Monitoring of Heart Rate Variability (HRV) is a crucial component of the periodization of an athlete’s training load in order to ensure the athlete is properly recovered and ready to achieve their optimal performance level. HRV is known as the time between heart beats and is a common measure of stress placed on the body. Decreases in HRV over long periods of time typically results in decreased performance, overtraining and chronic fatigue (Coker et al., 2016). This is especially important in dealing with increased injury risk in female collegiate soccer players during a season (“Women’s Soccer Injuries,” 2009) due to repeated high intensity matches with minimal time to recover over long periods of time. Soccer sport scientists who are seeking to reduce injury and improve performance of their athletes should focus on an individualized periodization centered around the biofeedback of the athlete.
In just one match, a female soccer player will perform 250 sprints and cover a distance of 8.6-11.3km (Turner at al., 2013). Female collegiate athletes must balance two games, practice, lift, school and travel, which adds to the chronic stress of the athlete ultimately decreasing the HRV (Schubert, 2009). By having a more comprehensive understanding about the relationship between training load, sleep and HRV, sport scientists will be able to better periodize training loads subsequently reducing injury and improving performance.
In previous research performed on division 1 football players, it was noted that linemen did not return to their HRV baseline before the next session held following day (Flatt et al., 2017). Furthermore, in a research performed by Coker at al., male collegiate soccer players running performance decreased throughout the season (Coker et al., 2016). Thus, indicating that taking HRV measurements can provide powerful information in creating an optimal human performance environment.
Purpose of the Study
The primary purpose of the study was to determine the effect of training load on heart rate variability. A secondary purpose was to determine the relationship between recovery and heart rate variability.
Research Questions
Does the training load effect heart rate variability? How does HRV change three days post soccer match?
Operational Definitions
In this study, biofeedback measurements are taken for analysis such as heart rate variability (HRV), resting heart rate (RHR), sleep hours, distance in meters and training load (TL). HRV is a combination of feedback from both the cardiorespiratory and autonomic nervous system. The autonomic nervous system comprises of the parasympathetic system which lowers the body’s heart rate and the sympathetic system which elevates the body’s heart rate, which is where the HRV score is derived from. TL is a measurement that is calculated based on heart rate, distance, age, weight, VO2max and training history (“Training Load,” 2018).
Methods
This was a non-experimental case study research design which analyzed the effect of training load and recovery on HRV.
Sample
Since this is a case study, the unit of analysis is the events that occurred throughout the season such as games, training sessions, recovery and off days. The participant observed was a National Collegiate Athletics Association (NCAA) Division 1 female soccer player (22 years old; 58.05kg; 167.64cm) who started all matches throughout the season. Her YYIRT score prior to the season was 48 (1920m). The participant participated in 1311 minutes of match time throughout the entire season and participated an average of 72.83 minutes per 90-minute match and covered an average distance of 11.47km per match.
Sampling Method
No sampling was performed as it was a case study research. All observations were performed on one participant.
Performance Instruments
The WHOOP strap 2.0 was used to assess the objective physiological state of the participant such as HRV, RHR, HR and the subjective data such as energy level, soreness, illness, injury, and stress level. The HRV and RHR is routinely taken while the participant is asleep to ensure accuracy. The Polar team pro was used to register HR, distance (m), duration and training load of games and training sessions. The WHOOP strap 2.0 was worn at all times where the polar chest strap was worn during training and games.
Results
Part 1: Descriptive Statistics
Frequency Distributions. The participant competed in games 20% of the days (n=19). Trained 40% of the days (n=38) and performed recovery sessions (n=14) such as pool sessions, foam rolling, and stretching 14.7% of the days and was off 20% of the days (n=20). A total of 95 days was analyzed in the research (n=95). In summary, the participant was active in games, recovery and training for 78.9% of the days.
Table 1. Performance Type Distribution
Measures of central tendency and variability. The average HRV score of the participant throughout the season was 50.3 with a standard deviation of 26.8. Due to the existence of multiple modes, it is not recommended to provide estimate ranges in accordance to normal distributions proportions. However, it can be stated that 50% of the observed values are between 32 and 64, based on the percentile values. The mean of RHR is 55.3 with a standard deviation of 6.4. The participant acquired a median of 7:01 hours of sleep, meaning 50% of the time the participant got less than 7:01 hours of sleep. The H-Spread for polar training load is the difference between the third and first quartile which were 204 and 30, indicating a H-spread of 174 as an addition measure of dispersion. There is a large difference between the standard deviation and the H-Spread, 100.39 and 174, respectively. The mode reported for polar distance (m) in games and training sessions was 2190m and was also noted to be multi-modal.
Table 2. Participant Biofeedback and Performance
Boxplot. Figure 1 displays a boxplot with the HRV scores post matches from one day after to five days after a soccer match. The HRV one day post-match had a mean of 26.5, this increased by 81.1% by two days post-match. Three days post-match the HRV had a mean of 50.5, this number decreased on day 4, but increased significantly 5 days after the match. Five days post-match, the mean of the HRV had significantly increased by 169% from the day one post-match HRV. Day 2 post-match HRV have one outlier by cause of a score over 100. Box plots for one day and 3 days after a match are relatively compact and indicate that HRV values are similar, with the exception of the lower whisker three days post-match. The boxplots for 2,3,4 and 5 days after the match are taller and the lower upper quartile ranges do not overlap any part of the day 1 post-match range. All of the boxplots except day 1 post-match have sections that vary in size indicating skewed distributions.
Figure 1: HRV post-matches
Bar graph. Figure 2 represents the types of activities and their frequency over the 95 days observed in the soccer season. The nominal variables are trainings, games, recovery sessions and off days. Training sessions occurred most frequently (n=38). Off days occurred the second most frequently (n=20), and were comparably frequent to matches (n=19) at about 20%. Recovery days were the least frequent (n=14). The participant did not perform recovery after each game only having 14 recovery sessions in total. Overall, this resulted in an active days (training, matches) to less active days (recovery, off) ratio of 57:34 or 1.68, indicating 5 active days for every 3 less active days.
Figure 2: Type of Activity Distribution
Histogram. Figure 3 shows a slightly negatively skewed distribution of the hours of sleep. The mean is 6:53 with a standard deviation of 1:39 from a sample size of 86. The mode is in the interval from 7:30 hours of sleep and 8 hours of sleep. The Kurtosis value was -.429 indicating a fairly normal peak of the distribution. The skewness value was -.341 indicating more values were to the right of the mean. 68% of the time the participant would attain from 5:31 to 8:16 hours of sleep, under the assumptions of data distributed according to the normal distribution.
Figure 3: Histogram of Hours of Sleep
Part 2: Inferential Statistics
Correlation. Pearson Correlation Coefficients were calculated to resolve if there was a relationship between polar training load and the post load HRV. The independent variable was the polar training load, where the dependent variable was the HRV post training. The analysis was conducted at a .01 alpha level. The null hypothesis is that the Pearson coefficients are equal to 0, meaning there is no relationship between training load and HRV. The alternative hypothesis is that the Pearson coefficients are not equal to 0 meaning there is a relationship between training load and HRV. The null and alternative hypothesis are symbolled below:
HO:rxy=0
H1:rxy 0
Table 3. Pearson Correlation Coefficients (n=55)
A scatterplot was configured to test for the assumption of linearity. There was no evidence of non-linear or curvilinear relationship. The assumption of independence was not met.
There is a significant negative relationship in that when the polar load increases, the post HRV will decrease (r(53)=-.409, p=.002). Therefore, we reject the null hypothesis at the alpha level of .01. This implies that there is a moderate to large effect size. Post hoc G*Power was scored as a high power (.740).
Dependent T-test. The dependent t-test was administered in order to determine the effects of three days recovery between games on HRV. The independent variable is three days recovery, and the dependent variable is the HRV score. The alpha level was .01. The null hypothesis of a paired t-test is equal to 0, meaning no change of HRV. The alternative hypothesis of a paired t-test is not equal to 0, meaning there is a change of HRV. As such, the hypotheses are symbolized below:
HO:d=0
H1: d 0
Table 4. Paired Sample T-Test.
There is a statistically significant increase in three days after a game from 29.6910.77 to 55.0823.02 (t(12)=-5.340, p<.01). Therefore, we reject the null hypothesis at a less than .01 level of significance meaning there is a positive increase in HRV. The assumption of normality was met using the Shapiro-Wilk test. The Cohen’s d effect size is very large as the HRV improved by over 1 S.D after three days of rest.
Discussion
Over 95 days, the participant competed in 19 games. The participant should have had ample time to recover between each match, but improper periodization of matches makes this very difficult. Due to scheduling of NCAA matches, the participant was required to play games with few days to recover which led to a lower HRV. HRV scores post-match were 47.3% lower than the reported average HRV score. These scores improved with two days recovery, but not enough for the athlete to fully recover. However, from the post-match HRV to five days post-match there was a 169% increase in HRV scores allowing for better recovery for the athlete. This implies that one match per week would be more manageable for the athlete’s health.
There was an inverse relationship between polar training and heart rate variability. This signifies that a high training load may produce a lowered HRV. With women collegiate soccer players needing to be fully recovered after each match, lowering training load when possible may allow the athletes to increase HRV scores prior to the next match. Monitoring HRV scores prior to prescribing training load for that day should also be considered to better balance the athletes’ recovery.
There was also a positive change observed on the HRV score after three days recovery. This indicates that athletes with more recovery days between games may increase HRV scores, thus further preventing injury and improving performance. Consequently, more training between matches may be more detrimental than advantageous to the athlete’s performance.
Limitations
The main limitation of the study is the method of research being a case study. Firstly, the results can’t be used for the general population due to the sample size. Second, since the data was collected on the author, it is possible that there is researcher bias. Third, the subject was only a college aged female women’s soccer player and HRV can have a different range based several different factors such as age (Jandackova, Vera K. et al., 2016).
Future Research
Further research is necessary on HRV in providing proper periodization for women soccer players. With the NCAA schedule, research needs to be done to provide proper periods of recovery between matches to better the health of athletes. Secondly, more research needs to be done on the effects of a low HRV and how that may affect the student athlete outside of their sport, such as school performance. Lastly, further explanation is needed on ways of increasing HRV such as sleep, food, hydration etc.
References
Coker, N. A., Wells, A. J., Ake, K. M., Griffin, D. L., Rossi, S. J., & Mcmillan, J. L. (2016). Relationship Between Running Performance and Recovery-Stress State in Collegiate Soccer Players. Journal of Strength and Conditioning Research,31(8), 2131-2140.
Flatt, Andrew & Esco, Michael & R. Allen, Jeff & Robinson, James & Earley, Ryan & Fedewa, Michael & Bragg, Amy & M. Keith, Clay & Wingo, Jonathan. (2017). Heart Rate Variability and Training Load Among National Collegiate Athletic Association Division 1 College Football Players Throughout Spring Camp. The Journal of Strength and Conditioning Research. 14-15.
Jandackova, Vera K. et al. “Are Changes in Heart Rate Variability in Middle‐Aged and Older People Normative or Caused by Pathological Conditions? Findings From a Large Population‐Based Longitudinal Cohort Study.” Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease 5.2 (2016): e002365. PMC. Web. 15 June 2018.
Schubert, C., Lambertz, M., Nelesen, R. A., Bardwell, W., Choi, J.-B., & Dimsdale, J. E. (2009). Effects of stress on heart rate complexity—A comparison between short-term and chronic stress. Biological Psychology, 80(3), 325–332.
Turner, Ellena & Munro, Allan & Comfort, Paul. (2013). Female Soccer. Strength and Conditioning Journal, 35, 51-57.
Training Load | Let’s Talk Polar. (2018, April 26). Retrieved June 10, 2018, from https://www.polar.com/blog/training-load-lets-talk-polar/
Women’s Soccer Injuries: Data from the 2004/05-2008/09 Seasons. (2009). Retrieved June 15, 2018, from https://www.ncaa.org/sites/default/files/NCAA_W_Soccer_Injuries_WEB.pdf