The Influence of Sleep, Menstrual Cycles, and Training Loads on Heart Rate Variability: A Four-Year Case Study on an Elite Female Slalom Kayaker

Abstract

This study aimed to examine the influence of training load, performance, sleep, and menstrual parameters on heart rate variability (HRV) and to evaluate its potential as a predictor of sports performance. A four-year longitudinal case study was conducted on a female elite kayak athlete, involving daily monitoring of HRV, sleep quality and duration, menstrual cycles, illnesses, and acute training loads. Over this period, 1394 measurements were taken each morning immediately after waking up and before getting up. The results of four competitive seasons were analyzed using a performance index and were statistically processed with a linear mixed model. The analysis revealed a statistically significant positive association between rMSSD and both sleep quality (p < 0.001) and the follicular phase of the menstrual cycle (p = 0.003). In contrast, the training load (p = 0.94), sleep duration (p = 0.27), and illness (p > 0.05) showed no statistically significant effect on rMSSD. Additionally, neither rMSSD (p = 0.82) nor its trend (p = 0.70) were significant predictors of the performance index. Despite the lack of a statistically significant correlation between HRV and sports performance, the findings suggest that the pre-competition decrease in HRV observed in this case study may reflect anticipatory physiological changes, potentially linked to increased sympathetic activation, as suggested in the existing literature.

1. Introduction

The demand for top-level sports is continuously pushing beyond both physical and mental boundaries. With the advancement of scientific methods and technologies, new possibilities for monitoring athletes’ performance are opening. One method that is gaining popularity among both elite and recreational athletes, thanks to its simplicity and accessibility, is heart rate variability (HRV) monitoring [1]. Monitoring HRV provides direct insights into the state of the autonomic nervous system, which is constantly influenced by various psychological and physiological stressors as an indicator of body homeostasis [2]. Additionally, activation of the sympathetic nervous system (SNS) increases HR (heart rate) and decreases HRV, whereas parasympathetic nervous (PNS) activity decreases HRV and increases HR [3]. The interplay between SNS and PNS activity is dynamic; for instance, during stress, a shift from PNS dominance to SNS activation occurs, affecting HRV patterns [4]. Technological developments in the last decade have allowed us to measure HRV using sports watches or smartphones that operate according to the photoplethysmography (PPG) principle [5,6,7,8]. HRV is widely used in sports science and is a valuable tool for monitoring training responses, detecting signs of overtraining, and evaluating recovery processes [9]. The athlete’s physiological response to exercise is influenced by a training stimulus, and a suitable training load promotes optimum performance improvements [10,11]. Daily fluctuations in HRV are affected by different stressors and provide valuable insights into the body’s current state, including training load, illness, travel, injury, and sleep quality. High-intensity exercise results in a 24–48 h decrease in HRV, while low-intensity exercise results in a 24 h increase in HRV [12,13]. Several studies confirm that tailoring exercise to daily HRV values increases the effectiveness of an exercise program compared to a traditional approach. The HRV-guided exercise principle recommends low training intensity when HRV is reduced and higher intensity when it reaches typical values. This approach has led to more significant progress in endurance athletes [13,14,15].

The most suitable parameters for the general assessment of HRV are standard deviations of NN intervals (SDNNs) and the HRV index [16,17], while the percentage of successive RR intervals (pNN50) and the root mean square of successive differences (rMSSDs) are most suitable for assessing short-term components of HRV. rMSSD is an HRV time-domain measure which quantifies the amount of variability in measurements of the interbeat interval, which is the time period between successive heartbeats [17]. Additionally, rMSSD is the primary time measure used to estimate the vagally mediated changes reflected in HRV. In sports, rMSSD is most used to monitor HRV, as it is a good indicator of parasympathetic modulation in short-duration measurements (10 s to 1 min) [10,12]. In addition, rMSSD is a validated and most-used HRV measure in scientific research [10,13,18]. Additionally, it is less sensitive compared to respiratory rhythm, and therefore, the athlete can breathe spontaneously during the measurement [12,19]. In a study performed by Kiss et al. [20] involving 138 elite athletes, the average rMSSD was found to be 71.8 ms, with a CI of 67.6 to 76.2 ms.

However, the interpretation of HRV measurements is not always straightforward, as it is widely recognized that HRV varies during a training mesocycle based on the training load. Moderate exercise is linked to higher HRV, while intense training results in a decrease [21]. HRV in elite endurance athletes shows a distinct bell-shaped pattern throughout the training periodization cycle, with higher HRV noted during high-volume, low-intensity training phases, followed by a decrease during intensified training and pre-competition tapering [22]. This decrease in HRV during the tapering phase is associated with elevated pre-competition stress and increased training intensity commonly seen during this training phase [23]. In certain sports, elevated sympathetic activity enhances performance by enabling greater exercise intensity [11,24,25]. Pichot et al. [21] indicated that a three-week intense training program for endurance athletes resulted in a progressive decline in HRV, followed by a rebound after one week of low-intensity recovery. When managed appropriately, this temporary reduction and subsequent increase in HRV contribute to enhanced performance. Increasing the training load reduces HRV, and the degree of this effect varies according to the athlete’s fitness level. Athletes with higher fitness levels tend to exhibit a more stable HRV. A significantly elevated HRV measurement deviates markedly from baseline values and may indicate overtraining. This sudden increase in parasympathetic activity is the body’s compensatory response when it can no longer effectively manage accumulated stress. Therefore, when interpreting HRV measurements, it is important to consider the broader context, including exercise, one’s overall well-being, the environment, and other influencing factors [11].

However, HRV analysis has some limitations, as it does not cover all the physiological processes associated with post-exercise recovery. Processes involved in post-exercise recovery, such as glycogen replenishment and delayed-onset muscle soreness, indicated by heightened creatine kinase levels, do not affect HRV [26,27,28,29]. Monitoring HRV alongside HR provides a clearer insight into performance monitoring. A decrease in HRV, when paired with a low HR and without fatigue or overload, may suggest parasympathetic saturation, primarily seen in elite endurance athletes [14].

This study aims to analyze the factors influencing HRV in an elite female kayak athlete daily over several years and to examine the relationship between HRV parameters and athletic performance, considering physiological, training-related, and external determinants.

2. Materials and Methods

The study participant is a top-level athlete and a Slovenian national kayak slalom team member. She competes in women’s kayak slalom and kayak cross (K1W and WCSLX) and follows a training regimen of 693 ± 22 h per year. During the study period, she was between 25 and 29 years old. Her anthropometric and physiological characteristics were as follows: body height, 169.6 cm; body weight, 63.2 ± 0.26 kg; body fat percentage, 12.1 ± 0.84%; muscle body mass percentage, 49.1 ± 0.56%; BMI, 22 ± 0.07; VO₂ max, 58.06 ± 2.95 mL·min⁻¹·kg⁻¹.

HRV was measured using an iPhone 8 and an iPhone X (Apple Inc., Cupertino, CA, USA) and the HRV4Training app (A.S.M.A. B.V., Aa en Hunze, The Netherlands). HRV4Training is a commercially available mobile app that is enabled on devices with photoplethysmography (PPG) technology, which allows for non-invasive measurements of HRV and HR at rest [8,30]. Sleep duration was monitored with Apple Watch Series 4 and Series 6 smartwatches. (Apple Inc., Cupertino, CA, USA) [31,32]. Performance tests were conducted on a cycle ergometer twice per year, beginning with an initial load of 60 W and increasing by 15 W every minute until the criteria for determining the maximal oxygen uptake (VO₂ max) were met (Cosmed, città metropolitana, Italy). The body composition was evaluated using an InBody 720 device (InBody Co., Ltd., Seoul, Republic of Korea).

The subject continuously monitored heart rate variability (HRV) from November 2018 to December 2022. Measurements were taken each morning while lying down after waking up. Each measurement lasted one minute, while spontaneous breathing was maintained throughout. At the end of each measurement, the subject completed a standardized questionnaire through a mobile app. The questionnaire assessed physiological and subjective factors, including sleep quality, potential illnesses, injuries, and menstrual cycle phases. Alongside these self-reported variables, we analyzed the impacts of sleep duration and training load on HRV [30]. The acute training load (ATL) was evaluated by multiplying the average training intensity (I), assessed through subjective ratings on the Borg CR-10 scale (0 = minimal effort; 10 = maximal effort), with the total weekly training volume (V), represented by the duration of training sessions. The equation derives from a version of the original Banister equation for quantifying training impulses (TRIMPs), which considers RPE intensity ratings instead of heart rates [33].

$$\mathrm{A}\mathrm{T}\mathrm{L}={\sum }_{\mathrm{i}}{\displaystyle \frac{{\mathrm{V}}_{\mathrm{i}}·{\mathrm{I}}_{\mathrm{i}}}{\mathrm{n}}}·{\displaystyle \frac{\mathrm{n}}{100}}{\sum }_{\mathrm{i}}{\mathrm{V}}_{\mathrm{i}}$$

A performance index (PI) was calculated under the guidance of the International Canoe Federation (ICF) using results from international competitions to assess athletic performance. The PI formula includes competition quality, event significance, and achieved results. PI derives as the normalized sum of two indices (PI₁ and PI₂): the first evaluates results based on the average of the top three overall performances, irrespective of gender or category, while the second explicitly assesses results relative to the average of the top three performances in the women’s kayak category. This dual approach offers a more precise evaluation of performance, avoiding over- or underestimations due to smaller or less competitive participants in women’s kayak events.

$$\mathrm{P}\mathrm{I}={\mathrm{P}\mathrm{I}}_1+{\mathrm{P}\mathrm{I}}_2$$

$${\mathrm{P}\mathrm{I}}_1={\displaystyle \frac{\mathrm{p}\mathrm{h}\mathrm{a}\mathrm{s}\mathrm{e} \mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r} \mathrm{x} \mathrm{r}\mathrm{e}\mathrm{s}\mathrm{u}\mathrm{l}\mathrm{t} \left(\mathrm{s}\mathrm{e}\mathrm{c}\right)}{\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e} \mathrm{r}\mathrm{e}\mathrm{s}\mathrm{u}\mathrm{l}\mathrm{t} \mathrm{o}\mathrm{f} \mathrm{t}\mathrm{o}\mathrm{p} 3 \mathrm{a}\mathrm{t}\mathrm{h}\mathrm{l}\mathrm{e}\mathrm{t}\mathrm{e}\mathrm{s}}}-\mathrm{p}\mathrm{h}\mathrm{a}\mathrm{s}\mathrm{e} \mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}+\mathrm{q}\mathrm{u}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y} \mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}$$

$${\mathrm{P}\mathrm{I}}_2={\displaystyle \frac{\mathrm{p}\mathrm{h}\mathrm{a}\mathrm{s}\mathrm{e} \mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r} \mathrm{x} \mathrm{r}\mathrm{e}\mathrm{s}\mathrm{u}\mathrm{l}\mathrm{t} \left(\mathrm{s}\mathrm{e}\mathrm{c}\right)}{\mathrm{a}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e} \mathrm{r}\mathrm{e}\mathrm{s}\mathrm{u}\mathrm{l}\mathrm{t} \mathrm{o}\mathrm{f} \mathrm{t}\mathrm{o}\mathrm{p} 3 \mathrm{K}1\mathrm{W}}}-\mathrm{p}\mathrm{h}\mathrm{a}\mathrm{s}\mathrm{e} \mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}+\mathrm{q}\mathrm{u}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y} \mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}$$

Competition quality and event significance are directly related, with event significance increasing alongside the quality of competing athletes. The competition quality factor is calculated based on the average of the ICF ranking points of the top three competitors. For performance evaluations, the best result achieved during a single competition day was considered, calculated as the sum of the paddling time and penalty seconds incurred by gate touches. The performance index was analyzed in relation to HRV values on competition days and the HRV trend during the competition week. The HRV trend was measured as the relative difference between the weekly average HRV and the mean HRV from the previous 60 consecutive days, a sufficiently long period to capture meaningful physiological adaptations and to provide a stable reference for assessing variations in autonomic regulation.

Data analysis was conducted using Excel (Microsoft Ltd., Redmond, WA, USA) and MATLAB 2023b (MathWorks, Natick, MA, USA). A linear mixed-effects model, accounting for fixed and random effects, was applied, with statistical significance set at p < 0.05. The effects of individual factors on rMSSD were analyzed, considering the day as a random variable. The analysis of sleep duration and quality, menstrual cycles, and illness effects on HRV encompassed 1394 measurements collected daily over four years. The relationship between rMSSD, HRV trends, and the performance index were examined. Over four competitive seasons (2019–2022), HRV data from 83 kayak slalom competition days were analyzed. The data were presented using descriptive statistics (means and standard deviations) and visualized graphically.

3. Results

The average morning rMSSD was 52.43 ± 16.77 ms (95% CI: 51.54–53.31 ms), while the average resting heart rate (HR) in the morning was 63.68 ± 4.97 bpm (95% CI: 63.42–63.94 bpm). During the competition week, the rMSSD trend was, on average, 7.51 ± 18.37% lower than the preceding 60-day average. On competition days, the average resting HR was 68.79 ± 6.24 bpm, indicating a 3.25% increase from the overall average. The average rMSSD on competition days was 44.08 ± 16.13 ms, 4.4% lower than the general daily average.

The analysis indicated a statistically significant positive effect of sleep quality (Figure 1a) on rMSSD (p < 0.001). The model considered fixed effects and random daily variations, allowing for a better understanding of the relationship between sleep quality and heart rate variability. However, the duration of sleep (Figure 1b) did not have a significant effect on rMSSD (p = 0.267). The random effects parameters reveal substantial day-to-day variability in rMSSD (SD = 14.01), suggesting that additional daily factors influence heart rate variability.

The model examined the menstrual cycle in relation to HRV and identified a statistically significant effect of the follicular phase on rMSSD (p = 0.003). The analysis showed an increase in rMSSD during the follicular phase (Figure 2a). The average HRV (rMSSD) during the follicular phase was 53.76 ± 29.37 ms compared to 51.05 ± 27.90 ms in the luteal phase. Substantial daily variability was observed in both baseline rMSSD values and the impact of the follicular phase on rMSSD.

Illnesses were recorded within 60 days during a four-year monitoring period (Figure 2b). The average rMSSD during illness was 50.46 ± 21.51 ms compared to 52.50 ± 16.57 ms when illness was absent. The model indicates that illness has no statistically significant effect on rMSSD (p > 0.05). The random effects parameters suggest substantial day-to-day variability in the baseline rMSSD and its relationship with illness.

In this longitudinal case study, we analyzed the impact of the acute training load (ATL) on HRV over a competitive season (Figure 3). The model indicates that the ATL has no statistically significant effect on rMSSD (p = 0.94).

The results further suggest that despite daily variability in the competition performance index, rMSSD1 is not a statistically significant predictor of performance (p = 0.82). Similarly, the rMSSD trend does not significantly predict performance (p = 0.70).

On average, the HRV trend during competition weeks was 7.51 ± 18.37% lower than the preceding 60-day average, while the HRV on competition days was 4.4% lower than the overall average.

4. Discussion

The study findings reveal that heart rate variability (HRV) is a crucial indicator of vital physiological processes related to athletes’ recovery and fitness. In this study, HRV was measured using HRV4Training, a validated mobile application utilizing photoplethysmography (PPG) technology. This method has been shown to be highly accurate and comparable to traditional electrocardiography (ECG)-based measurements, making it a reliable tool for long-term physiological monitoring in elite athletes [32]. Sleep quality is the most significant factor, underscoring the need for regular monitoring and improvements of sleep patterns. In this study, sleep quality was evaluated using a standardized subjective questionnaire, which captured the participant’s perceptions of restfulness, disturbances, and overall recovery. While this method offers valuable insights into perceived sleep experiences, it does not account for objective physiological parameters, such as sleep duration or architecture. Nonetheless, previous research has demonstrated that both subjective and objective indicators of poor sleep are linked to parasympathetic inhibition and sympathetic activation, reducing HRV values [34,35,36,37]. Our findings align with previously published studies. Flatt et al. [26] have examined how sleep quality impacts HRV in elite swimmers, confirming that lower sleep quality adversely affects HRV. Werner et al. [38] compared both subjective and objective assessments of sleep quality with morning HRV values in healthy adults. Their findings indicate that higher-quality sleep, measured by both methods, was associated with increased morning HRV values. Furthermore, they observed that nocturnal HRV values were less reliable sleep-quality indicators than morning HRV measurements. However, none of the studies measured HRV for as many days as our study.

Existing research on the relationship between sleep duration and HRV has produced mixed results. Some studies indicate that sleep deprivation causes parasympathetic inhibition and sympathetic activation, which appear as reduced HRV [39,40]. These findings highlight the complex interaction between sleep parameters and the regulation of the autonomic nervous system and emphasize the need for further investigation of the mechanisms underlying these relationships [41]. However, a study by Bourdillon et al. [31] compared HRV in volunteers who slept for 5 consecutive nights for more than 7 h with those who slept for 5 consecutive nights for less than 5 h, which really happens to professional athletes. Our findings support previous studies showing that sleep duration has little effect on HRV [35,38]. In a study by Sajjadieh et al. [27], participants maintained their usual sleeping patterns, making this finding more comparable to our research than a study conducted by Bourdillon et al. [31], mentioned in the previous paragraph, which involved sleep restriction over five consecutive nights. It is important to note that the population measured in the study by Sajjadieh et al. [27] consisted of medical staff, which cannot be directly compared to elite athletes. Previous research has shown similar average sleep durations, which, along with low variability, generally suggests that adults are getting adequate sleep based on current guidelines [42].

We found that the menstrual cycle significantly influences HRV, reflecting the dynamic interplay between hormonal fluctuations and the regulation of the autonomic nervous system. Several research studies have consistently [12,43,44,45] shown that HRV changes throughout the menstrual cycle, with higher values noted during the follicular phase than in the luteal phase. This pattern is primarily attributed to fluctuations in estrogen levels, which modulate parasympathetic activity. Elevated estrogen concentrations during the follicular phase enhance vagal tone, leading to increased HRV, whereas the decline in estrogen during the luteal phase is associated with reduced parasympathetic influence and lower HRV.

Research has reported varying findings regarding the relationship between the training load and HRV in elite athletes. Some studies suggest that HRV progressively decreases with an increasing training load [21,26], while others reported statistically significant effects without a consistent pattern [46]

Despite these discrepancies, researchers generally agree that a stable or increasing HRV trend in response to higher training loads indicates a positive adaptation to training. At the same time, a decline in HRV may signal excessive training stress, which could lead to overtraining [47,48]. Our study found no significant variations in HRV trends related to the training load. This lack of differences may be attributed to individual differences in autonomic regulation, variations in methodology, or effectively managed training loads that limit excessive physiological stress. Additional research is essential to clarify the relationships between these findings further.

In previous studies, HRV has rarely been investigated in the context of competitive sport performance. Several studies reported a positive correlation between HRV and ability testing. DeBlauw et al. [49] reported that lower daily HRV values are associated with poorer abilities on bikes. Stepanyan and Lalayan [50] found that underperforming athletes have higher parasympathetic tone and lower HRV compared to performing athletes in a competitive context. The association mentioned above may initially seem surprising, as higher HRV is generally associated with better health and greater performance, but researchers have found that HRV usually decreases before competitions, as an increase in sympathetic activation is linked to pre-competition stress. An increase in sympathetic activation correlates with a higher cardiac output, elevated catecholamine release, and improved oxygen delivery to active muscles, all of which contribute to enhanced acute exercise performance [22,23,51]. Some research reports a decrease in HRV only before competitions, which athletes consider to be more important [23,52].

Our study indicates a similar pattern to previous studies, with HRV values showing a decrease during competition weeks compared to non-competition periods. This difference did not reach statistical significance, but this trend is consistent with previous findings. HRV during the competition week was lower than in the preceding period, with an additional decrease on competition day. Deblauw et al. [53] reported a reduced HRV during the week of competition, but there was no evidence of an effect on competitive performance in elite rowers. Coyne et al. [54] found that lower HRV appeared to be a good indicator of sports performance in elite long jumpers participating in the Rio de Janeiro Olympic Games.

While few studies have explored the effect of HRV on athletic performance, they indicate the possible value of HRV as a marker of athletic performance. Depending on the specific demands of each sport, future research should explore the value of an autonomic profile that enhances competitive performance. While our study did not establish a clear relationship between HRV and sports performance, the decline in HRV we observed before competitions indicates a possible advantage of increased sympathetic activation.

5. Conclusions

This study provides valuable insights into the factors influencing HRV in elite athletes and explored its potential as an indicator of sports performance. The analysis of HRV in relation to sleep quality, the menstrual cycle, training loads, and illnesses revealed that sleep quality is the most significant factor influencing HRV, with positive associations observed between improved sleep and higher HRV values. While no clear relationship between HRV and competitive performance was established in this study, the observed pre-competition decrease in HRV suggests potential benefits of sympathetic dominance for performance in kayak slalom. These findings align with existing research that shows a decline in HRV before competitions, particularly in endurance sports, where increased sympathetic activation can enhance performance. However, further research with a broader sample and controlled experimental conditions is needed to clarify the potential role of HRV as a predictor of sports performance. HRV monitoring offers a non-invasive, cost-effective, and practical tool for coaches and athletes, enabling daily assessments of an athlete’s physiological state. Regularly tracking HRV can help optimize training adaptation, identify signs of fatigue or overtraining, and adjust workloads to enhance recovery and performance. The primary objective of this study was to analyze factors influencing HRV and to determine if HRV can serve as an indicator of athletic performance. Our research indicates that the training load minimally affects HRV, potentially because of successful adaptation and the absence of overtraining. The observed decrease in HRV before competitions suggests a potential advantage of sympathetic dominance in kayak slalom, similar to trends observed in endurance sports. As a case study focusing on a single subject, the results cannot be extrapolated to the general population. Additionally, since HRV is affected by gender and age, there is a necessity for future research to explore a broader sample, including participants of various sexes and age demographics. This longitudinal study was conducted without any interventions during the training process. A larger dataset, more recorded observations, and improved control conditions are necessary to analyze factors influencing HRV, like injuries and illnesses, accurately. Moreover, specific data in this study rely on self-reported metrics and subjective evaluations, such as sleep quality and training loads. Future studies should implement more objective assessment techniques to validate these findings.

6. Limitations

The primary limitation of this study lies in its case study design, focusing on a single elite female kayak athlete. Therefore, the results cannot be generalized to the wider population of athletes, especially considering the variability in HRV responses due to gender, age, and individual training adaptations. Furthermore, HRV responses may differ across various sports due to differences in physiological demands, training intensities, and competition schedules. Additionally, the reliance on self-reported metrics for sleep quality and training loads introduces a level of subjectivity into the analysis. Although these subjective assessments provide valuable insights into the athlete’s perceived state, they may not fully reflect objective physiological conditions. Lastly, a larger sample size and more rigorous control conditions would help refine the conclusions and enhance the applicability of the findings to other competitive athletes. Future studies should explore these aspects to improve the accuracy of HRV as a marker for performance optimization.