25 ± 1 year)

al. (2016) [17]

↑ =

VT 130 min at 55% PPO vs. increased concentration, ↓ = decreased concentration, ↔ unaltered concentration; E-MPs = endothelial cell-derived MPs; Plt-MPs = platelets derived MPs; TF+ = tissue factor positive; = ventilatory threshold; HICE = high-intensity continuous exercise; HIIE = high-intensity interval exercise; HIIT = high-intensity interval training; MICT = moderate-intensity continuoustraining;PPO=peakpoweroutput.

 **on MPs**

 both exercise

↔ CD62E+

 tissue, sex, and

 ↓ CD61+

 and 24 h post

 CD31+↑

 CD42b+↓

 CD45+ CD31+/CD42b−

 CD62E+ and

 intensity

 and sex specific

### *8.1. The E*ff*ect of Physical Exercise on MPs with Angiogenic Potential in Athletes*

Endurance athletes have the characteristics to present greater adaptation of the cardiorespiratory and neuromuscular systems resulting in the increase of the delivery and the consumption of oxygen [118]. Two of the sports that highly rely on these adaptations are marathon and half-marathon. Bittencourt et al. compared the amounts of E-MPs and Plt-MPs in a group of half-marathon runners with a control group of gender-matched healthy controls showing that CD51<sup>+</sup> E-MPs and CD42+/CD31<sup>+</sup> Plt-MPs did not significantly differ between the two groups [52]. This data suggests that athletes exposed to chronic endurance training can benefit from exercise effects but that such type of training does not lead to further decrease in MPs compared with a healthy non-athlete population. On the other hand, E-MPs and Plt-MPs significantly increase following a marathon run and return to baseline values within 2 days; conversely, leukocyte and monocyte-derived MPs decrease after such run [114].

Like long distance running, road cycling and triathlon are two sports heavily relying on cardiorespiratory endurance [119]. In triathletes/cyclists, high-volume (2 h at 55% of peak power output) and high intensity (four sets of minutes at 90–95% of peak power output interspersed by 3 min of active recovery or four sets of 30 s at maximal effort separated by 7:30 of active recovery) trials on cycle ergometer, significantly decrease CD31+/CD42b<sup>−</sup> apoptotic E-MPs 60 and 180 min after the termination of the trials [117]. It seems that such decrease was due to an increased uptake of MPs by cells. Indeed, triathletes/cyclists' serum collected 180 min after high intensities exercises promote the uptake of CD31+/CD42b<sup>−</sup> apoptotic E-MPs in endothelial cells in vitro [117]. Another recent study involving male athletes showed that an incremental cycling protocol until exhaustion determined a significant increase of CD105+, CD146<sup>+</sup> E-MPs, and CD62P<sup>+</sup> Plt-MPs at the peak of the exercise session [108].

### *8.2. The E*ff*ect of Physical Exercise on MPs with Angiogenic Potential in Healthy Subjects*

The physical exercise performed in a regular manner elicits beneficial effects that include improved cardiorespiratory fitness, body composition, decreased risk of noncommunicable diseases and of all-cause mortality [76,120–124].

Data on the effects of aerobic exercise on E-MPs and Plt-MPs are controversial. After 1 h of treadmill run at 70% of VO2max, a decrease in TF<sup>+</sup> Plt-MPs and neutrophil-derived MPs in healthy men [110] has been observed; moreover, a single session of 30 min of moderate intensity aerobic exercise (60–64% of their VO2peak) reduces the number of CD62E<sup>+</sup> activated E-MPs and CD31+/CD42b<sup>−</sup> apoptotic E-MPs [115]. On the other hand, in prehypertensive male and female, aerobic exercise performed at 65% of predicted maximal heart rate for a duration of 6 months, with a frequency of three times per week, decreases the blood concentration of CD62E<sup>+</sup> E-MPs and CD31+/CD42a<sup>−</sup> E-MPs [111]. Some authors also evidenced a possible sex/gender effect on the MPs release: indeed, an acute bout of aerobic exercise at 60–70% of the VO2peak produced a significant increase in CD62E<sup>+</sup> E-MPs in men, and an increase in CD34<sup>+</sup> MPs in women [51]. However, results of study involving both male and female are inconclusive. Indeed, another study showed CD62E<sup>+</sup> E-MPs decrease both during and after moderate intensity continuous exercise (65% of VO2max) in women but not in men [116].

One of the factors that must be taken into account analyzing the effects of physical exercise on the MPs is the timing of the analysis. In response to low to moderate intensity (33% of VO2max) concentric or eccentric cycling, men show higher concentration of CD41<sup>+</sup> Plt-MPs 5 min following exercise, but these levels returned to baseline levels after 40 min of recovery, while CD62E<sup>+</sup> E-MPs do not show modifications [56]. Another experimental study performed in young males showed that CD41<sup>+</sup> Plt-MPs and CD62E<sup>+</sup> E-MPs were unaffected by 1 h of cycling at <sup>≈</sup>46% of VO2max, while when the intensity were set <sup>≈</sup>67% of VO2max an increase of CD41<sup>+</sup> Plt-MPs was observed [17]. Interestingly, the MPs released in response to such type of exercise enhanced endothelial proliferation, migration, and tubule-like formation in vitro, compared with MPs present in resting conditions [17].

Another modality of training that is of increasing interest in research is high intensity interval training (HIIT). HIIT is a useful exercise modality to enhance cardiorespiratory endurance, and such type of training can be performed also at strenuous intensities [125,126]. Supramaximal sprint cycling protocol consisting of 10 bouts of 15 s at 120% of peak power output separated by 45 s of active recovery lead to a significant increase of circulating CD105<sup>+</sup> and CD106<sup>+</sup> E-MPs 90 min after the exercise session, whereas at 180 min post-exercise MPs returned towards resting values [112]. On the other hand, no changes in endothelial MPs concentration was observed after a high intensity interval exercise (95% of VO2max) [116].

Physical inactivity and sedentary behaviors have detrimental effects on vasculature state, contributing to the development of cardiovascular and metabolic diseases [127–129]. After 5 days of reduced physical activity (<5000 steps/day), significant increase of CD31+/CD42b<sup>−</sup> apoptotic E-MPs was registered, whereas CD62E<sup>+</sup> activated E-MPs remained unchanged, suggesting that inactivity fosters vascular dysfunction increasing endothelial apoptosis in recreationally active men [53].

### *8.3. The E*ff*ect of Physical Exercise on MPs with Angiogenic Potential in Pathologic Subjects*

Physical exercise is a fundamental stimulus for improving metabolic and cardiovascular health. Among the various typology, HIIT has the same, or superior capacity to improve health status in pathological subjects (e.g., presenting diabetes, cardiovascular diseases, or cancer) [130,131]. In overweight/obese men, high-intensity continuous exercise (20 min of cycling at just above the subjective ventilatory threshold) and high-intensity interval exercise (10 bouts of 1 min at ≈90% peak aerobic power) reduce E-MPs concentration 18 h following exercise; different results were observed in overweight/obese women in which high-intensity continuous exercise increased activated CD62E<sup>+</sup> E-MPs, leaving unaffected CD31+/CD42b<sup>−</sup> apoptotic E-MPs, regardless the intensity [15]. However, the chronic exposure of overweight/obese subjects to exercise (2 weeks of HIIT or moderate intensity continuous training) lowered the concentration of activated CD62E<sup>+</sup> E-MPs [54]. The effect of exercise on MPs release may also depend on the intensity of the physical activity. Indeed, in a recent study involving both over and normal weight subjects, a bout of moderate intensity (60% of their VO2max) aerobic exercise, reduced the total EVs in the MPs range (130–700 nm); however, the release of MPs in normal weight subjects was higher than in obese and in females than in males. [113].

Polycystic ovary syndrome (PCOS) is an endocrine disorder in which endothelial dysfunction and altered angiogenesis are common. Eight weeks of aerobic training consisting of three sessions per week of 1 h of running on a treadmill at 60% VO2max led to a significant decrease in CD105<sup>+</sup> MPs in women with PCOS compared with a control group of healthy women, while did not have an effect on the amount of CD106<sup>+</sup> E-MPs [55].

Coronary artery diseases determine a diminished myocardial perfusion that may result in angina, myocardial infarction, and heart failure [132]. In stable coronary heart disease patients, neither moderate-intensity continuous exercise (28.7 min at 70% of peak power output) nor a single bout of high-intensity interval exercise (10 min session with work interval of 15 s at 100% of peak power output interspersed with 15 s of passive recovery) have an effect on the concentration of circulating E-MPs and Plt-MPs [109].

Overall, the numerous studies show contrasting results regarding the qualitative and quantitative changes of MPs in response to acute physical exercise and to exercise training as well. Such divergent conclusions could be due to the heterogeneity of the studies, as they involved different samples in terms of age, sex, training status, possible pharmacological therapies [133], and pathologies [93] (Figure 3). More standardized and harmonized procedures are needed to compare the studies and obtain definitive data on the release of MPs with angiogenic potential in response to physical exercise.

**Figure 3.** This figure shows the potential influencing factors affecting MPs following physical exercise and training.

### **9. Conclusions**

Various modalities of cardiovascular endurance-based exercise can modify blood concentration of E-MPs and Plt-MPs, known to affect angiogenic processes. It is well known that physical exercise, among different adaptations, can lead to a functional angiogenesis resulting in a wider vascular bed that increases the blood flow to exercising muscles. Results regarding the effect of exercise on MPs are divergent, but it seems that acute exercise results in an increase of MPs while chronic exposure to exercise results in a decrease in most cases. In addition, MPs sensitivity to exercise can be affected by the training status, sex, age, as well as mode of exercise (e.g., run on a treadmill, road run, cycle ergometer), pharmacological therapies, diseases, and method of detection. While only one study directly showed the pro-angiogenic effects of MPs produced following exercise [17], numerous studies highlight the potential of E-MPs and Plt-MPs in communicating with other cells and stimulating angiogenesis [69,105–107]. Thus, future evidence regarding the effect of both acute exercise and training on MPs affecting angiogenesis will be useful to add important insights on the mechanisms by which exercise elicits vascular adaptations and to design specific training programs for various populations.

**Author Contributions:** A.D.C., P.I., G.G., A.D.B. and B.G. conceived of and designed the outline of the manuscript and examined and ensured the consistency and validity of the contents. A.D.B. and B.G. revised the final version of the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by the Italian Ministry of Education, University and Research (Ministero dell'Istruzione, dell'Università e della Ricerca-MIUR), grant number PRIN 2017ATZ2YK\_003.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article*

## **The E**ff**ects of Wild Ginseng Extract on Psychomotor and Neuromuscular Performance Recovery Following Acute Eccentric Exercise: A Preliminary Study**

### **Hyun Chul Jung 1, Nan Hee Lee 2, Young Chan Kim <sup>3</sup> and Sukho Lee 2,\***


Received: 30 July 2020; Accepted: 21 August 2020; Published: 23 August 2020

**Abstract:** To examine the efficacy of wild ginseng extract (WGE) on psychomotor and neuromuscular performance recovery following acute eccentric exercise. This study was a double-blind, crossover, and placebo-controlled design with a 14-day washout period. Ten male adults, aged 27.1 ± 4.33 years old, voluntarily participated in the study. Subjects were assigned to one of two parallel conditions (WGE or placebo) in a counterbalanced manner. Subjects consumed two packs of WGE (350 mg/package) or placebo drink immediately after acute eccentric exercise and the following four days. The eccentric exercise consisted of 20 min of downhill running at 60% of VO2peak and five sets (of 20) of drop jump exercise. Computer-based cognitive function test and neuromuscular performance tests, including straight leg raise, vertical jump, isometric leg strength, and anaerobic power test were administered four times, at baseline, 2 h, 48 h, and 96 h after acute exercise. The interleukin-6 (IL-6), myoglobin, cortisol, total antioxidant capacity (TAC), and perceived muscle soreness were also assessed at each time point. A significance level was set at 0.05. No significant differences between the WGE and the placebo groups were observed in psychomotor and neuromuscular performance variables. Blood markers, including IL-6 (p = 0.013), myoglobin (p < 0.001), and cortisol level (p = 0.047) were changed significantly across the time. A post-hoc test revealed that a significant increase in IL-6 was observed only in the placebo group (p = 0.014), while no significant changes found in the WGE condition. The perceived muscle soreness was not different between the WGE and the placebo conditions. The administration of WGE immediately after acute eccentric exercise and the following four days have no benefits on psychomotor and neuromuscular performance recovery in healthy adults. However, the acute WGE supplementation may attenuate the eccentric exercise-induced inflammatory process, such as IL-6, but future study with a large sample size is required to clarify the anti-inflammation process in response to acute eccentric exercise.

**Keywords:** ginseng; performance; recovery; eccentric exercise; inflammation

### **1. Introduction**

Ginseng is a popular selling product among herbal medicines and it has a long history concerning its efficacy on adaptogenic properties, such as stress management and fatigue recovery among Asian countries [1,2]. Various ginsengs, such as Cultivated Ginseng, Mountain Cultivated Ginseng, and Wild Ginseng have been introduced based on the grown environment and harvest methods [3]. Wild ginseng is grown in nature for an extended period to propagate in the mountain, while other ginsengs cultivate within 5–6 years in a farmland or mountain [3]. Chemical components of ginseng are also varied, depending on cultivation methods and curing processes [4]. One of the active ingredients, known as ginsenosides, plays an essential role in anti-inflammatory agents [5], and supplementation of these ingredients has been used to address the risk factors of cardiovascular diseases (CVDs) [6]. Over 40 different kinds of ginsenosides, such as Rb1, Rb2, Rc, Rd, Re, Rf, Rg1, Rg2, Rg3, Rh1, Rh2 have been identified to date, and the ginsenosides are determined by their positions on thin-layer chromatograms [5]. It has been reported that ginseng extract improves brain function, neuromuscular strength, and endurance [7–10]. Additionally, a decrease in anxiety and depression [11], as well as an increase in cognitive function [12] have been observed following ginseng supplementation. For these reasons, ginseng has been administrated not only for natural herb medicines but also for health supplements [13,14]. Recently, it is believed that ginseng may act as an ergogenic aid or recovery facilitator [15]. However, the ergogenic effects on physical performance is still unclear due to inconsistent results from the previous studies [2,16].

Participating in regular exercise and physical activity improves physical performance and immune function [17,18]. However, acute strenuous exercise or intense eccentric exercise can cause the disturbance of intracellular pro-oxidant-antioxidant homeostasis that leads to an increase in the accumulation of inflammation [19]. For instance, Banerjee et al. reported that an acute bout of exercise and/or prolonged intense exercise may cause a transient reduction and damage of various body tissues due to the augment exercise-induced oxidative stress [20]. In other words, performing one-off intense exercise can cause physiological imbalances, and this imbalanced physiological function is linked to immune dysfunction, lipid peroxidation, and free radical production. One of the common symptoms following intense or eccentric exercise is muscle soreness. Particularly, repetitive eccentric contraction exercise causes severe muscular pain 2–3 days later, which is known as Delayed Onset Muscle Soreness (DOMS) [21]. This acute muscle soreness is initially identified at the muscle/tendon junction, and then is gradually widespread to the entire muscle group [21]. An increase in muscle damage and soreness following intense or eccentric exercise affects performance recovery [22] as well as cognitive function [23].

Nutritional supplementation, such as ginseng, may be one feasible nutritional option to attenuate exercise-induced inflammation, as well as to facilitate performance recovery. Previous studies have reported that ginseng supplementation improves endurance performance [24–26], but fewer studies have been examined in other performance areas, such as neuromuscular performance. Pumpa et al. reported that ginseng supplementation following downhill running exercise was not associated with jump performance recovery in well-trained males [27]. Recently, a meta-analysis study demonstrated that there was a significant effect on fatigue recovery following ginseng supplementation, but the author pointed out that insufficient literature may not guarantee its efficacy on fatigue recovery [2]. We believe that well-designed studies are necessary to clarify the efficacy of ginseng supplementation on performance recovery, especially neuromuscular performance in healthy adults. Therefore, the purpose of this study was to examine the effects of wild ginseng extract on psychomotor and neuromuscular performance recovery following acute eccentric exercise.

### **2. Materials and Methods**

### *2.1. Subjects*

Subjects were randomly recruited by flyer and oral presentation in the university. Inclusion criteria of subjects were a) male adults aged from 18 to 40 years old, b) have no health problems, including skeletomuscular injuries, cardiovascular or pulmonary diseases, and c) have no experience using any ginseng supplementations. Subjects who were a) not able to participate in the physical activity or b) take any herbal-related medications were excluded from the study. Initially, 12 subjects agreed to participate the study, but only 10 subjects completed all procedures. Two participants voluntarily dropped out of the study for personal reasons (i.e., lack of interest, personal reason). All subjects received oral explanations, including the study procedure, possible risk, and benefits. Participants completed a physical activity readiness questionnaire (PAR-Q) and a written risk stratification screening form. When participants reported no issues with exercise, resting heart rate and blood pressure were taken, according to the American College of Sports Medicine (ACSM) guidelines. A written consent form approved by the Institutional Review Board of Texas A&M University, San Antonio, (#2015-126), was obtained from each subject.

### *2.2. Experimental Design*

This study was a double-blind, randomized crossover design. All subjects participated in the pre-tests baseline measurements for their basic characteristics, including body weight, height, resting blood pressure, body composition, and peak oxygen uptake (VO2peak). Then, subjects visited the laboratory three times (1st, 3rd, 5th day) to complete the first supplementation trial. Subjects performed the acute eccentric exercise, including 20 min of downhill running and five sets of 20 drop-jump exercises. The wild ginseng extract (WGE) or placebo drink was given to subjects immediately after the acute eccentric exercise and the following 4-day period. The WGE and placebo drink were in the identical packs manufactured by the Korea Food Research Institute. The dependent variables, including computer-based cognitive function test, physical performance test, blood analysis, and perceived muscle soreness were measured four times, at baseline, 2 h, 48 h, and 96 h after the acute eccentric exercise. The second trial was conducted after a 14 days washout period. The procedure for the second trial was the same as the first trial, with different supplement. All tests were performed at the Human Performance Laboratory in the university. The time points of each measurement are presented in Table 1.



### *2.3. Supplementation*

The WGE used in this study was 350 mg of cultivated wild ginseng extract in 140 mL packages. Subjects took two packs of WGE (WGE powder 350 mg/package) a day for five days after acute eccentric exercise. The ingredients of WGE drinks were composed of 13.8% of Korea wild ginseng, 1.5% of soy powder, 1.5% of almond paste, 1.75% of isomaltooligosaccharide, 1.75% of fructooligosaccharide, 0.05% of refined salt, 0.05% of scented almond, 0.025% of xanthan gum, 0.19% of sodium bicarbonate, 0.05% of grapefruit seed extract, and 79.335% of water. The placebo drink was manufactured to be identical in appearance and content, except the ginseng extract. The placebo contained 0.25% of scented ginseng, 1.5% of soy powder, 1.5% of almond paste, 1.75% of isomaltooligosaccharide, 1.75% of fructooligosaccharide, 0.05% of refined salt, 0.05% of scented almond, 0.025% of xanthan gum, 0.19% of sodium bicarbonate, 0.05% of grapefruit seed extract, and 92.885% of water. The WGE and placebo drink were provided by Korea Food Research Institute. Supplements were given to subjects in a counterbalanced manner to minimize the order effect. Subjects were asked not to consume any caffeine-containing (i.e., coffee, energy drink) or herbal-containing beverages during the intervention.

### *2.4. Pre-tests*

The pre-tests included body weight, height, resting blood pressure, body composition, and VO2peak test. Subjects' body height and weight were measured by a wall stadiometer (PAT #290237, Novel Products, Rockton, USA) and a digital weight scale (HD-366, Tanita, Tokyo, Japan). Then, resting blood pressure was measured using an autonomic device (BP79IT, Omron, Japan). Systolic blood pressure, diastolic blood pressure, and resting heart rate were recorded. Three sites of skinfold thickness, including chest, abdominal, and thigh were measured with a skinfold caliper (PAT #3008239, Beta Technology, Santa Cruz, USA) to estimate body composition. Each site was measured three times by the trained technician and an average value was recorded. The body density and body fat percentage was estimated by Jackson and Pollock and Siri's equation, respectively [28,29]. The VO2peak test was performed to accurately prescribe the exercise intensity for downhill running. The subject sat on the chair and wore a mask that connected with a metabolic cart (TrueOne 2400, Medics, USA). When the subject's respiratory exchange ratio (RER) dropped below 0.75, the graded exercise test begun on the treadmill. The Bruce protocol was applied at a speed of 2.74 km/h and a 10% grade. The speed and grade gradually increased every 3 min until subjects reached exhaustion. The VO2peak test was terminated when subjects met at least two of the following conditions: (a) showed signs of intense effort (heavy breathing, facial flushing, unsteady gait, and sweating), (b) respiratory exchange ratio (RER), 1.15, (c) rating of perceived exertion (RPE) ≥ 19 (Borg 6–20 scale), or (d) volitional fatigue.

### *2.5. Acute Eccentric Exercise*

The acute eccentric exercise program consisted of downhill running and drop jump exercise. The exercise program was modified based on the previous studies [30,31]. Prior to exercise, subjects performed jogging for 5 min on the 0-graded treadmill. Then, subjects performed downhill running for 20 min on a treadmill with a grade lowered to −10◦. The speed of downhill running was set at 60% of the subject's VO2peak. The average speed of downhill running was 6.4 ± 1.11 km/hour. After the downhill running, subjects performed drop jump exercises—five sets of 20 maximal drop jumps from the height of a 60 cm box. There was a 2-min rest interval between sets.

### *2.6. Psychomotor and Physical Performance Tests*

Psychomotor performance test: to assess the subject's psychomotor performance, the computer-based cognitive function test was performed. Prior to the test, subjects performed one trial for the familiarization during the pre-test. The test included psychomotor vigilance task (PVT) and delayed-match-to-sample task (DMS). The PVT test was composed of 40 trials for 5 min to assess mean reaction time while the DMS test consisted of 20 trials 20 min to assess memory ability [32].

Physical performance test: the physical performance test included straight leg raise, vertical jump performance, isometric leg strength, and anaerobic power. The straight leg raise was measured with a goniometer (Baseline stainless steel goniometers, USA) to evaluate the range of motion. The subjects lay on the medical bed in a supine position. Then, the subject raised the dominant leg as high as possible with the knee fully extended while the opposite leg remained on the bed firmly without movement [33]. Each subject performed two trials and the best score was recorded. Vertical jump performance was measured to evaluate the explosive power. The subject raised the dominant arm, fully extended, on the measuring board, and the point was marked as an initial point. Then, subjects performed a countermovement jump as high as possible and marked the highest point. The vertical jump performance (VJP )was calculated from the maximal jump height minus the initial point. Each subject performed three times with one-minute interval between trials, and the best score was recorded [33]. Isometric leg strength was measured by a hand-held dynamometer (Lafayette Instruments, Lafayette, USA). The device was validated in previous studies [34–36] and has been used to measure isometric muscle strength in healthy subjects [37]. Subjects were seated on the table in a supine position, and the knee and hip were positioned 90 degrees. The handheld dynamometer force pad was placed just proximal to the ankle joint area and the subject's knee extensor and flexor strength was recorded in kilogram force. All subjects performed two maximal isometric leg extension and flexions, holding for 3 to 5 s with a minute rest interval between trials. The highest score was recorded. Lastly, subjects performed anaerobic power tests measured by an Automatic Power Cycle

(Power Cycle, Austin, TX, USA). The device has been validated in previous studies [36]. First, subjects performed 2 min of warm-up cycling at 100 to 120 rpm with power of 100 to 120 Watt. Then, subjects performed cycling maximally for approximately 5 s. Each subject performed four trials with one-minute intervals between trials, and the highest peak and mean power were recorded [37]. Standardized verbal encouragement was applied to all subjects for their maximal performance.

### *2.7. Blood Analyses*

Blood sample was collected four times at baseline, 2 h, 48 h, and 96 h after acute exercise to analyze the markers of inflammation (interleukin 6; IL-6) and muscle damage (cortisol, myoglobin) as well as antioxidant capacity (total antioxidant capacity). Moreover, 3 mL of blood sample was withdrawn from an antecubital vein at each time point by the trained phlebotomist. The collected blood samples were drawn into a tube containing ethylenediaminetetraacetic acid (EDTA) to prevent coagulation. Then, the blood samples were centrifuged at 3000 rpm for 10 min. The upper layer of plasma was transferred into a conical tube and stored at –80 ◦C. Inflammation and muscle damage markers, including IL-6, myoglobin, and cortisol were analyzed by the enzyme-linked immunosorbent assay (ELISA) method using a commercially available Human ELISA IL-6 kit (EMD Millipore, Germany), myoglobin kit (EIA, Biocheck, USA), and cortisol kit (ELISA, DRG, USA). Total antioxidant capacity (TAC) was analyzed with the ELISA method using an antioxidant assay kit (total antioxidant, USA). All measurements were performed in duplicate. The coefficient variation (CV) of all blood analyses were below 10%.

### *2.8. Perceived Muscle Soreness*

Perceived muscle soreness was measured by a Visual Analogue Scale (VAS). The scale was distributed from 0 to 10 scales where '0 indicated not sore at all, and '10 indicated maximal soreness. Subjects were asked to point the score four times, at baseline, 2 h, 48 h, and 96 h after the acute eccentric exercise.

### *2.9. Data Analysis*

All data analyses were performed by a computer software program (SPSS version 25.0, SPSS Inc., Chicago, IL, USA). Prior to the data analyses, a Kolmogorov–Smirnov test was applied to check the normality of data. All variables exhibited normal distribution. Descriptive statistics, including mean, standard deviation (SD), and 95% confidence interval were computed for all variables. The repeated analysis of variances (ANOVAs) for conditions (WGE vs. placebo) by time (baseline, 2 h, 48 h, 96 h after acute exercise) were applied to assess the effects of WGE on dependent variables. If any significant interaction or main effects were identified, paired test for condition effects and repeated ANOVA with least significant difference (LSD) post-hoc test for time effects were applied. The effect size was presented as a partial eta-squared (η<sup>p</sup> 2) value. The effect size can be considered as "large effect", if the value is 0.14 or higher [38]. The level of statistical significance was set at 0.05.

### **3. Results**

### *3.1. Basic Characteristics*

All participants underwent pre-examination, including height, weight, resting blood pressure, heart rate, body fat percentage, and VO2peak tests. The results are presented in Table 2.

### *3.2. Psychomotor and Neuromuscular Performance*

The results of psychomotor performance are shown in Table 3. No significant interaction effects for condition by time were observed in the psychomotor performance, including psychomotor vigilance task (PVT) and delayed-match-to-sample task (DMS). However, the DMS correction (F = 8.844, p = 0.019, ηp <sup>2</sup> = 0.317) and DMS time (F = 4.961, p = 0.025, η<sup>p</sup> <sup>2</sup> = 0.355) were significantly reduced across the time in both conditions. Regarding neuromuscular performance, no significant interaction effects for

group by time were also identified in all performance variables. However, there was a significant time effect on straight leg raise (F = 4.100, p = 0.030, η<sup>p</sup> <sup>2</sup> = 0.313) where the straight leg raise value was significantly greater at 96 h compared to 2 h after exercise. The results of physical performance are presented in Table 4.


**Table 2.** Participants' basic characteristics.



Note. \* p < 0.05, indicate a significant different compared to baseline. DMS, delayed-match-to-sample task; G, group; G × T, group × time; msec, millisecond; PVT, psychomotor vigilance task; T, time; WGE, wild ginseng extract.


**Table 4.** The results of physical performance following acute eccentric exercise.

Note. \* p < 0.05, significantly greater compared to 2 h after acute eccentric exercise. WGE, wild ginseng extract; G, group; T, time; G × T, group × time.

### *3.3. Blood Analyses*

In this study, IL-6, cortisol, myoglobin, and total antioxidant capacity were analyzed to evaluate the changes of inflammation and muscle damage markers as well as antioxidant capacity following the acute eccentric exercise. There were no significant differences in all blood variables between the WGE and the placebo conditions. However, IL-6 (F = 5.671, p = 0.013, η<sup>p</sup> <sup>2</sup> = 0.387), myoglobin (F = 23.309, p < 0.001, η<sup>p</sup> <sup>2</sup> = 0.744) and cortisol levels changed significantly across the time (F = 3.553, p = 0.047, ηp <sup>2</sup> = 0.282). Especially, a significant increase in IL-6 was observed only in the placebo condition (F = 5.995, p = 0.014, η<sup>p</sup> <sup>2</sup> = 0.400) while no significant difference found in the WGE condition. The results of blood analyses are presented in Figure 1.

Note. \*p < 0.05, significantly increased only in the placebo trial at post 2 h + p < 0.05, significantly different compared to baseline, 48h, and 96 h in both trials WGE, wild ginseng extract

**Figure 1.** The results of blood markers of muscle damage.

### *3.4. Perceived Muscle Soreness*

There was no significant interaction effect for group by time on perceived muscle soreness measured by VAS. However, the increased in perceived muscle soreness was observed both the WGE and the placebo conditions after 2 h of acute eccentric exercise (F = 9.236, p = 0.001, η<sup>p</sup> <sup>2</sup> = 0.506). The results are shown in Table 5.


**Table 5.** The results of perceived muscle soreness following acute eccentric exercise.

Note. \* p < 0.05, significantly different compared to baseline, 48 h, and 96 h in both trials. WGE, wild ginseng extract; G, group; T, time; G × T, group × time.

### **4. Discussion**

The present study examined the efficacy of WGE on psychomotor and neuromuscular performance recovery following acute eccentric exercise in male adults. Our findings showed that the administration of WGE immediately after acute eccentric exercise and the following four days have no benefits on psychomotor and neuromuscular performance recovery. However, a favoring trend has been identified that a significant increase in IL-6 was observed only in the placebo condition, while no significant changes were found in the WGE condition.

### *4.1. Psychomotor Performance*

We examined the computer-based cognitive function test including PVT and DMS task to evaluate psychomotor performance during recovery. The supplementation of WGE following acute eccentric exercise did not provide a benefit on psychomotor performance in male adults. Consistent results have been identified in the previous study that 14 days of Korean ginseng (GINST15) supplementation prior to resistance exercise were not associated with the improvement of cognitive performance (i.e., quick board reaction time test) in healthy male and female adults [39]. A single dose of Panax ginseng (G115) for eight days also did not improve the working memory process in healthy subjects [40]. In contrast to our findings, Reay et al. reported that a single dose of Panax Ginseng (200 mg, G115) improves cognitive performance, such as serial seven task in healthy adults, but high dosages (400 mg, G115) and placebo conditions did not improve the cognitive performance [41]. Other Ginseng studies also demonstrated that cognitive functions, such as speed of the attention task [42], and secondary memory performance of cognitive research battery [43], improved following a single dose of Panax Ginseng (200 mg, 400 mg, respectively). Some possible explanations, such as the modulation of blood glucose and nitric oxide (NO) production have been proposed as potential mechanisms that may contribute to the improvement of cognitive performance [41]. It was believed that one of the active ingredients, ginsenosides, contributes to the enhancement of NO synthesis [44], and it modulates the glucose metabolism and insulin secretion in patients with type II diabetes mellitus [45]. Although some possible mechanisms, as well as benefits on cognitive function, have been demonstrated in the previous studies, our study confirms that five days of ginseng supplementation following acute eccentric exercise does not provide benefits on psychomotor performance recovery in male adults. In the present study, the DMS task correction and total time were significantly reduced across the time both in WGE and placebo conditions. We assume that decreased DMS correction partially related to 'speed accuracy trade-off'. Therefore, it is difficult to confirm that a decrease in cognitive function, such as DMS correction, was due to acute eccentric exercise.

### *4.2. Neuromuscular Performance*

In the present study, straight leg raise, vertical jump performance, isometric leg strength, and anaerobic power were measured as indicators of neuromuscular performance. We found that WGE supplementation has no benefit on neuromuscular performance recovery. Ginseng supplementation has long been believed to improve stamina or fatigue recovery, but it is unclear whether the apoptogenic properties can facilitate the performance recovery or not. In particular, the discrepancy among literature to confirm the roles as an ergogenic aid may be limited. While the favoring results have been found mostly in animal studies [15,16], the literature with human studies are limited. Pumpa et al. reported that five days of Panax Notoginseng supplementation, before and after downhill running exercise, did not show significant improvement in jump performance [27]. A recent study showed the 14 days of Korean Ginseng supplementation (GINST15, high dose: 960 mg/day, low dose: 160 mg/day) following acute resistance exercise (5 sets of 12 repetitions at 70% one-repetition maximum (1RMmax)) had no benefit on peak power in healthy adults [39]. However, authors pointed out an important consideration about the responder and non-responder effect in the supplementation research. Although this approach is not a novel concept, Cadwell et al. reported that the responders among high dose trials showed greater peak power compared to low dose and placebo trials [39]. Relatively long-term treatment (> 8 weeks) also demonstrated no benefits on neuromuscular performance, such as anaerobic peak power and mean power in active women [46]. As opposed to the current study, Liang et al. reported that 30 days of Panax notoginseng (1350 mg/day) supplementation have contributed to improve endurance performance time by >7 min in young adults [24]. Supplementation of ginseng saponin complex mixed with taurine and other substances for 15 days improved free fatty acid utilization during exhaustive cycling tests, which resulted in improving endurance performance by average 2.94 min in male adults [26]. Interestingly, the efficacy of ginseng supplementation has been observed in the particular area of performance, such as endurance performance [24–26]. A recent study reported 12-week administration of high-dose ginsenoside complex (500 mg/day) improved maximal oxygen consumption, but the efficacy was not found in muscle strength [25]. It has been proposed that ginseng supplementation improves energy utilization [15]. Although it is difficult to explain the outcomes of the animal study directly into human studies, Ma et al. reported that the administration of Panax ginseng extract improves glucose uptake, fatty acid utilization, and lower the lactate concentration that possibly contributed to the improvement of endurance swimming performance [15]. We assumed that if ginseng or active ingredients, such as ginsenosides, modulate the energy metabolism, such as glucose and free fatty acid utilization during exercise, the improvement of neuromuscular performance that predominantly utilized the energy through ATP–PCr and glycolytic energy system may be limited. In summary, we confirm that the administration of WGE immediately after an eccentric exercise, and the following four days, have no benefit on neuromuscular performance recovery, such as muscle strength and anaerobic power. Further studies may be needed to clarify whether WGE could act as a potential ergogenic aid for endurance performance.

### *4.3. Inflammation Markers of Muscle Damage*

Acute intense or eccentric exercise changes markers of muscle damage and inflammation resulted in inducing delayed-onset muscle soreness [19]. Various nutritional approaches have been introduced to attenuate the inflammation process or facilitate the recovery from muscle damages [47]. In this study, blood markers, including IL-6, myoglobin, cortisol, and TAC were not significantly different between the WGE and the placebo conditions following acute eccentric exercise. However, a favoring trend has been found that the serum IL-6 level significantly increased only in the placebo condition during 2 h recovery, whereas this value did not significantly change in the WGE condition. Similar results have been reported in previous studies [48–50]. Jung et al. reported that 7 days of Panax ginseng supplementation (20 g/d) prior to the uphill treadmill running (45 min at 10 km/h speed with a 15 degree) attenuated the increment of IL-6 level during 2 h and 3 h recovery period in young adults [48]. Another study also demonstrated that administrations of ginseng-based steroid Rg1 (5 mg) one night and one hour before 60-min cycling exercise suppressed the exercise-induced expression of IL-10 mRNA in quadriceps muscle [49]. Recently, Pumpa et al. reported that the level of serum IL-6 declined after downhill running in the ginseng group whereas the placebo group increased up to 4 h [27]. However, their findings revered 24 h after downhill running where the placebo group demonstrated a rapid return to baseline compared to the ginseng group. Interestingly, the placebo condition in this study also returned the IL-6 level to near baseline 24 h after eccentric exercise. A question arises based on the present and the previous studies whether the benefits of acute ginseng administration only valid during the first 2–4 h after eccentric exercise. However, this question should be approached carefully in the future study because the present and the previous data demonstrated inconsistent patterns from 24 h to 96 h after acute eccentric exercise. Although our study revealed a favoring trend at the first 2 h after the exercise, the benefit is not guaranteed until future research with a large sample size is conducted to clarify the anti-inflammation process in response to acute eccentric exercise.

Myoglobin is a muscle protein that is released from the damaged muscle following high intense eccentric exercise [19]. In the present study, the myoglobin level increased two hours after acute eccentric exercise, but no difference was found between the conditions. Lin et al. reported 12 weeks of Panax ginseng supplementation did not change inflammation adaptation, including muscle damage, oxidative, and inflammatory biomarkers to eccentric exercise training [51]. Pumpa et al. applied 5 bouts of 8 min downhill running at 80% of maximal heart rate (HRmax) with lowered −10◦ and participants received Panax notoginseng capsule (4000 mg/day) or placebo for 5 times (at 1 h prior to exercise and the following four days) in a crossover design. The author reported that myoglobin significantly increased after downhill running, but no significant differences were observed between conditions that support our result [27]. Cortisol is a glucocorticoid hormone that regulates energy metabolism during exercise [52]. It also interacts with inflammatory markers and nutritional supplementation, such as carbohydrates, and has been shown to lower cortisol levels after intense exercise [53]. In the present study, there were no difference in serum cortisol level between the WGE and placebo conditions. Youl et al. reported that a single bout of ginseng root extract (20 g) administered following a standardized exercise did not make a significant difference in cortisol levels [54]. Thus, it is confirmed that cortisol response following eccentric exercise was not influenced by ginseng supplement.

There are some strengths and limitations in the present study that need to be considered when interpreting the data. This study was well designed with randomized, double-blinded, placebo controlled, and crossover design. It assessed a wide range of markers of muscle damage and neuromuscular performance variables assessing effect of WGE in response to the eccentric exercise. However, small sample size may impact the statistical significance of the findings. Secondly, the outcomes can result differently by exercise protocols. The exercise protocol in this study was modified from the previous literature that was performed with healthy adults [30,31]. Although the current eccentric exercise protocol demonstrated a change of markers of muscle damage, such as myoglobin, IL-6, and cortisol, no significant changes in neuromuscular performance were observed after acute eccentric exercise. These results may tell us that the exercise protocol used in the present study may be insufficient to induce the delayed onset of muscle soreness.

### **5. Conclusions**

Our findings indicate that the administration of WGE (700 mg/day) immediately after eccentric exercise and the following four days have no benefits on psychomotor and neuromuscular performance recovery in male adults. However, the acute WGE supplementation may attenuate the eccentric exercise-induced inflammatory process, such as IL-6, but a future study with a large sample size is required to clarify the anti-inflammation process in response to acute eccentric exercise.

**Author Contributions:** Conceptualization, H.C.J. and S.L.; Formal analysis, H.C.J. and S.L.; investigation, H.C.J., N.H.L., Y.C.K., and S.L.; methodology, H.C.J., N.H.L., Y.C.K., and S.L.; Project administration, S.L.; Supervision, S.L. writing—original draft preparation, H.C.J.; writing—review and editing, H.C.J., S.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** All authors wish to thank the participants who participated in this study.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Explosive Strength Modeling in Children: Trends According to Growth and Prediction Equation**

**Vittoria Carnevale Pellino 1,2,**†**, Matteo Giuriato 3,**†**, Gabriele Ceccarelli 4,\*, Roberto Codella 5,6, Matteo Vandoni 1, Nicola Lovecchio 1,5 and Alan M. Nevill <sup>7</sup>**


Received: 26 August 2020; Accepted: 11 September 2020; Published: 15 September 2020

**Abstract:** Lower limb explosive strength has been widely used to evaluate physical fitness and general health in children. A plethora of studies have scoped the practicality of the standing broad jump (SBJ), though without accounting for body dimensions, which are tremendously affected by growth. This study aimed at modeling SBJ-specific allometric equations, underlying an objectively predictive approach while controlling for maturity offset (MO). A total of 7317 children (8–11 years) were tested for their SBJs; demographics and anthropometrics data were also collected. The multiplicative model with allometric body size components, MO, and categorial differences were implemented with SBJ performance. The log-multiplicative model suggested that the optimal body shape associated with SBJs is ectomorphic (H = −0.435; M = 1.152). Likewise, age, sex, and age–sex interactions were revealed to be significant (*p* < 0.001). Our results confirmed the efficacy of the allometric approach to identify the most appropriate body size and shape in children. Males, as they mature, did not significantly augment their performances, whereas females did, outperforming their peers. The model successfully fit the equation for SBJ performance, adjusted for age, sex, and MO. Predictive equations modeled on developmental factors are needed to interpret appropriately the performances that are used to evaluate physical fitness.

**Keywords:** allometry; standing broad jump; children; growth; maturity offset

### **1. Introduction**

The relevance of muscular strength and power is well recognized in human performance [1,2] and contributes to bone health across different age groups [3]. Several reviews highlighted health outcomes associated with muscular strength and showed that poor muscular strength in children and adolescents is associated with cardiovascular disease, metabolic profiles, skeletal health, and adiposity [4,5]. Other studies have shown that high muscular strength in youths is associated with better lipid metabolic profiles, independent of cardiorespiratory fitness [3]. Ideally, muscular power (or explosive

strength, as it is known in practical context) should be measured in lab settings [6], yielding valid values of muscular outcomes. Nevertheless, sometimes this may lack feasibility, and therefore field tests like the standing broad jump (SBJ) are preferable, since they are recognized as acceptable alternatives to lab assessments [7]. In addition, musculoskeletal fitness evaluation may enrich individual feedback to students and athletes, providing comparisons to international references on performance and health. For this reason, several youth fitness batteries [7–9] incorporated SBJ testing into their evaluation panel. Normative values for children and adolescents were created from these validation studies, proposing 50th percentiles as a proper medium level of performance that one could refer to. However, this procedure seems to have some limitations. For example, in Tomkinson et al. [10], the SBJ showed a mean improvement of 9 cm between 9–10 and 10–11 years (both males and females), while no information was obtained about changes in body mass and height, which differ according to age group and sex.

This stratification is useful for primarily surveillance purposes [11], but not for accurately monitoring training or developmental aspects. Moreover, childhood is a crucial period for detecting sensitive increases in performance (+10–15% per year) [12,13]. In children 7–8 years of age, SBJ performance is strictly dependent on body mass and height [14]. Although Body Mass Index (BMI) is the parameter summarizing the two variables, it does not represent a proper predictor of authentic performance [15,16]. Indeed, growth does not reflect parallel increases in this developmental pattern; while height increases by 3–4% per year, body mass increases by 8–10% [17]. Furthermore, a commonly used ratio scaling—the power-to-body mass ratio—may be a misleading index [18]. An alternative could be represented by the stratification of performance according to social condition [19] or BMI categories [16]. However, as mentioned earlier, this could result in a partial analysis. To convey these conceptual bottlenecks to a convincing ground of discussion, it is necessary to scale for maturity differences and determine the role of body size on performance [20], which is the allometric model. The objective of scaling is to normalize the performance for anthropometric characteristics [21]. In this framework, it is crucial to consider an adjunct polynomial model to analyze data and appropriately interpret them, discerning the contribution of developmental growth and maturation, among other factors, considering the exponential trend of human growth [22,23]. In the performance evaluation, another critical assessment item during the child growing phase could be the maturity offset (MO) at the peak of maturity. To this end, the peak high velocity (PHV) is capable of capturing a peculiar significance of speed and power results throughout growth, as well as the MO in strength (explosive strength) [16] or endurance outcomes [15]. In particular, the strength outcome depends on the developmental stage of the physiological determinants, whereas the peak of strength corresponds to the PHV [24]. Indeed, absolute values of specific strength actions, such as jumping or changing directions, are differently positioned in peak performance velocity curves as to PHV [25]. As such, controlling maturity is of paramount importance [26].

This study aims to model strength test-specific allometric equations, considering multiple individual factors such as height, body mass, sex, and age. Predictive equations will enforce a robust reliable indication of fitness per the structure and maturation levels of children (8–11 years old).

### **2. Materials and Methods**

### *2.1. Experimental Approach to the Problem*

During childhood, manifold factors could ameliorate performance, especially age and sex, according to the MO. In particular, the strength outcome depends on the developmental stage of the physiological determinants, whereas the peak of strength corresponds to PHV, where absolute values of specific strength actions are differently positioned in the peak performance velocity curves as compared with PHV. In this study, we decided to investigate the muscular power related to actual children's maturity, taking into account multiple individual proxies such as height, body mass, sex, and age. Children performed an SBJ as per the field test guidelines. The biological maturation (MO) was estimated using the somatic maturation method proposed by Werneck et al. [27]. Then, the age of PHV was determined by subtracting the MO from the chronological age. Participants were classified as late, early, or on time through the one-SD method derived from the current sample.

### *2.2. Subjects*

This cross-sectional study involved 7317 children (3627 girls) aged 8–11 years, recruited from 119 Northeast Italian primary schools (third, fourth, and fifth grade). Children with known chronic cardiac, respiratory, neurological, or musculoskeletal disorders were excluded. All the described measures were taken during physical education classes as scheduled in the morning activities (8:00–12:00 a.m.). The study protocol, including each feature of the experimental design, was approved by the ethical boards of the enrolled schools in accordance with the World Medical Association Declaration of Helsinki, as revised in 1983. All participants were free to withdraw their participation at any time. Written informed consent was obtained from the parents or legal guardians, while verbal assent was obtained from the children prior to participation.

### *2.3. Procedures*

Data collection included demographic and anthropometric information (sex, age, mass, and height) measured before the test sessions using standardized techniques (Table 1). Height was measured using a portable stadiometer with a precision of ± 1 mm (Stadiometer Seca 213, Intermed S.r.l. Milan, Italy) with children in an upright position, with bare feet placed slightly apart, arms extended, and head positioned in the Frankfort plane. Body mass was assessed using a beam scale with a precision of ±100 g (Seca 813, Intermed S.r.l. Milan, Italy), with children in light clothing, without shoes, and stood upright at the center of the platform of the mass scale. We calculated the age of the children from the birth date and subsequently rounded down values.

**Table 1.** Demographic and anthropometric information by sex. Numerosity following age: 8–9 y (1175 males, 1074 females), 9–10 y (1240 males, 1200 females), and 10–11 y (1275 males, 1353 females). All values are showed as mean ± DS.


The explosive strength of the children was measured by the SBJ test (systematic error nearly to 0) [5], a practical, time-efficient, and low-cost field test widely adopted in school or gym contexts [7,11,16,28]. Every child jumped for distance from a standstill. During the performance of the jumps, the children were asked to bend their knees with their arms in front of them, parallel to the ground, then swing both arms, push off vigorously, and jump as far as possible. The test was performed three times, scored in centimeters, and the best value was recorded. The score was obtained by measuring the distance between the last heel mark and the take off line [29].

All tests were conducted by a team of three students of the sport science degree course during scheduled physical education classes in the morning. Previous training and calibration of the operators were performed to ensure the accuracy and repeatability of the procedure. The presence and collaboration of the curricular teachers were guaranteed at any time to meet the confidence of the students [30].

### *2.4. Statistical Analysis*

The biological maturation (MO) was estimated through the somatic maturation method proposed by Werneck et al. [27].

This method estimates the MO from stature and chronological age using an algorithm, providing the result offset in years to peak height velocity (PHV). In particular, this easy approach was adopted because, briefly, involving only one anthropometric measure reduces operator-dependent errors, especially because it meets the full compliance of children and parents and avoids other maturity scales (i.e., Tanner; Marshall & Tanner 1969, 1970). In brief, the MO was determined using a specific formula for girls and boys:

$$\text{Maturity offset for girls (years)} = -7.709133 + (0.0042232 \times \text{(age} \times \text{height))}$$

$$\text{Maturity offset for boys (years)} = -7.999994 + (0.0036124 \times \text{(age} \times \text{height)})$$

Then, the age of PHV was determined by subtracting the maturity offset from the chronological age. Participants were classified as late, early, or on time through the one-SD method derived from the current sample.

The multiplicative model with allometric body size components of Body Mass (M) and Height (H) was used to identify the most appropriate body size and shape characteristics (Equation (1)) associated with, as well as detect any maturity (using MO) and categorical differences (e.g., sex, age) in, the physical performance variable (SBJ). The model is an extension of the one used to predict the physical performance variables of Greek children [23]:

$$\mathbf{Y} = \mathbf{a} \cdot \mathbf{M} \mathbf{k}^1 \cdot \mathbf{H} \mathbf{k}^2 \cdot \exp(\mathbf{c} \cdot \mathbf{M} \mathbf{O}) \cdot \varepsilon \tag{1}$$

where k1 and k2 are the ontogenetic allometric coefficients; a, b, and c are allowed to vary randomly from child to child; and Log (ε), is assumed to have a constant error variance. The constant a is also allowed to vary for different populations, in this case the fixed factor: sex.

The model has the advantage of having proportional body size components and a multiplicative error that assumes they will increase proportionally with the physical performance variable Y (e.g., see Figure 1a,b).

(**a**)

**Figure 1.** *Cont*.

**Figure 1.** (**a**) The association between SBJ and height in 8–11-year-old male children, and (**b**) the association between SBJ and height in 8–11-year-old female children.

The model (Equation (1)) can be linearized with a log transformation (Equation (2)). A linear regression analysis or analysis of covariance (ANCOVA) on log(Y) can then be used to estimate the unknown parameters of the log-transformed model:

$$\text{Log(Y)} = \log(\text{a}) + \text{k1} \cdot \log(\text{M}) + \text{k2} \cdot \log(\text{H}) + \text{c} \cdot \text{MO} + \log(\varepsilon) \tag{2}$$

Further categorical differences within the population (e.g., sex and age) can be explored by allowing the constant intercept parameter log(a) to vary for each group by introducing them as fixed factors (plus possible interactions) within an ANCOVA. The significance level was set at *p* < 0.05.

### **3. Results**

The explosive strength expressed by the SBJ revealed a mean value (SD) of 128.29 cm (22.76) and 121.17 cm (21.49), while the MO was, on average, 3.36 y (0.72) and 6.35 y (0.44) for boys and girls, respectively. A significant difference between sex was found (*p* < 0.001). Figure 2 shows the results between sexes (with the covariates factored in), suggesting that females outperformed males in the SBJ; females had greater performances than males by up to 20 cm (on average). Furthermore, the parameter Log a that results from interaction between height, mass, MO, sex, and age was 0.973, 0.985, 0.982, and 0.995 for males and 1.199, 1.211, 1.208, and 1.211 for females, respectively, for 8, 9, 10, and 11-year-olds.

The estimated parameters from the multiplicative model relating the SBJ distance to the body size components in Equation (1), incorporating the MO, are given in Table 2.

**Figure 2.** Estimated marginal means of SBJ by sex. Covariates appearing in the model are evaluated as follows: Ln Height = 4.9147, LN Mass = 3.4928, and MO = −4.8206.

**Table 2.** Parameters of Equation (1) about SBJ performance. The parameter Log a for males is 0.973, 0.985, 0.982, and 0.995, and 1.199, 1.211, 1.208, and 1.211 for females, respectively, for 8, 9, 10, and 11-year-olds. Note that the baseline group is the 11-year-old boys from which all other sex and age groups are compared.


The model (Equation (1)) relating the SBJ distance to the body size components was

$$\text{[SB] distance (cm)} = \text{a} \cdot \text{M}^{-0.435} \cdot \text{H}^{1.152} \tag{3}$$

With a positive height (H) and negative mass (M), the model suggested that the optimal body shape associated with the SBJ is to be taller, but lighter (less body mass).

Fittingly, the model (Equation (1)) revealed significant differences in the constant Log a parameter (Table 2) due to sex (*p* < 0.001) and age (*p* < 0.001), together with the interactions of age and sex (*p* < 0.001, Figure 3).

In particular, the parameter Log a appears in the footnote of Table 2. Note that the maturity offset (MO) made significant positive contributions to predicting the log-transformed SBJ distance (*p* < 0.001).

**Figure 3.** Estimated marginal means of the SBJ according to sex and age group. Covariates appearing in the model are evaluated as follows: LN height = 4.9147, LN weight = 3.4928, and MO = −4.8206.

### **4. Discussion**

Since childhood, musculoskeletal fitness evaluation has been a crude indicator of both child performance and general health. To evaluate muscular power, the SBJ was included in all field test batteries. Researchers created reference values to characterize growth curves, not only in terms of anthropometrics (body mass and height) but also regarding physical performance. A close relation does exist, in fact, between the BMI and SBJ performance. In contrast, recent studies found that the BMI has limitations for measuring growth during childhood. From the age of seven to eight, body mass and height did not follow the same growth trends, making the BMI an improper predictor of SBJ performance [11,31]. In this light, this study aimed to investigate SBJ performance according to a robust scaling in which height and body mass are objectively valued, considering their actual trends owing to PHV assessment as a crucial individual factor of all physical performances. The data of this study confirmed the efficacy of the allometric approach to identify the most appropriate body size and shape characteristics associated with SBJ performance through the PHV. For boys and girls 8–11 years old, it is the ectomorph one (M = −0.435; H = 1.1152). In fact, in explosive strength, the MO contributed to predicting log-transformed SBJ distance since childhood. Without an allometric approach, consistent with other studies [13,14], child SBJ performance differs between boys and girls, with boys outperforming in a constant growth trend. Instead, the introduction of the size approach through allometric modeling to investigate SBJ performance with MO evaluation allowed us to refine the analysis, releasing different results. Males throughout the studied time frame (8–11 years) did not significantly augment their performances, whereas females showed a significant increase with a better SBJ performance (Figure 2). In line with other results [31,32], males did not outperform females in SBJ performance until reaching 11 years old. Furthermore, Newton's second law suggests that the force acting on a body is directly proportional to acceleration and shares its direction and orientation, with a proportionality constant given by the mass of the body. With that being said, Martin et al. [33] also showed that during maturation, female peak power is most likely to be determined by the quantitative properties of the muscle. Even in our study, the influence of body mass and related MO on strength performance in females was supported by the significant relationship that emerged between the SBJ and PHV offset. Moreover, Armstrong et al. [18] identified that, during the growth period, an index of lower limb power (assessed through the Wingate test) is strongly linked with both body mass and

fat-free mass. This is probably explained by a lower body mass of the participants, which allows them to perform more efficiently. However, the possibility of a genuine comparison is hard due to the lack of data investigated with allometric approaches in the same age group. A small number of studies investigated such a performance during the early stages of maturity. Dos Santos et al. [34] suggested a better SBJ performance in 10–15-year-old boys controlled by body and shape with respect to their female peers. On the other hand, Meylan et al. [21] suggested that, when maturity was adjusted, the horizontal jump length results were slightly better (*p* = 0.04) for males than females.

Overall, the results of this study highlighted that females outperformed males in SBJ performance. A putative explanation may reside in the lower mass and higher coordination level of the females, with respect to males. Although our study did not investigate coordination, D'Hondt et al. [31] found an inverse correlation between motor coordination and lower mass in children of 5–10 years, and Stodden et al. [35] reported that children with a higher level of motor coordination exhibit greater performance (explosive strength). In this sense, sex differences in lower limb power performance have been related to the expression of significant differences in fat-free mass [36]. Subsequently, changes in force production at the age of 13–14 years have been attributed to the dramatic increase in steroid concentrations in males [37]. From a body shape view, the explosive strength in the lower limbs is facilitated through an ectomorphic body shape [23]. This is of relevance since these values are similar to those found by Lovecchio et al. [11] in middle school students (11–14 years old, Log(M) = −0.357, Log(H) = 1.302). Interestingly, in the two studies the trend was comparable, as though the explosive force paralleled the growth factors. Furthermore, the values of the a parameter for males and females (Table 2) are very practical, as they effectively complete the predictive equation for a field test widely used to evaluate physical fitness and, possibly, related general health beyond the performance extent by itself. Future studies may concern other field tests very commonly employed in children, such as speed and agility tests.

### **5. Conclusions**

This study suggested some practical applications for young people in the early stages of maturation. First, coaches of young athletes must be aware of the sex differences because of maturity issues; therefore, females with lower body masses than males may be favored in lower limb performance. Second, coaches should be adopting tests and sample-specific scaling modeled after body mass, rather than using theoretical models based on the assumption of body dimension similarity. From this perspective, the use of the parameter a (Table 2) would be of meaningful assistance. Third, sex-specific training may also be needed, given the early PHV in girls. This should make coaches and physical education teachers aware that females can train strength earlier than males. Furthermore, this approach was aligned with the bio-banding rationale that employed a stratification per maturity offset, within a competitive context (and not only assessment procedure or in-training evaluation) [38,39]. Finally, age, maturity offset, body shape, and body mass should be carefully considered when interpreting children's lower limb performance, as they are relevant indicators to the growth curves.

**Author Contributions:** Conceptualization, N.L., M.G., and A.M.N.; methodology, R.C., G.C., and N.L.; validation, M.G., M.V., and A.M.N.; investigation, R.C. and M.V.; data curation, M.G. and M.V.; writing—original draft preparation, R.C. and G.C.; writing—review and editing, N.L. and M.G.; supervision, N.L. and A.M.N.; project administration, V.C.P., R.C., and M.V. All authors have read and agreed to the published version of the manuscript.

#### **Funding:** This research received no external funding.

**Acknowledgments:** The authors would like to acknowledge all the students, teachers, and staff of the schools, university students, and graduates. All authors disclose professional relationships with companies or manufacturers who will benefit from the results of the present study.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## **Contributions of Anthropometric and Strength Determinants to Estimate 2000 m Ergometer Performance in Traditional Rowing**

### **Sergio Sebastia-Amat, Alfonso Penichet-Tomas \*, Jose M. Jimenez-Olmedo and Basilio Pueo**

Department of General and Specific Didactics, University of Alicante, 03690 Alicante, Spain;

sergio.sebastia@ua.es (S.S.-A.); j.olmedo@ua.es (J.M.J.-O.); basilio@ua.es (B.P.)

**\*** Correspondence: alfonso.penichet@ua.es

Received: 29 August 2020; Accepted: 18 September 2020; Published: 20 September 2020

**Abstract:** The purpose of this study was to analyze the contribution of anthropometric and strength determinants of 2000 m ergometer performance in traditional rowing. Nineteen rowers competing at national level participated in this study. Anthropometric characteristics, vertical jumps and bench pull tests were assessed to determine conditional factors, whereas the 2000 m test was used to set rowing performance. Pearson correlation coefficient, linear stepwise and allometric regression analyses were used to predict rowing performance (R<sup>2</sup> > 50%). Height, body mass and body muscle correlated with rowing performance in male and female rowers. Similarly, power output for squat jump and countermovement jump power correlated with performance. Finally, mean propulsive velocity, mean power and maximum power in bench pull also correlated with the test. Stepwise multiple regression analysis identified body mass (R<sup>2</sup> = 0.69, *p* < 0.001) and mean propulsive velocity in bench pull (R2 = 0.76, *p* < 0.001) for male rowers and body muscle (R2 = 0.89, *p* = 0.002) and maximum power in bench pull (R<sup>2</sup> = 0.62, *p* = 0.036) for female rowers as the best predictors of rowing performance. These results determine the relevance of anthropometric characteristics and, in contrast to Olympic rowing, support the greatest importance of upper body power in traditional rowing training.

**Keywords:** bench pull; vertical jump; power; talent detection; training

### **1. Introduction**

Rowing is a cyclic, strength-endurance sport that requires high levels of aerobic and anaerobic capacity to displace a boat through the water [1,2]. Rowers use the whole body to perform the rowing stroke for a distance that differs according to the modality [3]. There are two different rowing modalities, traditional—or fixed—rowing and Olympic rowing, with different performance indicators: distance, race time, mean force, total number of strokes and power per stroke and velocity of the boat [4]. Traditional rowing is a non-Olympic modality that demands high physical condition to carry out between 35–40 strokes per minute throughout a 19–20-min race, slightly longer than in Olympic rowing [4]. Traditional rowers show a 250–350 W average force applied for an optimal stroke length [4,5]. The power-capacity at each stroke has been identified as a key factor of rowing performance [4,6], together with other factors like large body size, relatively large limbs, high muscular strength, high muscular and cardiovascular endurance and proper balance [6–9].

Whereas approximately 46% of the power produced in Olympic modalities is generated by legs, the remaining is produced by trunk and arms [10]. In traditional rowing, the contribution of legs is slightly lower (40%) and the role of trunk and arms are slightly higher (60%) [11]. This fact may be due, among other factors, to the semi-flexed position of legs during the recovery of the traditional rowing cycle [12]. This position must be adopted as a consequence of the fixed seat, which entails an increase in the degree of body extension [13].

Traditional rowing is experiencing a significant increase in athletes worldwide and has attracted the attention of sports scientists in the same way was as in Olympic rowing. International research interest has enhanced due to physiological, performance and championships differences [4], the professionalization of this modality and the rise of worldwide championships in Europe (Spain, Italy, United Kingdom, etc.), America (Canada, USA, etc.) and other regions like Saudi Arabia [13]. Traditional rowing studies has increased in the field of sports profile [14], championships performance [13], supplementation [15], physiological factors [4] and different training characteristics and methodologies [4,16].

Most studies have mainly focused on describing performance factors in Olympic rowing [11], mainly those related with physiological and anthropometric variables and, to a lesser extent, in traditional rowing [4,13,16,17]. In the same way, research conducted to date relates the importance of some anthropometric characteristics like height, body mass, muscle mass or body fat in Olympic rowing performance [1]. Most of these studies have investigated the determinants of both Olympic and indoor rowing using correlation and linear regression techniques, assuming a linear relationship between rowing performance and determinants [8,18,19]. For curvilinear relationships between various measures of power output, a proportional allometric model can also be used [20,21]. However, the contribution of these determinants to rowing ergometer performance in traditional rowing has not been widely demonstrated. Several studies showed the relationship between 1500 m or 2000 m rowing ergometer performance in Olympic modality and lower body power obtained from different types of jump [8,22,23]. Similarly, a link has been shown between rowing ergometer performance and power produced by the upper body during different protocols of bench pull (BP) tests [4,24,25]. Even peak power output sustained during maximal incremental testing is an overall index of physiological rowing capacity and rowing efficiency and allows predicting rowing ergometer performance [26]. However, to the knowledge of the authors, there are no studies that relate the rowing ergometer performance in traditional rowing, considering that the contribution of the upper body is slightly higher in traditional rowing than in Olympic rowing [11,27]. Likewise, no available studies in the literature have considered the use of proportional allometric modeling to predict 2000 m ergometer rowing performance in traditional rowing.

Therefore, the purpose of this study was to analyze the contribution of anthropometric and strength determinants of rowing ergometer performance in traditional rowing. To that end, the relation of anthropometric characteristics and upper/lower body contribution with 2000 m rowing ergometer performance test was carried out using correlation, linear and allometric regression techniques.

### **2. Materials and Methods**

### *2.1. Participants*

Nineteen sweep (board) rowers competing at the national level participated in this study: 12 males (7 port and 5 starboard, 10 heavyweight and 2 lightweight, age: 24.6 ± 3.9 years, height: 178.4 ± 8.9 cm, body mass: 77.3 ± 7.9 kg) and 7 females (3 port and 4 starboard, 2 heavyweight and 5 lightweight, age: 25.7 ± 4.4 years, height: 166.3 ± 7.5 cm, body mass: 59.9 ± 8.3 kg). The requirement to participate was to have classified for the national championship, to train regularly a minimum of five days per week (> 12 h/week) for the last 3 years and not to have any musculoskeletal or neurological disorders, heart or respiratory failures, or any circulatory disturbance that may influence the results of the investigation. Rowers were requested to abstain from caffeine and alcohol consumption for 24 h and to avoid high-intensity training for 48 h before testing. All participants gave their written consent after project information, which was previously approved by the research ethics committee of the University of Alicante (IRB No. UA-2019-07-23).

### *2.2. Procedures*

To determine the contribution of the different variables in rowing ergometer performance, the strength of lower/upper body and 2000 m rowing ergometer performance were tested. Data collection was conducted on three sessions carried out at the same time of the day in a controlled laboratory environment. Dynamic muscle strength and power tests appear to discriminate better between levels of rowing performance than isometric strength tests [28,29]; therefore, during the second session, rowers performed vertical jump tests and the bench pull test (BP). Finally, rowers completed a 2000 m test on a rowing ergometer in the third session. All athletes were familiarized with the testing protocols used in the present study. Pearson correlation statistical test and stepwise multiple linear regression calculations were used to establish strong common variances shared between predictors.

Anthropometric measurements were collected with an astra stadiometer with a mechanical scale to measure height (0.1 cm) and with a body composition analysis device for body mass (0.1 kg), body muscle (0.1 kg) and percentage of fat mass (0.1%) (TanitaBC-545N) [1], through the use of bioelectrical impedance analysis, which is based on the rate at which a weak electrical current travels through the body [30,31].

### 2.2.1. Vertical Jump Test

Three vertical jump types, squat jump (SJ), countermovement jump (CMJ) and repeat jump (RJ) [13], were used to evaluate the lower body power with a jump-mat system (Chronojump, Bosco-System, Barcelona, Spain), capturing 1000 samples per second. Each participant completed three trials of each type of jump with a rest period between actions of 2 min. The best performance was used for data analysis. In SJ, participants began from a flexed position with knees to 90 degrees and hold this position for 3 s before executing jump without any countermovement. The CMJ test started with the standing position, hands on hips and using the countermovement to jump as high as possible after descending to the half squat position. The RJ followed the same procedure that CMJ but with a continuous execution during 30 s. Elastic index (EI) was calculated from the difference between two jump types (SJ and CMJ), mechanical power (MchP) was calculated with Test time (*T* = 30 s), flight time (*ft*) and the number of jumps (*n*) as MchP <sup>=</sup> (*g*2·*T*·*ft*)/[4*n*·(*T*−*ft*)]. Finally, to calculate the resistance index (RI) to fast strength, the average height reached in RJ was related to CMJ height as RI = *h*RJ/*h*CMJ [13]. Power output prediction equations based on body mass, jump height in SJ (*h*SJ) and CMJ (*h*CMJ) were: predicted power SJ = 60.7·*h*SJ + 45.3·body mass—2055 and predicted power CMJ = 51.9·*h*CMJ + 48.9 · body mass—2007 [32].

### 2.2.2. Bench Pull Test

The bench pull test is a specific tool to assess the pulling strength of the upper torso in rowers owing to similar shoulder adduction that take place during rowing stroke [3,24]. BP data were recorded by an optoelectronic encoder (Velowin, Deportec, Murcia, Spain) capturing 500 samples per second from which a dedicated software calculated velocity, power and force output for each repetition. Mean and maximum values of velocity, power and force, both for the entire concentric action and for the propulsive phase during BP were recorded. In such a test, the rower is laying, face down on the bench, whose height from the floor is adjusted according to the length of the rower's arms so that both elbows are in full extension and the arms completely suspended. The barbell is held with hands apart at shoulder level or slightly wider [3,17]. A light load (30% from the 1RM) as the minimum load that can discriminate the different levels of traditional rowers [4], was used for this study. Although some studies did not find a relationship between BP test and rowing performance, it may be a consequence of having used too high intensity to perform BP. The test was cancelled when the rower was unable to flex the arms sufficiently to touch the underside of the bench phase [33].

### 2.2.3. 2000 m Rowing Ergometer Test

The performance test was carried out on a rowing ergometer (Model D; Concept 2, Inc., Morrisville, VT, USA) [11,24,34,35] with the aim of reducing external influences, such as wind, temperature and waves, that could influence the final result [36]. The control of these cofactors is advisable in traditional rowing, practiced in open waters or rough sea, and in Olympic rowing, which is practiced in flat waters, mainly rivers and lakes. Therefore, a rowing ergometer allows for individual testing in a controlled way, providing a valid proxy for rowing performance [4]. The rowers performed an all-out 2000 m test on a rowing ergometer, with the drag factor set to 130 for males and 110 for females [24]. The warm-up consisted of 10 min of moderate intensity (heart rate below 140 beats per minute and 18–20 strokes per minute). The rowers' coach was continuously motivating and giving feedback to the rowers so that they could carry out the test in the shortest time possible. The rowers could see all the power, stroke rate, distance and time information on the screen of the ergometer. Power output, stroke rate and time to complete 2000 m rowing ergometer performance test were recorded.

### *2.3. Statistical Analysis*

The Statistical Package for Social Sciences (SPSS) v.24 program was used to compare the means of variables (IBM, Armonk, NY: IBM Corp). Descriptive statistics (mean ± SD) were used to report the characteristics of conditional factors. Shapiro–Wilk statistical test was used to determine whether the quantitative variables fulfil the criterion of normality. Pearson correlation coefficient (*r*) with 95% confidence intervals (CI) via bootstrapping was used to establish relationships between anthropometric characteristics, jump test, and bench pull test results with rowing performance. The magnitude of the correlation coefficient was interpreted with the following thresholds: 0.0–0.09 (trivial); 0.1–0.29 (small); 0.3–0.49 (moderate); 0.5–0.69 (strong); 0.7–0.89 (very strong); 0.9–0.99 (nearly perfect); and 1.0 (perfect) [37]. A stepwise multiple regression analysis was used to predict the 2000 m test performance. Additionally, a proportional curvilinear allometric scaling of 2000 m test performance was considered to identify key determinants of rowing performance [38,39]. For both regression techniques, predictors were iteratively adding and removing to the predictive model to find the subset of variables resulting in models which explained 50% or more of the variance of the data (R2 > 0.5). The resulting best-fit equations for the best predictive models of rowing performance within the current population for male and female rowers were also shown, together with adjusted R2 to account for non-significant predictors in the regression models. Statistical significance was set at *p* < 0.05.

### **3. Results**

As shown in Table 1, height resulted in strong correlation with performance (*r* = 0.68, *p* = 0.014) in male rowers. Body mass (*r* = 0.83, *p* < 0.001) and body muscle (*r* = 0.81, *p* < 0.001) had a very strong correlation but percentage of body fat and Body Mass Index (BMI) were not correlated with performance test. Jump tests heights (*h*SJ, *h*CMJ and *h*RJ), RI, EI and MP showed low correlation with performance test. Besides, the power output prediction equation based on body mass for SJ showed strong correlation with performance (*r* = 0.58, *p* = 0.048). Slightly higher correlations were found for CMJ (*r* = 0.70, *p* = 0.012). Mean velocity (*r* = 0.84, *p* < 0.001) and mean propulsive velocity (*r* = 0.87, *p* < 0.001) in BP test showed very strong correlation with performance. Likewise, mean power (*r* = 0.85, *p* < 0.001) resulted in very strong correlation and maximum power (*r* = 0.73, *p* = 0.007) showed very strong correlation with performance.


**Table 1.** Relationship of anthropometric and strength determinants with 2000 m rowing ergometer performance.

BMI: Body Mass Index; HSJ: squat jump height; WSJ: squat jump power; HCMJ: countermovement jump height; WCMJ: countermovement jump power; HRJ: repeat jump height; RI: resistance index; EI: elastic Index; MchP: mechanical power; MV: mean velocity; MPV: mean propulsive velocity; Vmax: maximum velocity; MF: mean force; MPF: mean propulsive force; Fmax: maximum force; MP: mean power; MPP: mean propulsive power; Pmax: maximum power; \* statistical significance *p* < 0.05; † statistical significance *p* < 0.01.

Female rowers showed strong to nearly perfect correlation between the same anthropometric variables than males with performance: height (*r* = 0.67, *p* = 0.101), body mass (*r* = 0.66, *p* = 0.107) and body muscle (*r* = 0.94, *p* = 0.002). Similarly to male rowers, the power output prediction equation based on body mass for SJ (WSJ) showed strong correlation (*r* = 0.57, *p* = 0.179), and stronger correlation values were found for CMJ (WCMJ) with performance (*r* = 0.59, *p* = 0.159). Finally, female rowers showed strong correlation in mean velocity (*r* = 0.63, *p* = 0.126), mean propulsive velocity (*r* = 0.67, *p* = 0.102) and mean power (*r* = 0.64, *p* = 0.124) in BP test with performance. Furthermore, very strong correlation in maximum power (*r* = 0.79, *p* = 0.036) were found.

The results of the stepwise multiple linear regression analysis in male rowers indicated that body mass is the only predictor variable for anthropometric characteristics explaining 69% (R2 = 0.69, *p* < 0.001) of W2000m. Similarly, the only predictor variable for BP test were the mean propulsive velocity that explained 76% of rowing performance (R<sup>2</sup> = 0.76, *p* < 0.001) (Table 2). The rest of the anthropometric and power variables in jump and BP tests did not contribute significatively and were excluded from the prediction equation. The best predictor of rowing performance among anthropometric characteristics for female rowers was body muscle, accounting for 89% of variance (R<sup>2</sup> = 0.89, *p* = 0.002) and maximum power for BP measures, explaining 62% of rowing performance (R<sup>2</sup> = 0.62, *p* = 0.036). As with the male linear regression models, the inclusion of the remaining variables resulted in models predicting less than 50% of the variance so they were excluded from the equations.


**Table 2.** Stepwise multiple regression analysis and proportional allometric scaling to predict rowing performance in male and female rowers according to anthropometric and power determinants.

SEE: standard error of estimate; W: power; MPV: mean propulsive velocity in bench pull; Pmax: maximum power in bench pull; M: Male; F: Female.

The proportional allometric model relationships between rowing performance (W2000m) and determinants (anthropometric and BP) showed approximately linear associations, indicated by the exponent near unity. The only prediction equation showing power-function characteristics is the mean propulsive velocity for male rowers, which explained 76% of the variance. The proportional slopes of the rest of the predicting equations gave similar prediction power (R2) but with lower estimate errors. The same variables that explain most of the variance in the stepwise linear regression were the remaining predictor variables making a significant contribution to the proportional allometric model. The remaining predicting variables lead to models explaining less than half the variance of rowing performance, so they were excluded from the equations.

### **4. Discussion**

The main aim of the study was to analyze the relationship of anthropometric and strength determinants with 2000 m rowing ergometer performance in traditional rowing. According to several studies, high-performance rowers of both sexes are usually heavier and taller than low-performance counterparts [9,40,41]. Our results are in accordance with other studies in which height, body mass and body muscle correlated with better performance [34,42,43]. Furthermore, among all variables, body mass for male rowers and body muscle for female rowers were the best predictors of rowing performance. Akça [1] found taller and heavier Olympic male college rowers (185.8 cm and 80.2 kg) compared to traditional rowers of our study (178.4 cm and 77.3 kg). This difference can be due to traditional rowing requiring shorter and lighter rowers in some boat positions for hydrodynamic reasons, so the crew must be not homogeneous [11]. For that reason, some physical advantages of heavier and taller rowers in Olympic rowing could become a disadvantage in boat hydrodynamics and rowing technique in traditional rowing [4,11,27].

In line with our findings, Yoshiga and Higuchi [44] reported that rowing performance is highly influenced by body size, in such a way that large body size increased the rowing performance. However, these authors found that female rowers were slower than males when both groups were matched based on the body size, possibly due to the larger body fat of females which deteriorates the rowing performance. Nevertheless, the differences between sexes in rowing performance were reduced when the fat-free mass was taken into consideration. Therefore, the results support the idea that muscle mass and fat-free mass are key factors related to rowing performance [45], especially in female rowers [19].

High percentage of body fat negatively affects rowing performance because body fat contributes a metabolically non-productive load [46] and low body fat percentage was associated with higher aerobic capacity [47]. Nevertheless, it is difficult to combine high level of musculature with low percentage of body fat [9]. In this study, body fat showed a small correlation with rowing performance [1,34]. This finding was consistent with previous studies which showed significant differences between age categories, although the differences between elite and sub-elite categories were minimal [4,41]. Despite this fact, it seems accepted that rowers with low body fat percentage perform a shorter time in the 2000 m test [4,48].

In this study, traditional male rowers had a low percentage of body fat of 11.9%, which is in accordance with Majumdar et al. [34], who reported similar values (11.1%) in the combined group (light body mass and open category), although elite rowers showed lower percentage of body fat (7.8%). The same trend is observed when the female group (20.9%) were compared with female elite rowers (16.3%). The difference between sex showed higher body fat percentage in the female group compared with the male group (Δ56.8%), usually accumulated around hips and thighs due to physiological and hormone characteristics [49]. Hence, rowers with high height and lean body mass values as well as low percentage of body fat seem to contribute to a greater power output stroke [9].

In the present study, there was a strong correlation in both sexes between power output in SJ (WSJ) and CMJ (WCMJ) with performance when body mass was considered. Greater muscle volume of the vastus lateralis could explain variance in rowing ergometer performance, sprint, and endurance capacity [50]. Battista et al. [8] did not find correlations between jump test and endurance test, probably since body mass, an important variable for rowing performance, has not been considered as an additional factor to assess vertical jump height [50]. To solve this problem, different authors proposed methods to estimate the lower limb power output during squat jump adding body mass variable to the equations [32,51]. Considering the contributions of these authors, the correlations between rowing performance test and jumps increased considerably. Comparing sex categories, our results showed a strong to very strong correlation between lower limb strength and rowing performance, although results were only significant in the male group. Similarly, Ingham et al. [20] reported a higher correlation between lower limb strength and rowing performance in male than female rowers, although these differences in correlation were considerably larger than our results.

The BP test, in contrast to Olympic rowing, showed higher correlations with rowing performance for all variables compared to jump tests. This fact can be due to the major contribution of the upper body in traditional rowing stroke. High correlation values between BP power average and rowing performance suggest that upper body power is one of the most important factors influencing the performance of traditional rowing, possibly due to the use of lower limbs in a flexed position and the greater degree of body extension compared to Olympic rowing [13]. In the same way, our results showed that mean propulsive velocity in BP was the best predictor of rowing performance for male rowers and maximum power in BP for female rowers, both for linear and proportional curvilinear (allometric) models, suggesting approximately linear associations between these determinants and rowing performance. The fact that seats are fixed in traditional rowing reduces legs freedom of movement and consequently their intervention in the stroke, although legs still have an important role in the first phase of paddling (isometric contraction). Comparing sex categories, male group showed higher correlation between BP power average and 2000 m test. These differences could be related to higher values of body muscle and less percentage of body fat of the male group since BP, 2000 m test is strongly correlated with the ratio between power and body mass [4]. Moreover, Attenborough, Smith, and Sinclair [52] studied the upper contribution to rowing performance of female rowers compared to male rowers and suggest to spend time to specific upper body conditioning as a key factor to improve rowing performance in the female category.

### *4.1. Limitations*

In this study, the vertical jump test was performed to evaluate the power of the lower body by the similarity of the jumping movement gesture and the first phase of the leg drive. Further research could perform other measures, such as the Wingate test, to assess lower-body power, more specifically, peak power, mean power and fatigue index, and the possible relationship with traditional rowing performance.

Another limitation of this study is the small sample size and, therefore, the findings of the study should be interpreted with caution. Future studies are required to confirm our results in a larger

population, including other traditional rowing modalities and nationalities to avoid a possible bias derived from an exclusively Spanish sample.

### *4.2. Practical Applications*

The results of this study provide further insight regarding the influence of different determinants of anthropometry and strength on 2000 m rowing ergometer performance in traditional rowing. This study demonstrates that some anthropometric characteristics may influence rowing success and a higher correlation between the upper body and rowing performance than lower body. Therefore, coaches should consider it to perform effective talent identification programs for rowing and training planning.

### **5. Conclusions**

In summary, the data presented within this investigation suggest that large values of height, body mass and body muscle were highly correlated with 2000 m rowing ergometer performance in traditional rowing. Furthermore, body mass for male rowers and body muscle for female rowers were found to be good predictors. The main results showed a strong correlation of WSJ and WCMJ with 2000 m rowing ergometer performance in traditional rowing. However, BP variables were those that most strongly correlated with performance, highlighting the relevant role of the upper trunk in traditional rowing. Mean propulsive velocity in BP in male rowers and maximum power in bench pull in female rowers were found to be good predictors of rowing performance.

**Author Contributions:** Conceptualization, S.S.-A., A.P.-T. and B.P.; Data curation, A.P.-T. and J.M.J.-O.; Formal analysis, S.S.-A., A.P.-T. and B.P.; Investigation, S.S.-A. and J.M.J.-O.; Methodology, A.P.-T., J.M.J.-O. and B.P.; Project administration, S.S.-A.; Resources, A.P.-T. and J.M.J.-O.; Software, S.S.-A. and B.P.; Supervision, B.P.; Validation, S.S.-A. and B.P.; Visualization, J.M.J.-O.; Writing—original draft, S.S.-A.; Writing—review and editing, S.S.-A., A.P.-T., J.M.J.-O. and B.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was supported by a pre-doctoral grant (ACIF/2018/209) from the Generalitat Valenciana, Spain, and vice-rectorate program of research and knowledge transfer for the promotion of R+D+I at the University of Alicante (Ref. GRE18-19).

**Acknowledgments:** The authors would like to express their gratitude to Jennifer Guerra, from the Department of Chemistry and Biochemistry, and to Elena Gandía García, from the Department of World Languages and Cultures, at the University of Nevada, Las Vegas, NV, USA, for providing assistance in preparing this manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## **E**ff**ects of Achilles Tendon Moment Arm Length on Insertional Achilles Tendinopathy**

### **Takuma Miyamoto 1, Yasushi Shinohara 2,\*, Tomohiro Matsui 3, Hiroaki Kurokawa 1, Akira Taniguchi 1, Tsukasa Kumai <sup>4</sup> and Yasuhito Tanaka <sup>1</sup>**


Received: 21 August 2020; Accepted: 22 September 2020; Published: 23 September 2020

**Abstract:** Insertional Achilles tendinopathy (IAT) is caused by traction force of the tendon. The effectiveness of the suture bridge technique in correcting it is unknown. We examined the moment arm in patients with IAT before and after surgery using the suture bridge technique, in comparison to that of healthy individuals. We hypothesized that the suture bridge method influences the moment arm length. An IAT group comprising 10 feet belonging to 8 patients requiring surgical treatment for IAT were followed up postoperatively and compared with a control group comprising 15 feet of 15 healthy individuals with no ankle complaints or history of trauma or surgery. The ratio of the moment arm (MA) length/foot length was found to be statistically significant between the control group, the IAT group preoperatively and the IAT group postoperatively (*p* < 0.01). Despite no significant difference in the force between the control and preoperative IAT groups, a significantly higher force to the Achilles tendon was observed in the IAT group postoperatively compared to the other groups (*p* < 0.05). This study demonstrates that a long moment arm may be one of the causes of IAT, and the suture bridge technique may reduce the Achilles tendon moment arm.

**Keywords:** insertional Achilles tendinopathy; Achilles tendon moment arm; suture bridge method

### **1. Introduction**

The Achilles tendon, attached to the posterior process of the calcaneus, is the largest and strongest tendon in the body. It is composed of the gastrocnemius muscle, originating from the condyle of the femur, and the soleus muscle, originating from the upper tibia. The gastrocnemius and soleus muscles, collectively known as the triceps surae, flex the ankle joint during gait. In the gait cycle, balance is maintained against gravity during the stance phase, and flexing the ankle joint during the toe-off phase provides a forward propulsive force [1].

The contraction of the triceps surae is exerted as a rotational force (torque) around the ankle axis, which can be calculated theoretically as the product of the muscle cross-section and the moment arm (MA) (Figure 1) [2–4].

In general, the length of the MA is known to change depending on the angles of plantarflexion and dorsiflexion of the ankle joint, which is different in each individual [2–4]. However, the length of the MA of the triceps surae has been reported to change depending on the angle of plantar- and dorsiflexion of the ankle joint [5–7]. Yet, it has also been reported that motion does not cause any change in the length of the MA [8,9], and that the plantarflexion moment length, from the plantarflexion region to the dorsiflexion region, is different in each individual, with little change caused by motion [2–4]. Thus, the changes in the length of the MA in motion has attracted great attention. Moreover, the shape

of the foot has been shown to influence the Achilles tendon MA [10]. However, these studies examined the length of the MA only in healthy individuals and it has not been investigated in patients with Achilles tendon disorders.

**Figure 1.** The contraction of the triceps surae can be calculated theoretically as the product of the triceps surae muscle cross-section area and the Achilles tendon moment arm using the lever rule.

The center of rotation (COR) and tendon excursion (TE) methods are mainly used for the measurement of the lever arm of Achilles tendon [5], with MRI and ultrasound generally used for these methods, respectively. Manja et al. [10] proposed a new method using a weight-bearing X-ray. The method allowed the evaluation of the morphology and anatomical position of the bone and soft tissue. Therefore, we are now able to investigate in various patients with Achilles tendon disorders.

Insertional Achilles tendinopathy (IAT), an Achilles tendon disorder, is caused by the traction force of the Achilles tendon [11,12]. This traction force is influenced by the rotational torque of the ankle joint, and the moment arm length is proportional to the plantar flexion torque of the ankle joint [13]. Therefore, when force is applied to the ankle joint, the Achilles tendon moment arm length is inversely proportional to the traction force applied to the insertion of the Achilles tendon. On the other hand, with a short moment arm, the circular arc associated with ankle rotation is thought to shorten, reducing the length displacement of the muscle-tendon complex, and further reducing the consumption of elastic energy of the Achilles tendon [13,14]. Thus, the moment arm of the ankle joint is expected to influence the traction force applied to the insertion of the Achilles tendon, the length displacement of the muscle-tendon complex, and the elastic energy.

Surgical treatments for IAT include para-tendon excision, adhesion detachment, intra-tendon degeneration excision, and calcaneal wedge osteotomy [15,16]. Of these, we preferred surgical treatment of insertional Achilles tendinosis with reattachment of the Achilles tendon using the suture bridge technique (Figure 2). This method firmly fixes the enthesis by pressing the tendon with the surface, and good outcomes have been reported with this method [17].

However, due to the lack of basic research on this surgical technique, its effectiveness remains unknown. This study aimed to examine the moment arm length as an anatomical feature in patients with IAT. We hypothesized that the suture bridge method for IAT influences the moment arm length.

ķ ĸ Ĺ ĺ **Figure 2.** We performed a reconstruction of the Achilles tendon with a suture bridge technique. -**1** : Bone spurs, calcifications in the tendon, and Haglund's prominence is removed. -**2** : Two Swive Locks with Fiber Tape inserted in the proximal hole. -**3** : Fiber Tape is passed through the Achilles tendon and inserted into the prepared distal hole using Swive Lock. -**4** : The tails on the distal row flushed to the anchor. (**A**): Preoperative X-ray. (**B**): Postoperative X-ray for the same case.

### **2. Materials and Methods**

This is a retrospective study. Fourteen individuals (16 feet) who required surgical treatment for IAT at our hospital between 2014 and 2019 were included in the IAT group. Those with a history of collagen diseases or trauma other than Achilles tendon injuries, and those who could not be followed up for more than one year after surgery were excluded from the study. As a result, 8 individuals (10 feet; IAT group) were examined in this study. Fifteen healthy individuals (with no complaints of ankle joint injuries nor history of surgical trauma) were included for examination in the control group.

The reconstruction in all IAT cases was performed using the Achilles SpeedBridge system (Arthrex, FL, USA). After extending the Achilles tendon in the midline and partly separating it from the insertion, the osteophytes, intra-tendinous degeneration site, and posterior superior calcaneus at the enthesis were excised. Subsequently, fiber tape was used to fix the Achilles tendon to the calcaneus using four knotless anchors. We analyzed a clinical evaluation by the American Orthopedic Foot and Ankle Society ankle/hindfoot scale (AOFAS scale) and visual analog pain scale (VAS) 1 year after the surgery.

At 6 months to 1 year after the surgery, 1.5T MRI (Vantage Titan Ver3.0, Canon, Tokyo, Japan) images were taken to evaluate the morphology of the Achilles tendon insertion area and abnormal signal in the tendon before and after surgery.

Plain X-ray images were used to measure the lever arm. The images of the affected side were taken for patients in the IAT group, while those of the healthy side were taken for patients in the control group. As described by Manja et al. [10], who used plain X-ray images of the feet during weight-bearing, the weight-bearing plane was taken as the plantar axis from the lateral image; a straight line perpendicular to this plantar axis, passing through the center of the Achilles tendon, was set as the central axis of the Achilles tendon. The shortest distance from the central axis of the Achilles tendon to the center of the first metatarsal head parallel to the plantar axis was defined as the length of the sole. A perfect circle was drawn to fit the pulley surface of the talus, and the center point was used as the center of motion. The shortest distance from the center of motion to the central axis of the Achilles tendon was defined as the MA length. In addition, the angle between the Achilles tendon moment arm and the line from the center of the ankle joint movement to the insertion of the Achilles tendon was defined as the α angle (Figure 3). The central axis of the Achilles tendon at the time of postoperative measurement was the midpoint from the most dorsal surface to the bottom of the insertion (Figure 4). Also, the ratio of the MA length/foot length was used to standardize the difference in foot size. Furthermore, in order to take the influence of body weight into account, the force applied to the Achilles tendon at the time of heel raise can be calculated as the rotational force around the ankle joint axis. Then, the force applied to the Achilles tendon, F, can be estimated by:

$$\mathbf{F} = \mathbf{w} \times (\mathbf{L} - \mathbf{M}\mathbf{A}) / \mathbf{M}\mathbf{A}$$

where w is the body weight, MA is the Achilles tendon moment arm length, and L is the length of the sole [14,18,19] (Figure 5). In the present study, this formula was used to calculate the force applied to the Achilles tendon, and the average value of three measurements performed by one examiner was used.

Both the control group and the IAT group were examined, and all measurements were taken preand post-operatively. Statistical tests were performed using Microsoft Excel. Data were tested using unpaired student *t*-tests, and effects were considered significant at 95% CI and *p* < 0.05. All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Nara Medical University (Approval number: 2740).

**Figure 3.** *Cont*.

**Figure 3.** The measurement of the lateral weightbearing radiograph were (A) the centerline of the Achilles tendon insertion, (B) plantar axis, (C) the center of rotation of the ankle, (L) the length of the sole, (MA) the Achilles tendon moment arm (α). The angle between the Achilles tendon moment arm and the line connecting the insertion point and the center of rotation.

**Figure 4.** B: Plantar axis. Af: Front edge of the Achilles tendon insertion. Ar: Rear edge of the Achilles tendon insertion. The central axis of Ar and Af was defined as Ap (the central axis of post-operated Achilles tendon insertion).

**Figure 5.** (A) The centerline of the Achilles tendon insertion, (C) the center of rotation of the ankle, (L) the length of the sole, and (MA) the Achilles tendon moment arm. (F) Traction force of the Achilles tendon. (W) Reaction force by body weight. We calculated F using principle of lever.

### **3. Results**

The average height of subjects in the healthy group was 163 ± 7.4 cm, with average body weight and BMI of 64.5 <sup>±</sup> 12.1 kg and 24.3 <sup>±</sup> 4.8 kg/m2, respectively. The average height of subjects in the IAT groups was 163.6 <sup>±</sup>13.3 cm, with average body weight and BMI of 73.5 <sup>±</sup> 11.0 kg and 27.7 <sup>±</sup> 4.5 kg/m2, respectively. Thus, there was no significant difference between the groups for these items (Table 1).


**Table 1.** Patient Characteristics.

Abbreviations: IAT, insertional Achilles tendinopathy. BMI, body mass index. Values are given as mean ± standard deviation (SD).

The AOFAS scale and VAS improved in all cases (AOFAS scale: Preoperative averaged 71.7, postoperative averaged 92.3; VAS: Preoperative averaged 5.0, postoperative averaged 0.83).

MRI assessment after surgery for IAT revealed Achilles tendon insertion running from the proximal knotless anchor insert to the distal knotless anchor insert via the suture bridge technique in all cases (Figure 6). Also, MRI findings showed the disappearance of abnormal signals in the Achilles tendon in all cases.

**Figure 6.** The MRI images of T2-weighted images of the same patient. ↔: Length of Achilles tendon insertion. (**A**): Pre-operation of the Achilles tendon insertion. (**B**): Post-operation of the Achilles tendon insertion. We can see the postoperative Achilles tendon insertion presence from the proximal knotless anchor insert to the distal knotless anchor insert.

There was no significant difference in the α angle between the control and IAT groups. The ratio of MA length/foot length was 0.30 <sup>±</sup> 0.9 <sup>×</sup> 10−3, in the control group, 0.31 <sup>±</sup> 1.0 <sup>×</sup> 10−2, in the IAT group preoperatively, and 0.27 <sup>±</sup> 0.2 <sup>×</sup> 10−<sup>1</sup> in the IAT group postoperatively, showing significant differences between these groups (*p* < 0.05). The force applied to the Achilles tendon was 148.5 ± 24.9 kg in the control group, 166.2 ± 23.0 kg in the IAT group preoperatively, and 200.8 ± 25.4 kg in the IAT group postoperatively. Despite no significant difference in the force between the control and preoperative IAT groups, a significantly higher force to the Achilles tendon was observed in the IAT group postoperatively, compared to the other groups (*p* < 0.05) (Figure 7).

**Figure 7.** Radiographic Measurement. (**A**): Comparison of the α angle between the control and pre-operative IAT groups. (**B**): Comparison of the ratio of MA length/foot length between the control group, the preoperative IAT group, and the postoperative IAT group, showing significant differences between these groups (*p* < 0.05). (**C**): Comparison of the force applied to the Achilles tendon between the control group, the preoperative IAT group, and the postoperative IAT group, showing that a significantly higher force to the Achilles tendon was observed in the IAT group postoperatively, compared to the other groups (*p* < 0.05). Values are given as mean ± SD.

### **4. Discussion**

COR method and TE methods are mainly used for the measurement of the lever arm of Achilles tendon [5]. It has been reported that the TE method shows a shorter moment arm than the COR method [5]. However, it has also been reported that there is no difference in measurements between the two methods [20]. Thus, there is no unified view on this issue. Furthermore, the Achilles tendon moment arm has been shown to be influenced by weight-bearing [21], and an evaluation method using weight-bearing X-ray has been developed for taking images during weight-bearing [10]. It is also feasible to use radiographs to measure the Achilles tendon moment arm [10]. In the present study, we examined the group of patients with IAT before and after surgery, in comparison to a control group of healthy individuals. In order to measure the moment arm and evaluation of the morphology of the bone during weight-bearing in each group, we decided to use a weight-bearing X-ray for evaluation. Also, the usage of an X-ray is a simple method and allows us to perform anatomical calculations easily. Therefore, we determined that it was a good method in our study.

A study using this measurement method reported a moment arm length of 64.3 mm [10]. In the present study, the average moment arm lengths of the healthy and IAT groups were 51.7 mm and 53.7 mm, respectively, both of which were shorter than the previously reported value. However, the length of the sole of subjects in the previous report was 185.7 mm, in contrast to 171 mm and 170.1 mm in healthy and IAT groups, respectively, in the present study. This suggests that the moment arm length may be influenced by the difference in the physical constitution.

In addition, Manja et al. [10] showed that the moment arm length and α angle are influenced by the difference in the shape of the foot, such as flat and concave feet. In the present study, no significant difference in α angle between the control and experimental groups was observed, indicating that the results are not influenced by the difference in the shape of the subject's foot, even in flat and concave feet.

To date, there have been no reports comparing the Achilles tendon moment lengths between healthy individuals and patients with IAT, but many factors are known to influence the pathophysiology of IAT. Lyman et al. [22] stated that it is caused by the traction force of the Achilles tendon during ankle movement. In the present study, those with IAT had a significantly longer Achilles tendon moment arm than healthy individuals. However, there was no significant difference in the force applied to the Achilles tendon between two groups. This indicates that the traction force of the Achilles tendon, which is the cause of IAT, is not influenced simply by the force applied to the Achilles tendon.

It has been shown that the Achilles tendon exerts approximately 4 to 6 times its weight during running [23,24], stretching approximately 8% of the total length of the Achilles tendon [25]. It has also been shown that, despite the smaller force applied during walking than in running, approximately 7% of the total length of the Achilles tendon is stretched during walking [26]. This indicates that the insertion of the Achilles tendon is influenced not only by simple tension but also by the elastic energy due to the stretching of the Achilles tendon [27]. In fact, triceps surae eccentric exercise has been recommended as a conservative therapy for IAT [28,29], and the extension of the Achilles tendon is a major cause of IAT. Our data revealed that individuals with IAT had a longer moment arm length than healthy individuals. This suggests that, with a long moment arm, the arc length associated with the ankle rotation lengthens, increasing the length displacement of the muscle-tendon complex, and further increasing the elastic energy consumption of the Achilles tendon [13,14]. Therefore, the elasticity of the Achilles tendon is shown to be greatly responsible for the development of IAT.

In addition, any measurements of the moment arm length before and after surgery using the suture bridge technique have not been reported to date. As a surgical treatment for IAT, the calcaneal wedge osteotomy takes the influence of the moment arm into account and aims to reduce stress on the insertion of the Achilles tendon by shortening the calcaneus length and slightly lifting the insertion of the Achilles tendon [30]. The suture bridge technique used in the present study has also been reported to have good long-term outcomes [17,31]. This method can treat various pathological conditions that influence IAT with minimal invasion [31]. Furthermore, by comparing the preoperative and postoperative data in this study, the postoperative Achilles tendon moment arm was found to be shortened. On the other hand, the force applied to the Achilles tendon was significantly increased in individuals with IAT compared to healthy individuals. This indicates that this method may not reduce the force applied to the Achilles tendon. Rather, it may reduce the length displacement of the Achilles tendon by reducing the Achilles tendon moment arm, thereby reducing the energy applied to the Achilles tendon.

There are three limitations to this study. First, the number of cases was small. The study needs to proceed with an increased number of cases, and proper statistical analysis needs to be performed. Secondly, the Achilles tendon insertion is an area, but for the sake of simplicity, it was considered to be a point. Performing 3D analyses allows more precise analyses, and we will need to investigate the area of the Achilles tendon insertion. Third, the moment arm measurement method used in this study was the static evaluation. An X-ray or MRI scan is required to measure the Achilles tendon moment arm after the reconstruction of Achilles tendon insertion using the suture bridge technique. However, it is difficult to measure the dynamic change of the moment arm before and after surgery with the measurement methods reported to date. Therefore, a new measurement method needs to be developed for measuring a dynamic moment arm using X-ray or ultrasonic waves in the future.

### **5. Conclusions**

In this study, we examined the moment arm in patients with IAT before and after undergoing the suture bridge technique compared to that of healthy individuals. The Achilles tendon moment arm was significantly longer in patients with IAT. We hypothesized that the suture bridge technique for IAT influences the moment arm length. To test this, we examined the changes in the moment arm length of patients with IAT before and after the surgery by analyzing the X-ray images of the lateral side of their feet during under weight-bearing. This study demonstrates that a long moment arm may be one of the causes of IAT, and the suture bridge technique may reduce the Achilles tendon moment arm. These findings suggest that it is important to consider the influence of the Achilles tendon moment arm in the treatment of IAT.

**Author Contributions:** Conceptualization, Y.S.; methodology, Y.S. and T.M. (Takuma Miyamoto); software, T.M. (Takuma Miyamoto); validation, T.M. (Takuma Miyamoto), Y.S. and Y.T.; formal analysis, T.M. (Takuma Miyamoto); investigation, T.M. (Takuma Miyamoto); resources, Y.T., T.K., A.T., H.K. and T.M. (Tomohiro Matsui); data curation, Y.T., T.K., A.T., H.K. and T.M. (Tomohiro Matsui); writing—original draft preparation, T.M. (Takuma Miyamoto); writing—review and editing, Y.S. and Y.T.; visualization, T.M. (Takuma Miyamoto); supervision, Y.S. and Y.T.; project administration, Y.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Immediate E**ff**ects of an Inverted Body Position on Energy Expenditure and Blood Lactate Removal after Intense Running**

### **Moo Sung Kim <sup>1</sup> and Jihong Park 2,\***


Received: 5 September 2020; Accepted: 21 September 2020; Published: 23 September 2020

**Abstract:** We compared the immediate effects of a cool-down strategy including an inverted body position (IBP: continuous 30-s alternations of supine and IBP) after a short period of an intense treadmill run with active (walking) and passive (seated) methods. Fifteen healthy subjects (22 years, 172 cm, 67 kg) completed three cool-down conditions (in a counterbalanced order) followed by a 5-min static stretch on three separate days. Heart rate, energy expenditure, blood lactate concentration, fatigue perception, and circumference of thighs and calves were recorded at pre- and post-run at 0, 5, 10, 20, and 30 min. At 5 min post-run, subjects performing the IBP condition showed (1) a 22% slower heart rate (*p* < 0.0001, ES = 2.52) and 14% lower energy expenditure (*p* = 0.01, ES = 0.48) than in the active condition, and (2) a 23% lower blood lactate than in the passive condition (*p* = 0.001, ES = 0.82). Fatigue perception and circumferences of thighs and calves did not differ between the conditions at any time point (F10,238 < 0.96, *p* < 0.99 for all tests). IBP appears to produce an effect similar to that of an active cool-down in blood lactate removal with less energy expenditure. This cool-down strategy is recommended for tournament sporting events with short breaks between matches, such as Taekwondo, Judo, and wrestling.

**Keywords:** cool-down strategy; heart rate; fatigue perception

### **1. Introduction**

Progressive overload is an essential principle in developing physical fitness, and it often requires fatigue-inducing training [1]. Fatigue is defined as a physiological (and psychological) constraint that results in failure to maintain a certain amount or intensity of voluntary muscle contractions [2]. By overloading exercise-induced fatigue, athletes may experience a temporary (e.g., a few days) performance reduction, referred to as functional overreaching [3]. Nonfunctional overreaching is also considered as the normal process of training adaptation, although it may take a longer time (e.g., a few weeks) to restore function [4]. Overtraining results in performance decrease with persistent physical and/or psychological fatigue for an extended period of time (e.g., a few months) [5]. While athletes and coaches seek optimal training adaptations, inadequate physiological (and psychological) recovery between training sessions can lead to overtraining [6]. Therefore, cool-down strategies to balance training stimulation and physical recovery are just as important as the training program.

Since "active recovery", defined as submaximal exercise or movement immediately following a training session, was first introduced in 1975 [7], its relative superiority over passive recovery has been reported in many studies [8–11]. Specifically, faster decreases of blood lactate [9], intramuscular pH level (due to reduced H+) [12], and oxyhemoglobin [13] as well as faster phosphocreatine resynthesis [14] are known advantages although the most effective working intensity remains controversial [9,15]. Enhancement of oxidation due to a higher rate or volume of blood flow to the working muscles [15] enhances the lactic acid removal [16] and prevents H+ release [17], which allows faster recovery from exercise-induced fatigue. Because of the increase in blood flow, practicing active recovery is preferred, especially when multiple exercise bouts are performed.

If the rate or volume of blood flow is an important causal factor, an inverted body position (IBP) might expedite the recovery process. This gravity-independent position (elevation of the lower body) decreases vascular hydrostatic pressure, resulting in increased venous return and lymphatic drainage [18]. According to the Frank−Starling mechanism [19], an increased venous return increases myocardial stretch (preload), which results in increased stroke volume and cardiac output [20]. Therefore, IBP activity could increase peripheral circulation in the lower extremities and enhance the process of physiological recovery. Previously, a position of head-down tilt at 15◦ was shown to increase stroke volume, mean arterial pressure, and cutaneous vascular conductance [21,22]. Another study [23] used the same position for 5 min and reported a 36% increase in tibial microvascular flow. Although an acute increase of cardiovascular function from holding a head-down tilt position is evident [21–23], changes of blood lactate removal and fatigue perception have not been evaluated.

Hence, the primary purpose of this study was to examine how a 10-min cool-down protocol containing IBP affects physiological and psychological recovery after intense running compared to the same duration of active (walking) and passive (seated) cool-down strategies. Since the effects of increasing blood flow were achieved on the 15◦ head-down tilt [21–23], our IBP protocol had a maximal inclination (up to 90◦) in expectation of the greatest effects of circulation of blood and body fluids [18]. Considering results from previous head-down tilt studies [21–23], the effects of elevation [18], and the Frank−Starling mechanism theory [19], we hypothesized that the lactate removal rate at the end of IBP would be greater than that of passive recovery at the same time point. Due to more voluntary muscle contractions and movements, we further expected higher values of heart rate and energy expenditure in the active recovery condition than in the IBP and passive recovery conditions.

### **2. Materials and Methods**

### *2.1. Experimental Design*

We used a three (condition) × six (time) single-blinded cross-over design with repeated measures of time. To objectively increase exercise-induced blood lactate, a treadmill run was implemented. In this approach, we assumed that heart rate increase by exercise correlates linearly with the amount of blood lactate. To test the condition effect, dependent measurements among the conditions were compared at each time point. Each condition had a 10-min cool-down activity that consisted of 5 min of IBP, walking on the treadmill or sitting on a chair, and another 5 min of static stretching (same activity for each condition). Therefore, the post-run 5-min time point reflected the acute effect of the IBP whereas the post-run 20- and 30-min time points corresponded to the combined effect of IBP and stretching. The temperature and relative humidity in the laboratory were maintained at 25 ◦C and 50% during the data collection period. One researcher guided the treadmill run and conducted the measurements while being blinded to the cool-down condition by leaving the laboratory while another researcher conducted the cool-down protocol. The study was approved by the Institutional Review Board and was conducted in accordance with the Declaration of Helsinki.

### *2.2. Subjects*

Fifteen healthy active adults (5 females: 20.4 ± 0.6 years, 161.4 ± 3.3 cm, 57.8 ± 4.2 kg; 10 males: 22.9 ± 0.9 years, 176.5 ± 1.9 cm, 72.1 ± 0.7 kg) who were participating in club sports (>240 min/week) were recruited for this study. Subjects had no history of lower-extremity surgery and were free from musculoskeletal injury within the last six months. Subjects with current lower-extremity musculoskeletal pain, anemia, diabetes, or cardiovascular disease were excluded. Prior to participation, all subjects gave informed consent, as approved by the University's Institutional Review Board.

### *2.3. Procedures*

Each subject visited the laboratory three times (with a week wash-out period) on separate days at the same time of day (Figure 1).

**Figure 1.** Testing procedures, IBP: inverted body position. Heart rate, energy expenditure, fatigue perception, circumference of lower-extremity, and blood lactate were recorded at each measurement time.

Upon arrival at the laboratory for the first time, subjects read and signed the informed consent form. At each visit, subjects were fitted with a heart rate monitor, and asked to lie down on a treatment table for a 10-min rest to achieve cardiovascular stability before measuring their resting heart rate.

Pre-run measurements were recorded in the order of heart rate, energy expenditure, fatigue perception, circumference of the thighs and calves, and blood lactate (measurements were recorded in the same order at all other time points). Subjects then performed an exercise protocol on the treadmill (Jog Forma, Technogym S.p.A, Gambettola, Italy) where the run began at 4 km/h without treadmill inclination and the speed and inclination increased by 1 km/h and 0.5% every minute. The exercise protocol was ended when the subjects reached the target heart rate (90% of the maximal heart rate calculated using the Karvonen formula) [24]. The post-run 0-min measurements were recorded immediately. After a set of measurements, the researcher left the laboratory and a different researcher came into the laboratory to guide the cool-down conditions.

Subjects executed the three cool-down conditions (IBP, active, and passive) in a counterbalanced order. For the IBP condition, subjects performed five sets of continuous alternating positions on the inversion table. Each set was defined as a position change from A to B for 30 s (Figure 2A,B) followed by a position change from B to A (Figure 2A,B). The duration of inversion was determined to prevent discomfort due to increased intracranial pressure [25]. To change positions, the researcher asked a subject to press a button on the inversion table. For the active condition, subjects walked on the treadmill at speeds of 5 and 6 km/h for females and males, respectively, for 5 min. The walking treadmill speed was determined in our pilot study where we wanted it to be neither too fast (to avoid additional fatigue) nor too slow (no effect on active recovery). For the passive condition, subjects remained seated on a chair for 5 min. After the next set of measurements (post-run 5 min), a 5-min static stretch of the hip adductors, extensors, abductors, and flexors, and knee extensors with a 30-s hold on each side (the same protocol was applied for each condition) was performed. After completion of the static stretch, the post-run 10-, 20-, and 30-min measurements were recorded.

**Figure 2.** The inverted body position, once subjects moved from (**A**) a supine position to (**B**) the IBP, 30 s were counted and then they were returned to the starting position (**A**). It took 17 s to move from position A to B or vice versa. Subjects then spent about 13 s in the IBP (**B**) before returning to position (**A**).

### *2.4. Measurements*

The total running time, average treadmill speed, and inclination were recorded for each subject during each treadmill run.

For the blood lactate measurements, a finger-pricking lancet needle (26 G, Lancets, Moa, Korea) was administered to collect 0.7 μL blood samples on lactate test strips (Lactate Plus Lactate Test Strips, Nova Biomedical, Waltham, MA, USA). The strips were then inserted into a lactate meter (Lactate Plus, Nova Biomedical, USA) for analysis. The blood lactate analysis using this device was previously validated [26].

To measure heart rate and energy expenditure, the subjects wore a heart rate monitor consisting of a Polar M400 GPS watch on their wrist and a Polar H7 strap on their chests throughout the experiment (Polar Electro Oy, Kempele, Finland). After the device was wirelessly connected to the mobile application (Polar Beat), the subject's age, sex, height, and mass were entered. All data recorded throughout the experiment were downloaded from a website (http://flow.polar.com).

The perception of subjective fatigue level was quantified using a modified 10-cm visual analogue scale [27,28] with the terms "unfatigued" and "fatigued" at each end. Subjects were asked to mark their fatigue level at each measurement time.

Using a tape measure, the circumferences of the thighs and calves were recorded (to the nearest 0.1 cm) at the midpoint between the anterior superior iliac spine and the superior pole of the patella and the thickest circumference on the triceps surae, respectively. Both sides were recorded each time (left side first).

### *2.5. Statistical Analyses*

Our sample size was calculated based on a pilot study conducted in our laboratory. We expected a mean difference in blood lactate concentration of 3 mmol/L with a standard deviation of 3.5 mmol/L (an effect size of 0.86). These calculations estimated that 14 individuals would be necessary in each condition (an alpha of 0.05 and a beta of 0.2). This calculation was supported by a similar previous study (*n* = 14) that reported a difference of 3.17 mmol/L [10].

Descriptive data (means and 95% confidence intervals) were calculated and checked for normality using the Shapiro−Wilk test. To test condition effects over time, a mixed model analysis of variance (random variable: subjects; fixed variables: condition and time) and Tukey's test were performed on each dependent measurement. To determine practical significance, Cohen's d effect size [29] was also calculated when statistical significance was found. To test for a correlation between the blood lactate concentration and fatigue perception, Pearson's correlation coefficient was calculated. SAS software (SAS 9.3, Institute Inc., Cary, NC, USA) was used for all tests (*p* < 0.05).

### **3. Results**

The mean ± 95% confidence interval total running time was 7 min and 39 s ± 46 s at a treadmill speed of 7.3 ± 0.4 km/h and an inclination of 1.7 ± 0.2%. The mean treadmill speed and inclination at the end of each run were 11.1 ± 0.8 km/h and 3.6 ± 0.4%, respectively.

The blood lactate concentration was different among the three conditions at post-run 5 and 10 min (condition × time interaction: F10,238 = 3.04, *p* = 0.001; condition effect: F2,238 = 12.38, *p* < 0.0001; Table 1). Specifically, the IBP condition (6.7 mmol/L, 23%, *p* = 0.001, ES = 0.82) and the active condition (6.9 mmol/L, 20%, *p* = 0.01, ES = 0.75) showed lower blood lactate values than the passive condition (8.6 mmol/L) at post-run 5 min. At post-run 10 min, the active condition showed lower blood lactate values than the passive condition (4.4 vs. 6.0 mmol/L, 27%, *p* = 0.02, ES = 0.90). Regardless of the condition (time effect: F5,238 = 314.95, *p* < 0.0001), the blood lactate values decreased gradually after running and they did not return to the pre-run level (1.3 mmol/L) at the end of the experiment (post-run 30-min: 3.1 mmol/L, *p* < 0.0001, ES = 2.41; Figure 3A).

**Figure 3.** (**A**) Changes in blood lactate and fatigue perception regardless of conditions. Error bars are upper and lower limits of 95% confidence intervals. † Different from other time points in blood lactate (*p* < 0.0001). ‡ Different from post-run 0, 5, and 10 min in fatigue perception (*p* < 0.0001). (**B**) Correlation between blood lactate and fatigue perception. There is a high correlation (r = 0.75) between blood lactate and fatigue perception.

The heart rate values differed between the conditions at each time point (condition × time interaction: F10,238 = 8.39, *p* < 0.0001; condition effect: F2,238 = 17.61, *p* < 0.0001; time effect: F5,238 = 1243.89, *p* < 0.0001; Table 1). The IBP condition (96 bpm, 22%, *p* < 0.0001, ES = 2.52) and the passive condition (103 bpm, 17%, *p* = 0.002, ES = 1.92) resulted in lower heart rate values than the active condition (123 bpm) at post-run 5 min.

Energy expenditure among the three cool-down conditions at each time point did not differ (no condition × time interaction: F10,238 = 1.65, *p* = 0.09; time effect: F5,238 = 224.40, *p* < 0.0001; Table 1). However, the total energy expenditure between the conditions was different (condition effect: F2,238 = 4.81, *p* = 0.01) in that subjects in the active condition had a 14% higher total energy expenditure compared to the IBP condition (219 vs. 189 kcal, *p* = 0.01, ES = 0.48; Figure 4).

**Figure 4.** Changes in total energy expenditure regardless of time, IBP: inverted body position; Error bars are upper limits of 95% confidence intervals, † Condition effect (F2238 = 4.81, *p* = 0.01): Active was recorded at higher total energy expenditure values compared with the IBP (219 vs. 189 kcal, 14%, *p* = 0.01, ES = 0.48).


**Table 1.** Changes in heart rate, energy expenditure, blood lactate, and fatigue perception.

IBP: inverted body position; ICC: intraclass correlation coefficient, † Different from the IBP (23%, *p* = 0.001, ES = 0.82) and active (20%, *p* = 0.01, ES = 0.75) conditions at post-run 5 min; ‡ Different from the active condition (27%, *p* = 0.02, ES = 0.90) at post-run 10 min; # Different from the IBP (22%, *p* < 0.0001, ES 2.52) and passive (17%, *p* = 0.002, ES = 1.92) conditions at post-run 5 min.

Fatigue perception did not differ among the three cool-down conditions at each time point (no condition × time interaction: F10,238 = 0.13, *p* = 0.99; no condition effect: F2,238 = 0.87, *p* = 0.42; time effect: F5,238 = 245.80, *p* < 0.0001; Table 1). Regardless of the condition, fatigue perception was 87% higher after running (6.8 cm, *p* < 0.0001, ES = 4.63) and returned to the pre-run level (0.9 cm) at post-run 20 min (1.2 cm, *p* = 0.55; Figure 3A).

The circumference of the left (no condition × time interaction: F10,238 = 0.28, *p* = 0.99; condition effect: F2,238 = 3.72, *p* = 0.03; time effect: F5,238 = 3.88, *p* = 0.002; Table 2) and right (no condition × time interaction: F10,238 = 0.25, *p* = 0.99; no condition effect: F2,238 = 1.26, *p* = 0.29; time effect: F5,238 = 6.66, *p* < 0.0001; Table 2) thighs, and left (no condition × time interaction: F10,238 = 0.96, *p* = 0.48; condition effect: F2,238 = 22.32, *p* < 0.0001; time effect: F5,238 = 41.33, *p* < 0.0001) and right (no condition × time interaction: F10,238 = 0.79, *p* = 0.64; condition effect: F2,238 = 12.13, *p* < 0.0001; time effect: F5,238 = 44.57, *p* < 0.0001) calves did not differ between the conditions at each time point.

Blood lactate was highly correlated with fatigue perception (r = 0.75, *p* < 0.0001; Figure 3B). Good to excellent measurement consistency (ICC values from 0.75 to 0.99; Tables 1 and 2) was calculated for all measurements [30].


**Table 2.** Changes in circumference of lower-extremities.

IBP: inverted body position; ICC: intraclass correlation coefficient, No condition × time interactions in circumference of left (F10,238 = 0.28, *p* = 0.99) and right (F10,238 = 0.25, *p* = 0.99) thigh, and circumference of left (F10,238 = 0.96, *p* = 0.48) and right (F10,238 = 0.79, *p* = 0.64) calf.

### **4. Discussion**

The purpose of this study was to examine how the IBP condition consisting of continuous alternation between 30-s IBP and a supine position for 5 min would help physiological and psychological recovery after intense running relative to active (walking) and passive (sitting) recovery methods. While walking is a commonly recommended method [10,31] and sitting is considered as a control, we were interested in comparing these existing and frequently practiced methods to a recovery condition containing IBP. During the initial 5 min of the recovery period, the IBP was effective in reducing the

blood lactate and heart rate to levels similar to the active and passive conditions. In other words, the IBP condition produced an effect similar to that of the active condition in blood lactate with less energy expenditure. This observation suggests that the recovery condition including IBP could be an alternative recovery strategy after a short bout of intense exercise. Since blood lactate removal is dependent on the amount of blood flow to working muscles, and the process is facilitated by muscle contraction [17], performing IBP could be beneficial for individuals who need a fast recovery while minimizing energy consumption.

The IBP and active (walking at a speed of 5 to 6 km/h) strategies in this study reduced blood lactate by 32% (9.8 to 6.7 mmol/L) and 27% (9.4 to 6.9 mmol/L), respectively, at 5 min postexercise (post-run 0 min). The effects of recovery strategies with active movements [10,31,32], electrotherapy [33], and massage [31] were previously reported. Comparable corresponding values at similar time points in other studies included a 19% (14.9 to 12.0 mmol/L) reduction with a 5-min treadmill run (at 50% of maximal speed) [32] and a 49% (3.9 to 2.0 mmol/L) reduction with 6 min of motor level electrotherapy (at 9, 8, and 7 Hz for 2 min each via 8 electrodes) [33]. At 10 min postexercise (after subjects had finished a 5-min static stretch in our study), blood lactate was reduced by 45% (9.8 to 5.4 mmol/L) and 53% (9.4 to 4.4 mmol/L) in the IBP and active conditions, respectively. Other studies reported reductions of 40% (14.9 to 8.9 mmol/L) with a 10-min treadmill run (at 50% of the maximal speed) [32], 51% (11.7 to 5.7 mmol/L) with a 10-min free style swim (at 65% of each individual's best 200-m front crawl swim velocity) [31], and 39% (11.6 to 7.1 mmol/L) with a 10-min massage (using the effleurage, petrissage, tapotement, and compression techniques) [31]. As indicated by these indirect comparisons at 5 and 10 min postexercise, our IBP condition (a combination of the 5-min IBP and static stretch) appears to be as effective as other recovery activities. Although not directly measured, we believe that the effect of the IBP could have been attributed to hemodynamic changes similar to those of using compression garments [34,35]. According to men's and women's Taekwondo Kyorugi schedules at the Olympic Games in Rio in 2016, the rests between matches ranged from 360 min to less than 60 min. In the Asian Games in Jakarta−Palembang in 2018, the rest times in the men's fencing épée ranged from 80 min to 175 min. During such short time periods, a recovery activity enhancing a rapid blood lactate removal with minimal energy expenditure would contribute to the performance in the subsequent match. Because all of the comparable recovery strategies reported require an electrical device [33], massage personnel [31], or to expend more energy [31,32], our method with IBP could be the best option. Although the blood lactate values in the three conditions in our study became similar at 20 min postexercise, IBP is still beneficial because it resulted in a longer time to rest with a low blood lactate level.

Fatigue perception did not differ among the three recovery conditions at each time point. Previously, electrotherapy (motor level stimulation via 8 electrodes) and the passive (sitting on a chair) recovery method reduced perceived exertion by 46% and 38% compared with active recovery, respectively [33]. Another study [32] reported 18% less perceived exertion with passive (sitting on a chair) recovery than with active (treadmill run at 50% maximal speed) recovery [32]. Although the results of the Borg scales (0 to 10 points: [33]; 6–20 points: [32]) are comparable to our fatigue perception because both are measures of subjective physical condition, the differences of the measurement characteristics could explain why there was no difference across the conditions in our study. The Borg scale matches a numeric scale with descriptions of exertion, asking subjects to indicate their exertion level, whereas our measurement simply asked subjects to place themselves between "unfatigued" and "fatigued" on a visual analogue scale. Measurements were also taken during exercise using the Borg scale and after exercise in our study. Since subjects in previous studies felt less exertion in the passive condition than those in the active condition [32,33], providing different descriptions (e.g., fresh or recovered) or using a more detailed numeric scale may have produced a different result. Interestingly, blood lactate and fatigue perception values had a strong correlation in our study (r = 0.75, r2 = 0.57). Thus, at least 57% of an individual's perceived fatigue can be explained by blood lactate and vice versa. However, fatigue perception returned to the pre-run level at 20 min postexercise (post-run 20 min)

whereas blood lactate levels did not. This observation has several practical implications. First, people may feel complete recovery although their bodies are still in the process of physiological recovery. Second, a complete physiological recovery may be unnecessary for psychological (perceptual) recovery. Third, in addition to blood lactate, there could be other factors which contribute to the level of fatigue perception. A future study should examine the relationships between psychological, physiological, and athletic performance.

Similarly, there were no changes in the circumferences of lower extremities at any time point. Given that a tape measure was used to obtain circumferences in many previous studies [34–37], the measurement reproducibility in our study was strong (ICC value of 0.99 for between-session reliability in all measurements), measurement technique or errors were unlikely to contribute to the results. In fact, we did not even observe a circumference change at the post-run measurements, which could be explained by the exercise duration and mode (specific energy system). Previously reported postexercise extremity circumference was induced by resistance [34,35] or endurance exercise [36,37]. Therefore, an approximately 7-min all-out treadmill run may not have been sufficient to increase leg swelling. Exercise-induced transient hypertrophy, also known as "the pump", is thought to be caused by the accumulation of intracellular fluid from reactive hyperemia [38] or osmolytic activity [39]. In endurance activities, the expansion of extracellular volume [37] due to increased plasma volume and sodium retention [40] is thought to be responsible for oedematous swelling. While we assumed that our intense treadmill run mostly used the combined energy system of anaerobic glycolysis, it produced neither of the physiological mechanisms, resulting in no circumference change. Because body fluid removal is an important facet of physical recovery, a future investigation should examine how the IBP strategy affects the increased circumference caused by either resistance or prolonged endurance exercise. Appendicular body composition analysis along with skinfold thickness would provide better accuracy in the assessment of exercise-induced anthropometric changes.

There are several limitations and assumptions in our study that may benefit future research. Although good criterion-related validity and high reliability in the estimation of energy expenditure using the same heart rate monitoring device were established [41], the major contributing factors such as the physical activity and sedentary levels [42], and percentage of body fat [43] were not reflected. Additionally, a lower heart rate recorded during the IBP might have underestimated the whole-body metabolism and energy consumption. Therefore, we acknowledge the level of accuracy in estimating energy expenditure as a limitation. While accepted, our hypothesis was based on increased blood flow for a body position with 15◦ head-down tilt [21–23] and the lack of direct measures on blood flow parameters limit us to explain and interpret the observed benefits. As an elevated blood pressure during and after exercise is typical, performing IBP immediately after exercise may be harmful by applying additional pressure to blood vessels. When comparing the inversion duration (90◦ for 2 min) in previous reports [44,45] to ours (90◦ for 13 s: Figure 2), we assume that the possibility of additional blood pressure was minimal. Additionally, our subjects' heart rate at the last minute of the treadmill run and the beginning of IBP were 183 and 121 bpm, respectively. A 2-min interval for the post-run measurements resulted in a 34% reduction in heart rate (183 to 121 bpm). Considering that both heart rate and blood pressure are linearly related to exercise intensity [46], a reduced heart rate value indirectly supports the idea that our subjects performing IBP were not in a deleterious situation. However, future studies should attempt to compare the change of blood flow volume and pressure during the recovery period to resolve our limitations. Lastly, the contraindications of IBP should be addressed since the IBP may be unsafe for pregnant individuals or those with abnormal blood pressure or other cardiovascular diseases, eye or ear injuries, or herniated discs.

According to our data, a 10-min recovery strategy consisting of 5 min of IBP and a 5-min static stretch was the most effective of the tested protocols in removing blood lactate and minimizing energy expenditure during the acute recovery period (the first 30 min, postexercise). Therefore, the recovery condition including our IBP protocol is an appropriate strategy to enhance physical and psychological recovery. We particularly recommend this IBP protocol for athletes who complete several matches

(events) within a few hours on a single day (e.g., tournaments for Taekwondo, Judo, weightlifting, fencing, wrestling, sprint cycling, and swimming). In the final stage of matches (e.g., after quarterfinals), the fatigue level would be high. In this situation, physical fitness and stamina can play bigger roles than technical and tactical strategies. Therefore, we recommend practicing our IBP method between matches. If an IBP device is unavailable, peer-assisted IBP or a supine position with legs elevated against a wall could be used instead.

**Author Contributions:** Conceptualization, M.S.K. and J.P. Methodology, J.P. Formal analysis, M.S.K. and J.P. Investigation, J.P. Data curation, M.S.K. Writing—original draft preparation, M.S.K. and J.P. Writing—review and editing, J.P. Supervision, J.P. Project administration, J.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** This study received no specific grant from any funding agency. We thank Chae Rin Kim, Chaerin Yeom, and Gihum Kown for their data collection and reduction.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article*

## **Ultrasound-Guided Percutaneous Needle Electrolysis and Rehab and Reconditioning Program for Rectus Femoris Muscle Injuries: A Cohort Study with Professional Soccer Players and a 20-Week Follow-Up**

**Fermín Valera-Garrido 1,2,3,\*, Sergio Jiménez-Rubio 4,5, Francisco Minaya-Muñoz 1,2,6, José Luis Estévez-Rodríguez 5,7 and Archit Navandar <sup>8</sup>**


Received: 4 September 2020; Accepted: 3 November 2020; Published: 8 November 2020

**Abstract:** Rectus femoris muscle strains are one of the most common injuries occurring in sports such as soccer. The purpose of this study was to describe the safety and feasibility of a combination of percutaneous needle electrolysis (PNE) and a specific rehab and reconditioning program (RRP) following an injury to the rectus femoris in professional soccer players. Thirteen professional soccer players received PNE treatment 48 h after a grade II rectus femoris muscle injury, followed by a the RRP 24 h later. Assessment of recovery from injury was done by registering the days taken to return to train (RTT), return to play (RTP), and structural and functional progress of the injured muscle was registered through ultrasound imaging and match-GPS parameters. Also, adverse events and reinjuries were recorded in the follow up period of twenty weeks. The RTT registered was 15.62 ± 1.80 days and RTP was 20.15 ± 2.79 days. After fourteen days, the ultrasound image showed optimal repair. Match-GPS parameters were similar before and after injury. There were no relapses nor were any serious adverse effects reported during the 20-week follow-up after the RTP. A combination of PNE and a specific RRP facilitated a faster RTP in previously injured professional soccer players enabling them to sustain performance and avoid reinjuries.

**Keywords:** muscle strain; quadriceps; football; invasive physiotherapy; post-injury performance; reinjury; return to play

### **1. Introduction**

Professional soccer players face significant mechanical, metabolic, and physiological demands as a result of their participation in the sport [1]. Regarding to these demands, scientific evidence shows, for example, that elite soccer players cover total distances of 11,173 ± 524 m [2], with 600–831 m ran at speeds exceeding 7 m·s−<sup>1</sup> [3] and over 160 accelerations between 1.5–2 m·s−<sup>2</sup> [4]. Given these demands, soccer has a high risk of injury [5,6]; in fact, at the elite level, up to two injuries per player per season occur [7]. Muscle strains are among the most frequent non-contact injuries [8], and 81–92% of all muscle injuries in professional soccer affect four muscle groups: the hamstrings, the adductors, the quadriceps, and the calf muscles, with the quadriceps being the most affected after the hamstrings and the adductors [9,10]. The biggest muscle among the quadriceps, the rectus femoris, is the anatomical area that is most affected during a tear in the quadriceps [10,11], representing 19% of all non-contact muscle injuries [10,12]. Among all rectus femoris injuries, 46.1% are grade I injuries (representing an edema during MRI evaluation), 37.3% are grade II (representing a partial tear during MRI evaluation) and 4.9% are grade III (representing a complete tear during MRI evaluation) [10].

The rectus femoris is composed of fast-twitch type II fibers [13] and is vulnerable to an injury because of its biarticular nature spanning over the hip and knee joint [12]. It has an important relationship with the iliopsoas helping in the flexion of the hip [14,15], and previous research has shown that a weak or an under-active psoas can result in the rectus femoris generating a greater hip flexion force, resulting in a rectus femoris strain [14–16]. It is mainly injured during kicking or sprinting [15]. In kicking, the rectus femoris plays an important role where it is subjected to changes in muscular length during the reduction of moment of inertia at the knee in the leg cocking phase (when the hip and knee both flex) to the subsequent increase in knee extension velocity in the leg acceleration phase [15,17]. Similarly, in sprinting, specifically during acceleration and deceleration, the quadriceps absorb a large amount of energy [15,17] when they are activated eccentrically as knee extensors while the hip flexes simultaneously. In such a scenario, the adoption of a correct posture is key in order to maintain stability and facilitate force transmission across the lower-limb joints, thereby avoiding compensation generated by an absence of coordination of the lumbopelvic complex [14,15,17].

Moreover, its high recurrence rate in the weeks following return to play (RTP) is a cause of concern in professional soccer [18–20]. Its recurrence rate is between 5% and 6.4% in the short-term (the first week following RTP) [21,22], between 50% and 77% in the medium-term (between the first and second month after RTP), and 15% in the long term (between 5–6 months after RTP) [22]. For these reasons, a successful management of these injuries constitutes a challenge for clinicians in order to define specific protocols of treatment that integrate biological stimulus [23–25] and reconditioning programs adapted for elite competition [17].

Ultrasound (US)-guided percutaneous needle electrolysis (PNE) is a minimally invasive technique that consists of the application of a galvanic current through an acupuncture needle under direct ultrasound visualization [26]. PNE stimulates a local inflammatory response leading to increased cellular activity and repair of the affected area [27]. It has proven to be effective in injuries such as tendinosis (for example, in patellar tendonitis [28], lateral epicondylitis [29], subacromial pain syndrome [30], etc.), and plantar heel pain [31]. The histological and functional evidence observed in an animal model of muscle lesion [32,33] demonstrates that the application of PNE during muscle regeneration induces a decrease in pro-inflammatory mediators (TNF-α and IL-1β), and an increase in the expression of anti-inflammatory proteins (PPAR-γ) and vascular endothelial growth factor (VEGF). The recovery time of the damaged muscle tissue is also reduced. However, such studies in humans are limited to case studies analyzing acute muscle injury [34,35]. Previous studies [36] have used different strength programs in participants with previous injuries to the rectus femoris in order to assess the effectiveness of the program. A review on the rectus femoris injuries has stressed the need to propose interventions through criteria-based reconditioning programs for this injury in order to ensure an optimal RTP scenario to sustain performance and prevent reinjuries [37]. Such criteria-based reconditioning interventions have been shown to be successful in the case of hamstring injuries [38,39].

The identification of elimination of potential risk factors for an injury can be aided by following a cohort of players in a single center over a large span of time. If the same rehabilitation program is applied to all players of the cohort, this also allows the evaluation of a particular program and helps look at the medium to large term development of the players. This has been seen in the case of hamstring injuries, where following a cohort of professional and youth players helped discarding isokinetic strength (of the hamstrings and quadriceps) [40] and flexibility [41] as weak indicators for a potential hamstring injury. In terms of medium to long-term evaluation of the program, the application of a criteria based reconditioning program showed the absence of reinjuries in a span of twenty-four weeks following injury [38] and the players actually managed to improve their performance as a result of individualized training they underwent [39]. However, to the best of the authors' knowledge, such programs following an injury to rectus femoris have not been described in the literature.

Hence, the aim of this study was to assess the safety and feasibility of a combination of percutaneous needle electrolysis and a specific rehab and reconditioning program (RRP) in professional soccer players with an acute muscle injury to the rectus femoris. A secondary aim of the study was to analyze possible reinjuries following RTP in the short, medium and long term. Based on previous research, the hypothesis was that the PNE and RRP program would reduce the layoff time, but the players' physical performance at RTP would be at a lower level as compared to the pre-injury levels.

### **2. Methods**

### *2.1. Study Design*

A single exposure cohort study design was used where a single group of players was followed over the course of three seasons. In case of sustaining an injury to the rectus femoris, their rehabilitation and performance was monitored in the short, medium, and long term.

### *2.2. Participants*

Players from a single club playing in the top professional soccer division of the Spanish soccer league system, commonly known as *LaLiga* were followed over three seasons, i.e., the 2017–2018, 2018–2019, and 2019–2020 seasons. To be eligible for the study, participants were required to meet the inclusion criteria of being professional male soccer players, over 18 years of age and having suffered the acute onset of anterior thigh pain, confirmed via ultrasound imaging [42–45] as a grade II [46] injury to the rectus femoris muscle located at the myotendinous junction. The sports medicine physician confirmed the diagnosis 48 h after the injury incidence. Additional eligibility criteria are described in Table 1.



Thirteen soccer players (age = 27.92 ± 3.17 years; height = 180.92 ± 3.42 cm; mass = 74.07 ± 3.42 Kg; 61.5% of whom were injured while kicking) (Table 2) were finally included in the cohort and volunteered to participate in the study. All participants provided written informed consent, and procedures were conducted under the guidance of the technical and medical staff of the team, always adhering to the ethical principles for medical research involving human subjects under the Declaration of Helsinki.


**Table 2.** Profile of professional soccer players.

\* CoD: change of direction; D: dominant; ND: non-dominant; Kg: kilogram; cm: centimeter.

### **3. Intervention Protocol**

### *3.1. US-Guided Percutaneous Needle Electrolysis (PNE)*

The PNE technique was performed under ultrasound guidance on the muscle injury using an intensity of 1.5–2 mA during 3 s, five times (1.5–2:3:5), according to the protocol by Valera-Garrido and Minaya-Muñoz [47], 48 h after the injury because electromyographic recovery of endplate noise in muscles treated percutaneously with galvanic current is relatively quick (72 h), coinciding with the inflammatory reaction [47]. A specifically developed medically certified device (Physio Invasiva®, PRIM Physio, Spain) was used (Figure 1). The device produced a continuous galvanic current through the cathode (modified electrosurgical scalpel with the needle) while the patient held the anode (handheld electrode). A GE® LOGIQ™ E9 ultrasound machine with a ML6–15 linear transducer (GE Healthcare®, Wisconsin, USA) was used. During the PNE technique, the player was placed in the supine position on a reclining bed at an elevation of 45 degrees with both legs flat resting on the bed. Prior to inserting a needle, the underlying skin was cleaned with isopropyl alcohol and chlorhexidine (Lainco® 2%). The transducer, enclosed in a non-sterile rolled latex covered over non-sterile ultrasound gel, was placed on the target area. Subsequently, an acupuncture needle 0.30 mm <sup>×</sup> 30 mm (Physio Invasiva® needles, PRIM Physio, Spain; uncoated steel needle with rigid metal handle with guide, Korean type) was inserted using a short-axis approach at a 80-degree angle with skin and advanced toward the muscle injury according to the technique described by Valera-Garrido and Minaya-Muñoz [47] (Figure 2). A physiotherapist with more than 10 years of experience in ultrasound evaluation and over fifteen years of experience in invasive therapy applied the PNE technique. Oral paracetamol was used when necessary for the purpose of pain relief in the first 24 h after PNE intervention.

**Figure 1.** Medically certified device (Physio Invasiva®, PRIM Physio, Spain). (**A**) Device; (**B**) cathode (modified electrosurgical scalpel with the needle); (**C**) needle holder; (**D**) needle; (**E**) anode (handheld electrode).

**Figure 2.** Percutaneous needle electrolysis (PNE) technique for rectus femoris muscle injury. (**A**) Graphic scheme in transverse plane. Short-axis approach at an 80-degree angle into the area of muscle injury (dark red part). (**B**) B-mode ultrasound image of PNE technique in which arrowhead represents the tip of the needle in a short-axis approach in area of muscle injury. (**C**) B-mode ultrasound image of PNE technique in which the white area represents the hydrogen gas released during the electrolysis. RF: rectus femoris; VM: vastus medialis; VL: vastus lateralis; VI: vastus intermedius; IT: intramuscular tendon of RF muscle.

### *3.2. Rehab and Reconditioning Program (RRP)*

The rehab and reconditioning program was divided into two parts: an indoor rehab (IR) program and an on-field reconditioning (OFR) program. The IR program, focused on restoring the athlete's neuromuscular function [48], had a duration of approximately 8–9 days post-PNE. This program was further divided into two phases. The initial phase (Phase 1, Table 3) was based on the optimization of range of motion (ROM) and strength. Progressive muscular tensions at different ROMs and movement velocities [49] were prescribed to protect it against future demands and promote connective tissue repair (48 h, 72 h & 96 h post-PNE-Table 3). Hip, pelvic and thoracic spine mobility, single leg strength training, and dynamic tasks that combined isometric activation with stretch-shortening cycle of the lower limb muscles were included in this initial phase. The second phase (Phase 2, Table 4) began 5–7 days post-PNE and focused on tasks based on the absorption and production of force in different planes. This acted as a link toward the progression of movement skills in order to optimize lineal and multidirectional movements [38,39,50]. Players were instructed to perform the included exercises within a pain score of 2 on a numerical rating scale from 0 to 10, where 0 represented no pain and 10 was the most extreme pain possible. Players were encouraged to increase the load with minor pain corresponding to 2 of 10; that is, if pain was ≤1 of 10, they were encouraged to increase the load, and if pain was ≥3 of 10, the load was reduced [51].

The OFR program (Phase 3, Table 5) focused on an optimal reconditioning to prepare the player for the competition demands through technical and coordination drills (OF-1, OF-2, OF-3, and OF-4), which were increased in terms of complexity and demands of decision making [38,52]. In addition, global positioning system (GPS) and accelerometer data were used to training load monitoring during the OFR program. When the player successfully completed all item of the RRP and progression criteria, they were declared fit to train with the team.




ROM: range of motion; EQI: eccentric-quasi-isometric; ISO: isometric tension; SSC: stretch-shortening cycle; DOMS: delayed onset muscle soreness; PNE: percutaneous needle electrolysis.



CMJ: countermovement jump; BW: body weight.


 3.


OF-1: on-field 1 (days 9–10 post PNE); OF-2: day 11 post PNE; OF-3: days 13–14 post PNE; OF-4: day 15 post PNE; RSA: repeated sprint ability; CoD: change of direction; PNE: percutaneous needle electrolysis.

### **4. Variables and Data Analysis**

### *4.1. Return to Training and Return to Play*

The number of days between the injury and the return to training (RTT), return to play (RTP 1), and following match (RTP 2) were counted.

### *4.2. Ultrasound*

Ultrasound scanning of the injured area was performed every three days with the help of Compare Assistant LOGIQ™ software (GE Healthcare®, Milwaukee, WI, USA) to compare prior examinations with the current one for confidence.

### *4.3. GPS Match Variables*

GPS data (WIMU PRO™, Real Track Systems®, Almeria, Spain) was collected for all players from all matches in which they participated before (PRE: PRE 1 and PRE 2) and after (RTP: RTP 1 and RTP 2) injury. This pre-injury data would serve as a control group for this cohort, as it would permit the comparison with the aim of evaluating the effects of the intervention, on the performance of the injured player. This could only be registered on 10 players, since three players were injured in the preseason, and there were no data from official competitions prior to the injury.

The variables used to compare the performance represented actions where the quadriceps are known to be the most active: in accelerations, high intensity runs, and sub-maximal and maximal sprints [53]. The variables measured were (a) total distance covered in m (as a reference value); (b) distance covered at high intensities (between 18.1–21.0 Km·h<sup>−</sup>1) in m; (c) distance covered at very high intensities (21.1–24.0 Km·h<sup>−</sup>1) in m; (d) distance covered at sprint velocities (above 24.0 Km·h<sup>−</sup>1) in m; (e) the peak speed registered in Km·h<sup>−</sup>1; (f) the peak acceleration registered in m·s<sup>−</sup>2; and (g) the explosive distance (i.e., the distance covered when the acceleration exceeded 1.2 m·s<sup>−</sup>2) in m·min<sup>−</sup>1.

### *4.4. Statistical Analysis*

The GPS data recorded were compared, pre-injury vs. post-injury values, using a repeated measures ANOVA. Post-hoc corrections were made with Bonferroni corrections considering the following analyses only: PRE 2 vs. RTP 1, PRE 2 vs. RTP 2, PRE 1 vs. RTP 1, and PRE 1 vs. RTP 2. All calculations were carried out with Jamovi 1.2.17 (version 1.2) with α = 0.05.

### *4.5. Adverse Events*

The soccer players were asked to report any adverse events that they experienced during or after the PNE technique and RRP. Adverse events can be defined as sequelae of medium to long term in duration, with moderate to severe symptoms, perceived as distressing and unacceptable and requiring further treatment [54].

### *4.6. Follow-Up*

After RTP, the players were followed up in the short term (1 week), medium term (8 weeks) and long term (20 weeks) to assess a possible re-injury and adverse effects.

### **5. Results**

Participants returned to full team training (RTT) in 15.62 ± 1.80 days and returned to play (RTP 1) in 20.15 ± 2.79 days and played a total of 94.6 ± 1.20 min. The following match (RTP 2) was in 29.62 ± 3.95 days following the index injury with a match time 95.1 ± 1.14 min.

### *5.1. Ultrasound*

Initially, all players presented images showing a fibrillar defect in the myotendinous junction of rectus femoris muscle with surrounding edema, without fluid collection/hematoma drainable, consistent with a grade II strain injury (Figure 3A). In 92.3% of the cases, the ultrasound examination in the day 7 post-PNE showed a significant change in the extent of edema and fiber disruption (Figure 3B). One player needed one additional session of PNE (7 days post-PNE-1). Ultrasound follow-up at 14 days showed an almost complete disappearance of the injury pattern of the muscle in all players in the ultrasound image (Figure 3C).

**Figure 3.** Transverse B-Mode ultrasound images of the evolution of the injured muscle. (**A**) Initial. 48 h post-injury. Ultrasound imaging shows muscle fiber discontinuity at the level of tear (arrowhead), with surrounding echogenic muscle edema (asterisk). (**B**) Ultrasound imaging at seven days post-PNE. (**C**) Ultrasound imaging at 14 days post-PNE.

### *5.2. GPS Parameters*

On comparing the data from the GPS data (Figure 4), there were significant differences between the GPS recorded distances at high intensities (F (3,27) = 4.67, *p* = 0.009, η<sup>2</sup> = 0.087) and distances covered at sprint speeds (F (3,27) = 2.98, *p* = 0.049, η<sup>2</sup> = 0.108) but no significant differences for distances covered at very high intensities (F (3,27) = 0.699, *p* = 0.561, η<sup>2</sup> = 0.010). Post-hoc comparisons showed a significant decrease between PRE-1 and RTP 1 in distances covered at high intensities (*p* = 0.007). All other post-hoc comparisons between PRE and RTP data showed no significant differences between data registered at PRE and POST. A comparison of the total distance run showed significant differences between the matches, with the distances run during RTP 2 being significantly higher than PRE-2 (*p* < 0.001) and PRE-1 (*p* < 0.001); and the distance run at RTP 1 was higher than that at PRE-2 (*p* = 0.007).

The variation in peak speed (Figure 5) was not significant between the different instances (F (3,27) = 2.49, *p* =0.136, η<sup>2</sup> = 0.126). However, a significant difference was observed for the peak acceleration registered (F (3,27) = 4.74, *p* = 0.009, η<sup>2</sup> = 0.257), although post-hoc comparisons showed no significant differences between the different instances. The explosive distance showed significant differences between the instances (F (3,27) = 13.4, *p* < 0.001, η<sup>2</sup> = 0.481), and post-hoc comparisons showed significantly higher values for RTP 2 compared to PRE-2 (*p* < 0.001) and PRE-1 compared to RTP 1 (*p* < 0.001).

**Figure 4.** Absolute total distance and absolute distances run in different velocity ranges by the participants in matches before (PRE) and after (POST) a rectus femoris injury. \* indicates a significant difference between instances at *p* < 0.05. \*\* indicates a significant difference between instances at *p* < 0.001.

**Figure 5.** Peak speed, peak acceleration, and explosive distance covered by the participants in matches before (PRE) and after (POST) a rectus femoris injury. \*\* indicates a significant difference between instances at *p* < 0.001.

### *5.3. Adverse Events*

There were no serious adverse events reported, whereas minor adverse events were common. It included pain during PNE (100%) and local soreness after treatment (23.1%) that resolved spontaneously within 24–48 h without any intervention.

### *5.4. Follow-Up*

No player suffered a reinjury 20 weeks following RTP. Also, there were no dropouts during the development of the study.

### **6. Discussions**

This study describes the intervention of percutaneous needle electrolysis treatment along with a rehab and reconditioning program following an injury to the rectus femoris in professional soccer players. Ultrasound images during the reconditioning and performance parameters after return to play were registered to assess the effectiveness of the intervention. A short-term, medium-term, and long-term follow-up was also carried out, and the results showed that the injured muscle recovered completely and were no reinjuries or adverse effects reported in the 20 weeks following the index injury.

The apparent success of the program was highlighted by the relatively short times for RTT and RTP and similar physical performance parameters before and after injury. The duration for RTT (mean: 15.62 days, median: 14 days) and RTP (mean: 20.15 days, median: 19 days) was lower than previous studies, which reported a mean layoff duration of 33 (median: 22 days) [10] to 46.8 days [55] for a grade II rectus femoris injuries, although these results must be taken with caution, given the lack of a control group in the study. When comparing physical match parameters before and after injury, the distances covered at the different speed thresholds were similar, and so were the peak values of speed and acceleration registered, suggesting that the participant not only recovered from the previous injury but also did not suffer a substantial loss in terms of their conditioning. Moreover, the gradual improvements in the variable of explosive distance suggests an optimal evolution in the ability to continue competing without a risk of reinjury. Given that the match schedule and opponents were not under the control of the researchers, nor the fact whether the player would go on to play the full 90 min, the similarity in values indicate a successful short-term recovery of the participant. Such a type of follow-up with GPS data has been previously reported to indicate the success of the rehabilitation applied [56]. This positive progress of the players appears to have been facilitated by the structural and functional changes noted during the injury reconditioning process.

Considering the structural data, important changes in terms of the extent of edema and fiber disruption were noted in all cases presented in this study early in the muscle healing process. Different authors [32,33] have evaluated the effect of the application of PNE such as on a model of muscle injury in animal tissue such as the rat, reproducing a physiological environment very similar to that in humans. They concluded that the application of PNE in an animal model of muscle injury reduces the recovery time of the damaged muscle tissue because of the galvanic current facilitates the inflammatory reaction and muscle irrigation, thereby, helping muscular regeneration and, simultaneously, improving the recovery of the injured muscle tissue at an earlier time (from 72 h to seven days post-PNE). The PNE protocol applied was similar in this study, and the results obtained are comparable and evident from the ultrasound scans. This regeneration of injured muscle is a crucial feature enabling the progression on to the functional tasks described in the RRP.

The RRP was designed in such a manner that it followed and complemented the PNE applied 48 h earlier. The early start of the RRP (Table 2) showed a promising advantage when compared other interventions [55], where the rehabilitation and reconditioning program for rectus femoris consisted initially of absolute rest for the first week with the application of ice, compression and non-steroidal anti-inflammatory medication. Furthermore, in this study, the progression criteria were defined based on previous research that has stated the importance of reestablishment neuromuscular function from the initial phases to reduce the layoff period and the risk of re-injury, and thereby, avoiding the loss of training capacity [48,57]. Mobility, core stability, and strength training exercises were prescribed from the first days after PNE for this reason. Due to the biarticular function of the rectus femoris, eccentric-quasi-isometric (EQI) strength training for the muscle at different ranges of movement and movement speeds were used to stimulate damaged fibers, thereby obtaining different ranges of muscular tension in the muscle healing process [49]. In the final phase of the indoor rehab program, the rate of force development and the improvement of movement skills in different planes were key, specifically in acceleration and deceleration patterns where the rectus femoris is highly involved [37]. Previous research has shown that such tasks help restore coordinated movement patterns that are essential for athletic success in the future [58].

During the on-field reconditioning program, drills with repeated sprints [59], multidirectional drills, and soccer-specific technical skills were included, with special emphasis in kicking drills [37]. The load and decision-making progressively increased in order to optimize the functions of the rectus femoris and reconditioning the players in order to prepare them for the demands of competition and training [38,52]. The constant monitoring of the evolution of the repair of injured tissue through ultrasound, provided the researchers with important progression criteria to be taken into account in the progress of the RRP.

An important part of the study was that the participants did not suffer any re-injury or serious adverse effects in 20 weeks following RTP. Epidemiological research has reported that the recurrence rate of a reinjury to the rectus femoris in the weeks following the RTP varied between 5% [21] and 6.4% [22]. As in previous studies, only post-needling soreness was reported, which is a frequent effect after PNE, usually lasting less than 48 h [60]. These results reinforce the medium and long-term success of the PNE + RRP, the progression of which is based on criteria and not time, and this in turn not only might reduce the reinjury risk, but also prepare the athletes for what they would be exposed to during the season.

However, given the small sample size, the results of this study, although promising, must be taken with caution. The absence of a control group is one of the main limitations of the present study. But it is important to highlight that in an elite professional environment is very difficult establish control or placebo group. The reason of choosing a single exposure cohort [61] was to reflect the practical realities in an elite, professional environment where all injured players were exposed to the same PNE and RRP program as per the decision taken by the club's medical and technical staff team. Nevertheless, in this paper, the results were compared to those obtained in the few previous studies that reported on rectus femoris injuries and showed faster mean and median recovery times.

Another limitation to this study was the use of a combination of qualitative and quantitative criteria to determine progress, although this is, to the authors' knowledge, the first study to detail progression criteria in scientific literature on rectus femoris injuries. There was an absence of a purely objective, biomechanical test that could measure the functional capacity of the rectus femoris during or after the rehabilitation and reconditioning process. On such an occasion, tests such as electromyographic evaluation could not be conducted due to decisions beyond the control of the researchers when working with elite athletes. Incorporating such testing could probably help reinforce the complete rehabilitation of the rectus femoris muscle. Although ultrasound was used for injury diagnosis and is a valid tool for the evaluation and monitoring of muscle injury in future studies, it was used qualitatively, and decisions were made based on the operator's experience. Objective quantification of these images could be carried out in future studies, and this could be complemented with the information provided by magnetic resonance imaging. Also, such a study could be implemented in the case of non-elite participants, where the possibility of a randomized clinical study could explain the findings of this study with more detail.

### **7. Conclusions**

An early intervention with PNE and RRP in soccer players with injuries to rectus femoris appeared to allow a safe return to training and to competition without re-injuries in the short, medium, and long term. Assessment of the evolution of the injured tissue through a combination of ultrasound imaging and functional criteria give the impression that they were crucial to make better decisions during the process. In addition, the use of GPS provided information about the athlete's performance after the injury, where the injured player reached similar values when compared to those pre-injury, which possibly indicated the importance of the intervention.

**Author Contributions:** Conceptualization, F.V.-G., F.M.-M., S.J.-R., J.L.E.-R.; Methodology, F.V.-G., F.M.-M., S.J.-R., J.L.E.-R., A.N.; Validation, A.N.; Formal Analysis, A.N.; Investigation, F.V.-G. and S.J.-R.; Writing—Original Draft Preparation, F.V.-G., F.M.-M., S.J.-R., J.L.E.-R., A.N.; Writing—Review & Editing, F.V.-G., F.M.-M., S.J.-R., J.L.E.-R., A.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** Our thanks to the players who participated in the study. The authors also wish to express their gratitude to Javier Vidal Deltell for his involvement, dedication and active participation in the process of returning players to competition. Without him the results would not have been the same.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Review*

## **E**ff**ectiveness of Training Prescription Guided by Heart Rate Variability Versus Predefined Training for Physiological and Aerobic Performance Improvements: A Systematic Review and Meta-Analysis**


Received: 27 October 2020; Accepted: 26 November 2020; Published: 29 November 2020

**Abstract:** A systematic review and meta-analysis were performed to determine if heart rate variability-guided training (HRV-g), compared to predefined training (PT), maximizes the further improvement of endurance physiological and performance markers in healthy individuals. This analysis included randomized controlled trials assessing the effects of HRV-g vs. PT on endurance physiological and performance markers in untrained, physically active, and well-trained subjects. Eight articles qualified for inclusion. HRV-g training significantly improved maximum oxygen uptake (VO2max) (MD = 2.84, CI: 1.41, 4.27; *p* < 0.0001), maximum aerobic power or speed (WMax) (SMD = 0.66, 95% CI 0.33, 0.98; *p* < 0.0001), aerobic performance (SMD = 0.71, CI 0.16, 1.25; *p* = 0.01) and power or speed at ventilatory thresholds (VT) VT1 (SMD = 0.62, CI 0.04, 1.20; *p* = 0.04) and VT2 (SMD = 0.81, CI 0.41, 1.22; *p* < 0.0001). However, HRV-g did not show significant differences in VO2max (MD = 0.96, CI −1.11, 3.03; *p* = 0.36), WMax (SMD = 0.06, CI −0.26, 0.38; *p* = 0.72), or aerobic performance (SMD = 0.14, CI −0.22, 0.51; *p* = 0.45) in power or speed at VT1 (SMD = 0.27, 95% CI −0.16, 0.70; *p* = 0.22) or VT2 (SMD = 0.18, 95% CI −0.20, 0.57; *p* = 0.35), when compared to PT. Although HRV-based training periodization improved both physiological variables and aerobic performance, this method did not provide significant benefit over PT.

**Keywords:** autonomic nervous system; cardiac autonomic regulation; cardiorespiratory fitness; daily training; endurance

### **1. Introduction**

To maximize the physical fitness of athletes, correct management of training program variables (i.e., volume, intensity, frequency, density, etc.) is needed [1]. The optimal training dosage promotes physiological adaptations and reduces the risk of injury and overtraining syndrome, finally improving athletic performance [2]. The rational distribution of training sessions would be the pillar to obtain the correct physiological modifications in athletes [3]. Therefore, several training periodization strategies have been applied to manage the training load and obtain performance enhancement [4]. However, the relationship between training stimulus and physiological responses depends on the individual

**Juan Pablo Medellín Ruiz 1,\*, Jacobo Ángel Rubio-Arias 2, Vicente Javier Clemente-Suarez 3,4 and Domingo Jesús Ramos-Campo <sup>1</sup>**

and varies widely [5]. Thus, to provide correct feedback to the training process and its optimization, the physiological monitoring of the athlete's individual response to the training program plays an essential role [6]. This way, the physiological monitorization allows the correct management of the training load, according to the athlete's individual response [6].

Recently, the autonomic nervous system analysis has been commonly used to manage the training load [7,8] and the endurance training prescription [9–14]. Heart rate variability (HRV) has been widely used, since it reflects the balance between sympathetic and parasympathetic modulation, showing autonomic nervous system (ANS) regulation [15–19]. After physical exercise, the ANS decreases the sympathetic activity and produces a rapid restoration of vagal tone (parasympathetic component) that allows performance improvements [20]. However, due to the misbalance between intensity, volume, and density of training, nonfunctional training loads produce a nonfunctional overreach, promoting a reduction in vagal indices of HRV and impairing the recovery process [21]. Consequently, changes in ANS regulation, assessed by HRV, can identify the relationship between training (stress) and recovery status. Thus, HRV can support the training process as an internal load marker or a long-term monitoring indicator [2].

Lately, a systematic literature review (five randomized controlled trials) critically discussed the potential of heart rate variability-guided training (HRV-g) as an intervention to improve aerobic performance in athletes [2]. Limitations of this previous work include the analysis of the effect of HRV-g only on runners. Only in 2018 and 2019, three additional randomized controlled trials [10,11,22] using cyclists or skiers were published, representing almost half of the total number of studies that were available until then. A potential limitation of a systematic review is that it does not include a data synthesis and statistical analysis to determine the summary effect of the intervention on the outcome's measures; it implies that results obtained in the literature review [2] could be oversized without a specific statistical analysis that offers a more accurate and general picture of the HRV-guided effects on aerobic performance. This highlighted the growing interest in the HRV-guided potential and the need to conduct a large meta-analysis; hence, it is necessary to systematically analyze the effect of this type of training as an intervention to improve aerobic performance in trained and untrained participants.

The aim of this study was to perform a systematic review and meta-analysis to determine if endurance HRV-g maximizes aerobic performance and/or aerobic physiological adaptation, compared to a predefined training (PT) program.

### **2. Materials and Methods**

### *2.1. Study Design*

The methodological process was based on the recommendations indicated by the PRISMA (preferred reporting items for systematic review and meta-analysis) statement [23]. All phases of the meta-analysis were conducted in duplicate. For the meta-analysis, only randomized controlled trials that investigated the effects of training prescription guided by HRV on any physiological (i.e., maximum oxygen uptake—VO2max) or aerobic performance variables (performance at VO2max, performance at VT1 and VT2, or performance test) were considered. The study was registered in PROSPERO (International Prospective Register of Systematic Reviews) (www.crd.york.ac.uk/prospero/index.asp, identifier CRD42020204461).

### *2.2. Data Sources and Search Profile*

A comprehensive literature search was performed using PubMed–Medline, Web of Science, and the Cochrane Library databases. The search was performed without date restriction and was completed on 15 August 2020. The following combination of terms was used: "HRV or heart rate variability", "autonomic nervous system", "parasympathetic nervous system", "cardiac autonomic regulation", and "vagal activity". The Boolean operator "AND" was used to combine these descriptors with "training guided", "training periodization", or "exercise prescription".

### *2.3. Data Extraction and Selection Criteria*

The following inclusion criteria were considered: randomized clinical trials, studies examining the effects of endurance training prescription guided by HRV on physiological or performance variables, studies that include a control group with a PT program, studies published in English, and studies that should report information on variables in one baseline and one post-treatment measure. Conversely, studies were excluded if they were not an original fully published work, if they did not specify the tests utilized or detailed the training program, and if they did not provide numerical data.

The articles analyzed were reviewed separately by two authors (J.P.M.R. and D.J.R.C.). Studies that fulfilled the inclusion criteria were coded and recorded on an Excel spreadsheet. In addition, the substantive aspects were extracted for Table 1: authors, country, methodology, number of participants per group, age, gender, level of physical activity, and methodological aspects; similarly, for Table 2: HRV variable, decision-making algorithm, volume, intensity distribution, frequency, load, and duration of the experiment. Finally, pre- and post-intervention means and the standard deviation of the studies included in the quantitative analysis were recorded.

### *2.4. Outcomes*

The primary outcome was VO2max. The secondary outcomes analyzed were (1) maximum aerobic power or speed (WMax) as a performance indicator in the VO2max, (2) aerobic performance as an extrapolated value from a field test (i.e., 40 km time trial, 3 and 5 km running test), and (3) power or velocity at VT1 (WVT1) and VT2 (WVT2) as the performance variables at those points.

### *2.5. Evaluation of the Methodology of the Studies Selected*

The methodological quality of the selected studies was assessed with the Cochrane risk-of-bias tool [24] that includes the following parameters: (1) random sequence generation (selection bias), (2) allocation concealment (selection bias), (3) blinding of participants and personnel (performance bias), (4) blinding of outcome assessment (detection bias), (5) incomplete outcome data (attrition bias), (6) selective reporting (reporting bias), and (7) other bias. For each study, each item was described as having either a low, an unclear, or a high risk of bias. In addition, the Egger's test was used to assess publication bias.

### *2.6. Data Synthesis and Statistical Analysis*

The meta-analysis and the statistical analysis were conducted using the Review Manager software (RevMan 5.2; Cochrane Collaboration, Oxford, UK). A random-effects model was applied to determine the effect of endurance training prescription guided by HRV on physiological or performance variables. The effects of training on these outcomes between HRV-g and PT groups were expressed as mean differences (MD) or standard mean differences (SMD) and their 95% confidence intervals (CI). The inverse of variance model was used for the analysis. The heterogeneity between the studies was evaluated through the I2 statistic and between-study variance, using the tau-square (τ2) [25]. The I2 values between 30 and 60% were considered as moderate levels of heterogeneity. Additionally, a value of τ<sup>2</sup> more than one suggests the presence of substantial statistical heterogeneity. The publication bias was evaluated through an asymmetry test as estimated from a funnel plot. A *p* value of less than 0.05 was considered to be statistically significant.

### **3. Results**

### *3.1. General Characteristics of the Studies*

A total of 849 studies were identified from the databases and no items were included from other sources. After removing duplicated articles from the different databases, 605 titles and abstracts were screened, 593 were excluded, and 12 were screened as full texts. Finally, statistical analysis was performed on 8 studies [9–12,22,26–28] (Figure 1).

**Figure 1.** PRISMA flow diagram for studies included.

Table 1 provides an overview of the participants' characteristics of the studies included in the quantitative analysis. The total participants was 190 (males and females), mostly trained or active subjects. The mean age ranged from 20.5 ± 1.3 to 39.2 ± 5.3 years (men: 21.8 ± 0.3 to 39.2 ± 5.3; women: from 20.5 ± 1.3 to 35.0 ± 7.0). Training experience was reported in some articles and it ranged from 11.3 ± 3 to 15.0 ± 8 years. In addition, VO2max values were between 35.0 ± 5.0 and 66.7 ± 5.9 mL/kg/min.


**Table 1.** Characteristics of the studies included in the meta-analysis.

M = men; W = women; SH = sleeping in hypoxia; BP: block periodization; TP: training periodization. \* There were 9 dropouts, gender is not specified.

The intervention programs were eight weeks long [10–12,26–28], except for Kiviniemi et al. [9] (four weeks) and Schmitt et al. [22] (15 days); the frequency of training was between 3 to 6.3 sessions per week. Regarding the distribution of time by intensity zones, the studies that reported these variables ranged from 49 to 84% in zone 1, from 12 to 39% in zone 2, and from 3 to 13% in zone 3 [10–12,26]. The predominant method of analysis for HRV monitoring was time domain, followed by frequency domain. Remarkably, only one study used nonlinear measures (Table 2).



rMSSD7D: 7-day rolling average of vagal-mediated square; QRT: quick recovery test using rMSSD; SWC: smallest worthwhile change; TRIMPS: training impact; VT: vigorous training.

### *3.2. Heterogeneity and Risk of Bias Assessment*

Risk of bias assessment is shown in Figure 2. The high risk of bias specified that none of the studies blinded the participants (performance bias) or the evaluators (detection bias). Visual inspection of the funnel plots showed an absence of asymmetry. Moreover, the Egger test demonstrated an absence of significant asymmetry in PT and HRV-g in VO2max (PT: −0.369, *p* = 0.712; HRV-g: −0.752, *p* = 0.452), WMax (PT: −0.539, *p* = 0.590; HRV-g: 0.103, *p* = 0.918), Performance (PT: −0.273, *p* = 0.785; HRV-g: −0.07, *p* = 0.944), WVT1 (PT: 0.095, *p* = 0.924; HRV-g: 1.898, *p* = 0.058), and WVT2 (PT: 1.542, *p* = 0.123; HRV-g: 1.598, *p* = 0.110) (Figure 3).

**Figure 2.** Risk of bias in the included randomized controlled trials.

### *3.3. Meta-Analyses*

Regarding cardiorespiratory fitness, a significant improvement in VO2max in participants who trained using HRV-g (MD = 2.84, 95% CI 1.41, 4.27; *p* < 0.0001) and PT (MD = 1.80, 95% CI 0.24, 3.36; *p* = 0.02) was found after training (Figure 4a). Thus, HRV-g and PT led to a significant increase in maximum aerobic power or speed (HRV-g: SMD = 0.66, 95% CI 0.33, 0.98; *p* < 0.0001; PT: SMD = 0.48, 95% CI 0.12, 0.83; *p* = 0.009) (Figure 5a) after training. Moreover, aerobic performance increased after HRV-g (SMD = 0.71, 95% CI 0.16, 1.25; *p* = 0.01) and PT programs (SMD = 0.47, 95% CI 0.07, 0.86; *p* = 0.02) (Figure 6a); however, no significant differences in training were observed for VO2max (MD = 0.96, 95% CI −1.11, 3.03; *p* = 0.36), maximum aerobic power or speed (SMD=0.06, 95% CI −0.26, 0.38; *p* = 0.72), or aerobic performance (SMD = 0.14, 95% CI −0.22, 0.51; *p* = 0.45) (Figure 4b, Figure 5b, and Figure 6b, respectively).

Concerning power or speed at VT1 and VT2, performance at VT1 increased significantly after HRV-g programs (SMD = 0.62, 95% CI 0.04, 1.20; *p* = 0.04) but not after PT (SMD = 0.17, 95% CI −0.25, 0.59; *p* = 0.42) (Figure 7a). In addition, performance at VT2 improved significantly after HRV-g (SMD = 0.81, 95% CI 0.41, 1.22; *p* < 0.0001) and PT (SMD = 0.53, 95% CI 0.13, 0.93; *p* = 0.009) training programs (Figure 8a). Nevertheless, no significant differences were observed between both training methods in performance at VT1 (SMD = 0.27, 95% CI −0.16, 0.70; *p* = 0.22) and VT2 (SMD = 0.18, 95% CI −0.20, 0.57; *p* = 0.35) (Figures 7b and 8b, respectively).

**Figure 3.** Test for funnel plot asymmetry. (**a**) VO2max PT; (**b**) VO2max HRV-g; (**c**) WMax PT; (**d**) WMax HRV-g; (**e**) Performance PT; (**f**) Performance HRV-g; (**g**) WVT1 PT; (**h**) WVT1 HRV-g; (**i**) WVT2 PT; (**j**) WVT2 HRV-g.

**Figure 4.** (**a**) Effects of endurance training on VO2max pre-training vs. post-training; (**b**) effects of endurance training on VO2max HRV-g vs. PT.


**Figure 5.** (**a**) Effects of endurance training on WMax pre-training vs. post-training; (**b**) effects of endurance training on WMax HRV-g vs. PT.


**Figure 6.** (**a**) Effects of endurance training on aerobic performance pre-training vs. post-training; (**b**) effects of endurance training on aerobic performance HRV-g vs. PT.


**Figure 7.** (**a**) Effects of endurance training on WVT1 pre-training vs. post-training; (**b**) effects of endurance training on WVT1 HRV-g vs. PT.


**Figure 8.** (**a**) Effects of endurance training on WVT2 pre-training vs. post-training; (**b**) effects of endurance training on WVT2 HRV-g vs. PT.

### **4. Discussion**

This systematic review with meta-analysis aimed to determine if endurance HRV-g maximizes aerobic performance and/or aerobic physiological adaptation, compared to a PT program. The major findings indicate that HRV-g significantly improves VO2max, maximum aerobic power or speed, performance at VT1 and VT2, and aerobic performance in running or cycling field test; however, these adaptations were not significantly greater than PT programs. Although these findings do not appear to support the use of HRV-g over PT programs, this study also highlights notable differences in the methodologies used between studies, which may impact the potential efficacy of HRV-g. Despite that both forms of training promote physiological adaptations and improve performance, most parameters related to aerobic performance and aerobic physiological indexes were improved further following HRV-based training.

One of the key physiological variables that determines endurance performance is VO2max [29], showing that high-level endurance athletes can achieve a large VO2max [10]. Therefore, a common aim of endurance training programs is to improve VO2max. This way, it was found how both training programs significantly increase cardiorespiratory fitness (VO2max) (HRV-g: MD = 2.84, *p* < 0.0001; ~5%) and PT (MD = 1.80, *p* = 0.02; ~3%). It could be assumed that both types of training models led to improvements in VO2max, because there were no significant differences between PT and HRV-g (*p* = 0.36); but some factors could modulate the results obtained. Hence, the intensity training distribution and the fitness level of participants could play a key role in the VO2max improvements. Although VO2max has been improved in untrained [30], recreational [9,26], and trained athletes [10,22], the trainability and the increase of this parameter are limited in trained athletes [11]; specifically, only one study did not find an increase in VO2max [11] and, remarkably, it used a trained cyclist sample. In addition, the intensity training distribution used by Javaloyes et al. [11] was a pyramidal distribution (~60% VT1, 30% VT2, 10% VO2max).

In another study [26] with low fitness level (recreational athletes), participants spent much more time in zone 1 (below VT1) (~80% of total training program) and less time in zone 3 (VO2max or higher) (~3% of total training program). While only one study found an improvement in VO2max with trained athletes [10], that study applied an intensity distribution with a pyramidal distribution (~52% VT1, 35% VT2, 13% VO2max). Thus, it seems that high intensity training (Z3) has to be higher in trained athletes than in untrained or recreational athletes; it is a fact that new periodization models are highlighted and defined as a crucial factor to increase athletes' performance [3,4]. For this reason, the use of HRV-g training could optimize the improvement in VO2max (as the trend to higher increases in this training condition have shown; *p* = 0.36), because if the high-intensity training is individualized and it is performed when the athlete is in optimal autonomic homeostasis, it could lead to an improved adaptive response to training [11].

Regarding aerobic performance measured by field test and maximum aerobic power or speed, both training programs led to a similar increase. Nevertheless, one of the characteristics of the HRV-g is the individualization of the training, making this strategy less variable, with fewer nonresponders than PT programs. For example, Vesterinen et al. [26] found that the HRV-g had lower dispersion of results in a 3000 m test (−1 to +6%) than the PT group (−4 to +8%). Similarly, Javaloyes et al. [10,11] reported less variation in their two studies of 40 km time trial following HRV-based training, compared to PT program. Therefore, participants have a greater probability to increase their aerobic performance and reduce the risk of injury and overtraining by following a day-to-day training based on HRV-g. Additionally, in some of the studies included in the present review, the number of high-intensity training sessions of the participants in HRV-g programs varied according to the individual's recovery response of HRV. For example, the number of high-intensity training (HIT) sessions ranged from 11 to 21 in PT and from 5 to 24 sessions in the HRV-g program of Vesterinen et al. [26]. Although previous research reported a nonsignificant correlation between the training adaptation and the number of HIT sessions [26], this finding suggests that the use of HRV-g to manage the inclusion of a HIT session in the program could increase the effectiveness of the training and could diminish the variation in the adaptation.

Regarding performance (power or speed) at VT, results showed that WVT1 increased in the HRV-g but not in the PT; hence, performance at VT2 improved significantly after HRV-g and PT programs. Furthermore, the meta-analysis showed a trend towards higher WVT1 (*p* = 0.22; Δ 13%) and WVT2 (*p* = 0.35; Δ 10%) improvements after HRV-g than PT (Δ 2.2%; Δ 7%, respectively). These statistical trends that showed greater effects for HRV-g than PT were in line with the results reported by the studies included in the present review [11,12,31,32]. Some possible reasons to explain these findings could be related to the aforementioned training intensity distribution developed by each training group, which could affect the results.

In some of the studies included, HRV-g led to a lower proportion of moderate and greater intensity training (as did HIT), in comparison to PT [11]; while in another study, higher moderate intensity training was performed by HRV-g, compared to the PT [10]. Besides, as it was explained above, the individualization and adjustment of the training load by HRV-g reduced the number of nonresponders to training, increasing the number of athletes that improved their VT1 and VT2. Therefore, HRV, as a monitoring tool, would allow taking the principle of individualization of the load a step further.

Notably, the main findings of the present meta-analysis reported that the aerobic performance and aerobic physiological adaptations after HRV-g are not significantly greater than those observed after the PT, and only some trends were found. It could be considered that the HRV-g program, in comparison to the PT, may have a small impact, or some confounding variables may adjust its magnitude. This way, the duration of the training program is one variable to emphasize. It seems that the length needed to achieve meaningful increases in performance using the HRV-g training program could be shorter than the PT, due to the individualized training and the greater training quality, but the duration of the published studies (all of them with less than eight weeks) seems to be very short to obtain a significant difference. Therefore, longitudinal randomized controlled trials, using programs with more duration, are needed to obtain more conclusive results.

The HRV measurement protocol applied in each study included in the present review is another factor to highlight. Different variables to assess HRV were found: (a) time domain variables (root mean square of successive differences of RR intervals—rMSSD) [10,11,26,28]; (b) frequency domain variables (high frequency—HF, low frequency—LF) [9,22]; and (c) nonlinear variables (standard deviation of the intervals to the transverse diameter (short axis) of the ellipse—SD1) [27]. Therefore, frequency domain variables identify some types of fatigue [33], whereas rMSSD has been suggested as a global fatigue measurement [34]. In addition, the HRV assessment ranged from ultrashort records of 90 s [10,11] to 15 min [22], depending on the HRV variables analyzed. Thus, recording time shorter than five minutes was applied if the study used rMMSD as an indicator, while longer records were reported if the study used a frequency-based or a nonlinear variable [35,36]. Moreover, frequency domain variables were more influenced by breathing patterns than time-domain analysis [37]; there were divergencies in the participants measurement postures that included sitting [28], supine lying [10–12,26], standing, or a combination of some of them (lying + standing [22]; sitting + standing [9,27]). It seems that the supine position measures showed lower daily coefficient of variation than standing measures [38], but a standing position was recommended previously [2]. Therefore, the posture differences could also affect the HRV results obtained in the studies and, consequently, to the present meta-analysis.

There are several limitations in this meta-analysis related to the available randomized controlled trials (RCTs) and the divergent methodologies employed, including (i) the small number of studies; (ii) the different intensity training distribution, training programs, and modalities applied in the studies; (iii) the lack of systematic information about the training load performed in most of the study; (iv) the different methodologies applied to assess HRV; (v) the small number of studies using trained athletes to obtain a more specific picture about the effect of this type of training in this population; and (vi) the lack of longer studies to analyze the chronic effect of HRV-g (the duration of the studies was <8 weeks). Additionally, readers should take into consideration that the sport modality (running, cycling, or skiing) can influence the aerobic enhancement and this fact could modify the results obtained in the present review. In addition, it was found that the available evidence has high risk of bias primarily due to the low quality of available RCTs. Therefore, to develop further studies with a better-quality design, and before a more comprehensive training trend, trained athletes' samples are needed in order to analyze the effect of longer interventions (>8 weeks).

According to the results obtained in the present study, while no significant benefits were observed for HRV-g compared with PT, small effects were evident in the larger increases in aerobic performance and physiological adaptations following HRV-g. This suggests that some individuals may benefit more from HRV-g compared with PT, which would be important in well-trained athletic cohorts, where small changes in physical attributes are difficult to achieve and the individualization of the training plays a key role. Further research is required to investigate these responses in more detail, but it appears that the efficacy of HRV-g strategies have been affected by large variations in the structure of the training program (intensity distribution, duration, volume, etc.) performed and the methodology used to assess HRV. Hence, practitioners and coaches must use HRV with caution, due to the factors that affect its measurement. In terms of practical applications, HRV assessment should be carried out daily in the morning, standing or lying in a supine position, in a standardized condition (e.g., with an empty urinary bladder and spontaneous breathing) and using a validated sensor to assess HRV. In addition, it seems that the rMSSD should be the indicator of HRV, since it produces fewer disturbances in the athlete's daily routine and has several advantages, such as its quick and easy accessibility, and the lower sensitivity for the breathing pattern in comparison with spectral variables.

### **5. Conclusions**

The current systematic review with meta-analysis concludes that HRV-g produces significant improvements in endurance performance and aerobic physiological adaptations. However, these adaptations are not significantly higher than in PT. Nevertheless, the findings from this meta-analysis are likely affected by the divergent methodologies employed in studies, specifically, in the HRV assessment, the training program, and the participant's characteristics. Therefore, the results of this meta-analysis reinforce the importance of additional detailed studies to analyze the effects of this novel training method.

**Author Contributions:** Conceptualization, D.J.R.-C. and J.Á.R.-A.; methodology, J.P.M.R.; formal analysis, D.J.R.-C. and J.Á.R.-A.; investigation, J.P.M.R.; resources, J.Á.R.-A.; data curation, J.Á.R.-A.; writing—original draft preparation, J.P.M.R., D.J.R.-C. and V.J.C.-S.; writing—review and editing, D.J.R.-C., J.Á.R.-A. and V.J.C.-S; supervision, D.J.R.-C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*

## **Exercise Training of Secreted Protein Acidic and Rich in Cysteine** *(Sparc)* **KO Mice Suggests That Exercise-Induced Muscle Phenotype Changes Are SPARC-Dependent**

### **Abdelaziz Ghanemi 1,2, Aicha Melouane 1,2, Mayumi Yoshioka <sup>2</sup> and Jonny St-Amand 1,2,\***


Received: 1 December 2020; Accepted: 15 December 2020; Published: 20 December 2020

### **Featured Application: This work highlights secreted protein acidic and rich in cysteine (SPARC) and its pathways as pharmacological targets**/**tools for conditions and diseases in which muscle properties enhancement would provide therapeutic benefits.**

**Abstract:** We previously identified secreted protein acidic and rich in cysteine (*Sparc)* as an exercise-induced gene in young and elderly individuals. Via this animal experiment, we aim to identify selected implications of SPARC mainly within the muscle in the contexts of exercise. Mice were divided into eight groups based on three variables (age, genotype and exercise): Old (O) or young (Y) × *Sparc* knock-out (KO) or wild-type (WT) × sedentary (Sed) or exercise (Ex). The exercised groups were trained for 12 weeks at the lactate threshold (LT) speed (including 4 weeks of adaptation period) and all mice were sacrificed afterwards. Body and selected tissues were weighed, and lactate levels in different conditions measured. Expression of skeletal muscle (SM) collagen type I alpha 1 chain (COL1A1) and mitochondrially encoded cytochrome c oxidase I (MT-CO1) in addition to SM strength (grip power) were also measured. Ageing increased the body and white adipose tissue (WAT) weights but decreased SM weight percentage (to body weight) and MT-CO1 expression (in WT). Exercise increased SM COL1A1 in WT mice and MT-CO1 expression, as well as weight percentage of the tibialis anterior muscle, and decreased WAT weight (trend). Compared to WT mice, *Sparc* KO mice had lower body, muscle and WAT weights, with a decrease in SM MT-CO1 and COL1A1 expression with no genotype effect on lactate levels in all our blood lactate measures. *Sparc* KO effects on body composition, adiposity and metabolic patterns are toward a reduced WAT and body weight, but with a negative metabolic and functional phenotype of SM. Whereas such negative effects on SM are worsened with ageing, they are relatively improved by exercise. Importantly, our data suggest that the exercise-induced changes in the SM phenotype, in terms of increased performance (metabolic, strength and development), including lactate-induced changes, are SPARC-dependent.

**Keywords:** secreted protein acidic and rich in cysteine (*Sparc*); exercise; muscle performance; metabolic phenotype; lactate; ageing

### **1. Secreted Protein Acidic and Rich in Cysteine as an Exercise-Induced Gene**

The modern lifestyle, characterized by the lack of physical activity combined with an unhealthy diet, leads to an increase in health problems typical of our era such as obesity and diabetes. In addition, the improvement of health care systems increased the life expectancy and, therefore, geriatric health

problems, such as sarcopenia, are also increasing. Interestingly, exercise is considered as a "panacea" for many of these problems [1]. Whereas exercise benefits have been widely documented, many exercise-related molecular mechanisms are yet to be fully elucidated. In addition to its direct metabolic implications, skeletal muscles (SM) represent secretary organs producing myokine, such as secreted protein acidic and rich in cysteine (SPARC) [2]. Within the exercise context, functional genomics studies (mainly but not only in the energy metabolism context [3]) have identified genes related to physical activity among which we have *SPARC*/*Sparc*. Indeed, beyond the known implications of SPARC in wound healing and tissue repair [4,5], this gene was characterized (for the first time) as an exercise-induced gene [6] and also as an electrical pulse stimulation (considered as an in vitro model of exercise)-induced gene in muscular cells [7]. Moreover, SPARC secretion is induced by exercise [8–10] after which the concentrations of myokines (including SPARC, interleukin 6 and fibroblast growth factor) increase in the circulation [11]. Therefore, we hypothesize that at least some of the exercise benefits and biological consequences, mainly the muscular phenotype adaptation to exercise, would be mediated by SPARC or the pathways it controls. Therefore, in this study we aim to explore the implication of *Sparc* (via it knock-out (KO)) in mice with a focus on exercise effects on muscles. In addition, age was also introduced as a variable in this study. Therefore, we would find out the combinatory impacts of *Sparc* KO, exercise and age on selected patterns related to SM physiological properties and metabolic performance. We explore the lactate levels and their implications with the SM phenotype changes (both structural and metabolic) in a SPARC-dependent way.

### **2. Animal Experimental Design, Material and Methods**

Our study was carried out on male mice and involved both wild-type (WT) mice (C57BL/6J, the most commonly used strain for genetic and/or transgenic study that also consistently showed the highest level of voluntary wheel-running [12]) and *Sparc* KO mice (129/Sv-C57BL/6J) fed with chow diet (Teklad global 18% protein rodent diets [13]). Mice had access to food and water ad libitum during the whole experimental period (except for fasting periods during which they had access to water only). WT mice were from the Jackson Laboratory (https://www.jax.org/) and *Sparc* KO mice were generated via in vitro fertilization using *Sparc* KO mice sperm generously provided by Dr. Amy D. Bradshaw. *Sparc* KO mice of Dr. Amy D. Bradshaw were generated as previously described [14,15]. Each age-group of mice (young (Y) and old (O)) was divided based on the genotype (KO or WT) to obtain 4 groups: Y-KO, Y-WT, O-KO, O-WT. Finally, each of these 4 groups was further subdivided into two groups according to whether they were exercising (Ex) or sedentary (Sed) mice. Therefore, our experimental design included 8 groups: Y-WT-Sed, Y-WT-Ex, Y-KO-Sed, Y-KO-Ex, O-WT-Sed, O-WT-Ex, O-KO-Sed and O-KO-Ex. Each group had 11 to 12 mice (n). Mice were housed at the animal facility of the CHU de Québec-Université Laval Research Center (12-h light/dark cycle) and periodically checked by animal care technicians for health and wellness. The exercise groups were trained during the dark phase.

The exercising mice were trained during 12 weeks (starting at the age of 9 weeks for Y mice and 66 weeks for O mice) on running wheels (Lafayette instrument Co, Lafayette, IN, USA) placed horizontally (no angle adjustment). Whereas Y mice were sacrificed at the age of 21 weeks, old mice were sacrificed at the age of 78 weeks. Mice were sacrificed following a 12-h fasting (postprandial period) by cardiac puncture following isoflurane inhalation anesthesia. The coming sub-sections detail the measures performed before, during and after the training, as well as on and after the sacrifice day.

All animal experimentation was conducted in accord with the guidelines of the Canadian Council on Animal Care and approved by the Animal Protection Committee of Laval University (Identifications: 2014165 and 2014168). Mice with any type of illness were immediately euthanized by cervical dislocation and excluded from the study. Mice found with anatomical abnormalities (during the sacrifice) were also excluded from the study.

### *2.1. Mice Exercise Protocol and Running Speed Determination*

At the beginning of the training, mice had an adaptation period of 4 weeks. During those 4 weeks, mice performed the incremental exercise. They were trained through a progressive (gradual) increase in both running speed and duration (up to their maximum endurance) throughout this adaptation step, at the end of which we determined the speed at the lactate threshold (LT). The LT level [16] is a parameter indicating the level of physical activity corresponding to the metabolic point at which the muscle production of lactate starts to increase and overcome the blood clearance of lactate. This indicates that the energy produced via oxidative phosphorylation is insufficient to meet energetic needs and, therefore, the muscles trigger anaerobic energy production that generates lactate at a level superior to its blood clearance. At the end of the adaptation period, LT levels were determined following a measure of running speed-dependent blood lactate level curves, based on previously reported protocols [12,17]. Briefly, the mouse run at a determined speed for 4 min, after which we immediately measure the blood lactate level (within 1 min), after that it ran at the next speed (higher) and again the blood lactate speed was measured. We repeated this procedure until the mouse was not able to run (cannot maintain the speed). At the end, we obtained the curve representing blood lactate levels corresponding to the different running speeds, based on which we obtained the speed at the LT. The blood lactate levels were measured, as described in the next section (**1.2**). The LT speed was chosen as a parameter for our study, based on evidence showing that exercise at LT generates metabolic and functional benefits, including improved insulin sensitivity, peripheral glucose effectiveness, lipid profile, blood pressure, physiological fitness [18–21] and body fat weight percentage decrease [6]. The LT was determined for each mouse of the exercise groups. After that, for each set of mice trained at the same period, the running speed of the 8 remaining weeks of training was a value chosen among the average values (range) of all the mice of that set.

The training was at those values close to the LT levels (LT speed) because it was the LT level that was the speed used during the study in which *SPARC* has been characterised as an exercise-induced gene [6]. In addition, the exercise frequency of our study (60 min/day, five times/week) was also similar to the same study [6]. However, we extended the duration from 6 weeks to 12 weeks to be able to easily see the impacts with significant differences between groups. The literature reported studies exploring the effect of exercise in which mice were both trained for longer periods (over 12 weeks) with the same frequency (60 min/day, five times/week) [22] and also at least as young and as old (3 and 19 months of age) [23] as the mice in this study. In addition, the life span for C57BL/6 mice is around 104 weeks (26 months) [24,25]. Therefore, our choices of mice ages, exercise speed and frequency were within a range of the mice's abilities and did not damage their muscles, nor were they limited by physiological parameters. Importantly, since LT speed was chosen, based on evidence showing that exercise at LT generates metabolic and functional benefits, our study will provide additional data for molecular and biochemical explanations of such training benefits.

During mouse training, and unlike other protocols, no electrical [17] or any potentially harmful stimulations were used to force the mice to run. Only a light air stimulation (using a small hand air pump) was applied for mice in some cases to ensure, as much as possible, that all mice ran during the same period and at the same speed throughout the same training period (optimize the protocol for similar exercise amounts). Moreover, mice were always handled gently when taken from the cage to the training device and vice versa. In addition, sedentary mice were also transported to the exercise training room and kept in their cages, while exercising mice were trained so that all mice received a similar environmental (light, noise, etc.) stimuli. Thus, any possible impacts of stress or environmental stimuli on the performed measures were reduced to minimum.

The last training session was 48 h prior to the sacrifice so that the measures we obtained during the sacrifice and those performed on tissues afterwards did not reflect any possible acute effect of exercise, such as dehydration or neuroendocrine changes that could impact gene expression, post-exercise energy intake or expenditure, etc. Our study aimed to investigate the effects of the 12 weeks of exercise (chronic).

### *2.2. Fasting Lactate and Oral Glucose Gavage-Dependant Lactate Levels*

As a resting metabolic indicator in the muscle, blood lactate levels were measured before and after glucose (prepared from 45% solution, Sigma-Aldrich Canada Co., Oakville, ON, Canada) oral gavage (2 mg of glucose per 1 g of body weight). Mice were fasted for 6 h prior to the glucose gavage. Each measure had 5 time points (0, 15, 30, 60 and 120 min after glucose gavage) allowing us to obtain a curve and calculate the area under the curve (AUC). This test was performed three times (a total of three AUCs)—before the training, at week 5 and the end of week 12 of the training. In addition, we also measured blood lactate (single measure) at the sacrifice day (following 12 h fasting) and at the end of the last training session. Blood lactate levels were measured via tail pricking with a needle to collect blood samples on lactate test strips that were then inserted into the lactate meter (Lactate scout, Sports Resource Group, Inc., Minneapolis, MN, USA).

### *2.3. Grip Power Test*

At the end of the training period, the muscle strengths of all the mice were evaluated through performing a grip power test with a grip strength meter (Columbus Instruments International, Columbus, OH, USA). The grip strength was measured by allowing a mouse to grab (with four limbs) pull bar assemblies attached to the force transducer while the mouse was pulled horizontally by the tail away from the bars, similar to what has been previously described [26,27]. The peak force applied by the mouse (g) was then shown on a digital display.

This test was conducted five times (5 min apart) for each mouse, after which the forces (both mean and maximum) were calculated both as absolute values as well as being normalized to the body weights of the corresponding mice.

### *2.4. Body and Tissue Weights*

Body and selected tissues were weighted at the sacrifice day. The selected tissues were the brain, pituitary gland, hypothalamus, liver, heart, aorta, white adipose tissue (WAT), SMs (gastrocnemius, soleus, tibialis anterior and extensor digitorum longus muscles). The values are reported both as tissue weights as well as a weight percentages (normalized) to the body weights of the corresponding mice.

### *2.5. Western Blotting*

We measured the SM (tibialis anterior muscle) expression of two proteins—collagen alpha 1 type I (COL1A1) and mitochondrially encoded cytochrome c oxidase I (MT-CO1). Whereas COL1A1 is important in the structure of the muscle [28], MT-CO1 is an indicator of mitochondrial oxidative phosphorylation [29]. At the day of sacrifice, the tibialis anterior muscle was removed and quickly put in liquid nitrogen (snap frozen) then moved to −80 ◦C and kept until the protein extraction procedure. To measure the expression of both COL1A1 and MT-CO1, total proteins were extracted from the tibialis anterior muscle, using a radio-immunoprecipitation assay (RIPA) buffer and protease inhibitors cocktail (Sigma-Aldrich Canada Co., Oakville, ON, Canada) and followed by a protein quantification of each protein extract using Bio-Rad protein assay (Bio-Rad Laboratories Ltd., Mississauga, ON, Canada). Fifteen (MT-CO1) or ten (COL1A1) micrograms of proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) using the TGX Stain-Free FastCast acrylamide solutions (Bio-Rad Laboratories Ltd., Mississauga, ON, Canada), and the trihalo compound in the gels was activated under UV light. Then, total proteins were transferred to polyvinylidene fluoride (PVDF) membranes (Bio-Rad Laboratories Ltd., Mississauga, ON, Canada), and gels (before and after the transfer) and membranes were visualized under UV light by using the AlphaImager TM 1220 (Alpha Innotech Co., San Leandro, CA, USA).

Membranes were blocked using the Pierce™ Protein-Free (TBS) blocking buffer (Life Technologies Inc., Burlington, ON, Canada), incubated with 1/400 (sc-8784R for COL1A1 and sc-48143 for MT-CO1) dilution of primary antibodies (Santa Cruz Biotechnology Inc., Dallas, TX, USA) and secondary antibodies (sc-2004 for COL1A1 and sc-2350 for MT-CO1, 1/10000 dilution: Santa Cruz Biotechnology Inc., Dallas, TX, USA), and finally visualized with the Clarity™ Western ECL Blotting Substrate on a film (Bio-Rad Laboratories Ltd., Mississauga, ON, Canada). The visualized total proteins on the membranes and target proteins on the films were quantified using ImageJ software (ImageJ bundled with 64-bit Java 1.8.0\_172, U. S. National Institutes of Health, Bethesda, MD, USA) [30]. The methodology of lane and band quantifications, followed by expression evaluations, was performed according to Taylor et al. [31,32] as we have detailed in one of our previous works [33].

### *2.6. Statistical Analyses and Sample Size Determination*

The data were analyzed by three-way (age, genotype and exercise) and the four-way (for the lactate AUC) ANOVA. When the ANOVA revealed a significant interaction between two or three variables, the Tukey Kramer post hoc test was performed to identify the significant difference between the groups (*p* < 0.05). A trend corresponds to 0.05 ≤ *p* < 0.1. In the results section, all the effects are significant (*p* < 0.05), unless mentioned as a trend.

The number of mice (11–12 mice per experimental condition) was based on the results of power analysis by setting the statistical power at 80% (α = 0.05 and β = 0.2) with our previous study, which used the same strain of WT mice [34].

### **3. Results**

### *3.1. Exercise Patterns, Running Speed and Lactate Concentrations*

Tables 1 and 2 report the data collected during the 4 weeks of incremental exercise, the weeks of the LT speed training and the day of sacrifice. During the 4 weeks of incremental exercise, we have the effect of age on both LT speed (Y > O) and blood lactate level at rest (O > Y). However, for lactate at rest, this genotype effect is attributed to the *Sparc* KO mice since, for the WT, there is no difference between Y and O mice, but for the *Sparc* KO mice, O mice have a higher lactate concentration at rest than Y mice (significant effects of genotype×age interaction). We also have the effect of age (Y > O) for both the mean exercise speed as well as the total exercise distance during the 12 weeks of training. The same effect of age (Y > O) is observed during the last training session for both speed and lactate concentration, measured at the end of that last running hour (Table 1).

In Table 2, we notice effect of age (Y > O) for blood lactate level (trend) on the sacrifice day (measured after 12 h of fasting). For the curve of post glucose–gavage lactate concentrations at different time points (0, 15, 30, 60 and 120 min), we only have an effect of the age (Y > O). The value of the AUC was measured three times—before the training and at week 5 and the end of the week 12 of the training. Each time, mice had a 6 h fasting period prior to glucose gavage.



millimolar; O, old; wk, week; Y, young. -: No effect.

**Table 2.** Fasting and post glucose–gavage blood lactate levels.


 perexperimentalage;genotype;mM, millimolar; O, old; wk, week; Y, young. \*: Trend (0.05 ≤ *p* < 0.1). -: No effect.

### *3.2. Body and Tissue Weights*

Mice were weighed the morning of the sacrifice. During the sacrifice, tissues were removed and weighed as well (Table 3). Analyzed data are both as absolute values (weight) and percentages of the tissues weights to the body weight.

We found the effect of age (O > Y) on body weight as well as on the weights of pituitary gland, hypothalamus, liver, heart, and WAT, in addition to the weight percentage of WAT; the opposite effect of age (Y > O) on the weight percentages (to the body weight) of the brain, heart, aorta, SM and tibialis anterior muscle. We also found an effect (trend) of age (Y > O) on the weights of both SM and the tibialis anterior muscle.

We found an effect of genotype (WT > KO) on the body, aorta, WAT, SM and tibialis anterior muscle weights, and another effect (KO > WT) on the brain weight and weight percentage, liver and heart (both weight percentage). Coming to the last variable, exercise, we also report these exercise effects—Sed > Ex (trend) for body weight and liver weight percentage and WAT weight. Ex > Sed (trend) for weight percentages of both the brain and the hypothalamus. Ex > Sed for tibialis anterior muscle weight and Sed > Ex for liver weight.

### *3.3. Muscle Strength (Grip Power Tests)*

As a measure of muscle strength for the four limbs (simultaneously), the grip power tests results (Table 4) show effect of the age (Y > O) for the both the mean and the maximum grip power as well as for the percentage of each of these two values (mean and the maximum grip power) on body weight. We also have an effect of the genotype (WT > KO) for both mean and the maximum grip powers. For the effect of exercise, we have a trend (Ex > Sed) for the percentage of the maximum grip power to the body weight.

### *3.4. COL1A1 and MT-CO1 Expressions in Tibialis Anterior Muscle*

Protein expression of both COL1A1 and MT-CO1 was measured in the SM tibialis anterior muscle. The results (Figure 1) indicate an effect of genotype (WT > KO) on both proteins and an effect of exercise (Ex > Sed) on COL1A1 (trend) and MT-CO1. For the interactions, we found one between genotype and exercise for COL1A1 (Ex > Sed in WT) and one between age and genotype for MT-CO1 (high in Y-WT).



#### *Appl. Sci.* **2020** , *10*, 9108



sedentary; WT, wild-type; Y, young. \*: Trend (0.05 ≤ *p* 0.1); \*\*: Inguinal and abdominal adipose tissues; \*\*\*: Gastrocnemius, soleus, tibialis anterior and extensor digitorum muscles. %: percentage to the body weight. -: No effect.



Data are mean ± SEM. Number of mice: 11–12 mice per experimental condition. Abbreviations: A, age; BW, body weight; Ex, exercise; G, genotype; g, gram; KO, knockout; O, old;sedentary; WT, wild-type; Y, young. \*: Trend (0.05 ≤ *p* < 0.1). -: No effect.

 Sed,

**Figure 1.** Expression of both collagen type I alpha 1 chain (COL1A1) (**A**) and mitochondrially encoded cytochrome c oxidase I (MT-CO1) (**B**) in the tibialis anterior muscle. The results indicate an effect of genotypeG (WT > KO) for both proteins and an effect of exerciseEx (Ex > Sed) for COL1A1 (trend) and MT-CO1. For the interactions, we have one between genotype and exercise<sup>G</sup> <sup>×</sup> Ex for COL1A1 (Ex > Sed in WT) and one between age and genotypeA <sup>×</sup> <sup>G</sup> for MT-CO1 (high in Y-WT). All data are mean <sup>±</sup> SEM. The number of mice: 11–12 mice per experimental condition. Abbreviations: A, age; Ex, exercise; G, genotype; KO, knockout; O, old; Sed, sedentary; WT, wild-type; Y, young. \*: Trend (0.05 ≤ *p* < 0.1).

### **4. Discussion and Interpretation**

As per Table 1, there is no effect in the genotype for the LT speed (in all the performed measures both during the 4 weeks of adaptation and during the 8 weeks of LT training), exercise speed, exercise time, exercise distance and even lactate concentrations during exercise. This has a key importance, since it means that mice of the two different genotypes (KO and WT) had equal amounts of exercise training (speed, distance, time and frequency) and blood lactate levels. Therefore, genotypes effects seen for the other measures will be, indeed, due to the genotype itself (consequence of *Sparc* KO) rather than difference in the exercise amount.

SPARC (osteonectin or BM-40) is a three-modular-domain [35,36] calcium binding extracellular matrix-associated glycoprotein [37,38]. The *Sparc* gene localized to the central region of chromosome 11 in mice [39] and in the chromosomal site at 5q31–q33 in humans [40]. It is well known for its roles in extracellular matrix (ECM) organization, growth, cellular differentiation, cell–matrix communication, wound healing, cell cycle and tissue response to injury [35,36,41–43]. SPARC is also implicated in metabolism [44,45], cancer [46] and inflammatory [47] homeostasis. Importantly, for the SM, a key metabolic tissue and the key organ for the exercise performance, SPARC represents an important element for its development [28] and function [7].

Indeed, SPARC is known for its importance is SM development and regeneration (satellite cells/myoblasts. myotubes and muscle fibers) [48]. Moreover, whereas during embryogenesis SPARC is highly expressed, its expression is mainly restricted to tissues undergoing changes and remodeling during adulthood [35,49–51] which indicates its importance for exercising muscle; which does undergo remodeling as an adaptation to exercise [52,53]. Importantly, SPARC modulates actin cytoskeleton within the SM structure which results in defective force recovery following in vitro fatigue stimulation in muscle from *Sparc* KO mice [54]; but in normal and uninjured muscles, SPARC is not detectable [48]. This further indicates the importance of SPARC in the context of healing, repair, remodeling and development, especially that the ECM (important for the cellular remodeling, for instance) repair, disassembly and degradation is mediated by SPARC [55]. These impacts on regeneration and during embryogenesis suggest that SPARC deficiency could impact some tissue development and growth, as illustrated by the loss of bone mass (osteopenia) in *Sparc* KO mice [56].

Another structural importance for SPARC in SM derives from its ability to interact with collagens. It interacts with collagen I and procollagen I [57], binds to fibrillar collagens [58], maintains SM stiffness (collagen accumulation regulation) [59] and specifically binds several molecules, including collagen

types I, III and IV [60]. SPARC deficiency has also been shown to reduce the expression of different types of collagen such as collagen type I in mesangial cells [61], collagen in skin [62] and fibrillar collagen accumulation in tibialis anterior muscle [59].

Moving from SPARC-related structural muscle properties to metabolic implications, SPARC has been shown to be required for the expression of the exercise-induced (in vitro model of) mitochondrial enzymes (oxidative phosphorylation) [28] and is suggested to enhance the muscle mitochondrial biogenesis [63] as supported by the fact that small interfering RNA (siRNA) of *SPARC* reduces 5-aminoimidazole-4-carboxamide-1-β-Dribofuranoside (AICAR)-stimulated adenosine monophosphate-activated protein kinase (AMPK) phosphorylation [64], which is known to induce mitochondrial biogenesis via the activation/induction of peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PPARGC1A, also known as PGC1α) [65–67], a master regulator of mitochondrial biogenesis. Importantly, knowing the importance of the mitochondria during regeneration [68,69], SPARC would impact regeneration. In addition, SPARC regulates glucose transporter type 4 expression [64] and improves glucose tolerance [70]. These are selected illustrations of SPARC importance and implications for the metabolism, mainly for the SM that we focus on in our study.

For other tissues, the implication of SPARC in tissue regeneration and development (including tissue repair, cell turnover, cellular differentiation and remodeling) [35,38,44,71,72], especially with the known implications of SPARC in the functions of stem cells [73,74] and other types of cells such as erythroid progenitors [74], could indicate that SPARC-deficient mice could exhibit impairments in terms of development for certain tissues under selected conditions.

### *4.1. Lactate Concentrations among the Indicators of Muscles Metabolic Performance*

Lactate is not just produced by SM and WAT [75] but it is also consumed by muscles [76] with special metabolic patterns [77] and serves as a gluconeogenic precursor [78]. Therefore, the blood lactate levels represent the outcome of the balance between the production and the consumption (clearance) of lactate [79] mainly (but not only) by SM [80]. The production of lactate by SM does not always mean insufficient energy production through oxidative phosphorylation, but could also be due to the lack of oxygen [81], as illustrated by the production of lactate in the adipose tissue of obese subjects as a consequence of hypoxia in this adipose tissue [75]. Importantly, exercise-produced lactate both upregulates the expression cytochrome oxidase gene and protein expression and is a mitochondrial biogenesis activation signal [79]. All these changes seem to result from negative feedback, aiming to increase the oxidative phosphorylation ability and, therefore, reduce lactate production and increase its clearance (usage). The liver and heart also contribute to lactate clearance and, whereas myocardia oxidases lactate as a fuel, the brain also takes it when its levels increase in the blood and the liver uptakes it to form glucose [80]. The fact that no genotype effect has been seen for lactate levels, at similar amount of exercise indicates that *Sparc* KO mice are able to maintain the lactate concentrations at a homeostatic level (similar to that of WT mice) in spite of the impaired muscular functions (compared to WT mice), suggesting a compensatory effect of other tissues to re-balance blood lactate (as we detail below). This compensatory pathway highlights the importance of lactate blood homeostasis.

### *4.2. Body and Tissue Weights (Table 3)*

The importance of SPARC is tissue development, embryogenesis, regeneration, its interaction with collagen and ECM, in addition to its role in collagen accumulation [59] would explain why *Sparc* KO reduces body weight and SM weights, including the tibialis anterior muscle (correlated with what Omi et al. reported [59]) in addition to other tissues (aorta and WAT), as a result of regeneration and development deficiency, similar to the decrease in bone mass (osteopenia) as a result of SPARC deficiency [56]. However, the observed increased weights or weight percentages of other tissues, such as the brain, liver and heart in *Sparc* KO mice, could result from feedback signals. Indeed, the reduced development and metabolic deficiency in *Sparc* KO mice would lead to the production of

signals aiming to correct this developmental and metabolic deficiency (resulting from muscle low oxidation capacity, myokines secretion reduction, etc). Such signal effects would target selected tissues (those increased with *Sparc* KO, such as the brain, which is the center of numerous neuroendocrine signals and in which Compolongo et al., have shown that the neuronal activity levels of *Sparc* KO mice are increased in the brain region dentate gyrus [82] which could support the hypothesis of such signals in *Sparc* KO mice) and either be nonspecific or with insufficient impacts on other tissues (those for which *Sparc* KO does not reduce the weights). For instance, the increased heart weight percentage in *Sparc* KO mice could be adaptive to the fact that these mice have reduced oxidative phosphorylation ability (as shown by the low MT-CO1 expression) and would have more muscle-produced lactate. The developed heart could be an adaptation to the increased lactate production in order to increase the circulation and, therefore, increase lactate clearance, which could be taken by the liver to form glucose [80], which could also explain the increased weight (percentage) of the liver in *Sparc* KO mice. Therefore, although *Sparc* KO mice SM produced more lactate (weak oxidation capacity), they have the same blood lactate levels as WT mice because they would compensate via increased lactate blood clearance through enhanced blood circulation (increased heart weight percentage) combined to an increased intake by the liver (increased weight percentage), the brain (increased weight and weight percentage) and probably other tissues leading to that weight/weight percentage increase in those tissues in *Sparc* KO mice. This correlates with the liver weight percentage for which we have Ex-KO>Ex-WT, meaning that, in the exercised groups, the liver (weight percentage) of *Sparc* KO mice is superior to the liver (weight percentage) of WT mice (even though both had similar amount of training). This could indicate more tissue glycogen storage [83] (in a hydrated form that adds more water weight to the liver [84]) built from glucose made of the taken lactate because the *Sparc* KO mice SM would produce more lactate (weak oxidative phosphorylation reflected by the decrease in MT-CO1 expression in the *Sparc* KO mice) but clear it better through an increased blood circulation (increased heart weight percentage in *Sparc* KO) combined with lactate uptake (clearance) by the liver [80] and also by the brain [80] (that also increased in weight in *Sparc* KO mice) to compensate the low oxidation ability of the SM (supposed to contribute to lactate clearance but remains insufficient in terms of lactate clearance in *Sparc* KO mice). Overall, there is no genotype-related difference in lactate level because there would be compensation. Indeed, whereas WT mice have good muscle lactate clearance (with low lactate production), *Sparc* KO mice (although they have higher lactate production) have increased lactate clearance via the liver, brain, heart (that have increased weight percentage, compared to those in WT mice), etc.

Furthermore, the known implications of SPARC in the functions of erythroid progenitors [74] could suggest that *Sparc* KO mice would have reduced hemoglobin (low blood cells cancer) and, therefore, reduced oxygen transport ability. This would require one to increase the blood supply to different tissues to compensate low blood oxygenation via increased blood circulation that would require a developed cardiac pump and, thus, explains the increased weight (percentage) of the heart in *Sparc* KO mice; such low oxygenation further worsens the weak oxidative phosphorylation capacity in SM that *Sparc* KO mice already have.

The other tissues patterns (age- and exercise-dependant) are in accordance with the known effects of both ageing and exercise on diverse tissues. For instance, the increased brain and hypothalamus weight percentages (trend) with exercise fits with the ability of exercise to enhance neurogenesis [23,85,86], the exercise also reduces (trend) both body weight and liver weight percentage and WAT weight, whereas it increases the tibialis anterior muscle weight percentage. All these elements correlate with the ability of exercise to increase energy usage (WAT lipids and liver glycogen) as well as muscle weight. Regarding the tibialis anterior muscle, in addition to its increase (weight percentage) with exercise, it decreases with both age (weight percentage) and *Sparc* KO (weight). It is for these patterns in changes according to genotype, age and exercise that we have chosen the tibialis anterior muscle to measure the expression of COL1A1 and MT-CO1; which allowed us to make a correlation between the genotype-dependent changes in muscle weight and power and the corresponding changes in

the expression of these two proteins, depending on SPARC expression. Additionally, the decrease in brain weight percentage with ageing correlates with age-related neurodegeneration and related diseases [87,88], which are improved by exercise [89–91] and that, also, correlate with our data, showing an increase (trend) in the brain weight percentage with exercise.

Interleukin 6 (among other myokine) is produced by the muscles during exercise [92,93], which reduces appetite [94] and WAT [93]. This correlates with our results, indicating an effect (trend) of exercise on reducing the body weight and WAT weight percentage, but without any interaction effect on genotype and exercise. This indicates that SPARC absence would not impact the ability of exercise to reduce adiposity. The possibilities could be whether the effect of SPARC is partial since (in *Sparc* KO mice) the WAT weight is lower in Y mice compare to O mice, or there are other SPARC-independent pathways linking myokine to adiposity reduction, such as IL-6 or also because both WT and *Sparc* KO mice spent similar amounts of exercise, leading to similar exercise-induced energy expenditure (would have similar impacts of reducing the WAT).

For the WAT, both for the age and age × genotype (both in WT and *Sparc* KO mice), we always found an increase in adiposity (weight and weight percentage) with ageing. In addition, there is also a reduction in the muscle mass (percentage) with age and for both WT and *Sparc* KO mice, which corresponds to the classical ageing profile (decreased muscle mass and increased adiposity) along with increased body weight with ageing [95–98], as our data show. It is worth noting that, while looking into the effects of genotype × age, both for the decrease in SM mass percentage and the increase in WAT weight (as well as weight percentage), we notice that these ageing-induced changes (musculature decrease and adiposity decrease), are more important in O mice than in Y mice. This could be explained by the implication of SPARC in these changes. Indeed, since *SPARC* expression is downregulated by ageing [8], the consequences of its KO would be more important in Y mice compared to O mice, where its expression is already reduced by ageing.

This ageing effect on SM explains the results of Table 1, showing that ageing reduces the LT speed (adaptation phase), mean exercise speed, total exercise distance (12 weeks of training) and both running speed and the lactate concentration of the last running session at the end of the 12 weeks training. However, the lactate at rest level (week 4 of the adaptation phase), which increases with age, indicates a reduced aerobic metabolic performance of the muscle. Importantly, the effect of genotype × age interaction reveals that the age effect comes from the *Sparc* KO mice rather than WT mice, meaning that it is the *Sparc* KO in O mice that leads to an increase in the resting lactate compared to both WT mice and KO-Y mice. This also explains, in part, how ageing is both a risk factor for numerous diseases and health conditions [99–103].

Since Norose et al. reported that, when handled, *Sparc* KO mice reduced physical activity [104], we deduce that our *Sparc* KO mice had reduced energy expenditure (compared to WT mice), but with a lower body and WAT weights they most probably had less food intake compared to the WT mice. This could indicate an effect (direct or indirect) of *Sparc* KO on appetite. This appetite (in addition to the physical activity patter), both impacting the body weight, could be explained by the increased levels of anxiety and reduced depression-related behaviors in *Sparc* KO mice [82]. Such variations in mood states would impact food intake and energy balance [105,106] and, therefore, body and tissue weights.

### *4.3. Protein Expressions (Figure 1) and Muscle Strength (Table 4)*

The reduced expression of COL1A1 in *Sparc* KO mice fits with what Omi et al. reported [59] and confirms the importance of SPARC for COL1A1 expression, as we have previously shown [28], and as reported by Norose et al. [104] and Bradshaw et al. [107].

In addition, MT-CO1 decreased expression with *Sparc* KO highlights the implication of SPARC in mitochondrial enzyme expression [28] and mitochondrial regeneration [63], whereas MT-CO1 increases expression with exercise, which fits with our previous gene expression studies, showing an increase in oxidative phosphorylation genes with training at LT intensity [6], which further validates the choice of the exercise speed in this study.

*Sparc* KO mice have been reported as passive and with reduced physical activity compared to WT mice [104], this would indicate weak muscles and correlates with reduced grip power (both mean and maximum) in *Sparc* KO mice compared to WT mice of our study. The importance of SPARC in myoblast fusion [28] and, more important, the interaction of SPARC with actin in SM (actin cytoskeleton modulator) [54] also support our data, indicating a decrease in muscle strength with SPARC deficiency. This SPARC deficiency also reduced COL1A1 expression; indicating an impact on muscle structure (for which collage is a key element) and correlated with the *Sparc* KO-induced muscle strength decrease.

The effect of age (Y > O) on the muscle strength and the effect of exercise (Ex > Sed, trend of the percentage of maximum grip power) is an additional illustration of how ageing worsened the effects of SPARC deficiency (reduce the muscle power) while exercise improved it (increase in both SM strength and expression of both COL1A1 and MT-CO1 with exercise represents muscle adaptation to exercise).

### **5. Conclusions and Hypothetical Mechanisms**

*Sparc* KO effects are toward a reduced body and WAT weights with a negative SM phenotype (metabolism and strength). Such negative effects worsen with ageing but relatively improve through exercise (Figure 2). While exercise reduces risk factor for many diseases, ageing increases those risks [108].

**Figure 2.** SPARC-deficiency impacts. Our data highlight that *Sparc* KO effects are toward a reduced body and white adipose tissue weights with a negative skeletal muscle phenotype (metabolic and strength). Such negative effects worsen with ageing but relatively improve through exercise. Abbreviations: KO, knockout; SPARC, secreted protein acidic and rich in cysteine.

Within this context, Aoi et al. reported 24 genes (including *SPARC*) that are both upregulated by exercise and downregulated by ageing [8] suggesting, once more, that some of the exercise-induced benefits such as mitochondrial biogenesis [109] could be SPARC-dependent or partially mediated by SPARC. It is within this perspective that we developed our hypothesis.

The exercise-produced lactate induces both an up-regulation of the expression of the cytochrome oxidase gene and protein, as well as a mitochondrial biogenesis activation signal [79], and since *Sparc* KO did not induce any genotype effect on the different lactate levels we measured (both WT and *Sparc* KO mice had statistically similar lactate levels) but reduced the expression of the MT-CO1 (WT and *Sparc* KO mice have similar lactate levels but WT mice express more MT-CO1 than *Sparc* KO mice), we hypothesize that the exercise-produced lactate-induced cytochrome oxidase upregulation and mitochondrial biogenesis activation do require SPARC (Figure 3A). This would, at least in part, explain why SPARC deficiency reduces tumor growth. Indeed, cancer cells require glycolytic energy and produce lactate, leading to lower pH, compared to normal tissue extracellular pH [110,111], and lactate would be an attempt to increase oxidative phosphorylation capacity via improving mitochondrial biogenesis. However, in the absence of SPARC this lactate-induced mitochondrial biogenesis remains limited, which worsens the tumor bioenvironment and results in tumor progress inhibition. This concept of lactate-related signaling correlates with the theory presenting lactate as a signaling molecule "lactormone" in the context of lactate shuttle [112,113] of lactate formation, utilization and exchange between tissues [114].

**Figure 3.** Hypothetical mechanisms linking SPARC to the exercise-induced SPARC-mediated changes in the skeletal muscle phenotype. [**A**] Exercise induces the secretion and expression of SPARC as well as the production of lactate. Our data suggest that, whereas SPARC enhances collagen expression and muscle strength, the lactate increases mitochondrial enzyme expression in a SPARC-dependant manner (probably via mitochondrial biogenesis induction) and, therefore, the oxidative phosphorylation capacity. [**B**] The similarities between exercise benefits and effects shown to be regulated or modulated by SPARC, such as inflammation, cancer growth, metabolic and structural remodeling of the skeletal muscle, and even neurite outgrowth and neurogenesis, all suggest that part of exercise benefits would be mediated by or dependent on SPARC expression. Abbreviation: SPARC, secreted protein acidic and rich in cysteine.

The results of Figure 1B further support our hypothesis that SPARC-dependent exercise impacts SM. Indeed, whereas there is a significant effect of exercise (increase in COL1A1 expression) in WT mice (WT-Ex > WT-Sed), there is no such effect in *Sparc* KO mice (both KO-Ex and KO-Sed mice have statistically similar expression of COL1A1). This also correlates with the genotype-induced decrease in COL1A1 expression (WT > KO) although both WT and *Sparc* KO mice had equal amounts of exercise (as detailed in the introduction of Section 4), therefore, indicating that exercise-induced COL1A1 is SPARC-dependent (Figure 3A). Moreover, since exercise-induced COL1A1 seems SPARC-dependent, and based on the importance of collagens in the SM structure (and fibrillar collagen I is reported to bind SPARC [115]) and development [116], this could also explain, in part, the low tibialis anterior muscle weight in *Sparc* KO mice compared to WT mice, even though all mice had similar amounts of exercise. Thus, this also suggests that SM development (tibialis anterior weight percentage increase; Table 3), as a part to the adaptation to exercise, would also be SPARC-dependent (Figure 3A).

These conclusions are based on the fact that, as per Table 1, there is no effect of the genotype for LT speed, exercise speed, exercise time, exercise distance and lactate concentrations during exercise. This means that mice of two different genotypes (*Sparc* KO and WT) had equal amounts of training (similar speed, distance, time and frequency). Therefore, genotype effects seen for MT-CO1, COL1A1 (Figure 1) and grip power (Table 4) are, indeed, due to the genotype itself (*Sparc* KO) rather than the difference in the exercise amount; suggesting that these exercise-induced changes are SPARC-dependent/mediated. Such exercise-induced effects in the muscle represent a part of the adaptation via increasing the respiratory capacity, mitochondrial content [52] and contractile properties of the SM [53].

In addition to these SPARC-mediated effects, the similarity of exercise benefits and the effects shown to be regulated or modulated by SPARC, such as inflammation [22,47], cancer growth [8,46], metabolic and structural remodeling of the SM [53], and even neurite outgrowth [117,118] and neurogenesis [85,119], all suggest that some of the exercise benefits would indeed be mediated by or dependent on SPARC expression (Figure 3B).

### **6. Implications and Perspectives**

Our data suggest that the benefits of exercise would also be reduced in *Sparc* KO mice, not only because some of exercise benefits are directly mediated via SPARC, but also because physical performance and muscle performance is reduced as a result of *Sparc* KO (indirect effects). As an illustration, *Sparc* KO increases tumor growth [120,121], loss of SPARC increases cancer progression [122] and the tumorigenesis is prevented and suppressed by exercise-induced SPARC [8,123,124]. More generally, since SPARC is also a myokine secreted during exercise [2,8], exercise benefits including metabolic benefits and inflammation regulation would also be reduced with the *Sparc* KO since SPARC has been shown to play roles in metabolism [44], inflammation regulation [47] and cancer homeostasis [46]. Thus, its absence from circulation would impact tissues other than the SM. For instance, even some of the mechanisms underlying the beneficial effects of exercise that have been shown to involve factors other than SPARC, such as tumor growth suppression through interleukin 6 and epinephrine [92], would also be deficient in *Sparc* KO mice, since the absence of SPARC would reduce the ability of SM to correctly secrete the other myokines involved in the related pathways.

Overall, SM, both as a secretory organ [2] and a metabolic engine, represents the key tissue upon which exercise benefits depend. SPARC represents a "booster" of the SM, with beneficial effects on some other tissues as well. Therefore, SPARC and the pathways it governs would represent good targets to pharmacologically mimic the effects of SPARC, including improved muscle strength and metabolic performances. This is of a particular importance for individuals suffering from health problems, such as heart failure or physical handicap and, therefore, are unable to perform the required physical activity although they need it (reduce obesity, treat lipid disorders, etc.). In such a scenario we could imagine an "exercise pill", targeting SPARC-related pathways and inducing exercise-like effects that would also be of a high therapeutic importance for diseases, such as sarcopenia. Such a therapeutic goal still requires further investigations into the implications of SPARC, not only within the SM but also on other tissues and in the diverse aspects of homeostasis. This will both extend the expected benefits of such an "exercise pill" and anticipate the side effects to decide whether it would be better given as a systemic drug or rather target a specific tissue (that would be the SM based on this study) to optimize clinical efficiency. Importantly, the results obtained from studying SPARC/*Sparc* in mice would be expected to be valid in humans, due the high homology between mouse and human SPARC [104], which would reduce the bridge between results in animals and future clinical studies. Perspectives of the future studies on SPARC implications in SM metabolism and contractile properties, both for sedentary and exercise individuals, would be of great clinical importance not only for SM diseases but also for ageing-related health deterioration and, due to the importance of the SM in the energy homeostasis, energy balance-related pathologies such as obesity and diabetes.

Our study was conducted only on male mice. Therefore, we acknowledge the limitation of sex-determined factors. Similar studies in female subjects, as well as studies comparing male and female subjects, will add significant data of clinical importance, especially with the sex-related difference effects on SM and exercise patterns, such as exercise capacity [125], pinch force reproduction [126], maximal oxygen uptake [127], cardiac adaptation to exercise [125], as well as metabolism, including lactate levels [128,129], aerobic oxidation and anaerobic glycolysis [130]. The parameters illustrated by patterns, including differences between men/male animals and women/female animals in red pepper-induced metabolic phenotype (carbohydrate oxidation Vs lipid oxidation) [131,132], beta-oxidation [130], type I fiber percentage [133] and enzyme activities [134], explain beyond such sex-related differences in exercise and SM properties. Based on such sex-related differences, our results could also indicate that SPARC involvement in exercise-induced muscle phenotype changes could also be sex-dependent and points to a possible interaction between SPARC activities and sexual hormones based on the known impacts of sexual hormones on exercise patterns and the adaptation to exercise [135–137].

Our data indicate that the impacts of *Sparc* KO on body composition, adiposity and metabolic patterns point toward a reduced WAT and body weight, but with negative metabolic and functional phenotypes of SM. Whereas such negative effects on SM worsen with ageing, they are relatively improved by exercise. Importantly, we report, for the first time, evidence suggesting that the exercise-induced changes in the SM phenotype in terms of increased performance (metabolic, strength and development), including lactate-induced changes, are SPARC-dependent. Such important implications of SPARC highlight SPARC and its pathways as pharmacological targets/tools for conditions and diseases in which muscle properties enhancement would provide therapeutic benefits.

**Author Contributions:** Data curation, A.G., A.M., M.Y. and J.S.-A.; Formal analysis, A.G., A.M. and M.Y.; Funding acquisition, J.S.-A.; Investigation, A.G., A.M., M.Y. and J.S.-A.; Methodology, A.G., A.M., M.Y. and J.S.-A.; Project administration, M.Y. and J.S.-A.; Supervision, M.Y. and J.S.-A.; Validation, M.Y. and J.S.-A.; Visualization, A.G., A.M., M.Y. and J.S.-A.; Writing—original draft, A.G.; Writing—review & editing, A.G., A.M., M.Y. and J.S.-A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Canadian Institutes of Health Research (CIHR), Number: 201309MOP-311306-BCA-CFBA-40425.

**Acknowledgments:** We thank Amy D. Bradshaw for providing our research group with sperm of *Sparc* KO mice. Abdelaziz Ghanemi received a Merit scholarship for foreign students from the Ministry of Education and Higher Education of Quebec, Canada, The Fonds de recherche du Québec—Nature et technologies (FRQNT) is responsible for managing the program (Bourses d'excellence pour étudiants étrangers du Ministère de l'Éducation et de l'Enseignement supérieur du Québec, Le Fonds de recherche du Québec—Nature et technologies (FRQNT) est responsable de la gestion du programme). Figures 2 and 3 were created using images from http://smart.servier.com. Servier Medical Art by Servier is licensed under a Creative Commons Attribution 3.0 Unported License.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Effects of High-Impact Weight-Bearing Exercise on Bone Mineral Density and Bone Metabolism in Middle-Aged Premenopausal Women: A Randomized Controlled Trial**

**Sung-Woo Kim 1, Myong-Won Seo 2, Hyun-Chul Jung <sup>3</sup> and Jong-Kook Song 2,\***


**Abstract:** This study examined the effects of high-impact weight-bearing exercise on bone mineral density (BMD) and bone metabolic markers in middle-aged premenopausal women. Forty middleaged premenopausal women were initially enrolled, but thirty-one participants (40.34 ± 3.69 years) completed in the study. The subjects were randomly divided into two groups including the highimpact weight-bearing exercise group (HWE, n = 14) and control group (CON, n = 17). The HWE group participated in the exercise for 50 min a day, three days per week for four months, while the CON group maintained their regular lifestyle. The HWE program included 10 different high-impact weight-bearing exercises such as jumping and running. BMD was measured using DXA (Hologic, QDR 4500W, Marlborough, MA, USA). The bone metabolic markers including serum 25-(OH) D, intact parathyroid hormone (PTH), osteoprotegerin (OPG), osteopontin (OPN), receptor activator of nuclear factor κB ligand (RANKL), osteocalcin (OC), C-terminal telopeptide of type 1 collagen (CTX), and calcium were analyzed. The results showed that the BMDs of femur, lumbar, and forearm did not significantly change during the intervention period in both the HWE and CON groups. A significant decrease in bone formation markers such as OC (*F* = 10.514, *p* = 0.003, ηp<sup>2</sup> = 0.266) and an increase in bone resorption marker including CTX (*F* = 8.768, *p* = 0.006, ηp<sup>2</sup> = 0.232) were found only in the CON group, while these values did not change in the HWE group. There was a significant increase in serum 25-(OH) D (*F* = 4.451, *p* = 0.044, ηp<sup>2</sup> = 0.133) in the HWE group. Our findings suggest that four months of HWE is not sufficient to improve BMD and bone metabolic markers, but this impact exercise program may prevent the age-associated changes in bone turnover markers in middle-aged premenopausal women.

**Keywords:** high-impact weight-bearing exercise; bone mineral density; bone metabolic markers; serum 25-(OH) D; middle-aged premenopausal women

### **1. Introduction**

Physical inactivity adversely affects some health conditions including osteoporosis, neurological diseases, type 2 diabetes, obesity, cardiovascular diseases (CVD), and sarcopenia [1]. In particular, osteoporosis has been reported worldwide as a serious public health problem in recent years. It is estimated that over 200 million women suffer from this disease, and there are approximately 75 million people experiencing osteoporosis in the USA, Europe, and Japan [2–6]. Bone loss is commonly accompanied by aging, which slowly increases bone loss after the age of 35 years, and the process of osteoporosis rises at an exponential rate after menopause [7]. Elderly women with hip fractures have a higher

**Citation:** Kim, S.-W.; Seo, M.-W.; Jung, H.-C.; Song, J.-K. Effects of High-Impact Weight-Bearing Exercise on Bone Mineral Density and Bone Metabolism in Middle-Aged Premenopausal Women: A Randomized Controlled Trial. *Appl. Sci.* **2021**, *11*, 846. https://doi.org/ 10.3390/app11020846

Received: 16 November 2020 Accepted: 16 January 2021 Published: 18 January 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

mortality rate of 10 to 20% than their counterparts, and osteoporosis has a much higher prevalence rate in women over 45 years of age than other diseases including breast cancer, diabetes mellitus, myocardial infarction [8,9]. Furthermore, the previous study suggested that the incidence of hip fracture would increase by 310% in men and 240% in women by 2050 due to the growth of the aging population and vulnerability to hip fracture with age [10]. Due to these clinical prescriptions, the diagnosis, consequences, monitoring of treatment, and therapy for osteoporosis are of crucial importance [11].

Analyzing bone turnover markers, an index of bone metabolism, are novel tools that detect bone remodeling dynamics including bone formation and resorption [12]. Bone remodeling is regulated by the activation of osteoclasts and osteoblasts and these biomarkers reflect the current status of bone turnover rate [11]. Bone turnover rate can be examined by analyzing bone formation markers such as bone-specific alkaline phosphatase, procollagen type 1 N propeptide, and osteocalcin (OC) and bone resorption markers including the N-terminal telopeptide of type 1 collagen and C-terminal telopeptide of type 1 collagen (CTX) [13]. It is also actively regulated by various factors such as osteopontin (OPN), osteoprotegerin (OPG), receptor activator of nuclear factor κB ligand (RANKL), parathyroid hormone (PTH), and vitamin D involved in the bone formation process [14–16]. RANKL is produced in osteoblasts and stromal cells and is the critical factor for the activation of mature osteoclasts and differentiation of monocyte-macrophage osteoclasts, which are precursors to multinucleated osteoclasts [10]. In contrast, OPG, one of the primary regulators of osteoclast-mediated bone resorption, is a factor observed in bone tissue [16]. OPN is a glycoprotein of the extracellular matrix of bone tissues that bind to hydroxyapatite and calcium, and it has been proposed as a mediator of atherosclerosis pathogenic pathways [16]. Vitamin D deficiency may be cause hyperthyroidism, resulting in increased bone turnover and bone loss [17,18]. According to the National Academy of Medicine, vitamin D deficient is defined as a level of serum 25(OH)D of 12 ng·mL−1, whereas a level of 12 to <20 ng·mL−<sup>1</sup> is considered to be inadequate, and a level of ≥20 ng·mL−<sup>1</sup> defines vitamin D sufficiency [19]. Additionally, the elevated circulatory PTH levels may have negative effects [20]. However, several confounding factors determine the cause of the inverse relationship between serum 25(OH)D and PTH, and consistent results are not found in previous studies [20,21]. Calcium plays an important role as an essential nutrient for bone health [22]. However, previous studies have shown an excessively increased serum calcium relationship with an increased risk of cardiovascular disease [23–25]. These causes require an understanding of the potential impacts of serum calcium on bone health and care for health benefits. Thus, exploring the changes in bone turnover markers in response to exercise is essential for identifying mechanisms of the bone response [26].

The mechanical loading of high-impact physical activities (e.g., jumping and running) has been shown to increase bone mass in humans [27]. Additionally, impact activities such as high-intensity weight-bearing and jumping are more likely to be osteogenic than non-weight-bearing or low-impact activities such as swimming and walking [28]. Metaanalysis studies have shown that dynamic and resistance exercises are appropriate for increasing bone mineral density (BMD) in middle-aged women [29,30]. The majority of female participants that have been studied in impact exercise interventions were postmenopausal women because they represent the most popular group in which osteoporosis is manifested. Therefore, studies on pre-menopausal females are perceived to be in the greatest need [31–35]. However, a relatively small number of studies utilizing impact exercise training have targeted pre-menopausal women with combinations of supervised and home-based exercise sessions, which resulted in small positive effects [29,30,36]. Analyzing the bone turnover rate, an index of bone metabolism, plays an important role in understanding the biochemical activities in the bone and the changes in these markers can increase bone mineral density after weight-bearing exercise. It has been reported that acute, short-, and long-term exercise training increased the regulation bone remodeling markers such as OC, CTX, OPN, OPG, RANKL, PTH, and calcium [37–40]. Scott et al. [39] reported

that acute weight-bearing exercise training increased serum levels of OPG, calcium, and PTH. Lester et al. [40] reported that combined aerobic and resistance exercise training for eight weeks significantly increased serum levels of OC in young women. Nevertheless, impact exercise training has not been explicitly directed at adult pre-menopausal women, despite the significant risk of developing osteoporosis after menopause.

Therefore, our study aimed to examine the effects of four months of high-impact weight-bearing exercise on BMD and bone metabolic markers in middle-aged premenopausal women.

### **2. Materials and Methods**

### *2.1. Subjects*

The power test was performed using G\*Power 3.1.9.2 (Franz Faul, University of Kiel, Kiel, Germany) at the effect size of 0.3, the significant level of 0.05 (α = 0.05), and the power of 0.8 for all statistical tests. G\*Power showed that 24 subjects had sufficient power for this study. Forty middle-aged premenopausal women aged 33–47 years were initially enrolled in the study. The exclusion criteria were specified as follows: (1) participated in regular exercise within the last three months; (2) users of pharmaceutical agents that directly affect bone, hypertension, and hyperlipidemia; and (3) blood pressure ≥ 140/90 mmHg. The subjects were randomly divided into two groups including HWE and CON. However, nine dropped out due to personal reasons and pregnancy; thus, thirty-one subjects completed the study (HWE: n = 14, CON: n = 17) (Table 1). All proceedings of the study were approved by the Institutional Review Board of Kyung Hee University (KHGIRB-19-226) and were conducted according to the Declaration of Helsinki.

#### **Table 1.** Physical characteristics and nutritional intake of subjects.


Values are expressed as mean ± SD. HWE: high-impact weight-bearing exercise group, CON: control group, BMI: body mass index.

### *2.2. High-Impact Weight-Bearing Exercise Program*

The training group performed the HWE program for 50 min a day (10:30 a.m.– 11:20 a.m.), three days per week for four months. The training program was composed of 10 min of warm-up, 30 min of HWE, and 10 min of cool-down. The HWE comprised of the clap, walking, jumping burpee, jumping squat, running in place, wall press, bench stepping, jumping lunge, jumping jack, and push up [30]. The exercise intensity including exercise time increased progressively every four week period. The training intensity was set at 60–80% (mean: 136.9 ± 9.03 beat/min, range: 123–180 beat/min) of the heart rate reserve (HRR) and monitored every day by using a potable heart rate monitor (Polar RS400, Polar Electro Oy, Kempele, Finland). The detailed training program is shown in Table 2.


**Table 2.** High-impact weight-bearing exercise program for the four-months of the study.

HRR: heart rate reserve, FITT: frequency, intensity, time, type, HWE: high-impact weight-bearing exercise.

### *2.3. Anthropometric Measurements*

The body height and body weight were measured, and body mass index (BMI) was calculated by dividing body weight (kg) by the square of body height (m2). Body height was measured using a stadiometer to the nearest 0.1 cm (T.K.K. 11253, Takei Scientific Ins Co., Tokyo, Japan). Body weight was measured using a balance beam scale (Seca 700, Seca Co., Hamburg, Germany) to the nearest 0.1 kg.

### *2.4. Bone Mineral Density*

BMD was measured using dual-energy x-ray absorptiometry (DXA), with a Hologic QDR 4500W bone densitometer (Hologic, Marlborough, MA, USA). All participants were scanned at three different sites of the BMD (e.g., femur, lumbar, and forearm). The same technician performed all tests. The intra-class correlation coefficient (ICC) was measured (femur BMD: 0.997; lumbar BMD: 0.990; forearm BMD: 0.987) [41].

### *2.5. Bone Metabolic Markers*

Blood samples were collected before and after the four-month intervention period. Participants arrived at the laboratory in the morning between 08:30 a.m. and 09:30 a.m. after 12 h, overnight fasting, and avoiding severe physical activity or training the night before. Venous blood samples were taken 5 mL at the antecubital vein by a medical laboratory technologist and separated into each serum separate tube. The clotting of blood was separated using centrifugation with 3000 rpm for 15 min. The samples obtained were stored at −80 ◦C until analysis.

Serum 25-(OH) D was measured by a chemiluminescence microparticle immunoassay (CMIA) method (Architect i2000SR, Abbott, Singapore) with an ARCHITECT 25-(OH) vitamin D kit (Abbott Diagnostics, Lake Forest, IL, USA). The electrochemiluminescence immunoassay (ECLIA) method (E170, Roche, Germany) was used for intact PTH and CTX with an Elecsys PTH kit (Roche, Germany) and β-CrossLaps/serum kit (Roche, Germany). Bone metabolic markers including OPG, RANKL, and OPN were measured (VERSA Max, Molecular Device, Sunnyvale, CA, USA) by an enzyme-linked immunosorbent assay (ELISA) with OPG kit (ICL, Portland, OR, USA), RANKL kit (sRANKL ELISA, Immundiagnostik, Germany), and OPN kit (ICL, Portland, OR, USA). A radioimmunoassay (RIA) method (COBRA 5010 Quantum, USA) was used for analyzing OC, a marker of bone formation, with an OC kit (BRAHMS, Berlin, Germany). Calcium was measured by a homogeneous enzymatic colorimetric assay (HECA) method (c702, Germany) with a Calcium Gen.2 (Roche, Germany). The blood analysis was performed at the expertized

research laboratory (Green Cross LabCell, Korea). The company is certified by the Korean Board for Accreditation and Conformity assessment.

### *2.6. Nutritional Intake*

The dietary intake was assessed by three-day dietary records including twice on weekdays and once at the weekend in the first and the last weeks of the intervention. All data were analyzed by a computerized nutrient-intake assessment software program (CAN-PRO 4.0, Korean Nutrition Society, Seoul, Korea). The participants were instructed to maintain the same macronutrient distribution throughout the study.

### *2.7. Statistical Analysis*

Statistical analyses were performed by using SAS software version 9.4 (SAS Institute, Cary, NC, USA). The mean, standard deviation, and 95% confidence interval (CI) were calculated. The normality of distribution of all dependent variables was verified using the Kolmogorov–Smirnov test. Two-way repeated-measures ANOVA was applied to determine the interaction effect for the group by the time during the intervention. If any significant interaction or main effects were observed, the independent t-test and paired t-test were applied to analyze the significant differences within groups and between groups. The effect size was computed as partial eta-squared values (η<sup>p</sup> 2; small: ≥0.01, medium: ≥0.06, large: ≥0.14) [42]. The statistical significance level was set at 0.05.

### **3. Results**

### *3.1. Bone Mineral Density*

There were no significant interaction effects for group by time on femur BMD (*F* = 0.458, *p* = 0.504, η<sup>p</sup> <sup>2</sup> = 0.016), lumbar BMD (*F* = 0.009, *p* = 0.925, η<sup>p</sup> <sup>2</sup> = 0.000), and forearm BMD (*F* = 0.048, *p* = 0.827, η<sup>p</sup> <sup>2</sup> = 0.002) during the intervention period (Table 3).

**Table 3.** Changes of bone mineral density (BMD) between pre- and post-tests in middle-aged premenopausal women.


Values are expressed as mean ± SD. 95% CI: 95% confidence interval, BMD: bone mineral density, HWE: high-impact weight-bearing exercise, CON: control.

### *3.2. Bone Metabolic Markers*

There were significant interaction effects for group by time on serum 25-(OH) D (*F* = 4.451, *p* = 0.044, η<sup>p</sup> <sup>2</sup> = 0.133) and OC (*F* = 10.514, *p* = 0.003, η<sup>p</sup> <sup>2</sup> = 0.266) (Figures 1 and 2). The post-test showed that serum 25-(OH) D concentration was increased significantly following the four-months of high impact weight-bearing exercise (*p* < 0.01), while no significant change was found in the CON group. OC level was not changed in the HWE group, but this value decreased significantly in the CON group (*p* < 0.05). There were significant time effects on intact PTH (*F* = 4.447, *p* = 0.044, η<sup>p</sup> <sup>2</sup> = 0.133), OPN (*F* = 5.480, *p* = 0.026, η<sup>p</sup> <sup>2</sup> = 0.159), CTX (*F* = 8.768, *p* = 0.006, η<sup>p</sup> <sup>2</sup> = 0.232), and calcium (*F* = 7.986, *p* = 0.008, η<sup>p</sup> <sup>2</sup> = 0.216) (Table 4).

**Figure 1.** Serum 25-(OH) D before and after the four-month exercise program. For serum 25-(OH) D, statistical analyses revealed an increase between pre- and post-tests in the HWE group. HWE: highimpact weight-bearing exercise, CON: control. Significant difference between pre- and post-tests, \*\* *p* < 0.01. Significant interaction effect, # *p* < 0.05.

**Figure 2.** Osteocalcin before and after the four-month exercise program. For osteocalcin, statistical analyses revealed a decrease between pre- and post-tests in the CON group. HWE: high-impact weight-bearing exercise, CON: control. Significant difference between pre- and post-tests, \* *p* < 0.05. Significant interaction effect, ## *p* < 0.01.


**Table 4.** Changes in bone metabolic markers between pre- and post-tests in middle-aged premenopausal women.

Values are expressed as mean ± SD. 95% CI: 95% confidence interval, HWE: high-impact weight-bearing exercise, CON: control, PTH: parathyroid hormone, OPG: osteoprotegerin, RANKL: receptor activator of nuclear factor kB ligand, OPN: osteopontin, CTX: C-terminal telopeptide of type 1 collagen. Significant interaction or main effect, # *p* < 0.05, ## *p* < 0.01. Significant difference between pre- and post-tests, \* *p* < 0.05, \*\* *p* < 0.01.

### **4. Discussion**

The present study examined the effects of four months of HWE on BMD and bone metabolic markers in middle-aged premenopausal women. Our main findings showed that a significant decrease in OC and an increase in CTX were found only in the CON group while these values did not change in the HWE group. However, four months of HWE were not sufficient to improve BMD in middle-aged premenopausal women.

Substantial empirical evidence has shown that resistance and step exercise have several benefits, which work as non-pharmacological interventions for the improvement of bone health in children/adolescents, adulthood, and the elderly [31,43]. However, the effects of high-impact exercise on BMD in middle-aged pre-menopausal women remain unknown [36]. The present study showed no significant effects on femur, lumbar, and forearm BMD after four-month of high-impact weight-bearing exercise. The International and National Osteoporosis Foundation, and other organizations recommend weight-bearing exercises to prevent osteopenia and osteoporosis [44]. There are different types of body weight-bearing exercises including high-impact exercises such as running and jumping as well as low-impact exercises such as body weight training, water aerobics, and walking [44]. It has been recently reported that trochanter and femoral neck BMD were significantly improved after high-impact weight-bearing exercise in pre-menopausal women [30]. Additionally, high-impact weight-bearing and resistance training either in combination or alone, induced gains in BMD by 1–2% at the femoral neck and lumbar spine with a minimal effect on other body sites in pre-menopausal and post-menopausal women [45]. Similar results in other studies have found that high-impact exercise for 12 months (3 times/week) resulted in a significant improvement in femoral neck BMD by 0.6–1.1% in pre-menopausal women [46,47]. Although the present study did not show a significant change, there was a tendency where the femur BMD increased following four months of HWE (0.47%) without any improvement on other sites of the body. Foster and Armstrong [48] reported that a bone remodeling cycle takes about 3–4 months and a minimum of 6–8 months is required for the bone to reach a new, measurable steady-state bone mass. Therefore, the absence of the significant improvement in BMD is likely due to the insufficient duration of an exercise intervention to promote bone formation in the present study. In fact, combinedimpact exercise protocols lasting more than six months (duration: 30–60 min; frequency: ≥3 days/week) have been recommended for greater bone health benefits [49]. Moreover, adults might benefit from high-impact exercises, but previous studies have suggested that its effect might occur to a lesser extent than in children or adolescents [50]. In this study, the effect of four-months of high-impact weight-bearing exercise appeared to be none or small only in particular sites such as femur BMD in middle-aged premenopausal women.

Biochemical bone metabolic markers indicate changes in bone metabolism more quickly than changes in BMD, which are commonly used to monitor osteoporosis [51]. Additionally, bone metabolic markers could be used to evaluate the antiresorptive agents in the clinical diagnosis of osteoporosis and osteopenia [52]. One intended therapeutic aim of the antiresorptive therapy is to decrease levels of bone metabolic markers below the average reference range for pre-menopausal women [53]. Bone metabolic markers provide important information such as the rate of bone remodeling by changing bone formation and resorption, and the evaluation of these markers has been used in previous intervention studies to identify the effectiveness of exercise [54–56]. The results of our study showed that serum 25-(OH) D (60.22%, *p* < 0.01), and intact PTH (23.86%, *p* < 0.05) were increased significantly in HWE between pre- and post-tests, whereas OPN (−20.45%, *p* < 0.05) was decreased significantly in HWE between pre- and post-tests. Furthermore, OC (−7.89%, *p* < 0.05) and calcium (−2.99%, *p* < 0.01) were decreased significantly in CON between pre- and post-tests, whereas CTX (30.94%, *p* < 0.05) was increased significantly in CON between pre- and post-tests. However, no significant interaction effects were found on other markers including intact PTH, OPG, RANKL, and OPN.

Vitamin D plays a critical role in promoting the intestinal absorption of phosphate and calcium [57], and participating in regular physical activity is a well-known stimulator of serum 25-(OH) D [58]. In accordance with our results, Josse et al. [55] reported that the serum 25-(OH) D increased significantly following 12 weeks of resistance training in young women, but the level of PTH was not significantly changed. It is an interesting result in the present study where the HWE group increased both vitamin D and PTH level after four-months of exercise intervention. Generally, serum calcium level is mediated by vitamin D and PTH [59] and the changes in these levels may increase or decrease the mineralization of bone. It is assumed that an increase in PTH level may help to maintain serum calcium level. In particular, our participants were healthy and had no bone-related disease, thus it is possible that PTH is increased for calcium homeostasis. Moghadasi and Siavashpour [56] reported that resistance circuit training for 12 weeks significantly increased PTH in healthy and sedentary young women, which agrees with the present result. Authors assumed that this increase in PTH after resistance training may promote anabolic activities in bone metabolism. However, it is difficult to explain the direct relations or physiological mechanism between the variables in the present study and conflicting results have been observed in previous studies. Pilch et al. [60] reported that Nordic walking training for six weeks significantly decreased serum 25-(OH) D in post-menopausal women and Evans et al. [54] reported that a recruit training program for four-months significantly decreased serum 25-(OH) D in men, whereas these values remained at baseline levels in women. Vitamin D is an important factor for bone health, and its deficiency leads to osteoporosis and CVD in adults [61,62]. Moreover, physically inactive individuals have two times lower vitamin D levels than active adults [63]. Although it is difficult to determine the direct relations between HWE and increased serum 25-(OH) D, several studies support our results that increase in physical activity is an effective manner to maintain optimal vitamin D level in later life [64,65]. In this study, the serum 25-(OH) D increased from an inadequate level (12–20 ng·mL−1) to a sufficient level (≥20 ng·mL−1) in HWE.

Bone turnover markers provide information on important biochemical indicators such as bone resorption and bone formation in managing osteoporosis [66]. A significant association between bone turnover markers and the risk of fractures has been identified for females and males [67]. Mohr et al. [68] reported that soccer training for 15 weeks

significantly increased plasma OC (37 ± 15%) and CTX (42 ± 18%) in middle-aged premenopausal women. A significant increase in OC has also been observed in untrained young women following sixteen weeks of exercise intervention [69]. Additionally, exercise training caused augmentation in resting plasma OC and CTX in elderly sedentary men with no resistance training [70]. However, CTX has been shown to increase shortly or immediately after impact exercise [26]. Thus, impact exercise may have a high potential to improve bone health in the general population. Therefore, the above intervention studies and our study results indicate that impact exercise describes a feasible intervention modality in middle-aged premenopausal women.

Previous studies by Ferrucci et al. [71] and Fuller et al. [72] emphasized that changes in pro-inflammatory and anti-inflammatory cytokines have profound implications on agerelated bone loss. OPN acts through binding to integrin β-3, leading to a decrease in the cytoplasmic calcium concentration associated with osteoclast activation [10]. Humphries et al. [38] reported that participants who underwent whole-body vibration with or without the combination of resistance training for 16 weeks showed a significant decrease in OPN in young healthy women, which is consistent with the present findings. Thus, exercise training induces changes in bone metabolism faster than actual changes in BMD in middleaged premenopausal women. Therefore, it is important to monitor and stimulate bone metabolic markers to prevent osteoporosis before menopause in women, especially when considering a higher prevalence of osteoporosis after menopause.

### **5. Limitation of the study**

In this study, there are some limitations that need to be considered. Although the present study is well designed with the randomized controlled trials, the relatively shortterm intervention period (four-months) may be a limit to confirming the longitudinal effects of high-impact weight-bearing exercise on BMD in middle-aged premenopausal women. However, the positive changes in bone metabolic markers such as OC and vitamin D may provide a potential benefit on bone health. In future study, a longer period of exercise intervention (more than six months) is needed to examine how changed bone metabolic markers are linked to the BMD changes.

### **6. Conclusions**

The present study revealed that four months of HWE is not sufficient to improve BMD and bone metabolic markers, but this exercise program may prevent the age-associated changes in bone turnover markers in middle-aged premenopausal women.

**Author Contributions:** Conception and study design, S.-W.K. and J.-K.S.; Statistical analysis, S.-W.K. and J.-K.S.; Investigation, S.-W.K.; Data interpretation, S.-W.K., J.-K.S., M.-W.S., and H.-C.J.; Writingoriginal draft preparation, S.-W.K.; Writing-review and editing, H.-C.J., and J.-K.S.; Supervision, J.-K.S. All authors have read and approved the final manuscript.

**Funding:** This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2019S1A5A2A01049721).

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board of Kyung Hee University (KHGIRB-19-226).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


### *Review* **Physical Activity and Redox Balance in the Elderly: Signal Transduction Mechanisms**

**Daniela Galli 1,†, Cecilia Carubbi 1,†, Elena Masselli 1,2, Mauro Vaccarezza 3,4, Valentina Presta 1, Giulia Pozzi 1, Luca Ambrosini 1, Giuliana Gobbi 1,\*, Marco Vitale 1,2,\* and Prisco Mirandola <sup>1</sup>**


**Abstract:** Reactive Oxygen Species (ROS) are molecules naturally produced by cells. If their levels are too high, the cellular antioxidant machinery intervenes to bring back their quantity to physiological conditions. Since aging often induces malfunctioning in this machinery, ROS are considered an effective cause of age-associated diseases. Exercise stimulates ROS production on one side, and the antioxidant systems on the other side. The effects of exercise on oxidative stress markers have been shown in blood, vascular tissue, brain, cardiac and skeletal muscle, both in young and aged people. However, the intensity and volume of exercise and the individual subject characteristics are important to envisage future strategies to adequately personalize the balance of the oxidant/antioxidant environment. Here, we reviewed the literature that deals with the effects of physical activity on redox balance in young and aged people, with insights into the molecular mechanisms involved. Although many molecular pathways are involved, we are still far from a comprehensive view of the mechanisms that stand behind the effects of physical activity during aging. Although we believe that future precision medicine will be able to transform exercise administration from wellness to targeted prevention, as yet we admit that the topic is still in its infancy.

**Keywords:** ROS; physical activity; signal transduction; aging

### **1. Introduction**

In 2018, the World Health Organization proposed healthy aging as "creating the environments and opportunities that enable people to be and do what they value throughout their lives". In western countries, aged people are numerous, and the pathologies associated with aging are widely studied. Besides pathogens, excessive Reactive Oxygen Species (ROS) can be considered as an effective cause of age-associated diseases [1].

ROS (i.e., hydrogen peroxide, nitric oxide radicals, hypochlorite) are molecules naturally produced in several metabolic cell processes. However, if their levels are above the physiologic threshold, they cause oxidative stress and cell damage. In fact, if ROS are too high, the cellular antioxidant machinery rises to bring their levels back below the threshold [2]. Thus, to promote healthy aging, adequate interventions should constrain ROS in their physiological boundaries and prevent their deleterious effects [3].

Physical activity is a double-edged sword in terms of oxidant/antioxidant balance, and whenever sport is practised for health—which should be the most common condition it should be programmed on a personal basis and on the basis of validated evidence. Data in this field are quite contradictory, often making difficult their translation into practical

**Citation:** Galli, D.; Carubbi, C.; Masselli, E.; Vaccarezza, M.; Presta, V.; Pozzi, G.; Ambrosini, L.; Gobbi, G.; Vitale, M.; Mirandola, P. Physical Activity and Redox Balance in the Elderly: Signal Transduction Mechanisms. *Appl. Sci.* **2021**, *11*, 2228. https://doi.org/10.3390/ app11052228

Academic Editor: Mark King

Received: 9 February 2021 Accepted: 26 February 2021 Published: 3 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

settings, particularly for the elderly, where the effects of aging on the oxidant/antioxidant balance intersect those of exercise. In fact, during exercise, ROS are produced in at least two biochemical pathways: (1) the mitochondrial electron transport chain; and (2) the system of xanthine oxidase that is responsible in the formation of peroxynitrite. With the aim of contrasting ROS effects, all tissues use their reserve of antioxidants, vitamins and glutathione [2,4].

Keeping in mind that oxidative stress is lower in fertile women because of oestrogen protection [5], a high quantity of ROS is generally induced by exercise that provokes modifications of antioxidant activity, both in cardiac and skeletal muscle [6]. Interestingly, it seems that both aerobic and anaerobic training stimulate the antioxidant system with respect to what happens in untrained subjects. Thus, it seems there are no specific exercise effects on redox balance. Moreover, the activation of antioxidant capacity is the same between the rest conditions of the three groups (aerobically, anaerobically and untrained people), suggesting that the antioxidant system is activated only transiently by exercise [6]. From a molecular point of view, DNA analysis of oxidative damage has shown that immediately after exercise, regardless of intensity, there are no signs of oxidative damage. Additionally, acute or extended moderate exercise does not induce DNA damage. Instead, it seems to be associated with decreased levels of oxidation. Finally, extended intense exercise increase DNA modifications [7].

More recently, studies performed in hypoxic conditions have shown that low–moderate exercise exerts a positive influence to protect against altitude/hypoxia-induced oxidative stress, while higher-intensity exercise increases oxidative stress [8]. These observations should be considered for training adaptations in athletes, to hypoxic or highaltitude conditions.

Finally, moderate exercise has been shown to be important to reduce the risks of cardiovascular diseases. In fact, acute cardiovascular exercise increases the oxidative stress, while regularly performed cardiovascular exercise increases the antioxidant capacity, of the cells [9]. Thus, it seems that the body's antioxidant responses are proportional to exercise intensity, but too-high-intensity exercise produces inflammation and cell damage, both in young and elderly subjects. On the other hand, de Sousa et al. (2017) [10] showed that, independently of intensity, volume, type of exercise and population, physical exercise always has an antioxidant effect. In conclusion, moderate-intensity exercise should be the right compromise to balance oxidant effects.

Interestingly, different scenarios are present in athletes with different levels of ability: while in elite athletes, physical exercise is not enough to maintain the oxidant/antioxidant balance, making an anti-oxidant supplementation necessary, in amateur and master athletes, oxidant/antioxidant activity is kept balanced by exercise, protecting the muscle from damage [11].

Of note, physical exercise also shows positive effects on the redox environment in non-muscular systems such as the cerebral [12,13] and vascular [14] systems. In fact, de Souza et al. (2019) [12] studied the effect of exercise volume on antioxidant enzyme levels in various brain regions. They found that 30–60 min of exercise increased Catalase (CAT) activity, while 8 weeks of training increased Superoxide Dismutase (SOD). More than 8 weeks of training increased both ROS and antioxidant activity. Thus, four to eight weeks of moderate exercise promote a healthy balance of antioxidant enzymes in the rodent brain. Recently, Pinho et al. (2019) [13] reviewed the effects of resistance training to counteract oxidative stress in the brain. In particular, they showed evidence that Insulin Growth Factor-1 (IGF-1) in muscle activates the Protein Kinase B (Akt) signalling pathway at brain level. However, further investigations are necessary, since some observations are only speculative.

In 2016 Park et al. [14] studied the positive effects of exercise on redox balance in vessels. They found that arteries from trained mice presented higher levels of antioxidant markers such as: (i) Peroxisome Proliferative Activated Receptor-coactivator-1 (PPARgamma), (ii) Cytochrome-C Oxidase Subunit IV isoform 1 (COX4I1), and (iii) Isocitrate Dehydrogenase 2 (Idh-2). Moreover, the trained mice showed higher respiratory capacity and improved oxidant/antioxidant balance compared to untrained mice. These results sustain the hypothesis that physical activity can be vasculo-protective. Unfortunately, further studies are necessary.

The aim of this review is to unravel the relationship between redox balance, physical activity and aging, analysing what is known of the involved molecular pathways. It has therefore been organized into five sections: (1) Introduction, (2) Physical activity and redox balance in the young; (3) Physical activity and redox balance in the elderly; (4) Insights into signal transduction; and (5) Conclusions.

Figure 1 reports the diagram of the process that we apply to select papers to review.

### **WZ/^DϮϬϬϵ&ůŽǁŝĂŐƌĂŵ**

**Figure 1.** Flow diagram according to Prisma 2009 guidelines. A total of 500 records were identified for this study in three online databases (Pubmed, Scopus, Web of Science) using this search expression: "Redox AND Balance AND Physical activity OR Exercise AND Age AND Signal transduction". After removing duplicates, we found 430 articles, and after first screening, 130 were excluded. 300 full-text articles were assessed for eligibility, and 209 were excluded because results were partially redundant. Finally, 91 studies were included in qualitative synthesis. We did not use any date limit.

### **2. Physical Activity and Redox Balance in the Young and Adults**

In young people, the majority of papers correlated obesity with oxidative stress. In fact, obesity is often associated with high levels of free fatty acids. This increases the levels of

NADH that generate high oxidative stress. In obese young subjects (males of 20–26 years), ROS levels are significantly higher than in non-obese, while (SOD) antioxidant activity is significantly lower [15]. After training (aerobic: 40 min treadmill, three times/week for 8 weeks; 70% Heart Rate reserve), ROS decreases and SOD increases. Interestingly, the authors also found a significant increase in Brain-Derived Neurotrophic Factor (BDNF), but not in Glial-cell-line-Derived Neurotrophic Factor (GDNF) or in Nerve Growth Factor (NGF) after aerobic training. At the same time, blood markers of blood-brain barrier (BBB) damage decreased. Notably, BDNF is lower in obese subjects, compared to nonobese. In non-obese subjects, the authors did not find significant changes in the levels of neurotrophines. Finally, Di Liegro et al. (2019) [16] reviewed the effect of physical activity on brain health. They found that BDNF is produced in the periphery of the nervous system in a physical-activity-dependent manner.

A very recent review [17] showed that overweight/obesity does not affect exercise lipid-oxidation systems, suggesting that physical activity could have a positive effect, not directly on lipid oxidation, but on blood-brain barrier damage. Interestingly, three months of Crossfit training increased BDNF level in active men [18]. In particular, they observed improvements in endurance performance (expressed by VO2max) and the levels of plasma antioxidant markers such as SOD, glutathione reductase (GSH), Uric Acid (Uric Acid), the ferric reducing ability of plasma and BDNF. Altogether, these data suggest that physical activity exerts a positive effect on brain health, both reducing damages that can occur to the blood-brain barrier and increasing the levels of the antioxidant BDNF.

Besides physical activity, antioxidant supplementation is widely distributed to counteract oxidative stress. In fact, the antioxidant effect of polyphenolic compounds is related to the inhibition/inactivation of several pathways such as the Nuclear-factor kappa light-chain enhancer pathway of activated B cells (Nf-kB), and the kappa kinase/c-Jun amino-terminal kinase pathway (IKK/JNK), that are involved in oxidation [19]. Moreover, polyphenolic molecules inhibit the activity of the enzymes responsible for ROS production, such as cyclooxygenase, lipoxygenase and xanthine oxidase. Of note, the effects of polyphenols can affect the activation of antioxidant pathways by exercise. For example, Cavarretta et al. (2018) [20] studied the effect of cocoa polyphenols in elite football players practising intense physical exercise. After 30 days of chocolate intake, the elite football players showed a significant decrease in their levels of muscle-damage markers such as Creatin Kinase (CK), Lactate dehydrogenase (LDH) and Myoglobin and an increased antioxidant power, suggesting that polyphenols positively modulated redox balance and reduced muscle tissue injury in elite football players.

On the other hand, activation of antioxidant mechanisms in Glucose-6-Phosphate Dehydrogenase (G6PDH) patients (adults, 33–43 years old), with the Alpha-Lipoic Acid (ALA) supplementation, increases the antioxidant defence without modifying the effects of exercise. In fact, administration of ALA (600 mg/day) or placebo for 4 weeks did not affect their performance with 45 min of treadmill, at 70–75% VO2max and then 90% until exhaustion [21]. Thus, administration of compounds like polyphenols seems to reinforce the antioxidant effect of exercise.

Additionally, some authors have checked the possibility that training level could be associated with different effects on oxidative stress. For example, in adolescent swimmers, it has been shown that high-intensity exercise modifies redox balance without inducing prolonged oxidative stress [22]. Similarly, ultra-marathon swimming did not induce oxidative stress in well-trained swimmers [23], suggesting that well-trained swimmers are able to regulate redox balance. These observations could be important in scheduling training programs for ultra-marathon swimmers.

These results, however, apparently depend on sport intensity. In fact, in amateur women gymnasts, low–moderate-intensity training increases total antioxidant activity after 48 h [24]. Instead, 48 h' recovery after high-intensity training is not enough to restore redox balance. A modest ROS increase is necessary for normal force generation, while higher ROS levels induce damage in a dose- and time-dependent manner [25]. Thus, a diet rich in antioxidants and sessions of low–moderate intensity training should be recommended to amateur gymnasts after high-intensity training [24].

Interestingly, antioxidant supplementation produces divergent results on exercise performance. For example, vitamin C and E supplementation and resistance activity induce a decrease in protein ubiquitination [26]. Moreover, vitamin supplementation reduces the phosphorylation of p38 Mitogen-Activated Protein Kinase (p38MAPK), of Extracellular signal-Regulated protein Kinases 1 and 2 (ERK1/2), and of p70S6 kinase [26,27]. Similarly, in rats trained with aerobic exercise, the antioxidant vitamin C blunts exercise-induced adaptations such as mitochondrial biogenesis and degradation of worn-out mitochondrial proteins [28].

Thus, the combination of antioxidant and exercise can induce contrasting effects, suggesting that further investigations are required on this subject.

### **3. Physical Activity and Redox Balance in the Adult and the Elderly**

Aging is associated with a decrease in the activity of anti-oxidant systems, while exercise reduces the levels of oxidative stress markers, thus activating the antioxidant enzymes [29], suggesting that the greater the physical exercise, the smaller the effects of aging. Although this hypothesis could be very intriguing, it is generally supposed that low– moderate physical activity induces adaptations that increase resistance to oxidative stress, since exercise training decreases the risk for several diseases associated with oxidative stress. Many studies have been published on this topic, but conflicting results have been reported, likely due to different parameters and measurement methods used: many groups have reported protective effects of exercise on antioxidant systems, while others have reported negative effects of training.

What is commonly accepted is that markers of oxidative stress are significantly lower in trained subjects with respect to untrained ones, similar to the conditions observed in young people. For example, 8 weeks of walking does not induce oxidative stress in aged subjects [30]. In fact, Low-Density Lipoprotein (LDL) oxidation and nitration levels are not affected by walking. Instead, LDL nitration is modified by acute moderate activity. Thus, it seems important to practice intense exercise to stimulate antioxidant enzymes, rather than low–moderate activity. This condition appears also after acute oxidative challenge. In fact, fit aged subjects show lower oxidative stress than unfit age-matched subjects at baseline conditions and after an acute oxidative challenge such as forearm ischemia-reperfusion [29]. However, this is not due to lower levels of circulating antioxidant molecules. Moreover, reduced oxidative damage that has been shown in fit individuals cannot be attributed to physiological parameters like adiposity or High-Density Lipoprotein (HDL). Instead, it is likely related to differences in antioxidant enzyme activity. Thus, to be physically fit appears an effective strategy for fighting age-related oxidative damage [29].

In 2016, Done and Traustadottir [31] showed that sedentary middle-aged adults could increase their resistance to oxidative stress (forearm Ischemia Reperfusion (IR)) through whole-body aerobic training (45-min sessions three times per week, with an intensity of 70–85% maximal heart rate). In fact, the level of the oxidative marker F2-isoprostanes was significantly decreased after exercise, but not at the baseline pre-IR [31].

Age-related dysfunctions of redox balance also determine endothelial alterations that have been associated with cardiovascular disease. Ten days of treadmill training (1 h at 70% of VO2max) significantly increase flow-mediated dilation of arteries and Vascular Endothelial Growth Factor Receptor-positive (KDR+) circulating cells, while not affecting the level of SOD mRNA, intracellular nitric oxide (NO), ROS, endothelial Nitric Oxide Synthase (eNOS), NADPH oxidase 2 and neutrophil cytosolic factor 1 in healthy older adult subjects. Thus, these data suggest that short-term aerobic training can reduce risk factors for cardiovascular disease [32].

As for young subjects, the effect of antioxidant molecules, alone or in combination with exercise, has been tested for muscle physiology and function. For muscle tissue, a good antioxidant has not been found yet, although the effect of plant extract and molecules like resveratrol and vitamins have been investigated [33–37]. For example, a waste-water polyphenolic mixture (LACHI MIX HT) containing hydroxytirosol (HT), gallic acid and homovanillic acid induced functional amelioration in the skeletal muscles of 27-month-old rats [33]. Moreover, a decrease of oxidative stress marker levels was observed [33]. The resveratrol had been used, with positive effects, in 18-month-old mice in combination with exercise for 4 weeks [34]. In particular, decreases in blood lactate, free fatty-acid levels, and gastrocnemius-muscle lipid peroxidation were observed, as well as an increase in the activity of antioxidant enzymes such as SOD and catalase [34]. Interestingly, the upregulation of peroxisome proliferator-activated receptor gamma coactivator-1α (PGC-1α) mRNA was also observed [34], suggesting the involvement of mitochondrial biogenesis and function. Similar effects have been obtained with plant (Rhus Coriaria) extract in human myoblasts, where Najjaret et al. (2017) [36] observed an increase of the antioxidant enzymes SOD2 and catalase, together with increased viability and adhesion of human myoblasts. Finally, oral ingestion of ascorbic acid increases muscle blood flow and oxygen consumption in older adults practising rhythmic handgrip exercises [37]. Antioxidant supplements have been tested for the treatments of sarcopenia which is a muscle disease known to be associated with the oxidant/antioxidant balance in the elderly. Specifically, sarcopenia is a condition characterized by the progressive loss of muscle mass and function, a decrease in number of motor units, and wasting of muscle fibres. Among the complex biological mechanisms associated with sarcopenia, there is a pronounced imbalance between oxidant and antioxidant species. As yet, it is not clear if antioxidant supplements have a protective effect against the development of this disease [38]. On the other hand, physical exercise could have a role in counteracting this disease, because at a low–moderate intensity, it stimulates antioxidant pathways [39].

Of note, resistance exercise increases satellite cell number, and reverses muscle atrophy in aging by increasing the number of fast fibres that are reduced in the elderly [40,41]. For example, cycles of 12 weeks of resistance training, 12 weeks of de-training and 12 weeks of re-training induce amelioration of physiological and cellular parameters of muscle in aged men [42]. The training increases power and strength in knee-extension exercises, and reduces the number of both 2a and 2x fast fibres. The de-training induces a modest loss of power and strength. Finally, 12 weeks of re-training produce a significant increase in type-II fibre hypertrophy, satellite cell number and myonuclei [42]. On the other hand, 12 weeks of aerobic training in sedentary healthy subjects increase the cross-sectional area of both type-I and type-II fibres, while satellite activation and nuclei addition are found only in slow type-I fibres, suggesting that a differential regulation concerning myonuclear condition occurs [43]. Since the slow fibres have a higher oxidative content than do fast type-II fibres, resistance training could represent a better training method to counteract aging effects on muscle mass and function, but this point is still under debate. As the increase of antioxidant species was observed in many studies, and the reduction of oxidant species (malondyaldehyde) was observed in type-II diabetic patients performing moderate aerobic exercise three times/week [44], with no changes in the control group, it suggests that there is a specific effect of exercise in diabetic patients on lipid peroxidation levels and on the susceptibility of DNA to oxidative damage.

Although it is not directly influenced by physical activity, nervous tissue has the highest metabolic rates in the organism [45] and this elevated oxygen consumption promotes ROS generation, making the nervous tissue more susceptible to developing oxidative stress [45]. Moreover, antioxidant levels are lower in the brain than in other parts of the body, due to the low permeability of the blood-brain barrier to most endogenous antioxidant molecules [45]. It is well known that aging and neurodegenerative diseases show increased levels of ROS and reduction of BNDF and NGF: oxidative stress modulates the activity of neurotrophins that then become unable to save neurons from cell death and to promote neuroplasticity [46].

Lifestyle strongly influences the oxidative stress state, and as a consequence, it could directly affect the developmental and clinical aspects of neurodegenerative disease. It has been reported that diets rich in antioxidants may delay the development of neurodegenerative disease [46]. Polyphenols show potential positive effects against Alzheimer's Disease (AD), preventing ROS accumulation in brain tissue and clearing neurofibrillary tangles. For instance, curcumin can upregulate anti-tau BAG2, generating a putative beneficial factor against AD-associated tauopathy [47].

Several different physical-activity interventions in the elderly affected by degenerative diseases have been reported. However, the conclusions of the different extensive and systematic analyses reported in several reviews are not always optimistic [48].

Trials suffer from a small number of enrolled subjects, different and short follow-up times, and great differences in type, frequency, intensity and duration of physical-activity protocols. Moreover, one's sedentary status is defined and measured differently. Finally, many different cognitive tests and a plethora of fitness assays are used, making comparison more difficult.

### **4. Insights into Signal Transduction**

The role of physical activity as an antioxidant to counteract the negative effects of aging and chronic disease has been widely studied, and we summarize below the main results obtained so far. First, it is worth noting, however, that we are still far from a structured comprehensive view of this complex topic. Besides, in this field, several studies have been done in the mouse model, and the translation to humans is far away.

Several different signalling pathways have been implicated in the control of redox balance by physical activity, also depending on the analysed organs or systems.

### *4.1. Plasma and Adipose Tissue*

Increased oxidative stress is related to abnormal activation of the renin-angiotensinaldosterone system [49]. Resistance exercise increases the level of Triacylglycerol Lipase Activity (TGLA), the level of plasma glycerol and the level of oxidative stress in obese with respect to lean subjects [49,50]. This is associated with higher levels of serum non-esterified fatty acid and increased lipolysis in adipose tissue [50]. The relationship between exercise and obesity is under study in a clinical trial (current controlled trials, ISRCTN95488515). In particular, they are investigating the effects of endurance and high-intensity intermittent endurance exercise on plasma concentrations of glycerol, Atrial Natriuretic Peptide (ANP) and Brain Natriuretic Peptide (BNP). In fact, ANP and BNP are potent lipolytic agents. The results of this trial will be useful for optimizing training protocols for the prevention and treatment of obesity [51].

Contrasting results have been obtained on the role of exercise on plasma levels of cytokines [52–54]. For example, triathlon competitions like Ironman or Half Ironman significantly increase the levels of oxidative stress markers, Interleukin-6 (IL-6), Interleukin-1 (IL-1) and Tumour Necrosis Factor alpha (TNF-alpha). However, the training hours spent before the competitions were inversely correlated with the level of IL-6 found, suggesting that only IL-6 is related to the time of training [52]. In aged subjects affected by rheumatoid arthritis, 10 weeks of a walking-based high-intensity interval training (HIIT) increased ROS production, but there were no significant changes in the levels of inflammatory markers such as IL-1, IL-6, TNF-alpha and Interleukin-10 (IL-10) [53]. On the other hand, a study performed in post-menopausal women observed a significant decrease in IL-2, IL-4, IL-6 and TNF-alpha levels, both after 3 months of aerobic training (60–70% of maximal heart rate on treadmill, with low speed and without inclination) and after 3 months of resistance training with better results in terms of BMI and quality-of-life in subjects performing aerobic training [54]. In contrast, a single short bout of resistance exercise in healthy untrained men increased catecholamines, Epidermal Growth Factor (EGF), IL-2 and TNFalpha, with the levels of IL-1α, IL-1β, IL-6, IL-8 and IL-10 maintained constant levels [55]. Additionally, Accattato et al. (2017) [49] observed that a single bout of physical activity does not induce changes in plasma cytokine levels. Instead, they observed a decrease of EGF and an increase of the antioxidant enzyme Glutathione Reductase (GR).

Eccentric exercise of the elbow flexors does not induce significant change in plasma cytokine levels, even in the presence of significant muscle damage as demonstrated by the increase in Creatine Kinase (CK) activity and myoglobin [56]. In any case, a transient increase in oxidative stress and cytokines is considered relevant for improvements induced by regular exercise, in a process called mitohormesis, because mitochondria are necessary for energy production and associated with oxidative stress induction [57]. All these data suggest that, since the role of cytokines in response to oxidative stress and exercise has not been definitively clarified, the aerobic, resistance, endurance and strength training sessions should be carefully set up within physical-activity programs for adults and elderly people in order to avoid undesirable effects in inflammation-signalling pathways.

### *4.2. Nervous System*

In the nervous system, ROS are involved in neurogenesis, proliferation and differentiation of neuronal stem cells; however, if the ROS concentrations exceed threshold levels, it can lead to neurodegeneration [58]. Physical activity increases neurogenesis by inducing neurotrophic factors. Moreover, it protects DNA from damage by stimulating antioxidant systems [58]. A recent review by Vilela et al. (2020) [59] summarizes the effects of physical exercise on brain synaptic plasticity and memory through different pathways. In particular, studies performed in rat hippocampi suggest that aerobic training induces N-Methyl-D-Aspartate phosphorylate (pNMDA) and Postsynaptic Density protein 95 (PSD-95) activity, which increases memory. Instead, the strength training ameliorates the results of memory tests by upregulating Protein Kinase C alpha (PKC-alpha) and the cytokines IL-1 and TNF-alpha. Moreover, aerobic exercise ameliorates spatial memory through BDNF signalling [60]. Interestingly, in older adults affected by mild subcortical ischaemic vascular cognitive impairment, 6 months of aerobic exercise on the treadmill improved efficiency of the affected brain areas [60].

Feter et al. (2019) [61] investigated how different training models affect memory and redox balance. Resistance training induced a significant increase in lipid peroxidation and ROS levels, compared to the sedentary group. The moderate-intensity continuous-training group and physical-activity (running wheel) group showed a higher level of the antioxidant enzyme catalase, compared to the sedentary group. Moreover, the moderate-intensity continuous-training group showed better recognition memory, suggesting that performing physical activity, with a moderate but continuous intensity, could represent a nonpharmacological strategy to counteract the symptoms of Alzheimer's disease. Of note, the role of physical activity as a strategy for counteracting Alzheimer's Disease (AD) has been reviewed elsewhere [62]. In particular, although the pathogenesis of this disease has been clarified, the exact mechanisms are still under study. The role of physical activity seems to be correlated with microRNA alterations associated with impaired autophagy. This process is mediated by the Phosphatidyl Inositol 3-Kinase (PI3Kinase)/Akt-mammalian Target Of Rapamycin (mTOR) signalling pathway [62]. In detail, physical activity, through mTOR signalling activation, downregulates the production of abnormal micro-RNAs such as miR-130a, miR106b and miRlet7c, and stimulates neurogenesis, synaptic plasticity and memory, although the exact mechanism is still vague. The mTOR activation leads to dysfunctional autophagy, and consequently, to Tau hyperphosphorylation, β-amyloid (Aβ) accumulation and neurofibrillary tangles that are characteristics of Alzheimer's disease [62]. Discovering how physical activity can exert this effect in the central nervous system and interfere with the abnormal synthesis of miRNAs in AD pathogenesis is an intriguing challenge that requires further investigation.

### *4.3. Liver*

Nuclear factor erythroid-2-related factor 2 (Nrf2) and its antioxidant responsive elements play a prominent role in the reaction to oxidative stress. Acute aerobic exercise activates nuclear Nrf2 independently of exercise intensity, while higher-intensity physical activity increases the levels of oxidative stress and antioxidant enzymes [63]. Interestingly, a bout of moderate-intensity stationary cycling up-regulates Nrf2 in young, but not in older people. Nrf2 is involved in redox-balance modifications induced by exercise, but the underlying mechanisms are still unclear [64]. Of note, both aerobic and resistance training, although with different molecular pathways, converge on activation of the mitochondrial biogenesis through Nrf2. Although Nrf2 is ubiquitously expressed, the effects of physical activity and redox balance on Nrf2 have been well characterized in the liver. In mice, sulphorane exerts a protective role against liver damage induced by acute exercise. In particular, sulphorane reduces inflammatory cytokines like IL-1beta, IL-6 and TNF-alpha increased by exercise, and upregulates the expression of the antioxidant enzymes catalase, SOD1 and glutathione peroxidase (GPx1) through the activation of the Nrf2/Heme Oxygenase (HO1) signal-transduction pathway [65]. Moreover, it has been shown that 6 weeks of aerobic training increase triglycerides, free fatty acids and blood glucose in Nrf2-*null* mice, showing that Nrf2 is involved in training-induced adaptations of glucose homeostasis. Moreover, the levels of oxidative stress markers in the liver and in adipose tissue are increased, with respect to the control group, in Nrf2-*null* mice [66].

### *4.4. Skeletal and Cardiac Muscle*

The importance of Nrf2 has also been evidenced in muscle. In fact, Huang et al. (2019) [67] showed that the level of Nrf2 mRNA in skeletal muscle decreases with aging, and concomitantly, a frailty phenotype arises. Muscle function and mass are similar in young wild-type and young Nrf2-*null* mice. In middle-aged and old Nrf2-*null* mice, muscle function significantly decreases with enhancement of frailty in old Nrf2-*null* mice [67]. This phenotype hinges on a decrease of mitochondrial biogenesis and alteration of mitochondrial morphology in skeletal muscle. In skeletal muscle, other pathways also seem to be involved in redox-balance signalling. In fact, skeletal muscle is a primary source of ROS during exercise, because during muscle contraction, there is a massive production of superoxide radical (O2 −) [67]. In rats, a combination of antioxidants (such as resveratrol) and exercise significantly increases muscle mass and grip strength in sarcopenic old mice, and reduces abnormal sarcomere length [68]. Concomitantly, the authors observed a significant increase in the levels of the phosphorylated form of 5 AMP-activated protein Kinase (pAMPK) and of Sirtuin1 (Sirt1), suggesting the involvement of the AMPK/Sirt1 pathway [68]. This last pathway is also involved in AdipoRon function that is used to reduce oxidative stress and inflammation in dystrophic mice [69]. AdipoRon increases the levels of pAMPK that reduces NF-Kb and increases the dystrophin analogue utrophin [69]. AdipoRon also increases muscle strength and endurance, leading to better physical performance [69]. The antioxidant and antiaging effects of resveratrol have been studied also in the kidney [70]. In particular, the authors suggest that resveratrol can activate both Nrf2 and AMPK/sirtuin1 pathways in the kidney, to counteract the pathologic aging effect in this organ. In fact, transfection with Nrf2 and SIRT1 siRNAs in HK2 cells blocks the antioxidant effect of resveratrol [70].

The Akt signal-transduction pathway has been evidenced in skeletal muscle, in association with exercise and redox balance. For example, in mice, swimming exercise reduces ROS, and prevents high-fat diet-induced insulin resistance through the decrease of NAPDH oxidase 4 (Nox4), as well as the increase of Akt signal transduction in skeletal muscle [71].

Akt signalling is also required for the correct regulation of muscle stem-cell (MuSCs) homeostasis during aging. In fact, Lukjanenko et al. (2019) [72] found WNT1-Inducible Signalling Pathway Protein 1 (WISP1) as a fibro-adipogenic progenitor protein that is strongly downregulated during aging. This protein is important for muscle regeneration and the asymmetric division of muscle stem cells through Akt signalling. In fact, systemic treatment of aged mice with WISP1 restores the proliferation and differentiation of aged muscle stem cells [72].

Peroxisome Proliferator-Activated Receptor gamma (PPAR-gamma) and Peroxisome proliferator-activated receptor Gamma Coactivator-1 alpha (PGC1-alpha) have been proposed as important players in muscle remodelling induced by exercise through Mitogen-

Activated Protein Kinases (MAPKinases), and are also correlated with redox balance. In particular, p38 MAPKinase activates PPAR-gamma. Sirt1, in turn, depending on physical exercise, activates PGC1-alpha [73]. PGC1-alpha is implicated in aging-associated mitochondrial diseases dysregulating ROS production and mitochondrial network structure in skeletal muscle, during aging and exercise training. PGC1-alpha is also activated by ROS through NF-kB, AMPK, and finally, by nitric oxide (NO) produced during muscle contraction [73,74]. PGC1-alpha-*null* aged mice show lower running endurance, higher mitochondrial damage, increased ROS and higher oxidative stress than do young mice. Exercise training increases maximal respiratory capacity, both in PGC1-alpha-*null* and wild-type mice. Instead, the rescue of mitochondrial homeostasis occurs in a PGC1-alphadependent manner [74].

Regarding cardiac muscle, aerobic exercise training on the treadmill ameliorates myocardial infarction symptoms. In fact, TNF-alpha, NADPH-oxidase activity and p38 phosphorylation are diminished in infarcted trained mice, with respect to infarcted sham-trained mice [75]. The positive effect of exercise is produced also by increasing the antioxidant defence through neuronal Nitric-Oxide Synthase (nNOS) [76]. In fact, cardiac-specific overexpression of this enzyme mimics the effect of exercise on maximal oxygen capacity, while aerobic activity increases cardiac dysfunctions in nNOS-*null* mice, confirming the protective effect of this enzyme.

Interestingly, Hao et al. (2014) [77] showed that exercise is a powerful instrument for protecting the heart from ischemia, and that the molecular mechanism at the base of this phenomenon involves Protein Kinase Cε. Specifically, Sprague-Dawley rats underwent exhaustive aerobic exercise on the treadmill, to induce myocardial injury [77]. In such exhaustive conditions, PKCε translocated to the intercalated disks of the heart. The exercise pre-conditioning increased the level of the phosphorylated/activated isoform of PKCε in the cytoplasmic membrane, suggesting that PKCε is involved in the signal transduction for heart protection induced by pre-conditioning [77].

Finally, Díaz-Ruíz et al. (2019) [78] evidenced a role for redox signalling in ischemic post-conditioning protection through PKCε and Erk1/2, which indirectly regulate Nrf2 activation. In fact, the authors showed that PKCε and Erk1/2 are activated in a redoxdependent manner. It was also shown that neither the PI3K inhibitor nor the Erk1/2 inhibitor reduces Nrf2 activation, suggesting that these kinases have other direct targets [78]. Notably, Buelna-Chontal et al. (2014) [79] found that Nrf2 activation is dependent on PKC in post-conditioned hearts. In fact, the use of PKC inhibitors reduces the level of Nrf2 phosphorylation and the activity of antioxidant proteins that are regulated by Nrf2 [79].

The PKC family includes 12 isozymes divided into conventional (alpha, beta and gamma), novel (delta, epsilon, eta and theta) and atypical (zeta and iota). PKCε belongs to the novel group which is activated by the conventional signalling of Ca2+ and diacylglycerol. In particular, the receptor, once activated, hydrolyses Phosphatidyl-Inositol-4,5-bisPhosphate (PIP2), generating Inositol trisPhosphate (IP3, responsible for Ca2+ mobilization) and diacylglycerol which, as the second messenger, triggers PKC-kinase function, leading to phosphorylation and activation of other trans-membrane or intra-cellular proteins. PKCs are generally considered as oncoproteins. However, many different roles have been proposed for PKC proteins in cell physiology. For example, PKCε is involved in many processes that maintain cell homeostasis and proliferation in several tissues such as the vessels, blood, heart and skeletal muscle [80–85]. For example, PKCε is differently expressed in blood cells suggesting multiple roles for this PKC isoform [84]. Moreover, Martini et al. [85] evidenced that PKCε controls the migration of centrosome during mitotic spindle assembly. Recently, D'Amico and Lennartz (2018) [86] showed a role in vesicle formation for PKCε. They demonstrated that PKCε is linked to the Golgi apparatus through the interaction between the Golgi's lipids and the regulatory domain of the kinase. Moreover, the mechanism seems to be independent of the kinase activity of PKCε, although it requires the binding of PKCε to actin and cytoskeleton for budding and fission of the vesicles [86]. Finally, PKCε is important for autophagy in the breast-cancer cell line [87]. In

fact, PKCε siRNA induces a significant decrease in autophagy by reducing the levels of Raptor and Rictor that are required to form mammalian Target Of Rapamycin Complex-1 (mTORC-1) and -2 (mTORC-2), which are important regulators of autophagy [87]. On the other hand, autophagy is important to counteract aging, although this process declines in the elderly. Exercise and diet downregulate mTORC-1, that acts as a negative player in autophagy and stimulates autophagy in several other tissues [88]. Moreover, mTORC-1 induces muscle atrophy in the elderly, while mTOR is essential for muscle hypertrophy. Most likely, a mechanism that involves Akt and mTORC1 is important in sarcopenia, although the definition of a clear signalling pathway is still under study [89]. Based on these considerations, we could speculate that PKCε is involved in exercise protection during the aging process.

Additionally, ROS can affect Ca2+ signalling at the level of channels, pumps and exchangers [90]. Alterations in ROS signalling can modify Ca2+ communication system, likely contributing to disease onset. In the heart, for example, Zima and Blatter [91] showed that the ROS produced during reperfusion of cardiac ischemic injury can affect ischemiarelated Ca(2+) overload. More recently, Cabassi and Miragoli [92] reviewed the important role of local environment (namely ROS and Ca2+) in mitochondrial re-organization and fusion in failing cardiomyocytes. In particular, a correct interplay between Ca2+ and ROS is important to avoid intracellular Ca2+ increase during diastole that leads to cardiac arrhythmia [92]. The relevance of interaction between Ca2+ signalling and ROS has also been described in skeletal muscle aging [40]. In fact, the increase of ROS in skeletal muscle can alter the Ca2+ signalling pathway that is necessary for the fibre contraction, leading to a decrease in muscle strength. Interestingly, physical activity can help to contrast strength loss by reducing ROS [40].

Finally, Peroxiredoxins (PRDXs) are important antioxidant enzymes that remove H2O2 to reduce oxidative stress. PRDXs are responsible for the discarding of 90% of cellular peroxides, they are ubiquitous and function as regulators of local H2O2 concentrations [93,94]. There are more than 3500 sequences of Peroxiredoxins, and these have been divided into six subfamilies with different expression levels in mammals, plants, yeasts and bacteria [95]. Considering the conservation of phylogenetic rhythms, the fact that PRDX proteins autoregulate independently from the transcriptional circadian clock and the highly conserved structures of these proteins, many researchers are now studying the links between aging, circadian rhythms and redox systems [96].

PRDX-3-*null* mice present a significant reduction in physical strength (measured by swimming performance), compared to their wild-type littermates at the age of ten months [97]. Moreover, PRDX-3-*null* mice show a higher number of apoptotic cells in the brain, higher expression of Nrf2 and a lower level of mitochondrial DNA (mDNA) at the age of ten months, when compared to wild-type, suggesting that the lack of PRDX-3 increases oxidative stress and mitochondrial impairment, which are characteristic of the aging process [97]. To study redox-enzyme concentrations after different types of exercise, a group of active men (mean age 28 years old) performed moderate-intensity exercise with one bout of High-Intensity Interval Exercise (HIIE) or an eccentric resistance exercise [98]. PRDX-4 (Peroxiredoxin-4) and SOD3 increased after High-Intensity Interval Exercise (HIIE), while Thioredoxin Reductase (TRX-R) decreased. Notably, resistance exercise did not determine any significant changes in redox enzymes, but induced skeletal muscle damages. Altogether, these results suggest that PRDX-4 and SOD3 could be considered as biomarkers of oxidative stress [98]. Interestingly, HIIE showed similar effects of endurance exercise in muscle mitochondria, independently from the total workload, suggesting that exercise prescription could be accommodated into individual prescriptions, generating comparable molecular effects [99].

Although the exact mechanism is not yet clear, it has been proposed that PRDX translates oxidative stress signalling by the activation of MAPKinases [100].

Figure 2 provides a schematic representation of the pathways involved in the effects of physical activity and redox balance to counteract aging. In plasma or serum, the proposed markers of oxidative stress are the increase of PRDX-4 and SOD3, while the level of EGF should decrease, given the above studies. Cytokines are inflammation markers, and contrasting results have been obtained in association with physical activity. In neural cells, PI3K/Akt/mTor seems to have a prominent role, especially in the treatment of Alzheimer's Disease. The Nrf2 pathway is central in the liver, but it also is an important effector in heart and skeletal muscle where many players have been evidenced, such as MAPKinases (p38 and Erk1/2), Akt and PPARγ/PGC1α. Finally, in adipose tissue, there are emerging results on the role of physical activity and redox balance to increase the levels of BDNF and ANF (lipolytic agents) in the elaboration of protocols to contrast obesity.

**Figure 2.** Schematic representation of the molecular pathways involved in redox regulation by physical activity.

In plasma and serum, the effect of physical activity on redox balance is exerted through the increase of SOD3 and PRDX-4 and the decrease of EGF; in neural cells, a decrease of PI3Kinase/Akt-mTOR is observed; in liver cells, Nrf2 pathways are activated; in the heart, the effect of physical activity is associated with a decrease of TNF-alpha and NADPH oxidase, and p38 phosphorylation; in skeletal muscle, Akt, MAPKinases, PPARgamma, PGC-1 alpha and Nrf2 are up-regulated; in adipose tissue, lipolysis is activated by physical activity.

### **5. Conclusions and Perspectives**

Redox balance is very important for cell life. As with other cell processes, oxidant/antioxidant balance is affected by aging, with a decrease in the efficiency of antioxidant systems. To counteract this effect of aging, antioxidant foods have been shown to be useful. Physical activity, as well, has been involved in the regulation of redox balance in several tissues such as skeletal muscle, cardiac muscle, liver, as well as neural cells. Although there are several pieces of evidence of the involvement of different signalling pathways, a comprehensive view of the molecular players, modulated by physical activity, that are able to counteract the effects of ageing on redox balance, is still lacking.

The trivial perception that reactive species of oxygen are exclusively detrimental molecules, is definitively substituted with mounting evidence that exercise-induced perturbation of redox balance is the upstream signal for the activation of transcription factors and the induction of gene expression associated with exercise effects involving total body adaptation to training. However, the same signalling should be analysed on the specific tissue and organ, as well as in terms of age, fragility state and comorbidities.

At the moment, although we believe that future precision medicine will be able to transform exercise administration from generic wellness to targeted prevention, we must admit that the topic is still in its infancy.

**Author Contributions:** D.G., C.C., E.M. contributed to the bibliographic research, writing of the manuscript and produced the figures. M.V. (Mauro Vaccarezza), V.P., G.P., L.A. reviewed the manuscript. M.V. (Marco Vitale) reviewed the manuscript and supervised the work. G.G. and P.M. conceived the original idea and supervised the work. All authors have read and agreed to the published version of the manuscript.

**Funding:** D.G.: C.C.: M.V. (Marco Vitale), G.G. were supported by Fondi Locali per la Ricerca 2019—Quota Prodotti di Ricerca—Parma University. P.M. was supported by "Programmi di ricerca di Rilevante Interesse Nazionale"—Italian Ministry of Education, University and Research (MIUR-PRIN) 2017 grant entitled "ACTLIFE: IS ACTIVE LIFE STYLE ENOUGH FOR HEALTH AND WELLBEING?" codice 2017RS5M44\_004.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** We are grateful to Cristina Micheloni, Luciana Cerasuolo and Vincenzo Alberto Piero Palermo for technical support. We thank Devahuti Chaliha for English proofreading.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **Abbreviations**


### **References**


### *Review* **Pulmonary Effects Due to Physical Exercise in Polluted Air: Evidence from Studies Conducted on Healthy Humans**

**Oscar F. Araneda 1,\*, Franz Kosche-Cárcamo 2, Humberto Verdugo-Marchese <sup>2</sup> and Marcelo Tuesta 2,3**


**Abstract:** Physical inactivity has caused serious effects on the health of the population, having an impact on the quality of life and the cost of healthcare for many countries. This has motivated government and private institutions to promote regular physical activity, which, paradoxically, can involve health risks when it is carried out in areas with poor air quality. This review collects information from studies conducted on healthy humans related to the pulmonary effects caused by the practice of physical activity when there is poor air quality. In addition, several challenges related to the technological and educational areas, as well as to applied and basic research, have been identified to facilitate the rational practice of exercise in poor air quality conditions.

**Keywords:** air pollution; physical exercise; lungs

### **1. Introduction**

The increase in the world population and the use of polluting energy sources in the modern world have led to changes in the quality of the air we breathe. Pollutants floating in the air cause harmful effects on health, which have been extensively studied [1–3]. As a result of these studies, this type of pollution has been linked to cancer, cardiovascular disorders, acute pulmonary disorders (infections, acute bronchial obstruction), and chronic disorders (asthma, chronic obstructive pulmonary disease) [4]. According to data provided by the World Health Organization in 2018, about 90% of people are exposed to polluted air, particularly in poor or developing countries on all continents, which produce a total of around seven million deaths annually at a global level [5]. Although we breathe a mixture of substances in the air, on an individual level, it has been possible to individually identify some pollutants that cause health disorders, such as tropospheric ozone (O3), particulate matter (PMx), carbon monoxide (CO), nitrogen oxides (NOx), and sulfur oxides (SOx) [6,7]. These are usually concentrated in urban centers, and are mainly, though not exclusively, caused by activities typical of human beings, such as transport, heating, cooking, and agricultural and industrial activities [8].

Our bodies come into contact with atmospheric pollutants through the wide area of exposure of the lung tissue, making this organ particularly susceptible to damage from the components of the air we breathe, such as particulate matter, cigarette smoke, various gases, and pollen [9,10]. Consequently, this interaction will cause alterations in the functionality of this organ. The amount of polluting substances that reach the lungs during exercise will depend on both their concentration in the air and the magnitude of the pulmonary physiological phenomena typical of physical effort: bronchodilation, increased ventilation, mouth breathing, and increased diffusing capacity. This implies that, during exercise, a

**Citation:** Araneda, O.F.; Kosche-Cárcamo, F.; Verdugo-Marchese, H.; Tuesta, M. Pulmonary Effects Due to Physical Exercise in Polluted Air: Evidence from Studies Conducted on Healthy Humans. *Appl. Sci.* **2021**, *11*, 2890. https://doi.org/10.3390/app11072890

Academic Editor: Daniela Galli

Received: 20 February 2021 Accepted: 18 March 2021 Published: 24 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

higher load of tissue pollutants will be observed [11], which impacts deeper lung areas and may even be associated with a greater passage of these substances into the bloodstream, as in the case of gases and ultrafine particulate matter (Figure 1).

**Figure 1.** Systemic and localized effects at the lung level generated by air pollutants. SOx = sulphur oxides; NOx = nitrogen oxides; PMx = particulate matter; UFP = ultrafine particles; SOD = superoxide dismutase; CAT = catalase; GPx = glutathione peroxidase; GR = glutathione reductase; GSH = reduced glutathione RONS = reactive oxygen and nitrogen free radicals; NOX-2 = NADPH oxidase type 2; XO = xanthine oxidase; XR = xanthine reductase; NOS = nitric oxide synthase; PMNs = polymorphonuclear leukocytes. VCAM-1 = vascular cell adhesion molecule 1; ICAM-1 = intercellular adhesion molecule 1.

> In consideration of the processes described, for some years now, the question has been what the pulmonary effects due to physical exercise in areas with poor air quality are, in the short and long term, both in healthy people and in those suffering from pathologies [12–15]. The issue is becoming relevant nowadays, since, in many countries with poor air quality, increased participation in outdoor sports activities, such as massive urban races and the use of bicycles, have been observed. In addition, there is little information about the effects and potential risks for the exposed population, as well as future measures to help to reduce exposure for both people who exercise and healthcare professionals. To address this matter, the interaction mechanisms and effects resulting from the contact between the air and lung tissue are presented in this study. In addition, some challenges in the field of research are described, and recommendations for the safe practice of physical exercise under these conditions are made. The papers used for this narrative review were original or review manuscripts, in which an exercise protocol under laboratory or field conditions was carried out, including the study of pulmonary function parameters, in the presence of air pollutants in healthy humans. The papers were mainly taken from the PubMed search engine, and there was no limitation regarding the year of publication.

### **2. Alteration of Ventilatory Function and Appearance/Exacerbation of Respiratory Symptoms during Physical Exercise with Polluted Air: Short- and Medium-Term Effects**

By doing physical exercise in an environment with polluted air, increased airway resistance is induced. Initially, this was described in protocols that set O3 at different concentrations [16–19]. Thus, in these studies, it is a common result to find lower values than those expected in forced expiratory volume in the first second (FEV1), forced vital capacity (FVC), and forced mid-expiratory flow rate (FEF25–75) [19,20]. Regarding the pollutant particles in suspension, particulate material less than 10 microns (PM10) promotes the establishment of chronic lung function problems, mainly associated with decreased FEV1 [21]. Likewise, PM2.5 (<2.5 microns) has been related to a drop in FEV1 and FVC in subjects who lived in the vicinity of a highway [22]. Physical exercise increases these effects as their intensity and duration increase as a consequence of the increased work of breathing. Minute ventilation, together with greater exposure to the pollutant, will determine their inhaled load in the airways [16,23,24]. Several authors have observed that, during moderate and maximum exercise in an environment contaminated with O3, there are more symptoms, such as dyspnea, respiratory distress, and a feeling of tightness in the chest, which quickly lead to the cessation of exercise [25–28]. Several authors have shown an alteration in lung function parameters (FEV1, FVC, and FEF25–75) in both healthy trained and untrained subjects exposed to high concentrations of O3 while performing physical exercise [15], with some of them even experiencing the effects a few hours after exercising [29]. Abnormal respiratory symptoms have also been observed after one hour of exercise and one hour of rest in an environment contaminated with PM10 and PM2.5, however, no spirometric changes were found [21]. In another study, Brant et al. [30] observed decreased mucociliary clearance, decreased pH in the expired air, and increased symptoms of respiratory distress in motorcyclists who were chronically exposed to NO2 during heavy traffic for 5–8 h a day, five days a week. Likewise, the practice of aerobic exercise in the open air at an average PM2.5 concentration of 65.1 μg/m<sup>3</sup> over a period of five days also altered mucociliary clearance [31]. Yet, the study conducted by Kubesch et al. (2015) demonstrated improvements in lung function with physical exercise in a highly polluted environment; however, an acute pulmonary inflammatory effect was also observed [32]. Other studies on healthy subjects who performed strenuous longduration exercise (between ~2–5 h) in a polluted environment (PM2.5 and O3) did not show deleterious effects on lung function parameters [33,34]. This should be interpreted with caution, since the systematization of this behavior, which constantly promotes an oxidative/inflammatory lung environment, could mean permanent lung damage in the future. Children are especially sensitive to contact with atmospheric pollutants, which leads to a higher rate of school absenteeism and an increase in high and low airway infections [35], and alters lung function, thereby increasing the probability of death [36]. Infant lungs show an incomplete respiratory tree where the pulmonary epithelium is being formed, which has been associated with greater permeability of this structure. Similarly, some studies have suggested that long-term exposure to air pollutants could potentially affect children's lung development [37–39]. Modification of lung function in children who are active in polluted air has been associated with high concentrations of O3 [40], MP10, NOx, and CO [41]. McConell et al. [42] demonstrated that there was an association between the prevalence of asthma in children and prolonged and intense exercise done outdoors with high levels of O3. In the case of adults, moderate exercise increases the amount of particulate material entering the lungs by 4.5 times versus resting [12], where similar results are expected for the other environmental pollutants. One group of subjects who can be particularly affected by polluted air is the group of amateur long-distance athletes (e.g., marathons, cross-country cycling). Thus, a group of cyclists exercised in open-top field chambers for dispensing O3, showing a decrease in FEV1 after moderate exercise for sixty minutes [24]. Furthermore, a group of cyclists in competition (75 min) increased their respiratory symptoms (wheezing, dyspnea) after inhaling a mixture of pollutants. Respiratory distress and spirometric alterations were correlated with the concentration of O3 [43]. Korrick et al. [44] found

a decrease in FEV1 and FVC in subjects who took prolonged walks in an environment with low levels of O3 and particulate matter, demonstrating that the changes in lung function were caused by exercise, even with low levels of pollutants. Gong et al. [45] conducted an intermittent exercise protocol (two hours) in healthy subjects and subjects with asthma using an environmental test chamber with ultrafine particulate matter seven to eight times higher than average air levels in unpolluted areas. Thus, a decrease in arterial oxygen saturation and FEV1 was observed the day after the test, with no differences between healthy subjects and subjects with asthma. A total of 23 studies evaluated the effect of O3 on the decline of FEV1, which were carried out under standardized conditions in a test chamber. Here, relationship models were observed in various conditions of O3 concentration, exposure time, and minute ventilation; in addition, the thresholds for the appearance of the broncho-constrictor phenomenon were attained [46]. Kim et al. [47] observed a decline in FEV1 in an exercise protocol at 0.06 ppm O3 and stable ventilation of 35 L/min in six cycles of 50 min. The obstructive modification caused by O3 is the result of a bronchoconstrictive response induced by a vagal reflex activated by irritation [48]. The release of acetylcholine (parasympathetic response) will increase the activation of the submucosal glands, leading to greater release of mucus [49]. The results found in relation to NO2 on lung function have been variable; Bauer et al. [50] and Strand et al. [51] demonstrated a broncho-constrictor effect in asthmatics. Likewise, Kulstrunk and Bohini (1992) observed decreased FEV1 after a maximal exercise test (~8 min) with high NO2 levels was performed by healthy subjects [52]. However, Jorres et al. [53] did not observe any changes in FEV1 after exercising for 1.5 h with 3 h of exposure to 1 ppm of NO2. In relation to CO, due to its high diffusion capacity in the alveolar capillary membrane and its great affinity with carboxyhemoglobin (COHb), systemic effects have mainly been recognized in healthy subjects who exercise. Due to COHb, there will be less availability of hemoglobin to bind oxygen (hypoxemia), thus altering the transport capacity of this gas to active muscles [54] and limiting energy production and physical effort. Decreased physical performance in healthy subjects due to CO exposure alone was observed with maximum intensity exercise [55], but not when it was performed at submaximal intensity and a lower concentration of this gas [56].

All previously mentioned studies describe the short- and medium-term effects of exposure to air pollutants, with little evidence to clarify the long-term effects. However, subjects exposed for years to outdoor jobs on the streets, but who did not exert great physical effort, showed a decreased ability to perform physical work, as well as a decreased maximum voluntary ventilation (MVV) [57].

### **3. Lung Inflammation and Oxidative Stress Due to Exercise in Polluted Air**

In the lung region, there is a complex microenvironment where the interaction of the cells of this organ (pneumocytes, endothelial cells, alveolar macrophages) and inflammatory cells (various types of leukocytes) from the bloodstream takes place. Due to the gas exchange function between the lungs and the environment, this tissue organization is constantly exposed to the action of pollutants present in the air. The interaction between lung tissue and air pollutants largely depends on the physical and chemical characteristics of the latter. In this way, some air pollutants will cause direct damage at the cellular level (O3, SOx, NOx), others will be deposited in the airway (particulate matter), and some will diffuse through the alveolus capillary membrane into the bloodstream (ultrafine particulate matter and gases), as shown in Figure 1.

Lung inflammation and oxidative stress play an essential role in altering lung function with exposure to O3, especially when exercising [58]. In this sense, the low solubility of O3 in water suggests that the main interaction of this molecule will be with the surface of the epithelium, and, in particular, with the epithelial lining fluid. This contact will promote reactions with thiol groups (-SH), antioxidants (glutathione, ascorbic acid), and macromolecules (lipids, proteins, carbohydrates), among others found in the epithelial fluid, promoting an oxidative/inflammatory environment. A higher number of molecules

that are attracted due to the increased ventilatory flow induced by exercise will enhance these effects, especially for long-duration exercises (e.g., marathons). A study conducted by Kinney et al. [59] observed increases in lung inflammation markers with low levels of O3 in the air during moderate exercise in a group of amateur runners. The predominance of this phenomenon occurred during the spring and summer seasons in comparison to the winter season. Both seasons are characterized by a higher production/concentration of ozone in the environment due to a higher level of ultraviolet radiation. The effect of the time of exposure to O3 was also addressed by Aris et al. [60], who observed an increase in polymorphonuclear leukocytes (PMNs or neutrophils) 18 h after exposing healthy subjects and athletes to 0.20 ppm of O3 during 4 h of moderate exercise. A study conducted by Gomes et al. (2011) showed a lung inflammatory/oxidative effect in subjects who ran 8 km in a hot environment (31 ◦C) with increased levels of O3, but this was not the case in a cold environment (20 ◦C); nevertheless, no changes in lung function were observed in any environment [61]. This suggests that O3 has a high capacity to promote inflammation and damage in the respiratory epithelium, affecting lung function, even more so when intense and long-term exercise is done. Other environmental factors, such as ambient temperature, must also be considered.

In the case of particulate matter, exposure to larger particles, such as PM10, has also shown lung inflammatory effects [62,63]. However, the main target structures of the smallest particles (PM2.5 and UFPs) will be the pulmonary vascular and cardiovascular components [64,65]. Some inflammatory responses against contamination are the infiltration of PMNs and macrophages that release proteases with degradative and pro-oxidant activity, favoring cell damage [9,11,66]. Depending on the degree of exposure, this phenomenon can occur on multiple occasions, and can even become a chronic process [39], similar to that seen in subjects who smoke [67]. In this way, it is necessary to detail these processes when exercising during exposure to a polluted environment. Ghio et al. [68] observed an increase in PMNs in the bronchial and alveolar fractions of bronchoalveolar lavage (BAL) after 1 h of intermittent light exercise in an environment with PM10 (23.1–311 μg/m3) and PM2.5 (47.2–206.7 μg/m3). Larsson et al. [69], 14 h later, observed increases in alveolar macrophages and lymphocytes in BAL after one hour of light exercise in an environment with PM10 and PM2.5. In asthmatics, Pietropaoli et al. [70] found increased macrophages in induced sputum when they performed two hours of intermittent exercise in an environment with ultrafine particulate matter (10 μg/m3), unlike healthy subjects (10, 25, and 50 μg/m3). A recent study highlighted the effect of outdoor endurance training as an immune protection factor against exposure to particulate matter by controlling lung inflammation in this group, but not in exposed sedentary subjects [71]. This reinforces the importance of controlling air quality, the type of exercise, and the level of physical condition in order to take advantage of its benefits and thus avoid the deleterious effects of pollution. Cavalcante de Sá et al. [31] observed higher pH values in the exhaled condensed air of athletes who ran (45 min/day for five days) in an unpolluted environment versus athletes who ran in a polluted environment with high PM2.5. Some studies have focused on the pro-inflammatory effects of NO2 [56,72,73]. Jorres et al. [53] found increases in pro-inflammatory mediators in BAL in asthmatics, but not in healthy subjects, after 1.5 h of exercising and 3 h of exposure to NO2 (1 ppm). Devlin et al. [74] showed that light/moderate exercise for two hours at a concentration of 2 ppm of NO2 increased the levels of PMNs, IL–6, and IL–8 in BAL after 24 h. The action of pro-oxidants released directly by PMNs or produced secondarily from reactions involving metals contained in particulate matter are a challenge for enzymatic (catalase, superoxide dismutase, and glutathione peroxidase) and non-enzymatic (glutathione, vitamins C, vitamin E, and uric acid) lung antioxidant defenses. If they are overcome by pro-oxidants, an imbalance called oxidative stress is established, causing damage to structural molecules, in which lipoperoxidation is the most studied phenomenon. The association between oxidative damage and the modification of lung function has previously been described and confirmed by observing the negative correlations between plasma lipoperoxidation and glutathione, each with

FEV1, as well as the positive correlations between the activity of the enzyme glutathione peroxidase and FEV1 [75]. Therefore, it is possible to think of a potential benefit on lung function in subjects exposed to polluted air [76–78] with the consumption of antioxidants in the diet (β-carotene, vitamins C and E), for example, through the consumption of fresh fruits and vegetables [79].

### **4. Recommendations to Consider When Prescribing Exercise in Areas with Low Air Quality**

Our biological design requires us to perform physical activity to stay healthy. Thus, the study of physical exercise has been the concern of researchers from the most diverse disciplines of knowledge. In this regard, one of the most relevant results, which is derived from research in this area, indicates that low levels of physical activity are associated with obesity, metabolic diseases, cancer, cardiovascular diseases, stress, anxiety, and depression [80–83]. In addition, low levels of physical activity have an impact on the quality of life of the population [84,85], and may be the cause of economic losses [86]. Hence, countries should promote exercise, but they should also be responsible for ensuring that it is done in a safe environment. Exercising with polluted air will be a problem for the foreseeable future, particularly since the concentration of pollutants in the air in many cities in developing countries will continue to be high. Hence, it is important for health personnel and the general population to be aware of the effects of pollution on the lungs, and to know both the guidelines regarding potential preventive measures and the status of issues that are not yet resolved in the research arena. In relation to prevention matters, one of the first pillars of the exercise/polluted air interaction consists of efforts to reduce air pollution. Firstly, we must think that the result of a high concentration of pollutants is a multifactorial outcome, thus, it depends on geographical factors (location and design of urban centers), environmental factors (temperature, humidity, luminosity), and derivatives of human activity (population growth, energy matrix used in industry/means of transport). Regarding these determinants, there are few factors that can be changed in the short term; however, both state and private organizations can devise plans for the re-evaluation of air quality and emission standards. At the same time, programs of general environmental education focused on air quality can be created. One measure with a direct impact on those who exercise under these conditions is to reduce the load of pollutants they are exposed to. To address this objective, it is necessary to favor access to air quality information by establishing monitoring points in cities where there are none, and by increasing their number in cities where they already exist. In addition, it will be advantageous to increase the number of mobile stations at training points and during sporting events, such as massive urban races. Along this line, there is a challenge for technological innovation in order to reduce the size and costs of the monitoring units, or, ideally, for them to be individual devices. The information obtained from the monitoring should also be integrated into applications for smartphones that can suggest routes with better air quality at the time of exercising. In the same direction, the development of new technologies focused on physical measures to reduce the entry of pollutants into the airway is pending [87]. In this regard, the greater social tolerance to the use of masks in times of the COVID-19 pandemic must be an advantage [88]. Finally, the identification of pollutants that are typical of indoor sporting environments should be improved by promoting measures to transform them into healthy environments and by using technology for the development and installation of filters in these places [89].

Regarding the biological phenomena of this problem, from our review, we conclude that it is necessary to increase the number of studies conducted on humans in relation to the acute effect of exposure to polluted air, and particularly in relation to chronic changes, where the information is practically nil. Likewise, the most solid information on biological effects refers to the action of individual pollutants, and progress must be made in the study of the interaction between pollutants and the influence of environmental conditions, such as humidity and temperature. The above objective is essential to optimize current models of inhalation load exposure [90,91] and the recommendations arising from it in the future. Another pending task in research is to improve and establish biological monitoring, which should be done by optimizing non-invasive methods (spirometric tests, analysis of expired air, and sputum analysis) and in ideal conditions carried out by the users themselves. One of the objectives of this biological monitoring is the monitoring of at-risk populations (patients with chronic lung diseases, professional athletes, urban long-distance runners or cyclists) and to carry out preventive interventions in these populations. Once these groups have been identified, targeted environmental education programs should be carried out. Another relevant aspect both for this group and for the general population is to optimize exercise planning and to be flexible regarding the time and place it will be done and its intensity, balancing this programming with the health objectives of the general population and competitive athletes. Finally, from the studies regarding the damage mechanisms, the dietary and pharmacological recommendations are alternative solutions, among which the maintenance of a healthy diet with an important content of antioxidants (fruits and vegetables) is noted, as well as the search for new strategies to complement these substances, since evidence has not shown any effectiveness in lung diseases [92]. From the perspective of pharmacological agents, the production and administration of anti-inflammatory agents focused on airway epithelial damage should be studied (a summary of the topics discussed in this chapter is presented in Figure 2).

**Figure 2.** Overview of future goals and strategies to optimize the performance of exercise in polluted conditions.

**Author Contributions:** O.F.A., H.V.-M., F.K.-C., and M.T. worked on the design, methodology, and drafting of the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


## *Article* **Augmentation Index Is Inversely Associated with Skeletal Muscle Mass, Muscle Strength, and Anaerobic Power in Young Male Adults: A Preliminary Study**

**Dongmin Lee 1, Kyengho Byun 2,3,4, Moon-Hyon Hwang 3,4,5 and Sewon Lee 2,3,4,\***


**Abstract:** Arterial stiffness is associated with an increased risk of cardiovascular disease. Previous studies have shown that there is a negative correlation between arterial stiffness and variables such as skeletal muscle mass, muscular strength, and anaerobic power in older individuals. However, little research has been undertaken on relationships in healthy young adults. This study presents a preliminary research that investigates the association between arterial stiffness and muscular factors in healthy male college students. Twenty-three healthy young males (23.9 ± 0.5 years) participated in the study. The participants visited the laboratory, and variables including body composition, blood pressure, arterial stiffness, blood parameters, grip strength, and anaerobic power were measured. Measurements of augmentation index (AIx) and brachial-ankle pulse wave velocity (baPWV) were performed to determine arterial stiffness. There were significant positive correlations among skeletal muscle mass, muscle strength, and anaerobic power in healthy young adult males. AIx was negatively associated with a skeletal muscle mass (r = −0.785, *p* < 0.01), muscular strength (r = −0.500, *p* < 0.05), and anaerobic power (r = −0.469, *p* < 0.05), respectively. Likewise, AIx@75 corrected with a heart rate of 75 was negatively associated with skeletal muscle mass (r = −0.738, *p* < 0.01), muscular strength (r = −0.461, *p* < 0.05), and anaerobic power (r = −0.420, *p* < 0.05) respectively. However, the baPWV showed no correlation with all muscular factors. Our findings suggest that maintaining high levels of skeletal muscle mass, muscular strength, and anaerobic power from relatively young age may lower AIx.

**Keywords:** skeletal muscle mass; muscular strength; anaerobic power; arterial stiffness

### **1. Introduction**

Arterial stiffness, one of the factors indicating the function of blood vessels, is a main independent risk factor that can predict cardiovascular diseases [1]. Especially, this indicator is a leading cause of myocardial infarction in male adults [2]. Stiffening of the arterial wall causes an increase in the blood pressure by reducing the storage and buffering function of the arteries as well as by increasing the rate of pulse transmission regardless of the causes [3]. Usually, arterial stiffness worsens with aging [4]. Aging leads to thickening of the lining of blood vessels by increasing the amount of collagen in the smooth muscle layer and destroying the elastin structure [5]. These alterations result in increased arterial stiffness, leading to cardiometabolic diseases such as hypertension, diabetes, and

**Citation:** Lee, D.; Byun, K.; Hwang, M.-H.; Lee, S. Augmentation Index Is Inversely Associated with Skeletal Muscle Mass, Muscle Strength, and Anaerobic Power in Young Male Adults: A Preliminary Study. *Appl. Sci.* **2021**, *11*, 3146. https://doi.org/ 10.3390/app11073146

Academic Editor: Daniela Galli

Received: 5 March 2021 Accepted: 28 March 2021 Published: 1 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

dyslipidemia [6]. Body mass index (BMI), systolic blood pressure (SBP), low-density lipoprotein cholesterol (LDL-C), and vascular dysfunction are independent risk factors for cardiovascular disease [7–9]. According to a previous study, regarding the relationships among these factors, young adults with a greater number of risk factors for cardiovascular diseases showed significantly lower dilatory capacity of the brachial artery [7]. This finding suggests that vascular function is related to the risk factors for cardiovascular disease even in the younger generation, and managing the blood vessel function from a relatively younger age may be effective in preventing future cardiovascular diseases.

Pulse wave velocity (PWV) and augmentation index (AIx) are widely used as valid methods for non-invasive measurement of arterial stiffness to evaluate the function of blood vessels in human [10–13]. AIx is estimated as the ratio of waveform reflection amplitude to central pulse pressure and has a high correlation with carotid-femoral pulse wave velocity (cf-PWV), a golden standard that can directly measure central arterial stiffness [14]. The measurement of PWV is accepted as the most reliable technology through a recent consensus statement and is the most widely used technique to assess arterial stiffness [15,16].

Recent studies have reported that physical fitness-related factors such as cardiorespiratory endurance and muscular strength are correlated with arterial stiffness [17,18]. Improving cardiorespiratory endurance with regular aerobic exercise lowers arterial stiffness and reduces the risk factors for cardiovascular diseases, consequently reducing mortality due to cardiovascular disease [19]. In addition, increasing skeletal muscle mass and muscular strength with resistance exercise has been known to prevent sarcopenia caused by aging process, and the American Heart Association recommends resistance exercise for prevention and treatment of cardiovascular diseases [17,20]. A study by Fahs et al. on 79 young male adults reported that upper body muscular strength was negatively correlated with aortic stiffness independent of cardiorespiratory capacity [21]. In addition, Ochi et al. study showed that the greater cross-sectional area of the thigh muscles in middle-aged male adults was associated with lower baPWV, suggesting that higher muscle strength and skeletal muscles may be correlated with arterial compliance [22]. It has been reported that an elevation in anaerobic power with anaerobic exercise has a positive effect on cardiovascular health as well as on the aerobic ability [23]. Moreover, improvement in anaerobic capacity with high-intensity interval training has a positive effect on releasing vasodilatory substances such as nitric oxide, suggesting improvement in anaerobic power has a positive effect on vascular function [24,25].

Taken together, anaerobic capacity such as skeletal muscle mass, muscle strength, and anaerobic power may be related to arterial stiffness. However, studies confirming this correlation among young adults are insufficient; especially, it is unclear whether there is a direct relationship between anaerobic power and arterial stiffness. Therefore, the present study aimed to investigate the correlation of skeletal muscle mass, muscular strength, and anaerobic power with arterial stiffness in physically-active healthy college young male adults.

### **2. Materials and Methods**

### *2.1. Participants*

Twenty-three healthy young male college students participated in the study. The exclusion criteria were specified as follows: (1) smokers; (2) high blood pressure ≥ 140/90 mmHg; and (3) reported history of any cardiovascular, chronic, and orthopedic disease. The participants visited the laboratory between 9:00 AM and 10:00 AM. All experimental measurements were conducted with the approval of the Institutional Review Board of Incheon National University (permission# 7007971-201904-006-01). The purpose and the procedure of the study were explained, and informed consent was obtained from all participants involved in the study prior to the experiment. The number of participations in the weekly exercise was recorded through a separate questionnaire to determine whether they participated in the exercise regularly. The weekly exercise questionnaire contained

self-report of habitual physical activity combined into leisure-time and sports-time. Basic clinical characteristics of the participants were presented in Table 1.



Values express means ± SEM. AIx, augmentation index; AIx@75, augmentation index corrected for heart rate 75; baPWV, brachial-ankle pulse wave velocity; HRrest, heart rate at rest.

### *2.2. Study Procedure*

All variables were measured in the following order: body composition, blood pressure in both arms at rest, arterial stiffness including AIx and baPWV, levels of lipid and glucose profiles in circulation, muscular strength, and anaerobic power. Participants fasted for at least 8 h before the measurement and were instructed to abstain from vigorous exercises or alcohol drinking for 24 h before visiting the laboratory. The order of measurements was arranged in such a way that the previous measure did not adversely affect the next one. Measurements for all participants were conducted at similar times and same place in a similar environment with controlled temperature 23–25 ◦C and humidity 40–60%.

### 2.2.1. Measurements of Anthropometric Parameters and Cardiometabolic Risk Factors

Height was measured using an extensometer (Sanwa, South Korea), and body composition was measured using an Inbody 720 (Biospace, Seoul, South Korea) machine utilizing the bioelectric impedance analysis (BIA) method. Body weight (kg), BMI (kg/m2), fat mass (kg), body fat (%), and fat-free mass (kg) were measured and obtained from BIA method. Skeletal muscle mass corrected by height (kg/m2) was used in this study [26,27]. Blood pressure was evaluated using Accuniq BP850 (Jawon, Seoul, South Korea). Participants measured blood pressure twice in a stable state, and the average of values was used in the analysis. For blood analysis, participants were seated in a chair and wiped their index fingers with alcohol cotton to draw whole blood. After placing the capillaries horizontally, 35 μL of whole blood was collected from index finger of each participant. Total cholesterol (Total-C), LDL-C, high-density lipoprotein cholesterol (HDL-C), triglycerides (TG), and fasting glucose (FG) were measured through Cholestech LDX system (Alere, Oslo, Norway), and hemoglobin A1c (HbA1c) was measured using Afinion AS100 Analyzer (Alere, Oslo, Norway). Anthropometric parameters and cardiometabolic risk factors of all participants were presented in Table 2.


**Table 2.** Anthropometric parameters and cardiometabolic risk factor of participants.

Values express means ± SEM. BMI, body mass index; DBP, diastolic blood pressure; FG, fasting glucose; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SBP, systolic blood pressure; Total-C, total cholesterol; TG, triglycerides.

### 2.2.2. Brachial-Ankle Pulse Wave Velocity (baPWV)

Peripheral arterial stiffness was measured with a non-invasive method using the Omron vp-1000 plus (Omron, Tokyo, Japan). Participants waited comfortably on the experimental bed for 10 min, placed cuffs at the same position of both upper arms and ankles, and prepared measurements by attaching electrodes to the left sternum. When the participant's resting heart rate was reached to stable condition, peripheral arterial stiffness was measured twice through an automatic waveform analyzer, and the value was obtained by calculating the movement distance and pulse wave propagation time between the limb arteries based on the height.

### 2.2.3. Augmentation Index (AIx)

AIx was assessed using SphygmoCor Xcel system (AtCor Medical, Sydney, Australia) to measure aortic stiffness in a non-invasive way. The participants lay in bed and rested for approximately 10 min with the cuff placed on the right upper arm. Before AIx measurement, participants' systolic and diastolic blood pressure were measured, and then AIx was measured using SphygmoCor Xcel software. The sphygmoCor device represents a pulse wave value calculated in consideration of age, heart rate, and height through a genderspecific equation [28]. AIx was obtained by dividing the augmented aortic pressure by the aortic pulse pressure. The augmented aortic pressure is made when the forward pulse wave generated from the left ventricle during the systolic period overlaps with the reflected pulse wave returned from the peripheral arterial trees. Since previous studies suggested that AIx and heart rate have a linear relationship, the value corrected to a heart rate of 75 bpm was used in the study [29,30]. AIx@75 was calculated through the following mathematical formula. AIx@75 = {−0.48 × (75-HR)} + AIx [31].

### 2.2.4. Muscular Strength

Muscular strength was measured using hydraulic hand dynamometer (Saehan, Seoul, South Korea). Hand grip strength was measured in a vertical position with 15◦ flexion at the elbow. After measuring each arm twice with an interval of 1 min, a higher value was selected and used as a parameter of muscular strength in the study.

### 2.2.5. Anaerobic Power

Wingate test (Monark 894E model, Vansbro, Sweden) was adopted to assess the anaerobic power. The participants adjusted the position of the saddle to bend 15◦ after the knee was completely extended from the bicycle ergometer saddle, and the participant lightly practiced pedaling at 50 rpm for 5 min. After a sufficient warm-up practice to relax their hip coxa and femoral muscle before the Wingate test, participants started pedaling maximum speed according to their willingness. The weight load was set to 0.075 kg per body weight. During the measurement, participants were encouraged orally to perform pedaling at maximum speed and power for 30 s. In this study, the absolute and relative peak anaerobic power, mean anaerobic power, and minimum anaerobic power were calculated and used.

### *2.3. Statistics*

Statistical analysis of all variables was performed using the SPSS version 23.0 software (SPSS Inc., Chicago, IL, USA). All measurements were presented as mean and standard error (mean ± SEM). Pearson correlation analysis was performed after the normality test of all variables. All statistical significance levels were set to *p* < 0.05.

### **3. Results**

### *3.1. Correlation between Skeletal Muscle Mass, Muscular Strength, and Anaerobic Power*

Pearson's correlation analysis was performed among skeletal muscle mass, muscular strength, and anaerobic power (peak power, average power, and minimum power). As expected, skeletal muscle mass adjusted by the height was positively correlated with muscular strength, peak power, average power, and minimum power (Table 3). In addition, muscular strength adjusted by the height was positively correlated with peak power, average power, and minimum power (Table 3).

**Table 3.** Results of Pearson's correlation analysis between skeletal muscle mass, muscular strength, and anaerobic power.


Correlation coefficient is in the upper line and *p* value in the lower one. The correlation analysis between all measurements.

### *3.2. Correlation between Cardiometabolic Risk Factors and Arterial Stiffness (baPWV, AIx, AIx@75)*

Pearson's correlation analysis between risk factors for cardiometabolic diseases and arterial stiffness revealed that both AIx and AIx@75 showed a negative correlation with BMI and FG and no significant correlation with SBP, diastolic blood pressure (DBP), total-C, HDL-C, LDL-C, TG, and HbA1c (Table 4). In addition, a positive correlation was observed between baPWV and DBP. However, baPWV showed no correlation with other variables, including BMI, SBP, Total-C, HDL-C, LDL-C, FG, TG, and HbA1c (Table 4).

### *3.3. Correlation between Muscular Variables and Arterial Stiffness (baPWV, AIx)*

Correlation between muscular variables and arterial stiffness was depicted in Figure 1. Pearson's correlation analysis revealed significant negative correlation between skeletal muscle, muscle strength, anaerobic, and AIx, respectively (Figure 1A–C). Likewise, AIx@75 (Figure 1D–F) revealed significant negative correlation with muscular variables. However, skeletal muscle mass, muscle strength, and anaerobic power showed no correlation with baPWV, respectively (Figure 1G–I).

**Variables r-Value BMI SBP DBP Total-C HDL-C LDL-C FG TG HbA1c** baPWV 0.247 0.256 0.382 0.072 0.476 0.022 −0.326 0.129 0.135 0.539 −0.309 0.151 0.235 0.281 −0.147 0.503 −0.015 0.946 AIx −0.636 0.001 −0.178 0.416 −0.084 0.702 −0.081 0.713 −0.151 0.491 −0.014 0.950 −0.513 0.012 −0.036 0.870 −0.068 0.757 AIx@75 −0.557 −0.168 −0.021 −0.073 −0.119 −0.014 −0.483 −0.001 −0.022

**Table 4.** Association between arterial stiffness (baPWV, AIx, and AIx@75) and cardiometabolic risk factors.

Correlation coefficient is in the upper line and *p* value in the lower one. AIx, augmentation index; AIx@75, AIx corrected for heart rate 75; BIA, bioelectrical impedance analysis; baPWV, brachial-ankle pulse wave velocity; DBP, diastolic pressure; FG, fasting glucose; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TG, triglycerides; Total-C, total cholesterol; SBP, systolic blood pressure.

0.589

0.951

0.020

0.995

0.920

0.741

0.006

0.444

0.924

**Figure 1.** Associations between arterial stiffness (AIx, AIx@75, baPWV) and muscular variables (skeletal muscle mass, muscular strength, anaerobic power). Pearson's correlations between AIx and skeletal muscle index (**A**), hand grip strength (**B**), and peak anaerobic power (**C**). Pearson's correlations between AIx@75 and skeletal muscle index (**D**), hand grip strength (**E**), and peak anaerobic power (**F**). Pearson's correlations between baPWV and skeletal muscle index (**G**), hand grip strength (**H**), and peak anaerobic power (**I**). \* *p* < 0.05, \*\* *p* < 0.01.

### **4. Discussion**

The present study aimed to examine whether muscular variables, including skeletal muscle mass, muscular strength, and anaerobic power, are related to arterial stiffness, an independent risk factor for cardiovascular disease. The major findings of the present study were that higher skeletal muscle mass, muscular strength, and anaerobic power were associated with decreased AIx, while baPWV was not associated with these muscular variables in physically-active young male adults.

This finding is consistent with the previous study, which reported a negative correlation between muscle mass and AIx in both males and females over 65 years of age [32]. In addition, according to study by Fahs et al., it represented a negative correlation between muscle strength and cf-PWV in 79 young male adults [21]; nevertheless, whether a correlation exists between AIx as a determinant of cf-PWV and muscle strength remains unknown. However, this study presented negative correlations between muscle strength and AIx, including corrected age, height, and heart rate 75 bpm. Interestingly, anaerobic power showed a negative correlation with AIx in the present study. Anaerobic power, which is an ability to express explosive energy, is an essential element for independent daily life [23]. It is well known that anaerobic power increases with elevated muscle mass and muscular strength and is improved with anaerobic training via short-term whole-body exercise [33]. Furthermore, it was reported that anaerobic power was reduced with aging and that the anaerobic energy system declined across the aging process without alteration in aerobic capacity in trained masters athletes [34]. For the most part, previous studies have examined anaerobic power in relation to sports-related performance in athletes [35,36], however, few studies have confirmed its correlation with cardiovascular risk factors [23,37]. According to previous studies, anaerobic exercise performed along with aerobic exercise has a positive effect on enhancing cardiovascular health by improving lipid components and BMI, and it improves vascular function by promoting vasodilators such as nitric oxide through high-intensity interval training [23,24,38]. Based on the results of this study and on those of previous studies, maintaining and increasing anaerobic power to a higher level may have a positive effect on AIx.

In the present study, skeletal muscle mass, muscular strength, and anaerobic power were not correlated with baPWV. Sugawara et al. have shown that baPWV was positively correlated with aortic PWV as well as with leg PWV [15]. Particularly, they suggested that baPWV was associated with central arterial stiffness, suggesting that baPWV may reflect central arterial stiffness as well as peripheral arterial stiffness [15,39]. However, Cortez-Cooper et al. indicated that baPWV was correlated with both central and peripheral arterial stiffness, but its correlation with peripheral arterial stiffness was stronger [40]. In addition, Laurent et al. suggested that it is clinically meaningful to evaluate central arterial stiffness rather than peripheral arterial stiffness while assessing cardiovascular diseases [16]. Taken together, central arterial stiffness may be an accurate indicator for the assessment of cardiovascular risk factors in a relatively younger age group [41]. Arterial stiffness shows a high incidence in patients with hypertension, and blood pressure is an index for predicting arterial stiffness even in young adolescents. Moreover, peripheral SBP was higher than central SBP even in adolescents, suggesting that peripheral arterial stiffness is greatly affected by blood pressure even in a relatively younger age group [42]. In case of young male adults in this study, the systolic (119.4 mmHg ± 2.1) and diastolic (66.9 mmHg ± 2.0) blood pressure measurements were within the normal range. Hence, baPWV may not show significant correlation in younger individuals unlike other age groups. In addition, an increase in skeletal muscle mass and muscular strength with resistance exercise may not affect peripheral arterial stiffness in young adults. For example, a study reported that lean body mass and muscular strength significantly increased after 11 weeks of high-intensity resistance exercise, but femoral-ankle PWV reflecting peripheral arterial stiffness did not ameliorate after the exercise [43]. The current study measured grip strength, an index reflecting muscular strength, and analyzed its correlation with baPWV. BaPWV accurately reflects arterial stiffness in the lower body when compared with that in

the upper body because it reflects peripheral arterial stiffness [44]. However, grip strength is better correlated with muscle mass and muscular strength of the upper extremities and is rather limited in assessing the ability of the lower extremities [45]. This fact may explain the lack of correlation between grip strength and baPWV in this study.

Several limits of this study should be mentioned to benefit future research. First, in the present study, we conducted in a small number of healthy young male adults as a preliminary research. Therefore, the results of our study cannot be generalized to female subjects and individuals of different age ranges. Second, causality between anaerobic power and arterial stiffness was not confirmed in the present study; therefore, future studies investigating the physiological relationship between these factors are warranted.

Although the present study did not suggest the exact mechanism of how skeletal muscle mass, muscular strength, and anaerobic power mutually affect the arterial stiffness, it showed that higher skeletal muscle mass and muscular strength might have a positive effect on AIx. These findings support the fact that maintaining or improving skeletal muscle mass, muscular strength, and anaerobic power to a higher level can produce a positive effect on AIx even in the young male adults.

### **5. Conclusions**

In conclusion, higher skeletal muscle mass, muscular strength, and anaerobic power are associated with decreased AIx in young male adults. However, baPWV is not associated with these muscular parameters. These findings suggest that maintaining or improving skeletal muscle mass, muscular strength, and anaerobic power to a higher level from a young age may have a positive effect on AIx.

**Author Contributions:** Conceptualization, D.L. and S.L.; Formal analysis, D.L., M.-H.H.; Investigation, D.L., K.B., M.-H.H., and S.L.; Methodology, D.L., K.B., and M.-H.H.; Original manuscript draft preparation, D.L. and S.L.; Review and editing, K.B., M.-H.H., and S.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** All experimental measurements were conducted with the approval of the Incheon National University Institutional Review Board (Permission# 7007971-201904- 006-01).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Data are not publicly available due to privacy or ethical restrictions.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


### *Article* **The Role of Cholinesterases in Post-Exercise HRV Recovery in University Volleyball Players**

**José Raúl Hoyos-Flores 1,2, Blanca R. Rangel-Colmenero 2, Zeltzin N. Alonso-Ramos 2, Myriam Z. García-Dávila 2, Rosa M. Cruz-Castruita 2, José Naranjo-Orellana <sup>3</sup> and Germán Hernández-Cruz 2,\***



**Citation:** Hoyos-Flores, J.R.; Rangel-Colmenero, B.R.; Alonso-Ramos, Z.N.; García-Dávila, M.Z.; Cruz-Castruita, R.M.; Naranjo-Orellana, J.; Hernández-Cruz, G. The Role of Cholinesterases in Post-Exercise HRV Recovery in University Volleyball Players. *Appl. Sci.* **2021**, *11*, 4188. https://doi.org/10.3390/app11094188

Academic Editor: Daniela Galli

Received: 25 March 2021 Accepted: 23 April 2021 Published: 4 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Abstract:** Some studies show interest in measuring heart rate variability (HRV) during post-exercise recovery. It is known that the parasympathetic system is relevant during this process, where one of the factors of this modulation is the interaction of acetylcholine and cholinesterases (ChE). However, the behavior of ChE and its relationship during recovery is little known; therefore, the objective of this study was to analyze the behavior of ChE and its relationship with recovery evaluated in HRV indicators in volleyball players. An exercise protocol with long-term and intermittent high-intensity phases was applied in nine volleyball players. HRV measurements were made, and blood samples were drawn to evaluate the ChE before exercise and after 24 and 48 h post-exercise. The results show a modification of the variables after exercises with respect to the baseline values (ChE: 1818.4 ± 588.75 to 2218.78 ± 1101.58; RMSSD: 42.64 ± 12.86 to 17.72 ± 12.55 (*p* < 0.05); SS: 8.76 ± 1.93 to 21.93 ± 10.05 (*p* < 0.01); S/PS Ratio: 0.32 ± 0.14 to 3.26 ± 3.28 (*p* < 0.01)), as well as recovery after 24 and 48 h with respect to postexercise (ChE: 1608.81 ± 546.88 (*p* < 0.05) and 1454.54 ± 580.45 (*p* < 0.01); RMSSD: 43.83 ± 24.50 and 46.18 ± 33.22 (*p* < 0.01); SS; 10.93 ± 5.16 and 11.86 ± 4.32 (*p* < 0.01); S/PS Ratio: 0.46 ± 0.32 and 0.50 ± 0.28 (*p* < 0.01)). ChE correlations (*p* < 0.001) were found with moderate (SS: *r* = 0.465) and large (RMSSD: *r* = −0.654; S/PS Ratio: *r* = 0.666) HRV indexes. In conclusion, ChE modifications are related to changes in HRV showing a very similar behavior in the case of the study subjects.

**Keywords:** cholinesterases; heart rate variability; autonomic recovery mechanisms; sympathetic– parasympathetic modulation; postexercise recovery

### **1. Introduction**

Training load monitoring has become of vital importance to achieve better exercise adaptations [1,2]. Postexercise recovery plays an important role in this working load monitoring where there is great interest in the use of heart rate variability (HRV), since it is an easily accessible noninvasive tool [3–5] for monitoring the state of the autonomous nervous system (ANS) through its sympathetic and parasympathetic branches [6]. It is also useful for athlete monitoring [7,8], since it allows observing adaptations to training by measuring the fatigue and recovery state [9]. One of the most analyzed indexes of HRV is the root mean square of successive RR interval differences (RMSSD) or its natural logarithm (lnRMSSD) to reduce interindividual variations [3,10]. Recently, Naranjo Orellana et al. [11] published two new HRV indexes, the stress score (SS) and the sympathetic–parasympathetic ratio (S:PS ratio), which provide another viewpoint regarding sympathetic activity and the balance between both branches of the ANS, respectively.

A large number of studies have focused on measuring these variables in postexercise recovery and in determining how the duration and intensity affect them [5,12–14]. However, there is a discrepancy in the use of these results, since, as far as we know, this tool has not been considered for evaluating the internal load [15]. Recovery is a complex interaction between the sympathetic and the parasympathetic system [5,12], where the sympathetic influence is measured by the release of catecholamines and the activation of adrenergic receptors, and the parasympathetic influence is controlled by the release of acetylcholine in the vagus nerve increasing the parasympathetic tone [6]. However, during recovery, parasympathetic activity becomes relevant in returning the organism to a stable state [12,16,17]; thus, afferent vagal stimulation will lead to a reflex excitation of efferent vagal activity that inhibits sympathetic activity [6].

Parasympathetic control depends on the release of acetylcholine from the vagus nerve, which will bind to muscarinic receptors that hyperpolarize the cardiac muscle to decrease the slope of depolarization, resulting in a slowing of the heart rate [18,19]. Efferent parasympathetic control is mediated by cholinergic signaling in the sinoatrial node, which is rich in acetylcholinesterase (AChE), which hydrolyzes acetylcholine after vagal impulses [6,20,21] that will maintain the parasympathetic tone [22,23]. Several studies have analyzed the inhibition of AChE using pyridostigmine bromide to increase the parasympathetic activity, measuring it by HRV [20,24,25].

Since the cholinergic system is involved in the process of ANS modulation, its measurement could provide information on the regulation of the parasympathetic activity. However, clinical measurements of ACh are complicated [26]; therefore, cholinesterase (ChE) measurements could provide this information, as plasma total free ChE has been reported to determine the hydrolysis capacity of ACh [27] and is effective in observing cholinergic events [28]. It seems logical to think that cholinesterases (ChE), divided into the isoenzymes AChE and butyrylcholinesterase (BChE), are the enzymes that hydrolyze ACh descending into muscles after physical exercise to favor parasympathetic activity, thus increase their concentration in the blood.

However, no studies have been found regarding the behavior of ChE in postexercise situations, relating it to HRV recovery. To our knowledge, two studies measure ChE after physical exercise but without associating it with postexercise recovery. One of these [29] assessed whether gender and a physical activity session could modify the AChE and BChE values in people not exposed to pesticides, concluding that physical activity modifies ChE levels after an exercise session. The other [30] analyzed the metabolic responses of traditional clinical markers of the liver; muscle; heart; and bone function (ChE, creatine kinase, lactate dehydrogenase, alkaline phosphatase, etc.) in response to a session of aerobic medium-long distance free running in football players. The authors concluded that these markers provided valuable information on postexercise muscle conditions. However, they mention that ChE should not be used as a marker of the metabolic response, possibly because it was analyzed as a marker of liver damage in this study. We assume at a theoretical level that it could be novel to study the behavior of ChE during recovery and its possible relationship with the ANS regulation process.

We hypothesize that ChE behavior is associated with the process of ANS modulation in postexercise recovery and may be a marker of internal loading. Therefore, the objective of this study was to analyze the influence of ChE behavior on ANS modulation, as assessed across HRV indicators after a combined long-duration and intermittent high-intensity phases exercise in collegiate volleyball players.

### **2. Materials and Methods**

### *2.1. Headings*

This was a quasi-experimental study with an explicit–correlative scope. The study began with a medical examination of the study subjects, measures of body composition, and a stress test, followed by the application of a physical exercise intervention protocol. After the intervention, the subjects had a recovery period of 48 h concentrated in the

installations, controlling rest, the intake of any supplements or drugs that could affect the results, and food. The study duration had a total of 15 days.

### *2.2. Subjects*

A total of nine volunteer volleyball players (21.44 ± 2.07 years; 77.38 ± 6.44 kg body weight; 185.82 ± 10.71 cm height; 25.48 ± 2.50% adipose mass; 48.07 ± 1.86% muscle mass; 18.33 ± 5.16% fat mass; and 49.47 ± 6.36-mL·kg−1·min−<sup>1</sup> VO2max) from the representative team of the Universidad Autónoma de Nuevo León were studied. Sample selection was non-probabilistic, and size was determined using the statistical package G\*Power version 3.1.9 (G\*Power, Heinrich-Heine, Universität Düsseldorf, Germany) with a probabilistic error of *α* = 0.05, a statistical power of the probabilistic error of 1 − *β* = 0.80, and an effect size of *d* = 1.20.

An informative meeting was held regarding the research objectives, and afterward, all the subjects signed written informed consent for their participation in the study. This study was carried out in accordance with the recommendations of the Ethics Committee of the Universidad Autónoma de Nuevo León (COBICIS-16.09/2012.01GHC), following the ethical standards of all the principles expressed in the Helsinki Declaration [31].

### *2.3. Procedure*

The research was carried out during a training recess of the volleyball players in the winter vacation period. The assessment began with a medical history, physical examination, a blood sample and an HRV measure as a baseline (BASELINE), a body composition analysis (subjects were asked as much as possible to come with the bladder and bowel emptied) by dual X-ray densitometry (DXA (Lunar Prodigy, GE Healthcare, Madison, WI, USA)), kinanthropometry using the complete profile proposed by the International Society for the Advancement of Kinanthropometry (ISAK) [32], and a stress test. These measures were carried out during the first 12 days of the study, as shown in Figure 1.

**Figure 1.** Experimental approach.

To determine the VO2max, an incremental stress test was performed starting at 8 km·h−<sup>1</sup> and increasing 0.5 km·h−<sup>1</sup> every 30 s until exhaustion with a COSMED Quark PFT ergospirometer (COSMED The Metabolic Company, Rome, Italy) and a TuffTread NF4616HRT treadmill (Tuff Tread, Conroe, TX, USA). VO2 was determined when a VO2 concentration plateau was reached; if this plateau was not reached, the value considered was when the respiratory coefficient was greater than 1.15, and the theoretical maximum heart rate was greater than 95%. In this way, the maximum aerobic speed (MAS) was established as the starting speed of the VO2 max plateau. The methodology proposed by Skinner and McLellan [33] was used to obtain the ventilatory thresholds.

Afterward, a high-intensity and long-duration intermittent exercise protocol was applied to produce elevated physiological stress. Then, blood samples were drawn, and HRV during the immediate recovery period (AFTER) and at 24 (24H) and 48 (48H) hours was measured. The subjects were concentrated in the installations of the university where they avoided stimulating food, supplements, or drugs that could alter the results. Food intake (they were given a balanced, proportional, and correct diet) and rest were controlled during the postexercise recovery process.

### *2.4. Blood Samples*

The extraction area was sterilized to obtain the blood sample, and a puncture was made in the median cubital vein with a double-bevel needle, collecting the sample in a 4-mL tube with the anticoagulant EDTA (K2 EDTA, BD Vacutainer, Franklin Lakes, NJ, USA). Afterward, the samples were placed in a Solbat J-40 (SOLBAT S.A. de C.V., Puebla, Mexico) centrifuge at 3000 rpm for 10 min to obtain plasma. Aliquots of plasma were placed in 1.5-mL Eppendorf tubes and stored at −20 ◦C.

### *2.5. Exercise Protocol*

To obtain a physical demand greater than at a competition and higher physiological stress in volleyball players, a combined exercise protocol was designed based on the Loughborough Intermittent Test [34]. The aim of this protocol was to equal the physical load on athletes to avoid unequal physical loads that commonly occur in team sports during competition, depending on the time of play of the athletes. The VT2 speed percentage reached was used to determine the mean intensity of work for each block.

The protocol consisted of 6 work blocks with 3 min of passive recovery between each. The sequence of the first five blocks was 3 rounds of walking, 3 jumps over a 50-cm hurdle, 1 maximum sprint round, three trotting rounds, and 3 maximum sprint rounds; each round was continuous without recovery with a distance of 20 m and repeating each of the blocks five times. Last, the sixth block consisted of performing a round trotting and a maximum sprint round, repeating these until no more rounds could be performed.

### *2.6. Cholinesterases*

An ELISA in 96-well microplates was used for total and overall concentration of AChE and BChE, following the protocol of the Human Acetylcholinesterase/ACHE ELISA kit (DuoSet ELISA Development System, R&D Systems, Minneapolis, MN, USA). This was done in duplicate, placing 100 μL of the standard solution and its dilutions for the calibration curve and the subjects' plasma samples. Afterward, concentration readings were performed with a Bio-Rad iMark Microplate Reader spectrophotometer (Bio-Rad Laboratories, Inc., Hercules, CA, USA) at a wavelength of 450 nm. To calculate the results, a four-parameter logistic (4PL) curve formula was used for logistic regression. For the BASAL level, two nonconsecutive measures were performed. The coefficient of determination of the linear regression of the real and the measured concentration was calculated with an *R*<sup>2</sup> = 0.995 and an ICC reliability = 0.998, *p* < 0.001. For the baseline measure, the determination coefficient was *R*<sup>2</sup> = 0.878 and the ICC reliability = 0.853 (95% CI = 0.731; 0.984); *p* < 0.01.

### *2.7. Heart Rate Variability*

The heart signal was recorded using the Polar Team 2 system (Polar Team2, Polar Electro OY, Kempele, Finland). Ten-minute recordings at rest lying down supine were obtained every day of the week (6:00 a.m.–8:00 a.m.). Data were transferred to the Polar ProtrainerTM software version 1.4.5 in its beat-to-beat version to extract the time series (R-R). This temporary series was exported to Kubios software version 2.2 (Kubios HRV Analysis Software, The MathWorks Inc, University of Eastern Finland, Kuopio, Finland) for filtration and analysis, obtaining the RMSSD, SD1, and SD2 parameters. Later, the SS and S:PS ratio values were obtained using the protocols proposed by Naranjo Orellana et al. [11].

### *2.8. Statistical Analysis*

Statistical analysis was carried out using SPSS statistical software version 25 (IBM Corp., Armonk, NY, USA) with a significance level of *p* < 0.05 for all analyses. Descriptive data

are presented as means and standard deviations. A normality analysis was performed with the Shapiro–Wilk test, and Friedman with a Wilcoxon post-hoc tests were used to compare means. Regarding ChE, a reliability analysis was performed using the intraclass correlation coefficient (ICC) and linear regression with a coefficient of determination (*R*2). Subsequently, the data were normalized using the Z-Score to adjust for a change in intrasubject values in inter-subject samples. To observe the relationship between variables, Pearson's correlation was used, and for the interpretation of correlation magnitudes (*r*), the following criteria were adopted: ≤0.10, trivial; >0.10 to 0.30, small; >0.30 to 0.50, moderate; >0.50 to 0.70, large; >0.70 to 0.90, very large; >0.90, extremely large; and 1.0, perfect [35].

### **3. Results**

The training load was similar for all study subjects with regards to the first five blocks. In relation to block 6, the individual workload varied a little because of the characteristics. Table 1 shows the workload performed by the participants.


**Table 1.** Training load of the exercise protocol.

VT2: second ventilatory threshold; M: mean; SD: standard deviation; CV: coefficient of variation.

Regarding the descriptive values of ChE, the RMSSD, SS, and the S:PS ratio for the different measures, these are shown in Table 2 with the means and standard deviation.



ChE: cholinesterase; RMSSD: root mean square of successive RR interval differences; SS: stress score; S:PS ratio: sympathetic–parasympathetic ratio; M: mean; SD: standard deviation.

Figure 2 shows the behavior of the analyzed variables, which shows that ChE, SS, and the S:PS ratio had a greater elevation in the AFTER measure compared to the BASELINE measure, with a significant change of the SS, as well as the S:PS ratio (*p* < 0.01). The RMSSD has a significant descent (*p* < 0.05) in the AFTER measure compared to the BASELINE measure. All significantly returns to baseline levels at 24H and 48H (*p* < 0.01), except for ChE, which decreases significantly more than its baseline level (*p* < 0.05 and *p* < 0.01) compared to AFTER; in the 48H measure, it continues to descend with a significant change (*p* < 0.05) compared to 24H.

**Figure 2.** Changes in cholinesterases and HRV at baseline, postexercise, and 24H and 48H after recovery. (**A**) Cholinesterases, (**B**) root mean square of successive RR interval differences, (**C**) stress score, and (**D**) sympathetic/parasympathetic ratio. \* *p* < 0.05, in relation to the BASELINE. \*\* *p* < 0.01, in relation to the BASELINE. ¤ *p* < 0.01, in relation to AFTER. § *p* < 0.05, in relation to 24H.

The relationship of the behavior between variables during the measures is shown in Table 3, where statistically significant mean and high correlations between all the variables can be seen. The correlations of ChE with the HRV indexes are presented graphically in Figure 3 by linear regressions to observe the trend line adjustment and the coefficient of determination (*R*2).

**Table 3.** Correlation between the variables ChE, RMSSD, SS, and the S:PS ratio.


\*\* Two-tailed bilateral significance at the level *p* < 0.001. ChE: cholinesterase; RMSSD: root mean square of successive RR interval differences; SS: stress score; S:PS ratio: sympathetic–parasympathetic ratio.

**Figure 3.** Correlation *R*<sup>2</sup> of the ChE with the root mean square of successive RR interval differences (**A**), the stress score (**B**), and the sympathetic/parasympathetic ratio (**C**).

### **4. Discussion**

The main contribution of our study is that ChE shows a behavior during postexercise recovery that is similar to the RMSSD indexes, SS, and S:PS ratio of HRV; therefore, we can consider these internal load markers.

It has been described in the literature that during a postexercise recovery, the interaction of the parasympathetic and sympathetic systems is complex [5,12] and mediated mainly by parasympathetic reactivation [36,37]. This includes multiple regulations such as the descent of circulating catecholamines, blood pressure, baroreflexes, and metaboloreflexes that cause a drop in sympathetic stimulation [4,5,17] due to an efferent reflex effect of vagal stimulation [6].

This effect is regulated by cholinergic signaling in the sinoatrial node, and in a certain way, ChE are mediators of the vagal impulse by hydrolyzing ACh [6,20,21]; thus, an increase of ChE in blood indirectly reflects an increase in the rate of ACh hydrolysis leading to ACh descent [27,38,39]. Our results support this idea since. in postexercise measures, the increase of ChE widely correlated with an increase in sympathetic activity (SS and S:PS ratio) and a decrease in RMSSD (Table 3). Afterward, in the 24H and 48H measures, ChE normalized, and RMSSD returned to her initial values, as well as the SS and the S:PS ratio. These changes may be because AChE facilitates the signaling of the onset and termination of the cardioinhibitory effect by hydrolyzing ACh in the synaptic cleft in a very short time (<1 s) [4,40]. Subsequently, the parasympathetic nerve will be stimulated in the phasic mode, where released acetylcholine will hyperpolarize the sinoatrial node by binding to its muscarinic receptors decreasing the depolarization slope and thereby delaying the next sinoatrial node impulse [18,19,41] and, thus, favoring postexercise HRV recovery.

This seems to indicate that these variables evaluate the internal load regardless of the type of exercise performed since, although some authors relate the recovery with duration [4,42] or with intensity [5,43], in our results, it can be seen that the workload was very similar for all subjects (Table 1), demonstrating that the changes in response show the individual assimilation of that load. Therefore, the changes in these variables would fit into the concept of internal load. The indexes, RMSSD, SS, and S:PS ratio have been used jointly in various studies to see the behavior of the sympathetic and parasympathetic systems, both in training control and in competition, concluding that these indicators are valid and reliable for training and competence monitoring [13,44–46].

On the other hand, cholinesterase values (Table 2) are within the range of acceptable clinical values for the technique used (1278 ± 338 pg/mL), so the changes observed in this variable are not at any time abnormal or pathological values. Thus, we observed that the postexercise ChE behavior coincides with the increase reported in the study by Zimmer et al. [29]. However, this does not seem to be the case with that reported by Chamera et al. [30], as the behavior of ChE in women is very similar to our results but not so in men. This could be because the impact of the session was not sufficiently demanding to provoke high rates of ACh hydrolysis. The values of RMSSD, SS, and the S:PS ratio coincide with those presented by other authors [13,44,47], and they would be within the 25–50 percentile of the reference tables reported by Corrales et al. [47]. The behavior of HRV indices following such an exercise is similar to what has already been well-documented in the literature [13,48], where it has been reported that HRV is a reliable indicator to measure fatigue and postexercise recovery with acceptable relations [3–5,8,9,49], establishing the average values for a good state of the athlete (recovered) of RMSSD above 30–40 ms, SS below 10AU and S:PS ratio below 0.5AU [11,44,50]. In addition, acceptable relationships have been shown with other classical markers of recovery (creatine kinase, testosterone/cortisol ratio, immune/inflammatory markers, and training load) [9,13,51–53].

To support the idea of the role of ChE during recovery, in our results, the correlation of ChE with HRV indexes was observed. This could indicate the modulation effect of the pacemaker potential of the sinoatrial node by parasympathetic innervations mediated by ACh and ChE [6,12,21,36,37,54,55]. This may be explained by the fact that, when central command and mechanoreceptor feedback cease, the arterial baroreflex is reset to a lower level, increasing the parasympathetic activity and rendering it phasic [36,37,40,56,57]. Since this parasympathetic activity is controlled by the release of ACh from the efferent vagal nerve discharge, its modulation will largely depend on the ChE activity [6,20,21], supporting this idea by the association found with HRV indices, since as mentioned above, plasma ChE could provide information on cholinergic events [27,28].

### *4.1. Limitations*

The main limitations of this study are the sample size, which has been set at this number to prioritize the homogeneity offered by being a team contemplating a statistical analysis to obtain the ideal sample size and not having the means that allow us to simultaneously contrast what happens in synaptic transmission. Nevertheless, the relationships found between the variables allow us to reasonably assume the conclusions found and to set a starting point for new studies in this direction.

### *4.2. Practical Applications*

Our results describe the behavior of ChE and how it is associated with the recovery of parasympathetic activity postexercise, which could be of useful application in understanding—more specifically, the behavior of HRV indices following a training load similar to that of this study.

### **5. Conclusions**

The modifications observed in ChE appear to influence the changes in sympathetic and parasympathetic activity, since they show an acceptable relationship with HRV indices, which could improve the physiological interpretation of the recovery of parasympathetic activity after a combined long-duration and intermittent high-intensity phases exercise. These findings might have relevance for specialists in the control of training areas for future research, since these results suggest that the ChE and HRV levels might be considered internal load markers.

**Author Contributions:** Conceptualization, J.R.H.-F., B.R.R.-C., and G.H.-C.; Methodology, J.R.H.-F., G.H.-C., R.M.C.-C., and J.N.-O.; Formal Analysis, J.R.H.-F., R.M.C.-C., and J.N.-O.; Investigation, J.R.H.-F., B.R.R.-C., Z.N.A.-R., M.Z.G.-D., and J.N.-O.; Resources, J.R.H.-F., B.R.R.-C., and G.H.-C.; Data Curation, J.R.H.-F. and M.Z.G.-D.; Project Administration, J.R.H.-F., B.R.R.-C., Z.N.A.-R., and G.H.-C.; Funding Acquisition, B.R.R.-C. and G.H.-C.; Writing—Original Draft Preparation, J.R.H.-F. and J.N.-O.; and Writing—Review and Editing, J.R.H.-F., G.H.-C., and J.N.-O. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Scientific and Technological Research Program (Programa de Apoyo a la Investigación Científica y Tecnológica (PAICYT)) of the Universidad Autónoma de Nuevo León, grant number CSA949-19.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of the Universidad Autónoma de Nuevo León (COBICIS-16.09/2012.01GHC).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

**Acknowledgments:** We thank the coach Jorge Azair and his team of athletes for their support and help in performing this research.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**

1. Bourdon, P.C.; Cardinale, M.; Murray, A.; Gastin, P.; Kellmann, M.; Varley, M.C.; Gabbett, T.J.; Coutts, A.J.; Burgess, D.J.; Gregson, W.; et al. Monitoring athlete training loads: Consensus statement. *Int. J. Sports Physiol. Perform.* **2017**, *12*, S2161–S2170. [CrossRef]


### *Article* **Changes in Plasma Bioactive Lipids and Inflammatory Markers during a Half-Marathon in Trained Athletes**

**Melania Gaggini 1, Cristina Vassalle 2,\*, Fabrizia Carli 1, Maristella Maltinti 2, Laura Sabatino 1, Emma Buzzigoli 1, Francesca Mastorci 1, Francesco Sbrana 2, Amalia Gastaldelli <sup>1</sup> and Alessandro Pingitore <sup>1</sup>**


**Abstract:** Background: Exercise may affect lipid profile which in turn is related to inflammation, although changes of ceramides, diacylglycerols-DAG and sphingomyelin-SM and their relationship with inflammatory parameters following a half-marathon have never been examined. Methods: Ceramides, DAG and SM, and markers of inflammation (soluble fractalkine-CX3CL1, vascular endothelial growth factor-VEGF, interleukin6-IL-6 and tumor necrosis factorα-TNFα) were evaluated in trained half-marathoners before, post-race (withdrawal within 20 min after the race end) and 24 h after. Results: IL-6 and CX3CL1 increased immediately after the race, returning to baseline after 24 h. Total ceramides and total DAG significantly decreased post-race. Several ceramide classes decreased after exercise, while only one of the DAG (36:3) changed significantly. Total SM and specific species did not significantly change. Conclusion: Some inflammatory parameters (IL-6 and CX3CL1) transiently increased after the race, and, being reversible, these changes might represent a physiological response to acute exercise rather than a damage-related response. The decrease of specific lipid classes, i.e., DAGs and ceramides, and the lack of their relationship with inflammatory parameters, suggest their involvement in beneficial training effects, opening promising research perspectives to identify additional mechanisms of aerobic exercise adaptation.

**Keywords:** exercise; ceramides; cytokines; diacylglycerol; biomarkers

### **1. Introduction**

Half-marathon (21.0975 km) is an increasingly popular recreational activity. Its diffusion is in part related to the fact that half-marathon requires less powerful training than the marathon [1]. Nonetheless, one reason for its importance in terms of clinical risk is that half-marathon is performed by a growing number of people in the world, although it still remains a powerful training, which might potentially retain adverse health repercussions. In fact, studies focusing on the role of exercise and inflammatory response have underlined that several cytokines increased after a competitive marathon [2]. In this context, it is interesting to assess the role of sphingolipids, bioactive lipids that regulate diverse cell functions, since the activity of sphingomyelinases (sphingolipid metabolizing enzymes) is increased under inflammation and oxidative stress [3]. Thus, ceramides have been associated with oxidative stress status and inflammatory processes [3]. However, changes of ceramide concentration after a half-marathon race have never been examined. Moreover, to the best of our knowledge, the relationship that may exist between bioactive lipids and markers of inflammation/immune response has never been studied in trained subjects following acute physical exertion such as a half-marathon. In this context, how changes in the balance between inflammatory mediators, immune response or lipid may contribute to the response in the post-race phase and affect the health status of subjects participating in such

**Citation:** Gaggini, M.; Vassalle, C.; Carli, F.; Maltinti, M.; Sabatino, L.; Buzzigoli, E.; Mastorci, F.; Sbrana, F.; Gastaldelli, A.; Pingitore, A. Changes in Plasma Bioactive Lipids and Inflammatory Markers during a Half-Marathon in Trained Athletes. *Appl. Sci.* **2021**, *11*, 4622. https:// doi.org/10.3390/app11104622

Academic Editor: Daniela Galli

Received: 23 April 2021 Accepted: 15 May 2021 Published: 19 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

events also remains unclear. Thus, we aimed to evaluate levels of bioactive lipids related to inflammation signaling pathways, and to this purpose, we studied plasma ceramides, diacylglycerol and SM. These lipids can modulate the activity of intracellular enzymes (e.g., those involved in insulin signaling). In detail:


Thus, we aimed to evaluate the changes in plasma levels of these bioactive lipids in healthy runners performing a half-marathon, at the end of the race and after 24 h recovery, and their associations with new recently proposed and common biomarkers of immune activation, which are:


### **2. Materials and Methods**

### *2.1. Characteristics of the Participants*

The studied population included 13 healthy Caucasian trained runners belonging to the "Gruppo Podistico Rossini" enrolled during the 2018 edition of the Pisa half-marathon. Inclusion criteria of athletes were regular training and absence of cardiovascular disease or any other systemic disorder. Preliminarily to the race, each participant was submitted to a questionnaire to obtain demographic and clinical data and training history. Body mass index (BMI) was calculated by height and weight, measured in each subject. Body fat composition (fat-free mass (FFM); fat% and fat mass) was evaluated by TANITA. The study was conducted according to the guidelines of the Declaration of Helsinki, and was approved by the Pisa Ethics Committee, Italy (protocol number for study acceptance 2805). Informed consent was obtained from all subjects enrolled in the study.

### *2.2. Sample Collection, Preparation and Evaluation of Lipids and Inflammatory Markers*

Three blood samples were collected from the peripheral vein of the athletes (in fasting conditions): (1) the day before the race (baseline), (2) within 20 min after the end of the half-marathon (post) and (3) 24 h after the run (24 h, recovery period). Blood samples were immediately centrifuged at 2500 *g* for 10 min and stored at −80 ◦C until assayed. Ceramides, SM and DAG were evaluated from 20 μL of plasma (upon being deproteinized using 200 μL of cold methanol) in 13 healthy Caucasian trained runners.

An aliquot of 20 μL of a mix of internal standards SM (d18:1/17:0), ceramide (d18:1/17:0), (Avanti Polar Lipids, Alabaster, AL, USA); DAG (17:0/17:0) (Larodan Solna SE) was added to the sample before deproteinization. The extract was injected in high performance liquid chromatography (Agilent UHPLC 1290, Santa Clara, CA, USA), coupled with a quadrupole time-of-flight mass spectrometry QTOF (QTOF-MS, Agilent 6540, Santa Clara, CA, USA), equipped with electrospray ionization source (ESI). For liquid chromatography analysis, we used ZORBAX Eclipse Plus C18 2.1 × 100 mm 1.8-micron column at 50 ◦C (Agilent, Santa Clara, CA, USA). The mobile phase A was made by water with 0.1% formic acid and the mobile phase B was made by isopropanol/acetonitrile (1:1, *v*:*v*) with 0.1% formic acid. Injection volume was 1 μL. Lipids were measured in positive electrospray ionization and identified using an internal spectral library. The data were normalized by internal standard representative lipids present in the sample and the analysis of the peaks was performed with Agilent MassHunter software program [11]. DAG species founded in plasma were 32:0, 32:1, 34:1, 34:2, 36:3 and 36:4, while ceramide species evaluated in plasma sample were 18:1/16:0, 18:1/18:0, 18:1/25:0, 18:1/26:0, 18:1/22:0, 18:1/24:1, 18:1/24:0, 18:0/24:0. The metabocard for ceramide species was reported as supplementary material (supplemental data). Fractalkine CX3CL1, VEGF, IL6 and TNFα were analyzed in 25 μL of plasma by a specific assay (MILLIPLEX MAP Millipore corporation, Billerica, MA, USA) using an integrated multi-analyte detection platform (high-throughput technology MagPix system, Luminex xMAP technology) with combined analyst software (MILLIPLEX®) for the biomarkers quantification developing new curve fitting algorithms and optimizing mathematical methods to minimize fitting errors.

### *2.3. Statistical Analysis*

Data were expressed as mean ± SD. **Repeated-measures ANOVA** was used for compared data from the same subjects measured more than once (baseline, post, 24 h). Correlation analysis was performed by Spearman parametric test to assess the relationship between continuous variables. Post Δ and 24 h Δ values from baseline (differences) and % change at different successive time periods with respect to values observed at baseline were calculated for fractalkine (pg/mL), IL-6 (pg/mL), TNFα (pg/mL) and VEGF-α (pg/mL). Owing to skewness, log transformation of fractalkine and TNFα was used for statistical analyses. Log-transformed values were then back-transformed for data presentation.

Data statistical analyses were performed with the Statview statistical package, version 5.0.1 (SAS Institute, Abacus Concept, Berkeley, CA, USA). A *p* value of <0.05 was considered statistically significant. Heatmapping was performed using MetaboAnalyst R 1.0.3 (XiaLab at McGill University, Montreal, QC, Canada). The data in heatmaps were analyzed by t test, the algorithm used was the average method and the measure of distance was Euclidean.

### **3. Results**

### *3.1. Demographic and Training Characteristics*

The characteristics of the runners are summarized in Table 1. Each athlete was regularly engaged in marathon training for more than 3 years, 3–7 times/week in 1–2 h/session. Athletes performed a running race over the distance of 21.0975 km. Race time ranged between 1.33 and 1.46 h. No correlation of inflammatory parameters or sphingolipids with demographic and training characteristics was observed. No gender-related differences were observed for all the biomarkers evaluated.


**Table 1.** Anthropometric measurement of runners and physical activity.

Abbreviations: BMI: body mass index, PAS: systolic arterial pressure, PAD: diastolic arterial pressure, FFM: free fat mass.

### *3.2. Lipids Levels and Race-Related Trends*

### 3.2.1. Total Ceramides, DAG and SM

Total ceramides, total DAG and SM at baseline, post and 24 h after the race are reported in Figure 1 (panel A, B and C, respectively). When plasma levels of total ceramides were considered, a significant decrease was observed after the race and after 24 h compared to baseline (12.25 ± 3.0, 8.68 ± 0.62, *p* = 0.011 vs. baseline; 9.70 ± 0.51, μmol/L, *p* = 0.059 vs. baseline; baseline, post-race, 24 h, respectively), Figure 1A. DAG total analysis evidenced a significant decrease post-race compared to baseline (53.26 ± 21.83 vs. 36.07 ± 18.75 μmol/L, post-race vs. baseline, *p* = 0.04; and 51.56 ± 34.25 24 h μmol/L, not significant), Figure 1B. Instead, SM tended to decrease after the race and at 24 h, although not significantly (145.66 ± 44.83; 100.2 ± 8.9; 105 ± 7.31 μmol/L), Figure 1C.

### 3.2.2. Ceramides, DAG and SM Species

The specific DAG and SM ceramide species at baseline, post and 24 h after the race are shown by heatmaps that provide intuitive visualization of the data table. Each colored cell on the map corresponds to a concentration value in the data file, with samples in rows and compound in columns (supplemental data).

Trends of DAG classes at baseline and after the race are reported in Figure 2. The greatest changes were attributed to the DAG 36:3 that decreased significantly post-race (25.56 ± 10.6 basal vs. 14.89 ± 8.14 post-race *p* = 0.01; 22.44 ± 8.14 24 h μmol/L). The other DAG species followed the same pattern but did not significantly change: DAG 32:0 (1.06 ± 0.26; 2.32 ± 4.6; 1.21 ± 0.71 μmol/L, baseline, post-race, 24 h, respectively); 32:1 (2.07 ± 0.71; 1.64 ± 0.94; 2.24 ± 1.62 μmol/L); 34:1 (8.22 ± 2.68; 5.85 ± 3.32; 8.66 ± 4.54 μmol/L); 34:2 (12.1 ± 4.85; 8.45 ± 4.76; 12.76 ± 9.39 μmol/L) and 36:4 (4.24 ± 4.13; 2.92 ± 2.04; 4.24 ± 3.95 μmol/L) (Figure 2).

**Figure 1.** Bar chart reporting mean and SD of total ceramides, total DAG and SM (panel **A**, **B** and **C**, respectively) at baseline, post and 24 h after the race.

**Figure 2.** Bar chart reporting the trend of DAG classes as mean and SD at baseline, post and 24 h after the race.

 Trends of ceramide species at baseline and after the race are reported in Figure 3. Ceramide species containing long chain fatty acids (18:1/16:0; 18:1/18:0) transiently increased after the race, to significantly decrease at 24 h (0.16 ± 0.06; 0.20 ± 0.05; 0.18 ± 0.04 μmol/L, baseline, post-race, 24 h, respectively); (0.18 ± 0.07; 0.18 ± 0.03; 0.16 ± 0.03 μmol/L) (Figure 3). Instead, ceramides containing very long chain fatty acids, i.e., 18:1/25:0; 18:1/26:0; 18:1/22:0; 18:1/24:1; 18:1/24:0; 18:0/24:0 significantly decreased after the race and at 24 h when compared to baseline (0.82 ± 0.21; 0.55 ± 0.13; 0.62 ± 0.15 μmol/L, baseline, post-race and 24 h, respectively); (0.16 ± 0.05; 0.11 ± 0.03; 0.13 ± 0.03 μmol/L); (1.52 ± 0.35; 1.91 ± 0.50; 1.46 ± 0.43 μmol/L); (3.29 ± 0.91; 2.36 ± 0.60; 2.61 ± 0.38 μmol/L); (5.56 ± 1.4; 1.46 ± 0.43; 4.35 ± 0.95 μmol/L) and (0.17 ± 0.05; 0.12 ± 0.04; 0.13 ± 0.38 μmol/L), Figure 3. SM did not significantly change.

**Figure 3.** Boxplots with indication of the median, interquartiles and error bars reporting the temporal profile of ceramides are given.

### *3.3. Inflammatory Levels and Race-Related Trends*

Basal plasma levels, post and after 24 h of TNFα, IL-6, fractalkine CX3CL1 and VEGF-A in all runners are reported in Table 2.


**Table 2.** Plasma values of chemokine and cytokines in runners.

Cytokines and chemokines at baseline and after the race are reported in Figure 4. IL-6 and fractalkine CX3CL1 significantly changed soon after the race, decreasing towards baseline values after 24 h (Figure 4A,B). TNFα had the same trend but the increment after the race was not significant (Figure 4C). VEGF-A tended to decrease after the race (Figure 4D), but these changes were not significant when compared to baseline. Table 3 shows the exercise-induced change from baseline, as Δ values (% change) for each variable.

**Figure 4.** Bar chart reporting mean and SD of cytokines and chemokines at baseline, (**A**) IL-6 (pg/mL), (**B**) Fractalkine (pg/mL), (**C**) TNFα (pg/mL), (**D**) VEGF-A (pg/mL), post and 24 h after the race.


**Table 3.** Exercise-induced change from baseline, Δ values (% change) in fractalkine, IL-6, TNFα and VEGF-A.

Post Δ and 24 h Δ are differences from baseline. % change at different successive time periods with respect to values observed at baseline is reported in brackets.

Interestingly, a strong correlation between basal fractalkine CX3CL1 and TNFα was found at baseline and during recovery (Rho = 0.91 *p* = 0.0017), Figure 5.

**Figure 5.** Regression of fractalkine and TNFα in runners (Rho = 0.91 *p* = 0.0017 24 h). Green dots—baseline, blue dots—post-race, red dots—24 h post-race.

### **4. Discussion**

This is the first study which assessed levels of DAGs and ceramides, and evaluated their relationship with inflammatory parameters in athletes before and after a halfmarathon race. Data evidenced the decrease of specific lipid classes, and the lack of relationship between lipids with inflammatory biomarkers, suggesting their possible contribution to exercise-related beneficial effects.

### *4.1. Demographic and Training Characteristics*

The population was similar in demographic and training characteristics; likely, for this reason, we did not observe any correlation concerning inflammatory parameters and sphingolipids. Regarding gender-related differences, it is not clear in literature if ceramides are different in males and females, and neither wheter there is a differential effect of exercise on ceramides in the two sexes. In our population, we did not observe gender-related differences for any of the biomarkers evaluated. Moreover, there is still a lack of a shared consensus on the optimal cutoffs and reference ranges of ceramides to be used. In this context, our data might suggest a lack of need to establish gender-specific reference ranges. However, considering the low number of subjects enrolled in the study, this issue merits further investigation.

Notably, values of inflammatory biomarkers of athletes at baseline did not differ from those observed in a group of 15 sedentary subjects (corresponding to 0.6 ± 3.1 pg/mL for IL-6, 164 ± 145 pg/mL for fractalkine CX3CL1, 23 ± 38 pg/mL for TNFα), except for VEGF-A, which was significantly lower in sedentary subjects (28 ± 61 pg/mL, *p* < 0.001) (unpublished data).

### *4.2. Lipids Levels and Race-Related Trends*

To the best of our knowledge, this is the first report describing half-marathon consequences on circulating sphingolipid metabolism. It is known that exercise may induce modification in plasma levels of numerous circulating biomarkers, inflammatory parameters and lipid metabolites in humans [12] Thus, these changes may affect sphingolipid metabolites as well. In recent years, plasma sphingolipids have attracted attention for their role in pathophysiology of cardiometabolic diseases. In fact, high plasma ceramides may induce endothelial dysfunction, which is closely correlated with aerobic capacity [13], and promote cell growth arrest, cytoskeleton rearrangements, senescence and death (e.g., activation of caspases), impairment of nitric oxide synthase (eNOS) activity and insulin signaling, increasing vascular permeability, oxidative stress and inflammation [14], thus

contributing to onset and development of atherosclerosis [15,16]. Accordingly, sphingolipids may represent an independent risk determinant for ischemic disease [17]. In particular, Cer(d18:1/16:0) appears to be independently correlated with the presence of more vulnerable coronary plaques in CAD patients [18]. Conversely, the Mediterranean diet shows the potential to reduce the adverse effects of high ceramide levels in the PRED-IMED trial (Prevención con Dieta Mediterranea, a prospective case-cohort study), which is a study including nearly 1000 elderly subjects at high cardiovascular risk [19].

Ceramide metabolism also appears responsive to the exercise stimulus. Better cardiopulmonary fitness correlated with low ceramide concentration in elderly coronary artery disease patients has been observed [20], whereas muscle ceramide levels decreased after chronic aerobic exercise [21]. Moreover, experimental data suggests that total content of ceramides decreased in the muscle in trained rats, contributing to the elevation of the glucose uptake observed in skeletal muscles after training [22]. Previous data suggested that 12 weeks of aerobic exercise training in obese or diabetic subjects [23] and 16 weeks of exercise in overweight/obese subjects may reduce sphingolipids (e.g., C18:0, C20:0 and C24:1) [24].

A decrease in ceramide concentrations hints at increased insulin sensitivity, which contributes, at least in part, to the beneficial exercise effects [25]. According to these previous results, we also observed a decrease in total ceramides after the race in trained subjects. Thus, all together these data suggest that sphingolipid species can represent valuable mediators of cardiovascular risk. Studies related to cardiac lipotoxicity showed that the lipotoxic species were primarily driven by ceramides and DAG and not triacylglycerol [26]. Regarding DAG species, Luukkonen et al. showed that DAG (32:1, 34:1, 36:2, 36:3) were significantly increased in subjects with high peripheral insulin resistance versus low peripheral insulin resistance subjects [27] In our study, DAG 36:3 decreased significantly after the race, indicating an improvement of lipotoxicity, attributable to that species after the acute exercise. Experimental data suggest that total content of ceramides decreases in the muscle of trained rats, contributing to the elevation of the glucose uptake observed in skeletal muscles after training [22]. In overweight, obese subjects or type 2 diabetes (T2D), there is a decrease in total ceramides, indicating that endurance training reduces total content of ceramides, likely improving glucose tolerance [22,25,27–29]. However, ceramide metabolism in patients and obese subjects could be dissimilar to that in lean, trained individuals. In fact, the increased lipid metabolism following the physical exercise reduces the substrate availability required for ceramide synthesis (e.g., palmitate, myristate). Moreover, the analysis of the lipid classes in its total assessment may be unsatisfactory, as the specific and complex relationship of different chains of fatty acids in ceramides with chronic and acute exercise demonstrates the complexity of these biological pathways, which require the assessment of increased, decreased or unchanged specific species. Accordingly, our data, for the first time, associate specific molecular species of lipids to exercise. In particular, Cer18:1/16:0 and Cer18:1/18:0 were very low at baseline in athletes and, after the race, the same ceramides increased, indicating a lower consumption of fatty acids contained in these species (palmitic and stearic acid), making these ceramides a non-preferential species for energy purposes. Mardare, Kruger et al. suggest that endurance training in mice is able to reduce long-chain fatty acids ceramides [28]. In particular, this experimental study suggests that the exercise-induced decrease of very long chain fatty acids ceramides (C24:0 and C24:1) may account, almost in part, for the reduced expression of blood inflammatory markers, as well as the increased glucose tolerance. An elevation in long chain fatty acids (C24:0, C24:1, C26:0, C25:0, C22:0) was associated with mitochondrial damage, apoptosis and cell necrosis [29]. It is not completely clear which are the mechanisms able to prevent the elevation in plasma ceramide level correlated with aerobic exercise. In addition to the possibility that lipid utilization during exercise reduces ceramide production by decreasing the availability of substrates required for ceramide synthesis, there are a number of hypothesis-generating ideas [25]; for example, aerobic exercise may accelerate ceramide degradation and clearance by increasing the expression of genes responsible for ceramide

clearance (e.g., acidic and alkaline ceramidase 1 and 3, glucosylceramide synthase and sphingosine kinase 1) [30].

### *4.3. Inflammatory Levels and Race-Related Trends*

IL-6, a pro-inflammatory cytokine, is known to increase after endurance exercise [31]. However, in this setting, IL-6 may have a role as a myokine, and as such, with antiinflammatory properties, rather than as an inflammatory facilitator [32]. In fact, the observed post-exercise elevations may be in line with exercise-related metabolic IL-6 effects (regulation of glucose homeostasis and fat oxidation) and adaptation to training [33]. Moreover, the release of IL-6 during exercise induces an increase in circulating anti-inflammatory cytokines (e.g., IL-10 and IL-1 receptor antagonist) and decrease of TNFα, or the release of cortisol from the adrenal glands, suggesting that the beneficial effects of exercise (IL-6-mediated) can be expressed, almost in part, through protection against TNFα-induced insulin resistance [34]. These responses, which probably mediate autocrine and paracrine benefits of training, are likely related to training levels, intensity and type of exercise and individual characteristics (e.g., sex and age), and thus may be different in different categories of subjects. Instead, relatively low information is known on the balance between IL-6 and other cytokines and inflammatory biomarkers under such conditions. TNFα did not increase significantly in our population, and this may be the effect of the anti-inflammatory cytokine cascade, which may oppose TNFα increase, giving protection against TNFα-induced damage, as previously observed [35].

CX3CL1 exists in two forms; one form anchors to the membrane, acting as an adhesion molecule, whereas the other form acts as a soluble chemoattractant. CX3CL1 A acts in acute skeletal muscle damage and regeneration through recruitment of macrophages and other immune cells involved in repair and growth of skeletal muscle, influencing the adaptive response to exercise [36,37]. As in our population, previous data suggested an increase of CX3CL1 related to exercise (e.g., cycling) [38]. The variation of circulating CX3CL1 is closely related to changes in muscle gene expression, and as such, might have significance in the adaptive response to exercise [38,39]. Moreover, local CX3CL1 synthesis and expression depend on many factors, including inflammatory cytokines (e.g., TNFα), giving evidence of the relationship that we observed between these two biomarkers [40]. Experimental data suggested that CX3CL1 stimulation of monocytes is associated with a marked increase of TNFα, which is known to stimulate satellite cell proliferation [41,42]. Accordingly, we observed a strong correlation between fractalkine and TNFα, suggesting that this effect could represent an indirect mechanism by which CX3CL1 acts as a mitogen for muscle cells. Moreover, CX3CL1 seems to have beneficial effects in muscle regeneration through a direct effect on myogenic cells [43]. Thus, CX3CL1 likely induces the monocytic and myogenic expression of different factors known to increase in human skeletal muscle after exercise, with a role in the adaptive response following an exercise burst.

Interestingly, in our population, inflammatory biomarkers did not influence the relationships between sphingolipids and exercise, not supporting the existence of a link between inflammation and sphingolipids in half-marathon response, although further studies are needed to verify this possibility. In this context, we also did not observe any relationship between total and species of SM, ceramides and DAG and reactive oxygen species (ROM, a biomarker of oxidative stress) (unshown data) [44].

### *4.4. Strengths and Limitations*

This study has limitations. First, the number of athletes enrolled is not high. However, a strength is that studied athletes were very similar according to training characteristics (day/week of training, Km/week, years of training). Moreover, each subject served as a control for him/herself, increasing the statistical power, and reducing the effects of confounding factors.

Second, insulin was not directly assessed in these subjects. This important issue merits further investigation in future studies to understand the link between species of lipids, inflammatory markers and insulin resistance (IR) during exercise. However, all our subjects had no history of IR and type 2 diabetes (T2D) and they showed normal body mass index; therefore, this relationship could be negligible. Nonetheless, it will be interesting to assess exercise effects on lipid-related biomarkers in IR, T2D or obese subjects in future studies.

### **5. Conclusions**

IL-6 and CX3CL1 transiently increased soon after the half-marathon. As they are promptly reversible, these changes might represent a physiological response to acute exercise rather than a damage-related response. The decrease in plasma ceramide concentration observed after the race and the lack of their relationship with inflammatory mediators suggested the involvement of lipids in exercise adaptation, as well as their possible role in mechanisms underlying beneficial effects of regular physical activity on disease prevention, especially regarding cardiometabolic disease. This opens up a new area of investigation for future research to establish whether the measurement of specific plasma ceramides could provide new biomarkers useful to assess exercise adaptation and to evaluate specific exercise interventions in different categories of healthy subjects and patients.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/app11104622/s1.

**Author Contributions:** All authors discussed the results and contributed to the final manuscript. M.G., C.V. and A.P.: study concept; L.S.: certified English revision; F.S., F.C., M.M., E.B., F.M. and L.S.: acquisition, measurement and analysis of data; M.G., C.V., A.P. and A.G.: interpretation of data; M.G. and C.V.: drafting of the manuscript. All authors contributed to the manuscript's intellectual content and gave approval for the final version. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the Pisa Ethics Committee, Italy, (protocol number for study acceptance 2805).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Data are available from authors upon reasonable request.

**Acknowledgments:** The authors acknowledge the athletes of "Gruppo Podistico Rossini (www.podisticarossini.it accessed on 15 May 2021)" for the fundamental contribution.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Applied Sciences* Editorial Office E-mail: applsci@mdpi.com www.mdpi.com/journal/applsci

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18

www.mdpi.com

ISBN 978-3-0365-2503-7