**Predictors of Discordance in the Assessment of Skeletal Muscle Mass between Computed Tomography and Bioimpedance Analysis**

**Min Ho Jo 1, Tae Seop Lim 1,2, Mi Young Jeon 1,2, Hye Won Lee 1,2, Beom Kyung Kim 1,2,3, Jun Yong Park 1,2,3, Do Young Kim 1,2,3, Sang Hoon Ahn 1,2,3, Kwang-Hyub Han 1,2,3 and Seung Up Kim 1,2,3,\***


Received: 20 February 2019; Accepted: 1 March 2019; Published: 7 March 2019

**Abstract:** Computed tomography (CT) and bioimpedance analysis (BIA) can assess skeletal muscle mass (SMM). Our objective was to identify the predictors of discordance between CT and BIA in assessing SMM. Participants who received a comprehensive medical health check-up between 2010 and 2018 were recruited. The CT and BIA-based diagnostic criteria for low SMM are as follows: Defined CT cutoff values (lumbar skeletal muscle index (LSMI) <1 standard deviation (SD) and means of 46.12 cm2/m2 for men and 34.18 cm2/m2 for women) and defined BIA cutoff values (appendicular skeletal muscle/height2 <7.0 kg/m2 for men and <5.7 kg/m2 for women). A total of 1163 subjects were selected. The crude and body mass index (BMI)-adjusted SMM assessed by CT were significantly associated with those assessed by BIA (correlation coefficient = 0.78 and 0.68, respectively; *p* < 0.001). The prevalence of low SMM was 15.1% by CT and 16.4% by BIA. Low SMM diagnosed by CT was significantly associated with advanced age, female gender, and lower serum albumin level, whereas low SMM diagnosed by BIA was significantly associated with advanced age, female gender, and lower BMI (all *p* < 0.05). Upon multivariate analysis, age >65 years, female and BMI <25 kg/m2 had significantly higher risks of discordance than their counterparts (all *p* < 0.05). We found a significant association between SMM assessed by CT and BIA. SMM assessment using CT and BIA should be interpreted cautiously in older adults (>65 years of age), female and BMI <25 kg/m2.

**Keywords:** sarcopenia; bioimpedance analysis; computed tomography; discordance

#### **1. Introduction**

Sarcopenia is a syndrome characterized by the loss of skeletal muscle mass, strength, and performance [1–3] that results in an increased risk of fracture, dysfunction, reduced quality of life, and increased mortality [4,5]. Due to the varying diagnostic cutoff values for muscle mass and the varying diagnostic tools used in previous studies, the reported prevalence of sarcopenia has been inconsistent [6,7]. Several studies of sarcopenia have been performed, and multiple guidelines have been proposed; these have enhanced our knowledge of the condition. Sarcopenia is now officially recognized as a disorder in some countries, with an ICD-10-MC diagnostic code [8].

The measurement of skeletal muscle mass (SMM) is of paramount importance to diagnose sarcopenia. Several imaging techniques that can assess SMM are currently available: Dual-energy X-ray absorptiometry (DXA), computed tomography (CT), magnetic resonance imaging (MRI), and bioimpedance analysis (BIA) [2]. DXA has several advantages over the other methods, such as safety, accuracy, and non-invasiveness, but it can overestimate muscle mass in cases of muscle edema or intramuscular fat deposition [9]. CT can accurately measure the quantity and quality of SMM, but it is costly and exposes the patient to radiation [2,10]. MRI has no radiation exposure for the patient and accurately measures SMM, but its clinical application is significantly limited due to its high cost [11]. BIA has been recognized as a rapid, inexpensive, portable, and safe methodology but, because BIA measures the resistance to a current that is applied through a body of water, the assessment of SMM may be inaccurate if the patients are dehydrated, overhydrated fluid status or obese [12]. BIA tends to overestimate SMM because it cannot discriminate among appendicular, non-appendicular fat, and non-fat mass [13].

To date, measurement of SMM by BIA has typically been performed using the Kyle, Jassen, Ergi, and Scafoglieri prediction models [14–17]. SMM can now be assessed directly using vertical, eight-point analyzers. Several studies have reported the accuracy and reproducibility of direct segmental multi-frequency BIA and the strong correlation between its results and SMM measured by DXA [18–23]. CT provides an accurate measurement of SMM, with a significant correlation to whole-body muscle mass [24,25]. Accordingly, CT has been considered to be the gold standard for measuring SMM [2,10]. Despite its drawbacks, BIA is more easily applied in clinical practice. Thus, investigating the prevalence of discordance in the assessment of SMM between CT and BIA and identifying predictors of this discordance is valuable. This investigation can ultimately help physicians select the optimal candidates for each modality to diagnose low SMM and interpret the results appropriately.

The primary aim of this study was to identify predictors of discordance between SMM measured by BIA and by CT. The secondary aims were to investigate the prevalence of low SMM by CT and BIA and the correlation between SMM measured by CT and BIA in apparently healthy subjects undergoing comprehensive medical health check-ups.

#### **2. Methods**

#### *2.1. Study Subjects*

A total of 1191 subjects who visited the health promotion center in Severance Hospital, a university-affiliated tertiary care hospital, for a comprehensive medical health check-up from June 2010 to April 2018 were included, see Figure 1. Severance Hospital is a 2000-bed academic referral hospital in Northwestern Seoul, Republic of Korea. Severance Hospital is supported by Yonsei University College of Medicine. Exclusion criteria were as follows: (1) no BIA data, (2) limited access to BIA data due to personal privacy, (3) poor CT quality, and (4) major operation in the lumbar area.

The study's protocol adhered to the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of Severance Hospital. Informed consents were waived due to the retrospective nature of the study.

**Figure 1.** Correlation between SMM assessed by CT and BIA. Crude (**A**) and BMI-adjusted SMM (**B**) assessed by CT were significantly correlated with those by BIA (all *p* < 0.001, correlation coefficient = 0.898 and 0.858, respectively). The correlation between crude SMM assessed by CT and BIA was significant in men and women (**C**) (all *p* < 0.001, correlation coefficient = 0.724 in men and 0.645 in women, respectively). LSMA (cm2) = <sup>−</sup>4.366 + 6.920 \* ASM (kg), Standard error = 0.099 LSMI adjusted by BMI = 0.212 + 6.424 \* (ASM adjusted by BMI), Standard error = 0.113. SMM, skeletal muscle mass; LSMA, lumbar skeletal mass area; AMS, appendicular skeletal mass; CT, computed tomography; BIA, bioimpedance analysis; BMI, body mass index; LMSI, lumbar skeletal muscle index. Regression equations and standard error are as follows.

#### *2.2. Data Collection*

A medical health check-up was performed, and collected data included age, gender, height, body weight, body mass index (BMI), and laboratory test results. Histories of hypertension, diabetes, and viral hepatitis were collected from the medical record and individual questionnaires.

#### *2.3. Fibrosis-4 Index Calculation*

Recent studies have shown that fibrotic burden in the liver is independently associated with sarcopenia. Therefore, the fibrosis-4 index (FIB-4) was calculated using the following formula: Age (years) × aspartate aminotransferase (AST) (U/L)/(platelets (109/L) × alanine aminotransferase (ALT) (U/L])1/2 [26].

#### *2.4. Measurements of Skeletal Muscle Area*

Skeletal muscle area was measured at the mid-body level of the L3 vertebra in a supine position by a dual-source 128-slice CT scanner (Somatom Definition Flash, Siemens Healthcare, Forchheim, Germany), a 64-slice CT scanner (Somatom Sensation 64, Siemens Healthcare), a Discovery 710 PET-CT 128-slice scanner (General Electric Medical Systems, Milwaukee, WI, USA), a Biograph

TruePoint 40 PET-CT 40-slice scanner (Siemens Medical Solutions, Hoffman Estates, IL, USA), or a Discovery 600 PET-CT 16-slice scanner (General Electric Medical Systems, Milwaukee, WI, USA). The muscle area was identified using attenuation values between −29 to −150 Hounsfield units. Total lumbar skeletal muscle area (psoas, erector spinae, quadratus lumborum, transversus abdominus, external and internal obliques, and rectus abdominus) (cm2) was defined as a region with density ranging from −29 to −150 Hounsfield units using Aquarius Intuition Viewer software, version 4.4.12 (Terarecon, San Mateo, CA, USA). Boundaries were corrected manually, as necessary. To minimize measurement error, the CT instruments are periodically tested and calibrated for spatial resolution, length measurement, alignment, and linearity of attenuation (CT number) using a standard phantom (AAPM CT Performance Phantom, 76-410). All tests are performed in compliance with the regulations of the Korean Institute for Accreditation of Medical Imaging. The lumbar skeletal muscle index (LSMI) was defined as 10,000 × lumbar skeletal muscle area (LSMA, cm2)/height2 (m2). Based on previous studies [2,10], we assumed that measurement of SMM by CT is more accurate.

The InBody 770 (Biospace Co., Seoul, Korea) measured body composition. Participants fasted for 12 h prior to testing. Participants wore a t-shirt and short pants on the day of testing, and provided their age, gender, and height at the time of measurement. Testing was conducted according to the manufacturer's instructions. Data were uploaded to the electronic medical record. The measurement was comprised of two combinations: z-axis at frequencies of 1, 5, 50, 250, and 500 kHz for impedance, and x-axis at frequencies of 5, 50, and 250 kHz for reactance. Impedance was measured for five body segments: Trunk, right and left arms, and right and left legs. We reviewed the medical records and measured appendicular skeletal muscle (ASM) (kg) through direct segmental multi-frequency BIA [27].

Muscle mass was determined by measuring electrical resistance [28] using four surface tactile electrodes placed on the dorsal surface of the hand and foot. Whole-body resistance (Rsumx) was calculated by summing the segmental resistances at frequency x, according to the following equation:

$$R\_{\text{sum}\infty} = R\_{\text{RA}} + R\_{\text{LA}} + R\_{\text{T}} + R\_{\text{RL}} + R\_{\text{LL}} \tag{1}$$

The index of Rsumx (RIsumx) is calculated by using the following equation:

$$\text{RL}\_{\text{sum}\,\text{x}} \text{Height (cm)}^2 / \text{R}\_{\text{sum}\,\text{x}} \text{ (\text{\textdegree})}\tag{2}$$

Appendicular muscle mass = 0.236 × Height2/RRA + 0.0109 × Hright2/RT + 0.121 × Hright2/RRL +1.554 (3)

Using the formula above, the muscle mass is automatically calculated in InBody.

#### *2.5. Definition of Low SMM*

The CT diagnostic criterion for low SMM was a lumbar skeletal mass index (LSMI) <1 standard deviation (SD) below the sex-specific mean of the study group. The BIA diagnostic criterion for low SMM was adopted from the Asian Working Group of Sarcopenia [6]: ASM/height2 <7.0 kg/m2 for men and <5.7 kg/m<sup>2</sup> for women [29].

We also used additional CT and BIA diagnostic criteria for low SMM. The additional criterion for CT was an LSMI ≤52.4 cm2/m2 for men and ≤38.5 cm2/m2 for women [30]. The additional criterion for BIA was adopted from The Foundation for the National Institutes of Health: ASM/BMI <0.79 for men and <0.51 for women [7]. We attached the relevant analysis using Supplementary Data.

#### *2.6. Statistical Analysis*

Statistical analyses were performed using Statistical Package for the Social Science (SPSS) version 23.0 for Windows (IBM Corp., Armonk, NY, USA). Continuous and categorical variables were expressed as mean ± standard deviation and *n* (%), respectively. *p*-Value < 0.05 was considered statistically significant. Simple and partial correlation analyses were used to analyze the relationship between CT and BIA muscle mass. The distribution between muscle mass by BIA and quartile stratification of muscle mass by CT was evaluated using the Mann-Whitney U test. The comparison between subjects with and without low SMM was performed using the chi-square test for categorical variables and Student's t-test for continuous variables. Multivariate analysis using binary logistic regression analysis was performed on variables that showed a *p*-value <0.05 and was used to determine the predictors of discordance in defining low SMM between CT and BIA.

#### **3. Results**

#### *3.1. Patients*

A total of 1191 subjects who underwent a comprehensive medical health check-up were considered eligible. However, 19 subjects were excluded due to a lack of BIA data, and an additional nine subjects were excluded due to poor-quality CT scans and a history of a major operations around the lumbar or appendicular skeletal muscle area. As a result, 1163 subjects were included in the statistical analysis, see Supplementary Figure S1.

Baseline characteristics of the study population (641 men and 521 women) are summarized in Table 1. The mean age of the patients was 57 years; 41.0% were over 60 years of age. The mean BMI of the patients was 24.0 kg/m2. Of the study population, 41.0% of subjects (*n* = 488) had hypertension, 29.4% (*n* = 314) had diabetes, and 4.9% (*n* = 57) had viral hepatitis. Using CT scans, the mean whole-body fat-free mass and LSMI were 45.3 kg and 46.9 cm2/m2, respectively. Using BIA, the mean ASM, ASM index, and ASM/BMI ratio were 20.1 kg, 7.1 kg/m2, and 0.82, respectively. The mean FIB-4 was 1.17.


**Table 1.** Baseline characteristics (*n* = 1163).

Variables are expressed as median (interquartile range) or *n* (%). INR, international normalized ratio; ASM, appendicular skeletal muscle mass; ASMI, appendicular skeletal mass index.

#### *3.2. Association between SMM Assessed Using CT and BIA*

The crude and BMI-adjusted SMM assessed by CT were significantly associated with those assessed by BIA (*p* < 0.001, correlation coefficient = 0.898 for crude SMM; *p* < 0.001, correlation coefficient = 0.858 for BMI-adjusted SMM), see Figure 1A. The association between crude SMM assessed by CT and BIA was statistically significant, regardless of gender (*p* < 0.001, correlation coefficient = 0.724 in men; *p* < 0.001, correlation coefficient = 0.645 in women), as shown in Figure 1. Linear regression results comparing CT and BIA assessed SMM were added to the Supplementary Table S6.

We divided the patients into four groups according to quartiles of SMM assessed by CT and BIA. SMM as assessed by BIA significantly increased according to the CT-assessed SMM quartile (*p* < 0.001), see Supplementary Figure S2.

#### *3.3. Comparison between Subjects with and without Low SMM Assessed by CT*

The baseline characteristics of subjects with and without CT-defined low SMM in Table 2. The cutoff value of low SMM was defined as less than one standard deviations sex-specific mean value of the participants. The sex-specific cut-off values of LSMI were 46.12 cm2/m2 in men and 34.18 c 2/m2 in women.


**Table 2.** Comparison between subjects with and without low SMM assessed by CT.

Variables are expressed as median (interquartile range) or *n* (%). SMM, skeletal muscle mass; CT, computed tomography; INR, international normalized ratio; ASM, appendicular skeletal muscle mass; ASMI, appendicular skeletal mass index. \* CT cutoff indicates <1 standard deviation (SD), sex–specific mean value of the participants.

When CT-defined cutoff values were used, subjects with low SMM were significantly older (median 63 vs. 57 years) and female gender (48.8% vs. 44.1%). Subjects with low SMM had significantly lower serum albumin levels (median 4.2 vs. 4.3 mg/dL), lower total cholesterol (median 177 vs. 188 mg/dL), higher high-density lipoprotein (HDL) cholesterol (median 50 vs. 48 mg/dL) and lower low-density lipoprotein (LDL) cholesterol (median 100 vs. 111 mg/dL) than those of subjects without low SMM (all *p* < 0.05). In addition, various muscle indexes were unfavorable in subjects with CT-defined low SMM.

We also analyzed additional diagnostic criteria for low SMM defined by CT (≤52.4 cm2/m2 for men and ≤ 38.5 cm2/m2 for women), see Supplementary Table S1.

#### *3.4. Comparison between Subjects with and without Low SMM Assessed by BIA*

The baseline characteristics of subjects with and without BIA-defined low SMM are shown in Table 3. The cutoff value of low SMM was defined previous study [6]. The Asian Working Group of Sarcopenia defined cutoff values appendicular lean mass (ALM)/height2 of <7.0 kg/m<sup>2</sup> in men and <5.7 kg/m<sup>2</sup> in women.


**Table 3.** Comparison between subjects with and without low SMM assessed by BIA.

Variables are expressed as median (interquartile range) or *n* (%). BIA Cutoff indicates AWGS index. SMM, skeletal muscle mass; BIA, bioimpedance analysis; INR, international normalized ratio; FNIH, The Foundation for the National Institutes of Health; ALM, appendicular lean mass; ASMI, appendicular skeletal mass index; BMI, body mass index; AWGS, Asian Working Group of Sarcopenia.

When BIA-defined cutoff values were used, subjects with low SMM were significantly older (median 60 vs. 57 years) and had a higher proportion of female subjects (67.5% vs. 40.0%), lower BMI (median 21.8 vs. 24.2 kg/m2) (all *p* < 0.05). In addition, various muscle indices were unfavorable in subjects with BIA-defined low SMM.

We also analyzed additional diagnostic criteria for low SMM defined by BIA (ALM/BMI <0.79 for men and <0.51 for women), see Supplementary Table S2.

#### *3.5. Prevalence and Predictors of Discordance in Defining Low SMM Assessed by CT and BIA*

The proportion of non-discordant and discordant subjects, when different measuring methods were applied (CT vs. BIA), is described in Table 4. The proportion of subjects without low SMM by both CT and BIA was 72.3%, and that of subjects with low SMM ranged was 3.9%. The overall proportion of non-discordant subjects was 76.2%. The results of analysis using additional diagnostic criteria for low SMM are given in Supplementary Table S3.


**Table 4.** Distribution of subjects with and without low SMM assessed by CT and BIA.

Variables are expressed as *n* (%). \* BIA cutoff indicates AWGS index (ASMI, ALM/height2) of <7.0 kg/m2 in men and <5.7 kg/m2. \*\* CT cutoff indicates <1 SD, sex-specific mean value of the participants. SMM, skeletal muscle mass; CT, computed tomography; BIA, bioimpedance analysis; FNIH, The Foundation for the National Institutes of Health; ALM, appendicular lean mass; ASMI, appendicular skeletal mass index; BMI, body mass index; AWGS, Asian Working Group of Sarcopenia.

To identify the predictors of discordant results by CT and BIA, univariate analysis was performed, see Table 5. Older age (HR = 1.05), female sex (HR = 1.48), lower BMI (HR = 0.73), lower serum albumin level (HR = 0.58), and higher GGT (HR = 1.01) were significantly predictive of discordance between CTand BIA-defined low SMM (*p* < 0.05). The results of analyses using the additional diagnostic criteria for low SMM are listed in Supplementary Table S4. The results of basic demographic characteristics, specificity and sensitivity of the low SMM defined by the BIA compared to the low SMM defined by the CT as diagnostic standard criteria, added to Supplementary Table S5.

Among selected independent predictors of the presence of discordance, age, female gender, and BMI were selected for multivariate analysis. Thus, we stratified our study population into two groups according to these three independent variables to check the prevalence of discordance, see Figure 2. Older age (>65 years) (22.3% vs. 12.2%), female gender (20.9% vs. 9.8%), and lower BMI (<25 kg/m2) (20.1%% vs. 3.5%) had a significantly higher risk of discordance than the counterparts (all *p* < 0.001). The results of analysis using additional diagnostic criteria for low SMM are given in Supplementary Figure S3.


CT cutoff indicates <1 SD, sex-specific mean value of the participants. BIA cutoff indicates AWGS index (ASMI, ALM/height2) of <7.0 kg/m2 in men and <5.7 kg/m2. SMM, skeletal muscle mass; CT, computed tomography; BIA, bioimpedance analysis; OR, odds ratio; CI, confidence interval; INR, international normalized ratio; FNIH, The Foundation for the National Institutes of Health; ALM, appendicular lean mass; ASMI, appendicular skeletal mass index; BMI, body mass index; AWGS, Asian Working Group of Sarcopenia.

**Figure 2.** Percentage of subjects with non-discordance and those with discordance in diagnosing low SMM using CT and BIA according to identified independent predictors. Participants with age > 65 years, female gender and BMI < 25 kg/m<sup>2</sup> had a significantly higher proportion of discordance than the counterparts (all *p* < 0.001). CT cutoff indicates < 1 SD. BIA cutoff indicates AWGS index (ASMI, ALM/height2) of <7.0 kg/m2 in men and <5.7 kg/m2. BMI, body mass index; ALM, appendicular lean mass; ASMI, appendicular skeletal mass index; AWGS, Asian Working Group of Sarcopenia.

#### **4. Discussion**

The diagnostic criteria for sarcopenia have not yet been definitively established, even though it is one of the most important public health concerns [31]. Varying diagnostic criteria for sarcopenia based on several assessment modalities, which include CT and BIA, are available [2,3,10], and the criteria are different between Asian and Western countries [6,7,32]. Ethnicity is an important factor for the diagnosis of sarcopenia [33]. Several recent research groups have published diagnostic guidelines for sarcopenia, which have emphasized the importance of ethnicity [34,35]. The BIA and CT diagnostic criteria differ according to ethnicity [3,6,7,35]. According to our knowledge, no comparison of BIA and SMM measured by CT at the L3 level has been performed. Therefore, our findings will facilitate the establishment of diagnostic cutoff values for Asian patients.

Our data show a significant association in crude and BMI-adjusted SMM assessed by CT and BIA, although the area assessed was different for each method. Similar to previous studies [3,11,32,36–39], the proportion of subjects with low SMM in our study varied from 15.1% to 16.4% when CT or BIA was used to assess SMM, and the risk factors for discordant results between the methodologies were advanced age, female gender, and lower BMI.

We believe the identified risk factors for discordant results can be explained in several ways: Total fat mass tends to be higher in older adults [40], and BIA can overestimate SMM when the subject has a high fat mass [41]; assessment of SMM using BIA can be overestimated in female subjects who have a higher probability of increased body fat [41]; and there is a weaker correlation between SMM in the limb and L3 area among subjects with a lower BMI [42,43]. All of these factors suggest that CT may be required for a more accurate assessment of SMM in subjects with advanced age, female gender, and low BMI.

Our study has several strengths. First, the overall sample size was over 1100, which ensures the statistical power and precision of our results. We adopted several cutoff values for CT and BIA when defining low SMM, and we found the three factors of age, gender, and BMI to be associated with discordance between CT- and BIA-based SMM assessments. Second, we focused on the general population instead of medically vulnerable subjects, such as only older adults, or those with liver cirrhosis or cancers for whom sarcopenia already showed clinical implications. Similar to our study, several recent studies proved the clinical significance of assessing sarcopenia in the general population and non-alcoholic fatty liver disease (NAFLD) subjects [44,45]. Thus, our study provides information that helps to identify optimal subjects for CT-based assessment of sarcopenia. Third, in contrast to most previous studies [7,11,27,46], we adopted several cutoff values for SMM assessed by CT and BIA. Although the predictors of discordance were not exactly the same, we obtained relatively consistent results regardless of the cutoff value used. Fourth, several studies have compared DXA and BIA, but few have directly compared CT and BIA to assess SMM [46,47]. In our study, SMM using CT and BIA was measured on the same day, in contrast to most previous studies [9,25,46]. As a result, any bias caused by different time points of SMM assessments may have been prevented. Lastly, because there are significant differences in SMM between Western and Asian populations, focusing on a single ethnicity may be important. Thus, the results of our study could be optimized for an Asian population.

Several issues remain unresolved in our study. First, although we adopted several known cutoff values for CT and BIA, the results of our study should be further validated based on existing diagnostic criteria for sarcopenia. Second, recent studies have insisted that other factors, such as muscle strength and walking speed, should be considered when diagnosing sarcopenia. However, our study was retrospectively performed based on the clinical information of the subjects who underwent a comprehensive medical health check-up, and we only used SMM to define sarcopenia. Further studies with additional markers of sarcopenia should validate our results. Third, our study only included subjects who were willing to receive and could afford a medical health check-up. The prevalence of hypertension (29.1%) and diabetes (11.3%) in the general Korean population in 2016 (Korean Center for Disease Control and Prevention; Ministry of Health and Welfare) [48], were lower than those in this study (41.0% and 26.4%, respectively). In Korea, individuals >40 years of age are eligible for basic health check-ups; those with chronic diseases such as hypertension and diabetes receive health checkups more frequently. The mean age of our patients was 57 years, and 40.9% were >60 years of age. Because of this potential selection bias, our results may not be fully applicable to the general population, but this can be resolved in future studies. Fourth, SMM measured by BIA is affected by the hydration status [12]. The patients were admitted to the health check-up unit and fasted overnight. Thus, the hydration status of all of the patients should have been similar. Lastly, when discordant results between CT- and BIA-based SMM assessments were obtained, we did not know which diagnostic modality to accept. For patients with discordant CT and BIA results, it is important to decide which results should be used. However, a definitive diagnostic method for sarcopenia has not been established. This issue should be explored in future longitudinal follow-up studies that use solid end-points, such as mortality. This issue should be explored in future longitudinal follow-up studies that use solid end-points such as mortality, which might propose the right direction toward CT or BIA.

In conclusion, the significant association between CT and BIA for SMM assessment suggests that BIA could be used to assess sarcopenia in clinical practice. However, because advanced age, female gender, and low BMI were risk factors for discordant results between CT and BIA, BIA assessment should be interpreted cautiously in subjects with these risk factors and, if possible, CT or other modalities should be considered as an alternative diagnostic tool to assess SMM to define sarcopenia.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2077-0383/8/3/322/s1, Supplementary Figure S1: Flow chart depicting selection of the study population. Supplementary Figure S2: Distribution of skeletal muscle mass assessed by BIA according to quartile stratification of skeletal muscle mass assessed by CT. Supplementary Figure S3. Percentage of subjects with non-discordance and those with discordance in diagnosing low skeletal muscle mass assessed using CT and BIA according to identified independent predictors. Participants with Age>65 years (A), female gender (B) and BMI<25 kg/m2 (C). Supplementary Table S1: Comparison between subjects with and without low skeletal muscle mass assessed by CT. Supplementary Table S2: Comparison between subjects with and without low skeletal muscle mass assessed by BIA. Supplementary Table S3: Distribution of subjects with and without low skeletal muscle mass according to CT and BIA. Supplementary Table S4: Predictors of discordance in defining low skeletal muscle mass assessed by CT and BIA Supplementary Table S5: Basic demographic characteristics, specificity and sensitivity of the low SMM defined by the BIA compared to the low SMM defined by the CT as diagnostic standard criteria. Supplementary Table S1: Linear regression results of comparing SMM evaluated by CT and BIA.

**Author Contributions:** Conceptualization, M.H.J. and S.U.K.; methodology, M.H.J., T.S.L., M.Y.J. and S.U.K.; participation in patient management and data collection; H.W.L., B.K.K., J.Y.P., D.Y.K., S.H.A., K.-H.H. and S.U.K.; contribution to the data acquisition, responsibility for writing the paper, and statistical analysis: M.H.J. and S.U.K.; All authors reviewed the paper and approved the final version.

**Acknowledgments:** This study was supported by the Liver Cirrhosis Clinical Research Center, in part by a grant from the Korea Healthcare technology R & D project, Ministry of Health and Welfare, Republic of Korea (no. HI10C2020). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

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

#### **References**


© 2019 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* **Abuse of Licit and Illicit Psychoactive Substances in the Workplace: Medical, Toxicological, and Forensic Aspects**

### **Ricardo Jorge Dinis-Oliveira 1,2,3,\* and Teresa Magalhães 1,2,4,\***


Received: 27 January 2020; Accepted: 10 March 2020; Published: 12 March 2020

**Abstract:** About one-third of adult life is spent in the workplace. The use of psychoactive substances is a major preventable cause of morbidity and mortality. The consumption of psychoactive substances during or outside working hours greatly increases the frequency and severity of labor accidents, as well as the workers' poor general state of health and productivity, implying higher costs for enterprises. It is the responsibility of organizations to ensure the safety and health of their workers. These cannot be limited to traditional routine clinical exams, as other aspects also have an impact on health. Thus, prevention and intervention in the consumption of psychoactive substances (e.g., ethanol, opioids, central nervous system stimulants or depressants, hallucinogens, *Cannabis* derivatives, dissociative substances, and inhalants) in labor activity should be considered as an investment of organizations and not as a cost, in view of the professional, personal, and family advantages for workers and employers, with a potential impact on productivity, security, health, and quality of life at work. Despite the extensive literature on the subject, each article generally focuses on one or another aspect of a very specific nature, not tackling the problem in a holistic way by confronting clinical, safety, and legal issues. This article presents a reflection on the legal, laboratorial, clinical, ethical, forensic, and safety concerns related to the consumption of psychoactive substances in the workplace, and can be a cross-cutting contribution to occupational medicine, forensic medicine, and insurance medicine, as well as for entrepreneurs, lawyers, judges, workers, and technicians from the public and private sectors that develop projects in this area. This discussion is based on general principles established internationally and highlights the role of the occupational healthcare system and other decision-making actors in the prevention and supervision of workplace psychoactive consumption.

**Keywords:** occupational medicine; forensic medicine; insurance medicine; psychoactive substances; safety; clinical; forensic; law; ethics

#### **1. Introduction**

Workplaces reflect, to a certain extent, the widespread presence of ethanol and other psychoactive substances in society and the type of work, but sometimes both. There has been global growth in the use of workplace drug testing as a response to drug-related risks to safety and productivity [1,2]. Depending on the countries, the labor sectors, and the professions, statistics suggest high percentages of workers suffer from or are at risk of becoming dependent on ethanol [3]. Civil construction workers,

transport workers, hotel staff, barmen, catering workers, farmers, and workers in the primary sectors are particularly affected by this problem, especially in terms of ethanol [3]. Additionally, ethanol-related indicators are often inversely proportional to the level of literacy, meaning that the higher this level is, the less likely workers are to become drunk [4]. However, in some countries, excessive ethanol or other substance consumption is clearly observed among some higher professional groups such as doctors and managers [5,6]. Regarding the consumption of other psychoactive substances in the workplace, the information is scarcer, with *Cannabis* being more prevalent among younger workers, whereas cocaine is more prevalent among highly skilled workers such as managers [7]. Indeed, because workplace drug testing is often performed on-site by occupational physicians, a global statistic spectrum is hard to obtain [8].

The consumption of psychoactive substances at work depends on the combination of multiple factors, some of them linked to individual characteristics and lifestyles, and others being of a professional nature, related for example to the typology of work, rhythms and cadences, shift work, and stress, among others [9,10].

It is obvious that the prevention and deterrence of problems associated with the use of psychoactive substances should be a global intervention involving the participation of all decision-making actors in the organization bodies, namely, occupational medicine, occupational safety and health, human resources, social action services, intermediate and direct leadership, workers' representatives, and workers themselves. Detection may or may not be part of the organization's health and safety policy procedures. In order to create a program that contemplates psychoactive substance testing, the underlying policy, objectives, and rights and responsibilities of all parties involved should be made explicit. Moreover, the analysis of results and the relationship with the worker must respect confidentiality, cooperation, mutual commitment, and capacity building. The final goals are beneficial both for employers and employees and include the increase of workplace productivity by reducing absenteeism, presenteeism, and workplace accidents, as well as raising awareness of the toxic effects of psychoactive substances and their consequences in workers' performance by adopting a more healthily behavior during and outside working hours [9,11,12].

This article presents a reflection on the legal, laboratorial, clinical, ethical, forensic, and safety concerns related to the consumption of psychoactive substances in the workplace, and can be a cross-cutting contribution to occupational medicine for businessmen, lawyers, juries, workers, and laboratory technicians that act in this area. We also aimed to discuss the best practices to be followed by laboratories providing workplace drug testing services.

#### **2. Methods**

A narrative review was performed by searching articles in English, French, Spanish, and Portuguese in PubMed, Scopus, Web of Science, and PsycINFO concerning legal, laboratorial, clinical, ethical, forensic, and safety concerns of workplace drug testing, without time limit. Besides these inclusion criteria, additional reports were obtained from the references of the articles identified in the original search. International reports from the World Health Organization; European Union; International Commission on Occupational Health; and legislation on psychoactive substance workplace consumption, testing, and prevention were also reviewed. Specifically, Portuguese legislation, under the European Union regulation, was used as a starting point and, regardless the jurisdiction, broad and transversal aspects general and useful for occupational medicine were discussed. The effects in workplace performance of ethanol, opioids, central nervous system stimulants or depressants, hallucinogens, *Cannabis* derivatives, dissociative substances, inhalants, and examples of new psychoactive substances were also reviewed. From this search, 192 documents were obtained and 105 were ultimately considered for the final version of the manuscript. The remaining documents were excluded if they focused on specific points of certain enterprises or on a very specific point of a country not considered useful for a broad discussion, or when those aspects do not fit well in the scope of occupational medicine, such as technical and analytical details of toxicological methods.

#### **3. Psychoactive Substances and Occupational Risks**

Psychoactive substances are those that act mainly on the central nervous system, where they alter brain function, resulting in temporary changes in perception, mood, consciousness, and behavior. The effect depends not only on the specific substance but also on the dose, previous consumption, tolerance, comorbidities, mixture with other substances, and route of administration, among other reasons. In a general overview, all psychoactive substances have, to a higher or lesser extent, a dysfunctional effect on work capability. Considering the clinical usefulness and their most important effects at the level of the central nervous systems at usual doses, psychoactive substances are classified into several groups, which are discussed below regarding their repercussions in terms of workplace performance.

**Opioids** are mainly used as analgesics and antitussives [13–17]. Physical work and high psychosocial work demands, excessive repetition of tasks, awkward postures, and heavy lifting are all known workplace risk factors for musculoskeletal pain and consequently lead to administration of opioids, according to prospective studies [18,19]. Interestingly, a recent randomized trial demonstrated that treatment with opioids was not better to treatment with nonopioid medications for improving pain-related function over 12 months in moderate to severe chronic back pain or hip or knee osteoarthritis pain [20]. Part of this group are natural compounds extracted from opium such as morphine, codeine, and thebaine; semisynthetic compounds such as heroin, oxycodone, hydrocodone, oxymorphone, etorphine, and hydromorphone; and the synthetics tramadol, tapentadol, meperidine (or pethidine), methadone, fentanyl, and pentazocine. Work-related injuries have been identified as a factor in the rise of opioid dependence and opioid-related overdoses [21,22]. Opioid use has also been associated with the risk of motor vehicle accidents in commercial drivers [23]. Indeed, several adverse effects are possible risk factors for workplace performance, namely, those mediated by μ-receptors, such as sedation, respiratory depression, suppression of cough reflex, sweating, euphoria, dysphoria, confusion, insomnia, agitation, fear, hallucinations, drowsiness, motor decoding and mood swings, miosis, and dependence [17,24,25].

Cocaine, amphetamines (i.e., *d*-amphetamine, *d*-methamphetamine, methylphenidate, 3,4-methylenedioxymethamphetamine), caffeine, nicotine, and ephedrine are **stimulants of the central nervous system** [26–28]. Xenobiotics belonging to this group exhibit, in general, a pronounced stimulating effect in the central nervous system. They reduce the feeling of mental (i.e., increased alertness) and physical fatigue (i.e., increased motor activity) and cause dependence [29,30]. Of these, the most consumed psychoactive substance is caffeine, which is present in coffee, tea, and chocolate, as well as in numerous foods, drugs, and beverages, as it increases energy and concentration, with no major negative labor effects having been reported [31], and it moreover has been linked to a reduced suicide risk [32]. In addition, considerable research suggests that nicotine enhances cognitive control-related processes (e.g., attention, memory) among nicotine-deprived smokers, both in terms of behavior and neural indices [33]. Indeed, smokers deprived of nicotine (e.g., 12 h smoking abstinence) exhibit reduced cognitive-attentional functioning [34]. Employees who tested positive for cocaine were four times more likely to be categorized as absentee employees and two times more likely to be terminated from employment than those who tested negative [35]. Comparing the consumers of amphetamines with those of other illicit psychoactive substances, it was found that they had higher absenteeism due to disease and/or accidents, as well as a higher incidence of behavioral risks. If consumption occurs outside working hours, the period of depression and asthenia and cognitive alterations may arise when they resume work activity. Low-to-moderate doses of stimulants (e.g., amphetamine, caffeine, modafinil) have been reported to be effective countermeasures for mood and performance decrements caused by sleep deprivation and fatigue [36,37].

More than 61 **cannabinoids** have been identified among more than 400 compounds documented in *Cannabis sativa* [16,38–40]. The most important forms are delta-9-tetrahydrocannabinol (Δ9-THC; main psychoactive constituent), Δ8-THC (almost as active as the previous one but at lower concentrations), cannabinol (low activity, but at high concentrations), and cannabidiol (not psychoactive, but in

high concentrations). Δ9-THC is the most widely abused illicit psychoactive substance and the potency of the derivative (e.g., marijuana, hashish, hashish oil) depends on the percentage of this xenobiotic [16,39–41]. Δ9-THC produces euphoria followed by relaxation, distortion in space and time, hallucinations (in higher doses than those normally found clinically), changes in short-term memory, motor incoordination, behavioral disinhibition, concentration and learning issues, decreased appetite, and in high doses paranoid psychosis. It possesses a low risk of psychological dependence and its abstinence is typically characterized by insomnia [42]. Marijuana cigarettes containing low-to-moderate Δ9-THC concentrations can decrease some night shift-related performance and mood disruptions [43,44]. Occupational medicine additional concerns come from the use of medicinal *Cannabis* [45–47]. Workers who have been authorized to use *Cannabis* should be required to report any change in product, dose, frequency, and timing of use or route of administration, and an occupational physician trained and knowledgeable on the impact and evaluation of potentially impairing substances in the workplace should be included [47]. Moreover, new psychoactive substances, such as synthetic cannabinoids, are now an additional concern to deal with [48]. Indeed, these compounds also bind to cannabinoid receptors and abuse can cause anxiety; confusion; hypertension; psychosis; hallucinations; tachycardia; seizures; and, in severe cases, it can lead to death [49–51].

**Hallucinogens**, such as the diethylamide of lysergic acid (LSD), mescaline, psilocybin, and psilocin, also have a pronounced effect on workplace performance [52,53]. Hallucinogens intensify sensory experiences and lead to behavioral alterations, delusions, as well as emotional lability at the time of and after consumption, confusional states, and *flashbacks* (i.e., reviving experiences). Chronic consumption can lead to depression, violent behavior, anxiety, and alteration in the perception of time, but the risk of dependence is absent.

The **central nervous system depressant** group mainly includes psychotropic drugs, notably the illicit use of anxiolytics, sedatives, and hypnotics, such as benzodiazepines; barbiturates; and non-benzodiazepine hypnotics such as zolpidem and ethanol [54,55]. Of all substances, ethanol is the one with the highest negative impact on work [9,56]. Globally, ethanol is the world's number one risk factor for ill-health and premature death amongst the 25- to 59-year-old age group, the core of the working-age population [57]. Moreover, in the USA, ethanol-induced impairment directly affects an estimated 15% of the workforce and causes more than 22% of the deaths as a result of injuries at work [58]. The main labor consequences of ethanol consumption are slow reaction time, motor incoordination, decreased visual acuity, emotional lability, reduced concentration, lower intellectual ability, behavioral changes, attendance/punctuality problems, lower productivity, absenteeism, presenteeism, workplace injuries, and higher employee turnover [59]. Several studies have shown that workers who drink the most are more often absent from work due to drinking [60], and they more often report alcohol-related presenteeism and inefficiency at work due to alcohol use [61]. Ethanol is anxiolytic/disinhibiting and has a high potential for abuse. Withdrawal leads to sweating, nausea, tremor, insomnia, decreased appetite, restlessness, aggression, anxiety, and eventually hallucinations. It is often used in conjunction with other substances to enhance the overall effect. Major international organizations such as the World Health Organization, the Council of the European Union, and the International Labor Office advertise the need as priority to actualize policies and programs focused on the issue of ethanol and work.

Among psychoactive drugs, benzodiazepines are the most prescribed drugs, especially as anxiolytic, sedative, or hypnotic drugs, and less often as antiepileptics and/or muscle relaxants [55]. Benzodiazepines have largely replaced barbiturates as they are safer drugs with fewer enzyme-inducing effects, and thus less severe interactions, less severe withdrawal symptoms, and a broad therapeutic margin [62]. However, persistent and longer than recommended use and self-medication is a reality. Benzodiazepines are distinguished between anxiolytic and hypnotic [24,55]. This distinction is somewhat artificial because all are anxiolytic, and all can change sleep as long as certain doses are reached [55,63]. What distinguishes them, however, is that so-called hypnotic benzodiazepines are potent drugs that can, therefore, modify sleep conditions at relatively low doses, whereas so-called anxiolytic benzodiazepines are less potent, allowing for a "therapeutic window" to exist where anxiolytic

action can be obtained without significantly interfering with sleep [55]. All benzodiazepines may induce tolerance, dependence, and addiction, but to a lesser extent than barbiturates [62]. Short-acting benzodiazepines have the highest potential to induce dependence, and withdrawal syndrome may even occur when discontinuation is abrupt. The abuse of these drugs is generally higher in females and older workers. The toxic effects of benzodiazepines on quality of work are mostly related to the central nervous system depressant effects, particularly sleepiness, motor incoordination, and impaired thinking [64]. They have a significant impact on motor vehicle driving, workability, and interpersonal relationships, and thus their use should be taken into account in these circumstances [65]. They can also cause short-term memory impairment, that is, diminished ability to learn new information, leaving already learned information intact.

Among **dissociative drugs**, phencyclidine and ketamine were originally developed as dissociative general anesthetics, capable of promoting sensory loss and analgesia, amnesia, and paralysis, generating an intense sensation of dissociation of the environment but without real loss of consciousness and protective reflexes [66,67]. Only ketamine is still available in therapy due to the high frequency of delusions and hallucinations observed postoperatively with phencyclidine [67–70]. The most classic adverse effects of ketamine are delirium, hallucinations, tachycardia, mild respiratory depression, confusion, irrationality, violent or aggressive behavior, dizziness, ataxia, slurred speech, delayed reaction time, euphoria, altered body image, analgesia, amnesia, and coma [66,67].

#### **4. Psychoactive Substances and Occupational Consequences**

Employers, for their part, have a broad range of responsibilities, and it is the employer's obligation to ensure the individual and collective health of workers for the proper functioning of all work activities. Psychoactive substance abuse is a major preventable cause of morbidity and mortality and has direct impacts on workability [71]. Staying in the workplace under the influence of psychoactive substances depends on the combination of multiple factors, some linked to individual characteristics and lifestyles, and others of professional nature, such as work typology, rhythms and cadences, irregular working hours as occurs in shift work, stress, and psychological harassment, among others. The following are some of the consequences of the occupational psychoactive substance consumption and thus justify why employers aim for a "drug-free" workplace [72–74]: (i) increase in the rate of presenteeism (i.e., being present at workplace in an impaired state) and absenteeism that affects professional performance promoting errors, and hence the competitiveness and productivity of the enterprise, as well as the country's own wealth; (ii) creation of a negative image, leading to discrediting and despising of the organization; (iii) negative effect on the equipment integrity and therefore a potential cause of financial losses; (iv) negative effect on the workers' physical, psychosocial, and behavioral integrity; (v) risk of neglect and reduced decision-making and motor coordination, consequently leading to a higher number of errors and accidents and therefore costs (e.g., insurance premiums); (vi) workers being more often involved in conflict, violent behavior, and theft, and being more frequently the subject of complaint by coworkers who may also see their physical integrity or even their own lives affected as a result of lack of care or discernment, decreased alertness, or altered perceptions or judgements from others being under the influence of psychoactive substances; and (vii) workers tending to be non-punctual (arriving at work later and leaving early) than the rest of the working population, putting a greater strain on coworkers by introducing additional tasks that still need to be done.

#### **5. Legal Aspects**

Aiming at a legal interpretation of this theme, it is important to make clear what is an occupational accident and workplace. Although, these concepts may vary with the different legislation of each country or state, regardless the jurisdiction, these are broad concepts in most legislations. We used Portuguese legislation as a starting point for this approach.

The Portuguese Law no. 98/2009 of 4 September regulates the regime of compensation for occupational accidents and diseases, including occupational rehabilitation and reintegration, pursuant to article 284 of the Labor Code, approved by Law no. 7/2009 of 12 February. In accordance with Article 8, an *occupational accident* is one that occurs at the place and time of work and that directly or indirectly results in bodily injury, functional disturbance, or illness resulting in reduced working or earning capacity, or death. Therefore, it is an event that has a professional factor for its occurrence, including acts of violence that occurred within the scope of the workplace concept. Its major differences from the concept of occupational disease, in which occupational factors are also determinant, include (i) very short time (at most a few minutes), usually sudden and unexpected (acute); (ii) easy identification of the cause (professional); and (iii) easy identification of the lesion.

The concept of the *workplace* means any place where a worker is or should be by virtue of his/her work and that are directly or indirectly under the employer's control, including road accidents during work-related activities. The term "working time beyond the normal working period" is defined in Portuguese law as the time that (i) precedes its commencement, including preparatory acts; (ii) follows, including related acts; (iii) resulted from normal or forced interruptions of work. Article 9 extends the concept, and also considers a workplace accident as one that occurs (i) on the way to or from the workplace in the routes normally used and considering the period of time usually spent by the worker in those routes: between any of their workplaces if they have more than one job (the destination employer being the responsible for the accident), between his/her habitual or occasional residence and the workplace or the place of payment or the place where they will receive any kind of assistance or treatment by reason of a previous accident, or between the workplace and the place of meal and between the new place defined by the employer for a specific work and the usual workplace or his/her habitual or occasional residence. It is also considered a work accident when the normal route has been interrupted or changed in order to satisfy workers' needs, as well as due to force majeure or due to a fortuitous event (ii) while performing spontaneous services that may result in economic gain for the employer; (iii) at and outside the workplace while representing workers; (iv) at the workplace, when attending a training course, or outside the workplace, if the express permission from the employer for such attendance was obtained; (v) at the place of payment of remuneration and while staying there for that purpose; (vi) at the place where the worker should receive any form of assistance or treatment due to a previous accident and while staying there for that purpose; (vii) while job searching during the hours granted by law to workers with process of termination of the current work contract; and (viii) off-site or working time when verified in the performance of services determined by the employer or consented by him/her.

#### *5.1. Workplace Drug Testing*

The legislation applied to workplace drug testing also has some differences among countries [75]. Indeed, a multinational company may not be able to implement the same procedure in all its offices. In this work, we focused upon and discussed general consentaneous aspects that need to be followed in the application of a program for workplace drug testing; the Portuguese legislation was used as a platform for critical reflection. This approach is even more interesting because by adopting a radical step, in July 2001 Portugal became the first country in the world to decriminalize the possession for consumption of all illicit substances (Law no. 30/2000 of 20 November). Rather than being arrested, those caught with a personal supply are obliged to undergo rehabilitation treatment.

Although there are no specific propositions referring to psychoactive substance testing in the workplace, the Portuguese law appoints a set of very strict general rules regarding the worker's health. Indeed, the human life (Article 24); the moral and physical integrity of persons (Article 25); the right to preserve the intimacy of private life (Article 26); the right to the protection of personal data and the use of computers (Article 35); and the rights of all workers (regardless of age, sex, race, citizenship, territory of origin, religion, or political or ideological convictions) to have a job and work under conditions of hygiene, safety, and health, as well as the rights to have assistance and fair rehabilitation when they are victims of workplace accidents or occupational disease (Article 59) are rights constitutionally enshrined in the Constitution of the Portuguese Republic. Moreover, in Europe, any regulation must

conform to the European Convention for the Protection of Human Rights and Fundamental Freedoms. The latter warranties for a person's right to privacy, which states that everyone has the right to his/her private and family life, his/her home, and his/her correspondence, and that public authorities must not interfere with the exercise of this right, except if the interests of national security, public safety, or the economic well-being of the country are at risk. On an international level, the matter might be covered by Universal Declaration of Human Rights (Article 12), which states that no one shall be subjected to arbitrary interference with his/her privacy.

The use of psychoactive substances in the workplace (or during working hours) is usually governed by health and safety laws that address the potential health and safety risks for themselves or co-workers [75–77]. Therefore, these laws usually legitimate the fact that the tests should be restricted to categories of workers whose activity may endanger their physical or third party integrity and make employers responsible for preventing psychoactive substance use in the workplace and impose them to carry out risk assessments and preventive measures [78]. The real question is how we can fulfil the purposes of the law, namely, the following Portuguese legislation: (i) Articles 15 (employer's general obligations), 16 (simultaneous or successive activities in the same workplace), and 17 (worker's obligations) of the legal regime promoting occupational safety and health (Law no. 102/2009 of 10 of September); (ii) Articles 19 (medical tests and examinations), 99 (internal company rules), 281 (general principles on safety and health at workplace), 282 (information, consultation, and training of workers), 283 (accidents at work and occupational diseases), and 284 (regulation of prevention and reparation) of the Labor Code (Law no. 7/2009 of 12 February); (iii) Decree-Law no. 4/2015 of 7 of January (Code of Administrative Procedure); and (iv) the General Data Protection Regulation 2016/679 of the European Parliament and of the Council of 27 of April of 2016. Besides these legislations, specific regulations of each enterprise should be also taking into account. The Portuguese State, through the Labor General Inspection, the Directorate-General of Health, and the National Protection Centre against Professional Risks, proceeds to inspections to see if the rules are followed.

5.1.1. The Legal Regime for the Promotion of Health and Safety at Work (Law no. 102/2009 of 10 September)

The legal regime for the promotion of health and safety at work (Law no. 102/2009 of 10 September) states that occupational safety and health must be based on a correct and permanent risk assessment and be developed according to principles, policies, standards, and programs, focusing on the promotion and monitoring of occupational health and the enhancement of technical and scientific research in the field of safety and health at the workplace, particularly the emergence of new risk factors (Article 5). In Portugal, the existence of in-house occupational safety and health services is mandatory in organizations with more than 400 workers or in those with more than 30 workers if they are exposed to high-risk activities. In cases where the enterprise does not have an internal occupational health and safety service, this must be assumed by the external entities that provide occupational health and safety services (Article 78).

There are several general obligations that the employer must continuously and permanently fulfil (Articles 15 and 18): (i) ensuring safety and health in the workplace, (ii) identifying all foreseeable risks in all activities of the organization, (iii) mitigating monotonous and repetitive work, and (iv) reducing psychosocial risks. These concerns should be balanced accordingly to the risks to which the workers are potentially exposed to and are recognized as causes for psychoactive substance use, which in turn is a risk factor for accidents and enhancer of work-related diseases. Nevertheless, it should be borne in mind that employers cannot legitimately invoke the obligation to carry out psychoactive substance screening tests to accomplish their duties of ensuring the health of workers (Article 108). Indeed, the obligation to perform clinical exams (i.e., for admission, periodic, or occasional) is duly specified in the legislation and its purpose is aimed at attesting and evaluating the physical and mental fitness of the worker to perform the activity, as well as possible repercussions. In other words, compulsory psychoactive substance testing undermines the rights, freedoms and personal guarantees enshrined in

the Constitution of the Portuguese Republic, namely, the right to personal integrity (Article 25) and the right to privacy reserve (Article 26). This means that there will be no justification for drug testing in all (or random) workers in an organization, but only for those whose job requires high skills or involves considerable risk to themselves or other workers and in all who show manifested and serious signs of being influenced by psychoactive substances [79]. Moreover, it should be mentioned that the existence of an internal regulation that considers a program for workplace drug testing cannot be in itself a just cause for dismissal because it is not provided by law and violates the principle of job security and the fundamental right of workers accordingly the Article 53 of the Constitution of the Portuguese Republic.

On the other hand, it is also the worker's obligation to comply with the occupational safety and health requirements provided by legislation and collective labor regulation instruments, as well as for instructions determined for that purpose by the employer (Article 17).

#### 5.1.2. The Labor Code (Law no. 7/2009 of 12 February)

Regarding the Portuguese Labor Code (Law no. 7/2009 of 12 February), the article 281 et seq. focuses on the prevention and reparation of accidents and occupational diseases in the workplace; workers have the right to provide work in safe and healthy conditions and should respect the occupational safety and health requirements laid down by law or collective labor regulation instruments, or those determined by the employer, and the employer should apply the necessary measures to provide such environments.

The carrying out of medical tests and examinations (Article 19) is within the scope of the organization of occupational safety and health services and must respect citizens' rights, freedoms, and guarantees. Regarding the detection of psychoactive substance use, this may or may not be part of the organization's health and safety policy. The most common procedure is to perform drug testing in those workers randomly nominated by the computer, as well as those appointed by the occupational physician, or to those who request according to the rules of procedure. As mentioned above, random drug testing has been prone to controversy because employers must ensure that every aspect of their policies is rooted in scientific evidence, linked rationally to the goal of workplace safety, and are ethically justifiable [79]. Some organizations also advocate testing following an accident of specific consequences (e.g., fatalities, injuries that require anyone to be removed from the scene for medical care, damage to vehicles or property above a specified monetary amount) in order to determine whether the abuse of psychoactive substances were a factor. Of course, a positive test for psychoactive substances cannot prove that this was the cause of the accident.

Considering the possibility of creating a program that includes toxicological analyses for detection of psychoactive substances, this must comply with the legal rules in force and be part of an internal regulation according to Article 99. This regulation should spell out the underlying policy, objectives, and rights and responsibilities of all parties involved, as well as issues regarding the protection of personal data, namely, assuring the right of the worker to privacy, the need for his/her consent to perform the tests, and the preventive and non-sanctioning character of the drug tests. These are important topics specially designed to increase the sensitivity of health and safety professionals to this problem and encourage them to raise awareness of them.

Screening tests for psychoactive substance use are restricted to the occupational physician or, under his/her guidance and control, to other health professionals obliged to professional secrecy (e.g., occupational nurses) and trained to use the kits for toxicological analysis. The same is true for the results of the tests because this represents health information. Therefore, clinical data can only be known to health professionals from the occupational medicine team, who are subject to confidentiality, and any hypothetic witnesses nominated by the worker.

In order to comply with professional secrecy and to guarantee the confidentiality of information resulting from medical examinations, the occupational physician must record results with "generic terms", namely, the worker (i) is fit (i.e., does not consume), (ii) is "fit with restrictions", or (iii) is temporarily unable (i.e., consumes) to perform his/her duties. According to no. 2 of Article 17 (on the protection of personal data) of the Labor Code, at no time should the physician report the test results to the employer. Only in this way is the adequacy of the preventive and deterrent measures ensured and it represents a very serious offence the disrespect of these directives. Accordingly to Article 195 of the Portuguese Penal Code, the disclosure of other people's secrets, known to someone on account of his/her state, job, employment, profession (e.g., physician), or arts, is punished by a term imprisonment of 1 year or fine until 240 days. Moreover, the physician also follows the Deontological Code of the Medical Association, which compels him/her to professional secrecy (Article 71).

It is also made clear by the Labor Code (Article 19), in addition to the situations provided in occupational safety and health legislation, that the employer cannot, for admission purposes or job maintenance, require the applicant or worker to perform or present tests or medical examinations of any nature (such as the results of psychoactive substance use tests), or prove physical or mental conditions and therefore health. An exception occurs when these examinations aim for the protection and safety of the worker or third parties, or when the demands inherent to the activity justify it. In these cases, the reasons must be provided in writing to the job seeker or worker. Therefore, conducting screening tests for psychoactive substance use will only be legitimate in exceptional cases where health, welfare, worker, employer, or third-party concerns are at stake. In cases of job admission, even if screening tests for psychoactive substance can be justifiably performed in candidates, the possibility of the job applicant stopping drug use several days before testing should be born in mind. However, the screening will no longer be legally acceptable on the principles of proportionality, appropriateness, and reasonableness when there are no objective grounds to perform them in order to assure safety to the workers, service users, or the wider community, or when, from this point of view, the risks are minimal.

It is also clear that a job applicant or worker who has provided personal information has the right to control his/her personal data, to be aware of its content and the intended purpose, and to ask for its correction and update.

#### **6. Diagnosis of the Influence of Psychoactive Substances and Toxicological Analysis**

It is fundamental to have the substances that will be monitored clearly defined in the rules of procedure. Suspicion of psychoactive substance consumption can be made at various levels, of a subjective and objective nature. The loss of productivity and decreased quality of work, lack of punctuality and absenteeism, indiscipline and inappropriate behaviors, and the increase in workplace accidents are warning signs that cannot be neglected. However, these signs of suspicion should be part of a broader clinical and laboratory evaluation under the responsibility of occupational medicine and, in some cases, with the contribution of insurance medicine as well as forensic medicine.

Occupational medicine is a medical specialty concerned with the maintenance of health in the workplace, including prevention and treatment of diseases and injuries. In other words, the aim of occupational medicine focuses on workers' health. Accordingly, the International Code of Ethics for Occupational Health Professionals published by the International Commission on Occupational Health (ICOH) states, "the aim of occupational health practice is to protect and promote workers' health, to sustain and improve their working capacity and ability, to contribute to the establishment and maintenance of a safe and healthy working environment for all, as well as to promote the adaptation of work to the capabilities of workers, taking into account their state of health" [80].

Many enterprises establish a drug policy with little or no structure for drug testing, namely, (i) quality control, (ii) systematic confirmation of positives, and (iii) procedures to accomplish the chain of custody. Indeed, procedures that ensure the chain of custody compliance are of utmost importance in toxicological analyses. Moreover, drug testing is usually performed on-site by occupational physicians, who are usually not familiarized with analytical toxicological aspects. In Europe, the European Guidelines for Workplace Drug Testing have been prepared and updated by the European Workplace Drug Testing Society (EWDTS) for different samples. These guidelines are designed to [81] (i) establish best practice and standard procedures whilst allowing individual countries to operate within the

requirements of national customs and legislation, (ii) ensure that the entire drug testing process is conducted to give accurate and reliable information about a donor's drug use, (iii) maintain the legal defensibility and scrutiny either by an employment tribunal or a court of law, (iv) protect the dignity of the specimen donors and the validity of the specimen, (v) define common and critical quality control procedures for laboratories, and vi) help in the interpretation of the analytical results. Guidelines for collection of biological samples for toxicological analysis were recently published [82] and have largely conformed to the EWDTS guidelines, focusing on sample collection and testing in urine, hair, and oral fluid. Our work focused only on specific points relevant to the interpretation of workplace drug testing [75,81]. Trained personnel, who do not need to specifically be healthcare professionals, are required [81]. All samples must be kept for an agreed period or in respect of the national legislation of each country to allow rebuttal if any judicial sues are made regarding the obtained results. After the agreed time, the laboratory may discard the sample if the customer did not request the laboratory to retain the sample for an additional period. Samples must be retained within the secure laboratory area until the disposal date agreed with the customer [83].

#### *6.1. Interpretation of Ethanol Results*

The alcoholaemia is usually performed by quantitative breath analysis, using an alcohol meter (duly calibrated). In Portugal, there is a model certified by the Portuguese Institute of Quality that is based on the theoretical relationship defined in Law no. 18/2007 of 17 May, assuming that 1 mg/L in breath alcohol concentration (BrAC) is equal to 2.3 g/L of blood alcohol concentration (BAC) [65]. Counterproof should always be available to confirm results and should be provided by a referenced toxicology laboratory and by using a different technique and chemical principle from that of the screen test in order to ensure reliability and accuracy. Real BAC can be quantified by blood collection and further analysis by using gas chromatography with a flame ionization detector and a headspace system. The direct determination of ethanol itself in hair is not possible due to its volatility and its potential absorption from external sources [54]. Instead, the minor ethanol metabolites ethyl glucuronide (EtG) and/or fatty acid ethyl esters (FAEE) can be measured in hair by GC or LC coupled to MS/MS as a direct alcohol consumption marker [84].

There are no legally defined values for ethanol blood concentration in the workplace and it should be noted that the maximum limit established for the Portuguese Highway Code should not be generalized to all professions or tasks [65]. Nevertheless, several Portuguese organizations are governed by the values defined in the Highway Code. For instance, the Collective Bargaining Agreement for the Construction and Public Works Industry that has been in force since April 1, 2010, in clause no. 78 (on the prevention and control of alcoholaemia) considers that it is under the influence of alcohol the worker who, under examination by BrAC, reveals 0.5 g/L or more. For workers under the Highway Code, the BAC provided in that Code is applicable.

#### *6.2. Interpretation of Other Psychoactive Substances Results*

Regarding the analysis of other psychoactive substances, toxicological analysis can be performed in various biological samples, being more commonly analyzed in oral fluid [83–85] and urine [81,83] or in hair [86] and nails [87], given that, although analytically possible, the toxicological results are not relevant to allow for a quantitative interpretation [65]. Indeed, in most cases, only by using blood, serum, or plasma is it possible to truly document the impairment. Therefore, the most correct procedure will be to consider as positive the test that reveals the presence of illicit psychoactive substances above the defined concentrations cut-offs for screening and confirmation tests [88], regardless of the quantitative interpretation the test may provide [89]. Although urine is an important sample for toxicological screening, it is only useful if freshly voided and collection is witnessed and supervised to avoid adulteration. Indeed, the collection facilities should be arranged to prevent adulteration of the specimen as much as possible and adding coloring agents to toilet water has been recommended to reduce the risk [81]. Some authors suggest that all urine specimens taken for drug testing from

both the workplace and court settings need to be tested for validity [90]. Nevertheless, urinating is a personal act and most people feel inhibited about being observed in such circumstances by close family members, medical staff, or even sexual partners, and the situation can never be less than humiliating. Furthermore, individuals may not want to disclose pregnancy or venereal diseases. A wide variety of collection devices are available in the market to collect oral fluid for toxicological analysis. A minimum validity test should be performed for oral fluid, such as visual inspection of the sample(s), measurement of oral fluid volume and testing on matrix authenticity through measurement of endogenous biomarkers such as salivary amylase and cortisol [83–85].

Hair samples have been gaining popularity. This is by far the most expensive method, but it is claimed to be more secure in the event of legal challenges, to provide conclusive evidence of rebuttal, and is an excellent indicator of addiction, as opposed to occasional use, which may be particularly suitable for pre-employment testing programs. Analysis of nail clippings may be a useful back-up for hair analysis when hair is unavailable [87,91,92]. Useful applications of nail analysis may be where contamination is highly unlikely (i.e., if the fingers do not become contaminated) as is the case during the consumption of tablets or capsules as medication or where the question of time or quantity of use rather than the fact of use is being investigated [92].

The most frequently analyzed substances are amphetamines and derivatives, cannabinoids, cocaine, opioids, and benzodiazepines. Regarding cocaine, immunoassays detect the metabolite benzoylecgonine quite accurately without much concern for false positive or negatives. Nevertheless, the metabolite is inactive and may be present up to 3 days after use, whereas most clinical effects of cocaine occur within 6 to 12 hours of use [26,27]. As for amphetamines, there are many false positives, including antihistamines, decongestants, antidepressants, or acid-blockers, and these may be detected 1 to 3 days after use. Indeed, methamphetamine only differs from pseudoephedrine in a single atom of oxygen. In fact, many methods flowed by clandestine laboratories use *l*-ephedrine (present in stimulant supplements and used for weight loss) or *d*-pseudoephedrine (present in several decongestants drugs) as starting reagents to produce *d*-methamphetamine [93].

Regarding cannabinoids, common over-the-counter analgesics such as ibuprofen and naproxen, as well as the increasing interest in the use of medical *Cannabis*, can increase the probability of cross-reaction with this assay, creating false positives. Many opioids can be missed during routine screening, generating significant false-negatives, and it has been reported that screening is also prone to many false-positive results in the presence of poppy seeds and quinolone antibiotics [13,94]. Indeed, poppy seed paste used in foods could lead to positive urine tests for opioids, even if increased cut-off levels are used and different biomarkers are considered for the differential diagnosis, such as the presence of thebaine to ensure justice for each individual [13,95]. Moreover, if an individual is currently taking a lawful medication containing codeine, the test can suggest that he/she has used heroin.

Nevertheless, it should be mentioned that there are several unscreened substances in existence, such as ketamine, chloral hydrate, gamma-hydroxybutyrate (GHB), psilocybin, mescaline, and cathinones. Moreover, qualitative or semi-quantitative immunoassays may cross-react with other substances, many of which are licit, namely, pharmaceutical drugs [89]. In this case, tests showing the presence of prescription drugs should not be considered positive.

On the basis of the technical limitation of the immunoassay analysis, whenever positive results (i.e., above a predefined cut-off level) emerge at the screening stage, confirmatory tests on the sample must be carried out by a referenced toxicology laboratory, normally by gas and liquid chromatography coupled with mass spectrometry. If the screen results are all negative, no further analysis is necessary. Results of confirmatory analysis are usually not readily available, and even if positive results are obtained, interpretation is required by skilled toxicological professionals in conjunction with a qualified occupation medicine physician [96].

Additionally, of note is that there are many other clinical situations with signs and symptoms similar to psychoactive substance poisoning, and thus definitive diagnosis is only possible through clinical examination. Therefore, a differential diagnosis is needed to find out whether the reason for the positive result can be explained by prescribed medication; some other acceptable reason; or whether it is, in fact, a question of drug abuse [97]. If an occupational physician is not present, the worker should be removed from activities that put his/her or others' safety at risk and should be properly referred to a health unit. It is underlined that in this case all labor rights, including remuneration, must be ensured for as long as the worker is away from work.

#### *6.3. Worker's Refusal to Undergo Toxicological Analysis*

Besides the health professional (or other) who is competent to carry out the sample collection for toxicological analysis, it should be made possible for the worker to request the assistance of a witness, having a defined time to do so, but the test cannot be stopped if the presentation of the witness was not feasible. Refusal to submit to screening tests for psychoactive substance use, although possible, does not imply that the worker is unsuitable for work. When unjustified, the worker may incur any disciplinary offence and, in some enterprises, the refusal leads to the assumption that the worker has a BAC of 0.5 g/L or higher and therefore the employee is unfit/inapt for work. Moreover, it should be highlighted that if a worker is declared unfit by the occupational medicine as a result of psychoactive substance use, it is not a just cause for dismissal or to affect career progression, as clinical results are confidential and as such cannot be made known to the hierarchical superiors or other entities outside the clinical sphere. Only the behavior that may result from the influenced state can be framed according to Article 351 (just cause for dismissal) of the Portuguese Labor Code and as such be subject to disciplinary sanction (and possibly dismissal) due to the unacceptable or breach of established rules.

As a precautionary measure, if a BAC of 0.5 g/L or greater is assumed or presumed, or if positive illicit psychoactive substances results are obtained, the worker will be immediately prevented from working during the remaining daily work period, with the consequent loss of remuneration for such period.

If the worker is found to abuse psychoactive substances in the workplace, it is possible and permissible to initiate disciplinary proceedings against the worker, especially in organizations with internal rules prohibiting the use of psychoactive substances in the workplace.

#### **7. Conclusions and Future Perspectives**

The use of psychoactive substances in the workplace is a public health problem that influences the safety and health of workers and extends beyond the workplace itself. It is a contentious issue that has moved up the human resources agenda in recent years, and different organizations take diverse approaches to psychoactive substance workplace testing, some of them following a zero-tolerance approach, whereas others have had to develop a more nuanced policy. Such cases may also have insurance and legal repercussions [98,99]. Indeed, workplace psychoactive substance testing is one of the latest components to be added to the discipline of forensic toxicology, which now comprises the triad fields of post-mortem, human performance, and workplace drug testing toxicology [88]. There are several studies that demonstrate that the psychic and motor disturbance due to the consumption of psychoactive substances is in the origin of several work accidents and related economic impacts [12,100,101]. Moreover, testing for psychoactive substances in the workplace, at random or by surprise, has a statistically significant preventive effect in overall professions [102]. Nevertheless, testing in the workplace is a complex topic because it is not often directly regulated by supranational or national law. Only a few countries report legislation that clearly and specifically address the issue of drug testing in the workplace [51].

The participation of workers and their representatives in the design of an occupational health promotion plan and in the definition of policies to be followed undoubtedly plays a decisive role in the implementation of prevention programs against the use of psychoactive substances in the workplace. Human resources policies aimed at promoting worker safety, health, and well-being that integrate worker assistance programs, information campaigns, and other interventions in this field reflect the level of organizational cultures that incorporate concepts and principles of corporate citizenship,

encouraging entrepreneurs and managers to good practices, the production of deontological ethics, and codes that value the image of the company or organization and its end products. Workers should also be aware that the employer is committed to creating a work environment that promotes safety and health and that policies and related measures should be applied to all elements of the organizational system so that the individual and collective rights of society will be fully accomplished. Indeed, if on the one hand the use of illicit drugs in the workplace raises issues pertaining to prevention and safety and the responsibility of the various members of staff, on the other hand it also brings into question the interface between work and private life [103].

As a prerequisite to testing, the company must have an antidrug program aiming to ensure health and safety, including a written drug testing internal regulation. In this regulation, problems related to the use of psychoactive substances should be considered as health problems and consequently should be treated like other health problems in the workplace regarding temporary disability, sick pay, and other social benefits, especially during periods of rehabilitation treatment.

It is important to clarify that under the law, the screening, counterproof, or confirmatory analysis should not be charged to the worker because they are preventive health and safety activities of companies and the costs must always be supported by the employer. All these measures are stimulating and represent facilitating factors for recovery, although under treatment, the employer must also ensure that the workstation is maintained or transferred to other duties that do not pose a risk to the safety of himself or others without loss of rights or other benefits. Nevertheless, an employee's failure to successfully complete treatment requirements can eventually result in the termination of their employment.

Nowadays, we are witnessing a paradigm shift from an evolution centered on the treatment and rehabilitation of proven psychoactive substance dependents to a cross-cutting approach through the implementation of a specific regulation that includes (i) prevention of consumption through information actions on health consequences of the use of psychoactive substances; (ii) early detection; (iii) treatment at the enterprise (if it has the means), or through referral to other specialist physician, or to local or regional specialist services (e.g., integrated response centers or alcoholology units) to recover workers in compliance with personal freedom, and these treatments cannot be imposed by coercion; and (iv) socio-professional rehabilitation of the worker, aiming at relapse prevention.

For the full implementation of these objectives, the availability of skilled human resources specifically in occupational medicine with toxicological background and skilled toxicologists in clinical, forensic, and analytical aspects of psychoactive substances, to correctly interpret toxicological results, are needed. Given this reality, fulfilling the purposes of the law, namely, Article 19 of the Portuguese Labor Code, is an almost impossible task. There should be a balance between prevention and repression.

Finally, policies and educational interventions to reduce psychoactive substance consumption in the workplace are needed [104]. These should address awareness to workplace hazards, physician education to promote best practices for treatment, and overdose prevention. In addition, the phenomenon of the new psychoactive substances are also motifs of concern in the workplace that should not be disregarded [51]. Indeed, the consumption of new psychoactive substances poses a significant risk to public health and a challenge to national and international drug policies for occupational medicine, and only very recently have laboratories begun to offer screening and confirmation analysis for new psychoactive substances at lower costs [105].

**Author Contributions:** R.J.D.-O. created and designed the study, wrote the first draft of the manuscript, and analyzed and interpreted the literature. T.M. also analyzed and interpreted literature and wrote and revised the final version of the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

**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/).

### *Review* **Role of Nutrition and Exercise Programs in Reducing Blood Pressure: A Systematic Review**

#### **Roman Jurik and Petr Stastny \***

Department of Sport, Faculty of Physical Education and Sport, Charles University, 162 52 Prague, Czech Republic; rjurik@centrum.cz

**\*** Correspondence: stastny@ftvs.cuni.cz; Tel.: +420-777-198764

Received: 30 July 2019; Accepted: 3 September 2019; Published: 5 September 2019

**Abstract:** The combined effect of diet and strength training (ST) on blood pressure (BP) seems to be very important for the treatment of prehypertension and hypertension (HT). Therefore, the aim of this study was to determine whether ST alone or combined with nutrition or supplementation has an impact on the arterial pressure reduction in normotensive and hypertensive populations. A systematic computerized literature search was performed according to the PRISMA guidelines using PubMed, Scopus and Google Scholar; only English language studies published from 1999 until 2018 were included. This systematic search identified the results of 303 individuals from nine studies. The ST program alone had a similar effectiveness as the nutrition program (NP) alone; however, their combination did not result in increased effectiveness in terms of a high BP reduction. The consumption of L-citrulline had a similar effect as ST on lowering BP; on the other hand, caffeine led to an increase in BP during the ST session. Our data suggest that a combination of ST 2–3 times a week at moderate intensity and a NP seems to be equally effective in terms of lowering BP (systolic and diastolic) as ST and NP alone.

**Keywords:** resistance training; hypertension; arterial pressure; disease prevention; caffeine

#### **1. Introduction**

The nonpharmacological approach to hypertension (HT) reduction is based on lifestyle changes using nutritional and exercise strategies; different training interventions [1–7] or nutritional plans [8–10] have been shown to decrease arterial pressure values. Previous reviews have summarized the approaches of aerobic training [11–13], anaerobic training [14–17] and nutrition [18–21], and those strategies were effective for HT reduction with expected decreases of 5 mmHg in systolic blood pressure (SBP) and 3 mmHg in diastolic pressure (DBP) [22] after three months. However, nonpharmacological strategies do not contraindicate each other, and their combination has been shown to be effective for other health improvements, such as bodyweight reduction [10,23–27]. Therefore, there is a question as to what kind of intervention or their combination has an improved effect on the HT decrease.

One of the well-documented interventions that has been shown to reduce arterial pressure is strength training (ST), which has already been reviewed to set optimal training loads, such as the number of sets [28–31], repetitions [32–35] and rest intervals during training sessions [32–34,36–38]. One of the ST effects is eliciting high muscle protein degradation followed by protein synthesis, which increases basal metabolism and is therefore usually accompanied by changes in nutritional requirements. In clinical practice, it is typical that ST is prescribed along with a low carb diet [39–41], specific protein intake [42–49] or another strategy that might support and magnify the arterial pressure decrease. However, there is currently no recommendation regarding whether the ST program should be accompanied by specific nutritional support that could result in a greater effect of HT reduction.

Numerous studies have shown that different nutritional programs (NP) and strategies might lead to nonpharmacological decreases in arterial pressure [50–55], while many strategies place a high demand on patients to make changes in their eating habits. One way to easily change food intake might be accomplished by using food supplements, which have a synergistic effect with the application of ST. Some food supplements have been shown to decrease arterial pressure when used with aerobic training [56–59] or alone. On the other hand, some food supplements might have a positive effect alone but when used simultaneously with exercise might cause side effects, such as post-exercise hypotension [60–66].

In the current literature, there is no current overview of whether ST or nutritional strategies have a stronger effect on HT reduction. Therefore, the aim of this study was to determine whether ST alone or ST combined with nutrition or a supplement has the greatest impact on arterial pressure reduction in normotensive and hypertensive populations. Additionally, this review aimed to summarize what kind of combined ST and NP might be effective for HT reduction without side effects. The main hypothesis of this article is that the biggest effect on HT should be observed when strength training is combined with energy intake restriction. Based on this result, practitioners can establish nonpharmacological treatment for the individuals with increased blood pressure (BP).

#### **2. Methods**

This systematic review is reported in accordance with recommendations as presented in the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement [67]. The protocol for this systematic review was published on PROSPERO under registration number CRD42019130631.

#### *2.1. Search Strategy*

A systematic computerized literature search was performed using PubMed, and Scopus and included studies published in English from 1966 until November 2018. The search formula included the following terms: Blood pressure AND hypertension OR cardiovascular disease OR hypotension AND resistance training OR strength training OR weightlifting OR bodybuilding OR exercise. The search was limited by article types, species, subjects, language, age, and text availability. A manual search was performed using the identified reviews, reference lists of the selected articles and Google Scholar.

#### *2.2. Types of Studies*

The review considered cohort studies, analytical cross-sectional studies, randomized control trials, nonrandomized control trials, intervention studies, case-control studies and others that included BP and HR measurements as well as data on NP and ST in all adult populations. The review studies were used for manual searches of their reference list. Dissertations, theses, conference proceedings, conference monographies, and other reviews were not included. Retrospective studies were not included because the area of interest requires performing experiments. The qualitative component also considered the type of ST and NP and methodological designs. All titles and abstracts were screened according to the above-mentioned inclusion criteria after removing the duplicates. Full texts of eligible articles were retrieved and assessed by two reviewers (R. J. and P.S.). Any discrepancies between the two reviewers were managed by a consensus discussion.

#### *2.3. Types of Outcomes*

The review considered studies that included the following outcome measures: Acute SBP and DBP variability before, during, and after exercise or a nutritional treatment; the HR variability before, during, and after exercise or treatment; and a mean arterial pressure before, during, and after exercise. The exclusion criteria were as follows: Full text was not available in English, the study did not contain an appropriate description of measurement devices and procedures, the study did not include a proper exercise or nutrition task, or the study did not report how raw data were processed.

#### *2.4. Data Extraction and Evaluation*

Data extraction included aspects of the study population, such as the average age and sex, specific aspects of the NP and ST intervention (sample size, type of exercise performed, presence of supervision, frequency, and, duration of each session, type of diet, type of supplements), outcome measures and results presented, and the values of BP or HR (Tables S1–S3); however, the studies were not rejected if any part were missing. The Physiotherapy Evidence Database (PEDro) scale was used to assess the methodological quality of a study based on general criteria, such as concealed allocation, intention-to-treat analysis, and adequacy of follow-up. These characteristics make the PEDro scale a useful tool to assess methodological quality [68]. Extraction was performed by two reviewers (R. J. and P.S.). The lack of clarity during the extraction process was resolved by the reviewer's discussion. The PEDro scale based on a Delphi list [69] was used for all articles even if the trials had already been rated by trained evaluators of the PEDro database (http://www.pedro.fhs.usyd.edu.au/).

#### **3. Results**

#### *3.1. Study Characteristics*

The systematic literature search through database searching identified 15,302 records. However, after duplicates were removed, 11,558 records were screened based on the title and abstract. The title and abstract screening resulted in 144 records for full-text eligibility. From these records, nine studies satisfied the quality and exclusion criteria and were selected for this systematic review after full-text screening (Figure 1).

**Figure 1.** Flowchart of the selection process. BP = blood pressure.

Table 1 provides a general description of each study, sample, and intervention characteristics. Only two studies included seniors [70] and [71]. One study included only men [72], and three included a mixed sample of men and women [70,71,73]. Five studies included only women [74–78]. In total,

five studies compared the effect of ST with the NP intervention effect on BP [70,73–76]. Another three studies compared the effect of ST and supplementation intervention on BP [71,72,78,79] and one study analysed the effect of ST and supplementation sesion in cross sectional design [71].

Based on the study results compilation, it can be stated that ST alone has a positive effect on the BP values (Figure 2A,B). However, the effect did not depend on the type of ST (resistance training, bodyweight training or training on vibration platforms, etc. Table 2). An important prerequisite for effective ST is the selection of suitable parameters and methods; otherwise, there is a risk of injury or side effects. All studies used training parameters (Table 2) in accordance with recommendations established by The American College of Sport Medicine (ACSM) and the Canadian Hypertension Education Program, according to Pescatello et al. [80]. However, each study chose different training parameters based on the performance and health state of their subjects.

**Figure 2.** Changes in systolic (**A**) and diastolic (**B**) blood pressure (mmHg) in nutrition and strength training program studies. Abbreviations: NP = nutrition program, NT = normotension group, PHT = prehypertension group, SD = standard deviation, SE = standard error, ST = strength training, NS = not significant change—no change; \* significant difference on reported *p* value.





baPWV, brachial–ankle pulse-wave velocity; BP, blood pressure; DBP, diastolic blood pressure; HR, heart rate; LIRET, low-intensity resistance exercise training, PWV, pulse-wave velocity; MAP, mean arterial pressure; RPP, rate pressure product; SBP, systolic blood pressure; y, year.


**Table**


#### *3.2. Strength Training Intervention*

In studies by Figueroa et al. [76], Arazi et al. [77], Wong et al. [78] and Astorino et al. [72], a reduction in SBP after ST alone was detected in prehypertensive individuals. Only Astorino et al. [72] observed a negligible improvement in the BP values in normotensive individuals. Figueroa et al. [76] also detected a reduction in DBP after ST. It seems that ST is one of the initiators of post-exercise hypotension in adults, regardless of the type of ST (see in Figure 2A,B, Figure 3 and Tables S1–S3).

**Figure 3.** Changes in systolic blood pressure (mmHg) in supplement studies. Abbreviations: SD = standard deviation, \* significant difference on reported "*p*" value. NSP = not significant according to

Body composition and strength parameters have been improved by ST alone in the study by Figueroa et al. [76]. This confirmed the assumption that ST has a generally positive effect on health status because it might reduce body fat and increase muscle mass and other conditioning values. Strength abilities were improved in participants of studies combining ST and nutrition, and these findings have been reported by Sales et al. [75], Figueroa et al. [76], Moraes et al. [70], and Lee et al. [73]. Strength abilities worsened in the nontraining group (only NP) in the study by Figueroa et al. [76]. In contrast, the Dietary Approaches to Stop Hypertension (DASH) diet group had no negative effect on strength according to Lee et al. [73]. A comparison of the effect of ST alone or ST and NP on absolute and relative strength bring very similar results as those shown by Figueroa et al. [76]. The aerobic parameters were improved in the ST group with NP in the studies by Moraes et al. [70] and Lee et al. [73], and the VO2max was improved in the study by Sales et al. [75].

#### *3.3. Nutrition Program*

the post hoc test.

Figure 2A,B show that SBP and DBP decreased after a nutrition program [74,76]. Only a study by Lee et al. [73] reported a negligible increase in SBP after eight weeks of a NP in individuals with prehypertension (PHT) and HT. Moraes et al. [70] showed that a higher intake of dairy products together with combined aerobic and anaerobic training can lead to a slightly larger drop in BP than that associated with a lower intake of dairy products. Moreover, the higher intake of dairy products resulted in a smaller increase in BP (i.e., return of BP to the initial values) after six weeks of nontraining [70]. It appears that the inclusion of more dairy products in the NP together with combined aerobic and anaerobic training has a positive effect on the reductions in the SBP and DBP values.

A study by Villani, Gornall [74], Figueroa et al. [76], and Lee et al. [73] revealed that a training program combined with a NP leads to a significant drop in BP. Moreover, one study showed that ST alone had similar effectiveness as that of the NP [76].

#### *3.4. The E*ff*ectivity of Food Supplements*

Caffeine is a very popular pre-workout stimulant among athletes. Astorino et al. [72] found that caffeine intake immediately before training increases SBP (*p* < 0.05). However, the effect on HR (*p* = 0.16) and DBP (*p* = 0.10) was similar for the caffeine and placebo. Values of a HR and BP were significantly higher in men with PHT than in normotensive men (*p* < 0.05).

A study by Arazi et al. [77] performed on middle-aged women with HT revealed practically the same reduction in BP after resistance training with or without green tea supplementation (Figure 3). There was no significant difference between the placebo and green tea intake groups (500 mg daily = 245 mg polyphenol, 75 mg epigallocatechin gallate, 25 mg caffeine) or placebo (490 mg maltodextrin). The participants performed circuit training consisting of two sets with a resistance of 50% one repetition maximum (1RM). BP was measured at zero, 15, 30, 45, and 60 min after training.

Wong et al. [78] observed the BP changes in postmenopausal obese women with a BMI <sup>≥</sup> 25 kg/m2. There was a similar decrease in the brachial and arterial BP in all groups (*p* < 0.05). Vibration training in combination with the placebo, vibration training with L-citrulline or L-citrulline alone can be used as effective ways to lower BP (Figure 3). However, there was no improvement in body composition between the study groups.

Romero et al. [71] observed the effects of folic acid supplementation for six weeks among seniors. The authors found that the folic acid intake immediately before training reduces HR but not mean arterial pressure. At the end of the six-week experiment, the HR values and mean arterial pressure were higher than the baseline values (*p* = 0.05 compared to acute folic acid ingestion). A significant positive feature is the fact that folic acid increases the blood flow to active skeletal muscles, mainly due to better local vasodilation.

#### **4. Discussion**

The present study is the first systematic review to analyze the evidence for the effectiveness of ST combined with the NP or supplementation on the BP values. This systematic research analyzed the results of 303 individuals from nine studies. Studies by Moraes et al. [70], Villani, Gornall [74], Sales et al. [75], Figueroa et al. [76], and Lee et al. [73] compared the effects of ST combined with a NP on BP. All of these studies used different methods and parameters for the ST protocol and the NP. A common feature of these five studies was the reductions in SBP and DBP or mean arterial pressure in all experimental groups, regardless of the type of training or diet. Only four studies with supplements met the criteria for inclusion in the systematic review: Astorino et al. [72], Arazi et al. [77], Wong et al. [78] and Romero et al. [71]. The effects of caffeine, L-citrulline, folic acid, and green tea were investigated. Only two supplements had a positive impact on BP: L-citrulline and green tea. On the other hand, caffeine and folic acid did not decrease BP, and caffeine has even been reported to increase BP.

In general, the issue of ST and diet or supplementation has been considered as completely different disciplines, which should be examined separately. However, in the case of treatment of PHT and HT, their combination may appear to be the most effective method. Although this study was not able to conclude whether ST alone or ST combined with the NP has the greatest impact on arterial pressure reduction, it successfully summarized current studies comparing different ST and NP interventions.

#### *4.1. Blood Pressure Reduced by Strength Training Alone*

It is generally known that ST increases strength, muscle mass, and bone mass and simultaneously helps reduce the symptoms of various chronic diseases, such as heart disease [81–85]. In 2005, the AHA (American Heart Association) started to recommended ST for lowering BP [2] because ST induces a post-exercise BP decrease, as supported by many reviews [14–16,86]. One of the first long-term ongoing studies conducted on cardiovascular disease was the Framingham Heart Study [24], where one of the findings reported was that a 2 mmHg reduction in DBP was associated with an estimated 17% decrease in the prevalence of HT. However, unsuitable training parameters such as the work load, reps per set, rest interval, etc., can increase BP [87] above the recommended values, i.e., 220/105 mmHg [1]. This study summarize that ST alone can decrease SBP from 132 ± 4 mmHg to 125 ± 2 mmHg [76] or from 141 ± 2 mmHg to 132 ± 16 mmHg [78], and can decrease DBP from 82 ± 3 mmHg to 77 ± 2 mmHg [76], which is much more than the smallest significant values of 2 mmHg. The one-time effect of ST induced a SBP decrease (from 136.88 ± 5.9 mmHg to 117.82 ± 6.09 mmHg) in the study by Arazi et al. [77] and in the prehypertensive group of the study by Astorino et al. [72] (from 143 ± 11.4 mmHg to 131.7 ± 16.6 mmHg). However, in a study by Arazi et al. [77], the biggest BP reduction was not observed immediately after training but after 1 h. There is great diversity between the training sessions in this systematic review. None of the included studies was identical in terms of the training parameters, methods, or exercises (Table 2), but all included studies set up the ST intervention according to the recommended parameters and methods published by Pescatello et al. [80]. Moreover, no study reported a dangerous increase of BP over 220/105 mmHg (according to the ACSM) or substantial post-exercise hypotension. According to the results of this review, ST alone can be recommended as an effective nonpharmacological intervention and prevention method for people with HT or PHT.

#### *4.2. Blood Pressure was Reduced by the Nutrition Program*

Nutrition-based approaches are recommended as a first-line therapy for the prevention of HT [72], where the AHA recommends a specific program called the DASH diet to treat and prevent HT. However, some food components such as alcohol, sodium, simple sugar, and saturated fat, have been shown to increase BP [9,51,88]. It has also been found that for HT reduction, weight loss is essential, which has been shown to reduce BP in overweight hypertensive and prehypertensive individuals [10,89–92]. This systematic review included only one study that used the DASH diet. The Korean variation of the DASH diet in a study by Lee et al. [73] improved SBP and DBP only in conjunction with ST (from 139.3 ± 13.2 mmHg to 135.7 ± 15.3 mmHg). Diet alone increased SBP (from 135.3 ± 11.8 mmHg to 135 ± 9.6 mmHg), but decreased DBP (from 86.7 ± 9.2 mmHg to 81.1 ± 8.2 mmHg [73]. Conversely, Villani, Gornall [74], Sales et al. [75] and Figueroa et al. [76] included hypocaloric diets, which led to a reduction in BP in both groups (ST group and NP alone). Moreover, it emerged that ST alone had similar effectiveness as the NP alone [76]. Based on this review result, we recommend a NP that takes into account the individual components of the food and that also leads to a drop in BP. Moraes et al. [70] showed that the inclusion of more dairy products in the NP, together with training, can have a positive effect on the lowering of the SBP (from 138.3 ± 4.6 mmHg to 135.2 ± 4.5 mmHg) and DBP (from 91.3 ± 5.3 mmHg to 88.3 ± 4.9 mmHg) values. The same higher intake of dairy products resulted in a smaller increase in BP (i.e., return of BP to the initial values) after six weeks of nontraining. It seems that a higher intake of dairy products prolongs training hypotension. Therefore, multiple effects of ST and a NP should not be expected for the BP lowering. Although their combination is not more effective, their combination might result in more health benefits (such as weight loss) than ST and NP alone.

#### *4.3. E*ff*ect of Supplements on Blood Pressure*

Caffeine is a central nervous stimulant whose physiological effects for increasing sport performance are extensive, but sometimes conflicting or contradicting [93,94]. Caffeine supplementation has been shown to decrease feelings of fatigue and promote mood and perceptual responses during any exercise, including ST [94–96]. It has been found that SBP and DBP increase after caffeine ingestion due to the vasoconstrictive effects of caffeine [93]. However, some studies have reported mixed results regarding caffeine intake and BP. In randomized controlled trials (RCTs), short-term caffeine intake caused an acute increase in SBP and DBP by 2/1 mmHg, respectively, compared with the effects of decaffeinated coffee or abstinence [8,97]. Caffeine supplementation, according to the results of Astorino et al. [72], cannot be recommended in individuals with HT or PHT because it not only increases the resting BP but also maintains the BP at a high level after the end of the workout. Therefore, the hypotensive effect

of ST is completely lost [72]. For that reason, we cannot recommend caffeine to individuals with HT or PHT although caffeine reduces body fat, increases sport performance, and delays fatigue [94–96].

L-citrulline is a precursor of L-arginine, a substrate for nitric oxide synthase, in the production of nitric oxide. Deficiencies in the L-arginine supply have been strongly implicated in cardiovascular diseases, including HT, atherosclerosis, diabetic vascular disease, hyperhomocysteinemia, heart failure, etc. [98–100]. In a study by Wong et al. [78], L-citrulline decreased BP in individuals performing ST or not performing ST. Both combinations led to a reduction, but slightly better results were achieved in the group that performed the ST alone. Here, again, this study confirmed that ST plays a primary role in the entire process of hypotension. The main and indisputable advantage of L-citrulline is that it reduces BP alone but also in combination with NP or ST. However, more studies are needed to explore this supplement.

A link between green tea and BP reduction has been explored for decades in Chinese populations [101]. However, there are few studies regarding the long-term effects of tea drinking on the risk of HT, and the results of the few studies investigating the relationship between tea consumption and BP were opposing. In epidemiological studies, a higher consumption of black tea in Norwegian individuals was associated with a lower SBP [102], while the green tea intake in Japanese self-defense officials was unrelated to BP [103]. Long-term effects of green tea consumption do not reduce the values of SBP or DBP after training in comparison with the effects of training alone. In contrast, a significant difference between the two groups was observed in the increase in HR after the end of the training. The green tea group performing RT showed a smaller increase in HR than the training group alone. These positive effects can be attributed, at least in part, to the antioxidant properties and vasodilating effects of the catechins in green tea [104]. Unlike caffeine, green tea is not dangerous for individuals with a high BP because it does not increase BP. For this reason, it can be included for its stimulating effects as a preworkout drink. A very significant drop in the systolic blood pressure was measured in the study by Arazi et al. [77], where blood pressure dropped from preintervention values from 133.12 mmHg (± 3.7) to 116.25 mmHg (± 3.71) in the group that combine ST and green tea. According to this information, green tea can be recommended as a suitable food supplement for the treatment of HT. However, supplementation with green tea together with ST does not lead to a bigger reduction in BP in comparison with the effects of the ST alone.

The primary risk factor for a stroke, is HT as stated by Meschia et al. [105]. For this reason, the effect of folic acid on the frequency of infarcts was investigated. Most of the relevant randomized trials were designed for secondary prevention and have not shown a beneficial effect of folic acid for cardiovascular disease prevention [106–109]. Based on this study result, folic acid had an effect on the HR and mean arterial pressure [71]. The acute folate intake before exercise induced a reduction in HR but not a long-term decrease. Even the long-term intake did not lead to reduction. For all measurements, a similar increase in HR and mean arterial pressure during isometric exercises was observed with the acute and long-term intake of folic acid [71]. Based on this study, folic acid supplementation in elderly individuals does not lead to a significant decrease in the mean arterial BP or HR during exercise compared to the measurements of the control subjects [71]. Similar results were observed in studies by Huo et al. [63] and Hernández-Díaz et al. [110]. According to these findings, we cannot recommend folic acid alone for lowering BP.

#### *4.4. Combined E*ff*ect of Strength Training, Nutrition Program and Supplementation on Blood Pressure*

Although BP lowering is effective using ST alone [76–78], NP alone [74,76] and some food supplements alone [77,78], the combination of ST, NP, and supplementation did not have additional systolic BP lowering effectivity (Supplemetary Table S1). On the other hand, the combination of ST and NP seems to be necessary to decrease the diastolic BP [70,78] (Supplemetary Table S2). Moreover, the combination of ST with green tea or L-citrulline decreased the systolic BP on a higher significance or effect size (Supplemetary Table S3) than the ST alone, although the significant difference between the ST and ST with supplementation has not been directly reported. This findings supports the notion

that total behavioral modification such as lifestyle modifications, rather than focusing on modifying a single behavioral target, is more important. Therefore, a combination of ST with the energy intake restriction NP, which include food supplement stimulants green tea and L-citrulline should be used for BP lowering and additional health benefits such as body composition changes.

#### *4.5. Changes in Body Composition*

Body composition improvements have been shown as a result of ST and NPs in studies by Villani et al. [74], Sales et al. [75], Figueroa et al. [76], Moraes et al. [70], and Lee et al. [73]. Slightly better results, in terms of improved body composition, were achieved by a NP alone in the studies by Villani et al. [74] and Figueroa et al. [76]. On the other hand, better results were not achieved by the DASH alone compared with the DASH diet and a training program in the latest study by Lee et al. [73]. Only a small improvement in the body composition was recorded by Figueroa et al. [76] in the ST group. Furthermore, Figueroa et al. [76] found that a hypocaloric diet decreased the brachial–ankle pulse-wave velocity mainly by reducing the leg pulse-wave velocity, and this reduction was related to fat loss. Although ST alone does not affect the pulse-wave velocity or body composition, ST combined with NP improves the brachial–ankle pulse-wave velocity and muscle strength while preventing the loss of lean body mass in obese postmenopausal women. The results of the pulse-wave velocity are considered an independent predictor of systolic hypertension [111].

In this case, a NP appears to be important for body composition transformation. However, the advantages of ST cannot be overlooked, and their combination seems to be the optimal variant. This systematic review reported some important findings concerning the different effects of various training, nutrition, and supplementary strategies on HT. Individuals with a high BP can improve their BP values, not only by using medication but also by utilizing a nonpharmacological method, which will not only have a positive impact on BP but also on other health components, such as body composition, muscle strength, and bone density. Future studies should more closely examine the effect of specific training, nutrition, and supplementation programs on individuals with HT, normotension, and PHT. Additionally, future studies should explore mechanisms of how different nutrition programs and supplements lower BP.

#### **5. Conclusions**

Presented data suggest that a combination of ST and NP seems to be an effective solution for lowering BP (systolic and diastolic). This method can be recommended for individuals with PHT or HT. However, an essential role can be attributed to the strength training program alone. In both aerobic and anaerobic forms, exercise causes post-exercise hypotension, but the effect depends on the selected training parameters. A NP based on the restriction of energy intake positively affects BP and body fat. However, this study cannot recommend caffeine supplementation because it increases BP during ST, while a more suitable stimulant is green tea or L-citrulline, which lower BP. For BP lowering in clinical practice, it is recommended to prescribe a combination of ST with the energy intake restriction NP, which include food supplements stimulants green tea and L-citrulline.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2077-0383/8/9/1393/s1. Table S1: Changes in the systolic blood pressure in strength training and nutrition program groups; Table S2: Changes in the diastolic blood pressure in strength training and nutrition program groups; Table S3: Changes in the systolic blood pressure in strength training and supplementary groups.

**Author Contributions:** Conceptualization, R.J. and P.S.; Methodology, R.J. and P.S.; Writing—original draft preparation, R.J. and P.S.; Writing—review and editing, R.J. and P.S.

**Funding:** This research was funded by Charles University grant no. UNCE/HUM/032, SVV 2017-2019 260466 and GACR GA19-12150S.

**Acknowledgments:** This study was supported by Charles University grant no. UNCE/HUM/032, SVV 2017-2019 260466 and GACR GA19-12150S.

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

#### **References**


© 2019 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*
