Use of Heart Rate Variability in Monitoring Stress and Recovery in Judo Athletes : The Journal of Strength & Conditioning Research

Secondary Logo

Journal Logo

Original Research

Use of Heart Rate Variability in Monitoring Stress and Recovery in Judo Athletes

Morales, José1; Álamo, Juan M.2; García-Massó, Xavier3; Buscà, Bernat1; López, Jose L.4; Serra-Añó, Pilar3; González, Luís-Millán5

Author Information
Journal of Strength and Conditioning Research: July 2014 - Volume 28 - Issue 7 - p 1896-1905
doi: 10.1519/JSC.0000000000000328
  • Free



High-standard sport performance requires high training loads and accurate strategies for its control. Volume and intensity are the most important elements of the training load, but the time and characteristics of the recovery process between sets and repetitions to reach optimal performance are also significant. For coaches and sports physicians, it is essential to establish an appropriate training load for each period of the season to allow for an effective tapering period to achieve better performance and prevent any potential overtraining (3). The main characteristic of the preparation periods is an increase of training load volume and frequency. In these periods, the balance between training and recovery is crucial for reaching an optimal adaptation for the next period and avoiding overtraining (4). One of the main reasons for overtraining in athletes is the lack of control over the load magnitude and the recovery time between loads. Nowadays, there are numerous tools to guide coaches to accurately quantify the efforts imposed on their athletes, and therefore, recovery times can be administered consistently. However, these developments have taken place mainly in so-called “noncontact sports,” i.e., sports in which direct interaction between athletes is scarce.

However, sports in which direct contact is permitted (e.g., rugby, water-polo, wrestling, etc.) have great difficulty in controlling the intensity to which athletes are subjected during each specific training session. To a large extent, this difficulty results from the interaction between different subjects, as depending on how they act, their loads may increase or decrease without those involved being aware of it.

Judo is clearly one of the contact sports in which interaction between opponents is greatest, because technical actions are directly targeted to the opponent's body. Obvious difficulties in monitoring and quantifying training load amounts are found because of the characteristics of the sport (including group organization, different opponents, and combat development). During the training process of a judo athlete, there are also individual conditioning sessions where it is much easier to quantify the training load. However, these sessions are not enough to monitor the global effects of training. Previous studies focused on the demands of judo competition (8,11) have attempted to understand the effects of training programs on athletes. For instance, Koga et al. (24) studied the physiological effects of high-intensity training in male judo athletes over 3 months, and observed a significant increase in neutrophil function and a decrease in creatine kinase levels. Umeda et al. (42) observed a negative association between mood disturbance and changes in myogenic enzymes during a training camp with female judo athletes.

The use of heart rate variability (HRV) as a training tool has progressively increased and deserves attention as a tool to monitor the possible states of overtraining and recovery after a training process (1,9,30,36). Heart rate variability is defined as the heart's ability to produce fluctuations in the beat-to-beat interval in response to different situations (39). The relationship between autonomic modulation and HRV is different during exercise and immediate recovery compared with rest periods (16). Because of the importance of stationarity for spectral analysis, it is recommended that HRV spectral analysis should not be performed during exercise and early recovery (39).

Several studies suggest that the quantification of HRV can be used as a noninvasive method for assessing autonomic cardiovascular control through the impact of HRV on beat-to-beat heart rate modifications (2,38,39). When athletes are training with heavy loads during a period of time, training can produce cumulative stress-related effects. The magnitude of this stress response can be observed by the activation of the sympathetic arm of the autonomic nervous system and the variations in autonomic balance, which can be indirectly assessed using HRV analysis (38).

From a psychological perspective, there are instruments that attempt to address stress regeneration in athletes. Kellmann and Kallus (20) developed the Recovery Questionnaire for Athletes (RESTQ-SPORT), which is frequently used in research to observe the balance between stress and recovery during training processes in rowing (19), kayaking (21), and athletics (17). This instrument is much appreciated by coaches and athletes for its ability to help in determining the appropriate preseason and high training period training loads and for its simplicity. However, we suggest that a simple physiological analysis, combined with the psychological data of perceived exercise exertion and recovery, could be a better integrated methodology in assessing the stressful effects of training loads in judo. Thus, the aim of this investigation was to examine the effect of different judo training loads on HRV measurements and to determine if they can be used as valid indicators in monitoring stress and recovery in judo athletes. We simultaneously measured changes in stress and recovery using RESTQ-SPORT (20) to determine their links with HRV in different training loads. We hypothesized that a low HRV may be a valid indicator in quantifying a high level of stress because of specific training loads in judo. We also expected higher sympathetic and lower parasympathetic activity in judo athletes with high training loads in the HRV analysis.


Experimental Approach to the Problem

A random, experimental study was conducted in which 2 groups of judo athletes were subjected to different types of training. The sample was divided randomly into 2 groups, with each group following a different type of training. Simple randomization was used, rather than techniques such as matched pair randomization or block randomization. The first group followed a high training load (HTL) program with daily judo sessions and additional endurance, strength and physical fitness training sessions twice a week. The second group followed a moderate training load (MTL) program with only 3 specific judo sessions per week. Although the latter was not exactly a control group, the training carried out by them was not expected to result in changes in dependent variables. That is, it could be seen as a placebo group (subject to an intervention with no effect on the variables of interest).

Pretesting was performed at the beginning of the training period (September) after a 1-month transition period without systematic judo training (baseline). The data collection sessions included HRV monitoring, RESTQ-SPORT administration, and strength measurements. Posttesting was carried out 4 weeks after the training program to evaluate the effects of the training. The researchers who conducted the training program did not know the objectives of the study, and therefore did not influence the training program.


Fourteen male national-standard judo players were recruited to participate in this study (age = 22.85 [19–25] years; height = 174.08 [6.89] cm; body mass = 76.85 [60–120] kg). The participants were divided into 2 equal-sized groups. The characteristics of the subjects in both groups are shown in Table 1. There were no statistical significant differences between the groups in these variables (p ≤ 0.05). All the participants were informed in advance of the nature of the research and provided their written informed consent. The ethics committee of Ramon Llull University of Barcelona approved the conduct of this study, which complied with the latest version of the WMA Declaration of Helsinki.

Table 1:
Subject's characteristics.*


Heart Rate Variability Measurements

The HRV was recorded at the beginning of the testing session. All subjects were asked to refrain from ingesting beverages containing caffeine or alcohol during the 48 hours preceding the tests. At 8.00 AM, after an adaptation period of 10 minutes in supine decubitus, data acquisition was initiated by a cardiotachometer and RR signal (on a beat-to-beat basis) for 10 minutes, with a temperature of between 20 and 22° C, in the absence of noise and with no negative atmosphere. The subjects were asked to remain silent and attempt to maintain their breathing rate as low as possible, without speaking or moving at all. The data collection was always performed by the same researcher, who maintained the same conditions for, and gave identical instructions to all the participants.

To record the HRV, a Polar S810 cardiotachometer and coded transmitter (Polar Electro Oy, Kempele, Finland) were used to record the RR signal. This series of heart rate monitors has been validated for HRV recordings. The reliability of the protocol and cardiotachometer used was ICC >0.81 (32).

The heart rate data were stored in a personal computer using Polar Precision Performance software (version 3.0). The signal processing was performed using Kubios HRV analysis software (version 2.0; Biosignal Analysis and Medical Imaging Group, Department of Physics, University of Kuopio, Finland) with time, frequency, and nonlinear domain analysis. The methodological criteria that were applied were those proposed by the Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology (39).

In the HRV time domain, the mean RR interval, the SD of RR, the heart rate, the SD of the heart rate, the square root of the mean squared difference of successive RR intervals (RMSSD), the number of consecutive RRs that differed by more than 5 milliseconds each, and the percentage of consecutive RRs that differed by more than 5 milliseconds each were obtained.

A Fast Fourier Transform of the RR signals was used for the HRV frequency domain analysis. The spectral response provided by the system was broken down into 3 bands: very low frequency (0.003–0.04 Hz), low frequency (0.04–0.15 Hz), and high frequency (0.15–0.4 Hz).

The nonlinear analysis techniques used in this study were the Poincaré plot and detrended fluctuation analysis. The Poincaré diagrams were obtained by plotting the RR values of n on the x-axis and the RR values of n + 1 on the y-axis. The SD1 axis indicates short-term variability, whereas the SD2 axis indicates long-term variability. A greater dispersion of the scatter plot is associated with relaxation and a well-balanced autonomic nervous system, whereas a narrower dispersion is associated with an imbalance with a predominance of sympathetic activity (27) (Figure 1). Superimposed on this diagram, the short-term variability and long-term variability were calculated by means of ellipse fitting. In the detrended fluctuation analysis, 4 steps were performed: (a) a data series was obtained by integrating the differences in the average scores, (b) the dataset was divided into nonoverlapping n segments, (c) the local trend of each segment (least-square fit) was calculated and the variance between the actual data and the local trend was obtained, and (d) the average value of the variance of the segments and the square root was calculated. This procedure was repeated for different segment sizes. Once these measurements were performed, the scaling exponent was obtained by calculating the slope of the log fluctuation to the log segment plot size. The scaling exponent was calculated for long range and short range.

Figure 1:
Example of Poincaré plot. The RR signal of one study participant in the pretest and posttest are plotted in layers (A) and (C), respectively. The Poincare plot corresponding to both RR signals is shown in layers (B) and (D). The large axis of the ellipse represents the long-range variability, whereas the little axis of the ellipse represents the short-term variability.

Stress-Recovery Assessment

After the HRV measurements, the RESQT-SPORT was administered to assess the frequency of perceived general and sport-specific stress, along with the frequency of recovery-associated activities. The questionnaire consists of 77 items and 19 scales, with 4 questions per scale and 1 warm-up question. The final score of each scale is the mean of the points obtained in each item. The subject's responses were established according to a Likert-type scale with values ranging from 0 (never) to 6 (always), which indicated how often the respondent participated in the different activities in the preceding 3 days/nights. In addition, the RESTQ-SPORT scores were classified as general stress scores (mean of 7 scales), general recovery scores (mean of 5 scales), sport-specific stress scores (mean of 3 scales), and sport-specific recovery scores (mean of 4 scales). The Spanish version of the RESTQ-SPORT was validated by González-Boto et al. (13), who reported a good internal consistency with Cronbach's alpha indexes that ranged from 0.77 to 0.94 with significant correlations between the RESTQ-SPORT scales and the profile of mood states scales. Also, this scale has good test-retest reliability for all of the scales (r > 0.79) (18).

Strength Testing

Strength testing included the measurement of 1 repetition maximum strength (1RM) and maximum power production with different loads in a bench press exercise, whereas isometric strength was measured using a hand-grip exercise. The subjects performed a free-weight bench press test of 1RM using standard methods described by Kraemer and Fry (25). Starting with a warm-up set of 5–10 repetitions, the subjects performed 40–60% of the perceived 1RM. After a 90-second rest period, a single set of 2–3 repetitions was performed at 60–80% of the perceived 1RM. Consequently, 3–4 maximal trials of 1 repetition were performed to determine the 1RM with a 3-minute rest between trials. The technique for the bench press exercise, as described by Zatsiorsky and Kraemer (44), was explained to the subjects and corrected as necessary.

We measured the maximum power during bench press exercise through a progressive load protocol using the linear encoder of the Musclelab System (Ergotest, Langesund, Norway) attached to the weight bar, giving 1 pulse approximately every 7.5 × 10−4 m of load displacement. Every 0.01 seconds (sampling frequency = 100 Hz), the total counts of pulses were read and the displacement was calculated. Data were transferred to a personal computer and analyzed using Musclelab Software (version 7.18). We used loads of 30, 40, 50, and 60% of 1RM. Subjects were asked to perform between 3 and 5 repetitions at maximum speed in a concentric phase, avoiding separating the chest from the bench. We used the most powerful repetition for analysis. A 5-minute rest period between sets was used.

Isometric hand-grip strength was measured using a digital dynamometer H500 E-link (Biometrics, Newport, United Kingdom). The subjects were asked to support their back on the wall, flex their legs 45°, with their arms at their side. In this position, they were required to attempt maximal grip pressure (right and left) for 10 seconds, while the dynamometer registered the time-force curve. The maximum measurements of the time-force curves were selected for further analysis.

The strength-related variables were expressed relative to body weight before statistical analysis. This was intended to prevent the weight of the participants in both groups affecting the comparison between the groups.

Training Program

The HTL group underwent a 4-week intensive program with 8 sessions per week with high loads and poor recovery periods. The MTL group underwent a 4-week extensive program with 3 sessions per week and longer recovery periods. The weekly schedule of training sessions is shown in Table 2.

Table 2:
Weekly training program for both groups.*

The load of the sessions was controlled according to their different natures. The judo randori sessions took place in groups. The Japanese term randori means “free practice,” and it is carried out in very similar conditions to those in judo competition. It is a very common training method in which both participants practice attack and defense techniques in real combat conditions. Contact with the opponent and the range of actions during randori, makes it difficult to quantify the training loads in these sessions.

The athletes started with a 15-minute warm-up (slow running and callisthenic exercises), which was followed by judo technique repetitions (uchi komi) for 15 additional minutes. Afterwards, the athletes performed 6 sets of 5-minute randori (1 minute rest) on the floor against opponents of similar standard and weight, and another block of 6 sets of 5-minute randori (1 minute rest) starting standing up against opponents of different weight and standard. Judo technique sessions consisted of technical and tactical exercises without combat. In particular, the athletes repeated sets of static and dynamic techniques of uchi komi and nage komi. These sessions lasted an average of 85 minutes.

The Foster training impulse method was used to monitor the exercise training (10). This method calculates the sum of the loads by multiplying the volume and the factor derived by 5 heart rate zones with respect to maximum heart rate (50–60% = 1, 60–70% = 2, 70–80% = 3, 80–90% = 4, and 90–100% = 5). A Polar S810 cardiotachometer (Polar Electro Oy) was used to monitor the heart rate during the judo randori and judo technique sessions. The average load of these sessions was training impulse = 317 ± 29 and 178 ± 11, respectively.

Strength training was focused on muscular resistance training, and consisted of performing 3 sets of 20 repetitions (3–4 minutes of rest between sets) of the following exercises: half back squat, bench press, shoulder press, leg press, triceps extension, deadlift, and high pull. Endurance training consisted of 40 minutes of running, where 70–80% of maximum heart rate was maintained.

Statistical Analyses

The researcher who performed the statistical analysis was blinded about the training followed by each group. All calculations were made with SPSS version 17, statistical software (SPSS Inc., Chicago, IL, USA), and the level of significance was set at p ≤ 0.05. Data normality and homoscedasticity were confirmed before inferential analysis through Kolmogorov-Smirnov and Levene's test, respectively. A mixed-model multivariate analysis of variance (MANOVA) (group [2: HLT and MLT] × testing time [2: pretest and posttest]) was then applied to assess the effects of the training model on the HRV parameters. In addition, we applied a mixed-model MANOVA (group [2: HLT and MLT] × testing time [2: pretest and posttest]) to evaluate the effects of the training mode on the stress-recovery parameters. Finally, a mixed-model MANOVA (group [2: HLT and MLT] × testing time [2: pretest and posttest]) was applied to assess the effects of the training mode on the strength parameters. The monitoring of the multivariable contrasts was conducted through univariate analysis contrasts. To track these contrasts, we performed a post-hoc analysis with a Bonferroni adjustment.


Heart Rate Variability Variables

The multivariate test indicated that there was an interaction effect between the testing time and group (F9,4 = 8.96, p = 0.025,

= 0.95) on HRV variables. The univariate contrasts are shown in Table 3. Pairwise comparisons revealed that there were no differences between groups in the pretest in the HRV variables. Nevertheless, the HTL group had lower short-term and long-term scaling exponents than MTL (p ≤ 0.05) in the posttest. Moreover, the HTL group showed lower RMSSD, very low frequency, high frequency, short-term variability, and short-range scaling exponent in the posttest than in the pretest (p ≤ 0.05). The HTL group showed higher low/high frequency ratio in the posttest than in the pretest. Finally, there were no differences between the pretest and posttest in the MTL group.

Table 3:
Intergroup and intragroup comparisons of heart rate variability variables.*

Stress/Recovery Variables

Multivariate analysis of variance indicated a testing time × group interaction effect (F4,9 = 89.42, p < 0.001,

= 0.97) on the stress/recovery variables. The univariate contrasts are shown in Table 4. The pairwise comparison showed that there were no differences between groups in the pretest in the stress or recovery variables. Nevertheless, in the posttest, the HTL group showed lower general recovery and sport-specific recovery than the MTL group (p ≤ 0.05). However, in the posttest, the general stress was higher in the HTL group (p ≤ 0.05). The general and sport-specific recovery was lower in the posttest than in the pretest only for the HTL group (p ≤ 0.05). Moreover, general stress was higher in the posttest than in the pretest in the HTL group (p ≤ 0.05). No differences between testing times were observed in the MTL group.

Table 4:
Intergroup and intragroup comparisons of stress/recovery variables.*

Strength Variables

There was an interaction effect between the testing time and the group (F4,9 = 7.7, p = 0.006,

= 0.77) in strength variables. The univariate contrasts are shown in Table 5. There were no differences between groups at any testing time. The HTL group showed higher maximum strength and maximum power in the posttest than in the pretest (p ≤ 0.05). The hand-grip strength of right and left hand did not show changes between pretest and posttest in any group.

Table 5:
Intergroup and intragroup comparisons of strength variables.*


The main finding of this study is that the higher stress and worse recovery conditions of the HTL athletes were caused by, among other possible mechanisms, an increase in sympathetic and a decrease in parasympathetic activity, as observed by HRV analysis. Furthermore, the physiological and psychological variables (strength measurements and RESTQ-SPORT scores) reinforce this finding. Consequently, our main hypothesis is confirmed, since the monitoring of HRV can distinguish between the stress produced by 2 specific training intensities in judo athletes.

Previous studies attempted to monitor the stress state from a psychological and physiological approach using RESTQ-SPORT or profile of mood states and hormone markers and performance tests, respectively (31,37). The cost-benefit of these combinations prevents coaches from using them. The concurrent administration of a psychological (RESTQ-SPORT) and a physiological assessment (HRV) during a training period could be an adequate and more accessible monitoring system of the stress-recovery state.

Our study indicated significantly higher stress levels compared with the baseline levels of the HTL group subjects after a 4-week training period. The general recovery and sport-specific recovery scores were significantly lower with respect to the baseline scores, and only specific stress scores showed no significant differences after a training period. Moreover, there were significant differences between both groups with a significantly higher general stress level and a lower general recovery and specific recovery rate in the HTL group compared with the MTL group. The highest stress levels and the lowest recovery levels of the HTL group matched a clear predominance of sympathetic activity. The results obtained in this study are in line with the work from Dupuy et al. (9), who, applying a similar approach, reported a decrease in cardiac parasympathetic control and an increase in cardiac sympathetic control after an overload period.

In particular, time domain analysis indicates a significant decrease of root mean square of the SD of the RR intervals. Moreover, the general trends, according to the frequency domain analysis of the HTL group, suggest that the autonomic nervous system increases its sympathetic activity together with an inhibition of parasympathetic activity. In addition, according to other authors (1,22,30,34), the increase of the low frequencies/high frequencies ratio and a significant decrease of high frequencies, demonstrating parasympathetic inhibition, indicates that a high training load combined with an insufficient recovery provoke higher sympathetic activity. Parrado et al. (33) studied the stress effects of competition load using the French Society of Sports Medicine Questionnaire and HRV analysis during the 2006 Field Hockey World Cup. They observed a significant correlation between the perceived stress markers and the low/high frequencies ratio of the HRV frequency domain and concluded that perceived stress could be an adequate parameter of fatigue and overtraining in long period competitions. The study by Tian et al. (40), with elite female wrestlers, denoted a lower parasympathetic tone in HRV time domain during training in transition to nonfunctional overreaching. In judo, Umeda et al. (42) and Koga et al. (24) described the effects of an intensive training program over the psychological and physiological parameters with the consequent decrease in performance, which is in accordance with our results.

Franchini et al. (12) demonstrated the importance of maximal strength and maximal power in explaining the performance of international-standard judo athletes. In our study, these strength variables were significantly lower after an HTL training period and coincided with higher stress and lower recovery perception, together with a higher sympathetic activity on the HRV analysis, which is in accordance with the results of Iellamo et al. (15) and Baumert et al. (1) after an intense training period. The study conducted by Chen et al. (7), showed that in weightlifters, the parasympathetic nervous activity decreased after training sessions. However, the time required for parasympathetic reactivation was much slower than previously reported by Iellamo et al. (15) after submaximal endurance training. This was probably because of the different type of muscle contraction. Nevertheless, we did not find any effects of the HTL on hand-grip strength. The strength results showed moderate evidence of a reduction in strength because of the HTL. It will be interesting in future work to corroborate these results and test other strength parameters, and so gain further information about this issue.

Hedelin et al. (14) suggested that a decrease in sympathetic activity of the autonomic nervous system at rest, caused by a depletion of low frequencies, is associated with improved muscular and aerobic performance. In contrast, our results indicate an increase in sympathetic, and a decrease in parasympathetic activity of the autonomic nervous system, together with lower muscular performance in strength tests and stress markers according to the RESTQ-SPORT. These findings could reflect the noncompleted phase of training adaptation where our subjects were located in the posttraining assessment. A possible explanation of the relationship between HRV and muscular performance could be that the baroreflex activity induced by muscular ischemia is inherently related to isometric efforts in judo techniques. The baroreflex initially increases sympathetic activity after a parasympathetic activation and, consequently, HRV increases (14). This argument could explain the higher sympathetic activity, the higher stress levels, and the poorer muscular performance observed in the HTL group.

High-standard judo combat requires intermittent short-duration and high-intensity complex actions that are characterized by the use of anaerobic glycolysis as a main energetic resource (8). However, the length of the training sessions involving these types of efforts, together with the associated rest periods, evokes the use of aerobic metabolism (12). Thus, the increasing sympathetic activity observed in our study through HRV analysis, after a period of training sessions with the intermittent efforts described in methods, is in accordance with other research (5,28).

The most used techniques of nonlinear analysis are the Poincaré plot and detrended fluctuations. In this analysis, the dispersion diagram of the Poincaré plot shows information about the sympathetic-parasympathetic balance. Short-term variability reports information about parasympathetic activity (41). In our study, the Poincaré plot analysis shows significantly lower parasympathetic activity considering the length of the short-term variability in the HTL group. The aforementioned changes in the sympathetic and parasympathetic activity modulation match the studies of Mourot et al. (30), Kiviniemi et al. (22), and Sartor et al. (36), reflecting changes in the autonomic modulation of the sinus node in the training period as seen using the Poincaré plot.

Detrended fluctuation analysis displayed differences between the analyzed variables. The results reinforce the idea that parasympathetic activity is lower and sympathetic activity is higher when the stress produced by high training loads is higher and the recovery perception is lower. Thus, our study supports the vagal-sympathetic balance observed by other authors (6,29,43), where values of α = 1 indicate a balance, values of 0.5 < scaling exponent <1 indicate a predominance of sympathetic activity, and values of 1 < scaling exponent <1.5 indicate a predominance of parasympathetic activity.

Short-term practical applications of HRV analysis are crucial for trainers, who need to know the real effects of high training loads to have better planning and load control. Several studies have used HRV analysis for short-term load control (22,23). They reduced the training loads when HRV decreased, and increased the load when HRV remained stable or increased. Plews et al. (35) used the tendency of the absolute values of the HRV and daily variations to control the training process of high-standard triathletes. Recently, Leti and Bricout (26) suggested the use of HRV analysis to monitor the fatigue modulation of the training process together with the administration of questionnaires as performed in our study.

This study has some limitations that should be taken into account. First, we did not use a control group that did not perform any training. Nevertheless, when studies are performed with elite athletes, it is difficult to have a control group. On the other hand, the researchers who performed the data collection were not blind to the group of participants. However, they followed a standard protocol and gave the same instructions to all the participants.

In conclusion, we observed that judo athletes enrolled in a high training load program displayed an imbalance of the sympathetic and parasympathetic activity of the autonomic nervous system, which led to a decrease of the vagal modulation. However, this imbalance seemed to be mainly provoked by an inhibition of the parasympathetic activity compared with sympathetic stimulation. At the same time, we observed a decrease in the maximal strength and muscular power parameters in the upper body, together with higher markers of stress and lower recovery perception.

Practical Applications

It is normal in contact sports, as in many others, to carry out general training sessions during which physical conditioning sessions are included using traditional methods. The loads associated with these sessions are usually quantified without any great problems as at a theoretical and technological level, a wide range of systems and methods can be employed that help coaches to make decisions during the training and later during recovery.

However, the real difficulty starts when athletes undertake training specific to their sport. It must be taken into account that, throughout a season, the importance of specific loads increases throughout the cycle, and these types of loads are applied in a very similar way to what happens in the competitive environment. Because of their importance, the coach has to know at the end of the session if the programmed load matches the load actually realized. It is important to bear in mind that the direct fight used during contact sport training makes the use of technology impossible, such as heart monitors or other devices, as these could hurt or injure the athletes themselves or others. As a consequence, real-time monitoring of a body on body training session is only a remote, unfeasible possibility. In summary, when 2 people fight, successful forecasting about intensity to which athletes will be subjected is very unlikely. Aspects such as motivation, level of fitness, psychological variables, or environmental factors to which the 2 contenders are subjected are so varied that in the majority of times training fights fall outside the expected ranges of pretraining form.

Therefore, the standard approach is for coaches to monitor the training through subjective systems, based on the experience of the judo athletes and coaches themselves. Although these systems have been relatively successful over the years, our work provides new evidence that may help coaches to quantify what happened during the specific training time.

Our basic proposal to the coaches would be: (a) conducting weekly HRV and RESTQ-SPORT tests to check the development of the training, (b) carrying out controls in the morning, the day after the specific training session, (c) analyzing the results immediately after carrying out the control so as to give immediate feedback to the athlete and make the relevant decisions, and (d) adjust the training load if they are outside of the planned ranges.

The analysis can be made easily, as free, high-quality software is widely available to the public. However, and despite that there are multiple parameters that can be calculated, what follows is a brief outline on what variable we consider to be the most important and how it can be easily interpreted.

In case (a), coaches should mainly concentrate on the RMSSD variable. If this variable starts to decrease throughout the weeks, it needs to be taken into account that intensity affects general performance, and as a consequence, overtraining could be happening. This would need to be corroborated by the RESTQ-SPORT scores. If these are increasing, it will be definitive proof that a high level of general stress exists.

In cases (b) and (c), it is very important to control the effect of the training load after a night's rest. The coaches should concentrate on the sympathetic-vagal balance of the autonomic nervous system. A predominance of the sympathetic system indicates a situation of physical or mental stress, which could be interpreted as an incomplete recovery.

Frequency domain parameters that indicate an increase in sympathetic activity are an increased high frequency, caused by the decrease in parasympathetic activity and a decrease in low frequency. There is no consensus in the scientific community as to the validity of the use of low frequency in this way, and so observing whether an increase in the low/high frequency ratio exists is sometimes used, interpreting it as a predominance of sympathetic activity.

Using the parameters of the nonlinear domain has more consensus within the scientific community in the interpretation of the resulting parameters. In our opinion, a good choice would be the observation of the Poincaré plot SD1 (short-term variability axis), and checking whether a significant decline has been produced, as it would indicate a predominance of sympathetic activity and therefore insufficient recovery.

It is important to give this information to the athlete, because in this way, if the training has been more intense than expected, the feeling could be internalized, and then the intensity of the fights with rivals in successive training sessions may be controlled to adapt the load to what the coach requests.

Finally, in case (d), the basic idea in adjusting the load is to watch the sympathetic-vagal balance of the autonomic nervous system. A predominance of sympathetic activity means that there is a situation of stress or poor recovery, so a decision can then be made to reduce the training stimulus or to maintain it if an overload is desired. If no changes are observed, or if there is an imbalance in parasympathetic activity, the choice may be made to maintain or increase the training stimulus.

This methodology is based on empirical data collected in our study and should be used with caution, as we have only tested it with judo athletes. However, we think that it is also applicable to other sports such as wrestling, rugby, or any other discipline in which the quantification of specific loads presents problems.


We would like to acknowledge the technical and material support of the Catalonian Judo Federation. We would also like to thank the work carried out by the 2 anonymous reviewers of this article, whose work has contributed to improving its quality.


1. Baumert M, Brechtel L, Lock J, Hermsdorf M, Wolff R, Baier V, Voss A. Heart rate variability, blood pressure variability, and baroreflex sensitivity in overtrained athletes. Clin J Sport Med 16: 412–417, 2006.
2. Bosquet L, Merkari S, Arvisais D, Aubert AE. Is heart rate a convenient tool to monitor over-reaching? A systematic review of the literature. Br J Sports Med 42: 709–714, 2008.
3. Bosquet L, Montpetit J, Arvisais D, Mujika I. Effects of tapering on performance: A meta-analysis. Med Sci Sports Exerc 39: 1358–1365, 2007.
4. Brink MS, Visscher C, Coutts AJ, Lemmink KA. Changes in perceived stress and recovery in overreached young elite soccer players. Scand J Med Sci Sports 22: 285–292, 2012.
5. Buchheit M, Al Haddad H, Millet GP, Lepretre PM, Newton M, Ahmaidi S. Cardiorespiratory and cardiac autonomic responses to 30-15 intermittent fitness test in team sport players. J Strength Cond Res 23: 93–100, 2009.
6. Casties JF, Mottet D, Le Gallais D. Non-linear analyses of heart rate variability during heavy exercise and recovery in cyclists. Int J Sports Med 27: 780–785, 2006.
7. Chen JL, Yeh DP, Lee JP, Chen CY, Huang CY, Lee SD, Chen CC, Kuo TB, Kao CL, Kuo CH. Parasympathetic nervous activity mirrors recovery status in weightlifting performance after training. J Strength Cond Res 25: 1546–1552, 2011.
8. Degoutte F, Jouanel P, Filaire E. Energy demands during a judo match and recovery. Br J Sports Med 37: 245–249, 2003.
9. Dupuy O, Bherer L, Audiffren M, Bosquet L. Night and postexercise cardiac autonomic control in functional overreaching. Appl Physiol Nutr Metab 38: 200–208, 2013.
10. Foster C, Florhaug JA, Franklin J, Gottschall L, Hrovatin LA, Parker S, Doleshal P, Dodge C. A new approach to monitoring exercise training. J Strength Cond Res 15: 109–115, 2001.
11. Franchini E, De Moraes Bertuzzi RC, Takito MY, Kiss MA. Effects of recovery type after a judo match on blood lactate and performance in specific and non-specific judo tasks. Eur J Appl Physiol 107: 377–383, 2009.
12. Franchini E, Del Vecchio FB, Matsushigue KA, Artioli GG. Physiological profiles of elite judo athletes. Sports Med 41: 147–166, 2011.
13. Gonzalez-Boto R, Salguero A, Tuero C, Marquez S. Concurrent validity of the Spanish version of the Recovery-Stress Questionnaire for Athletes (RESTQ-Sport). Sports Phycology Magazine 18: 53–72, 2009.
14. Hedelin R, Bjerle P, Henriksson-Larsen K. Heart rate variability in athletes: Relationship with central and peripheral performance. Med Sci Sports Exerc 33: 1394–1398, 2001.
15. Iellamo F, Legramante JM, Pigozzi F, Spataro A, Norbiato G, Lucini D, Pagani N. Conversion from vagal to sympathetic predominance with strenuous training in high-performance world class athletes. Circulation 105: 2719–2724, 2002.
16. Kaikkonen P, Nummela A, Rusko H. Heart rate variability dynamics during early recovery after different endurance exercises. Eur J Appl Physiol 102: 79–86, 2007.
17. Kalda J, Jürimäe T, Jürimäe J. Relationships between recovery-stress state and performance in sprinters and jumpers. Percept Mot Skills 99: 12–16, 2004.
18. Kellmann M. Preventing overtraining in athletes in high-intensity sports and stress/recovery monitoring. Scand J Med Sci Sports 20(Suppl. 2): 95–102, 2010.
19. Kellmann M, Günther KD. Changes in stress and recovery in elite rowers during preparation for the Olympic Games. Med Sci Sports Exerc 32: 676–683, 2000.
20. Kellmann M, Kallus KW. Recovery-Stress Questionnaire for Athletes: User Manual. Champaign, IL: Human Kinetics, 2001.
21. Kenttä G, Hassmén P, Raglin JS. Mood state monitoring of training and recovery in elite kayakers. Eur J Sport Sci 6: 245–253, 2006.
22. Kiviniemi AM, Hautala AJ, Kinnunen H, Nissilä J, Virtanen P, Karjalainen J, Tulpo MP. Daily exercise prescription based on heart rate variability among men and women. Med Sci Sports Exerc 42: 1355–1363, 2010.
23. Kiviniemi AM, Hautala AJ, Kinnunen H, Tulppo MP. Endurance training guided individually by daily heart rate variability measurements. Eur J Appl Physiol 101: 743–751, 2007.
24. Koga T, Umeda T, Kojima A, Tanabe M, Yamamoto Y, Takahashi I., Iwasaki H, Iwane K, Matsuzaka M, Nakaji S. Influence of a 3-month training program on muscular damage and neutrophil function in male university freshman judoists. Luminescence 28: 136–142, 2013.
25. Kraemer WJ, Fry AC. Strength testing: Development and evaluation of methodology. In: Physiological Assessment of Humane Fitness. Maud P., Foster C., eds. Champaign, IL: Human Kinetics, 1995. pp. 115–138.
26. Leti T, Bricout VA. Interest of analyses of heart rate variability in the prevention of fatigue states in senior runners. Auton Neurosci 173: 14–21, 2013.
27. Makivić B, Djordjević Nikić M, Willis MS. Heart rate variability (HRV) as a tool for diagnostic and monitoring performance in sport and physical activities. JEPonline 16: 103–131, 2013.
28. Millar PJ, Rakobowchuk M, McCartney N, MacDonald MJ. Heart rate variability and nonlinear analysis of heart rate dynamics following single and multiple Wingate bouts. Appl Physiol Nutr Metab 34: 875–883, 2009.
29. Morales J, Garcia V, García-Massó X, Salvá P, Escobar R, Buscà B. The use of heart rate variability in assessing precompetitive stress in high-standard judo athletes. Int J Sports Med 34: 144–151, 2013.
30. Mourot L, Bouhaddi M, Perrey S, Cappelle S, Henriet MT, Wolf JP, Rouillon JD, Regnard J. Decrease in heart rate variability with overtraining: Assessment by the Poincaré plot analysis. Clin Physiol Funct Imaging 24: 10–18, 2004.
31. Nederhof E, Zwerver J, Brink M, Meeusen R, Lemmink K. Different diagnostic tools in nonfunctional overreaching. Int J Sports Med 29: 590–597, 2008.
32. Nunan D, Donovan G, Jakovljevic D, Hodges L, Sandercock G, Brodie D. Validity and reliability of short-term heart-rate variability from the Polar S810. Med Sci Sports Exerc 41: 243–250, 2009.
33. Parrado E, Cervantes J, Pintanel M, Rodas G, Capdevila L. Perceived tiredness and heart rate variability in relation to overload during a field hockey World Cup. Percept Mot Skills 110: 699–713, 2010.
34. Pichot V, Roche F, Gaspoz JM, Enjolras F, Antoniadis A, Minini P., Costes F, Busso T, Lacour JR, Barthélémy JC. Relation between heart rate variability and training load in middle-distance runners. Med Sci Sports Exerc 32: 1729–1736, 2000.
35. Plews DJ, Laursen PB, Kilding AE, Buchheit M. Heart rate variability in elite triathletes, is variation in variability the key to effective training? A case comparison. Eur J Appl Physiol 112: 3729–3741, 2012.
36. Sartor F, Vailati E, Valsecchi V, Vailati F, La Torre A. Heart rate variability reflects training load and psychophysiological status in young elite gymnasts. J Strength Cond Res 27: 2782–90, 2013.
37. Schmikli SL, Brink MS, De Vries WR, Backx FJG. Can we detect non-functional overreaching in young elite soccer players and middle-long distance runners using field performance tests? Br J Sports Med 45: 631–636, 2011.
38. Seiler S, Haugen O, Kuffel E. Autonomic recovery after exercise in trained athletes: Intensity and duration effects. Med Sci Sports Exerc 39: 1366–1373, 2007.
39. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: Standards of measurement, physiological interpretation and clinical use. Circulation 93: 1043–1065, 1996.
40. Tian Y, He ZH, Zhao JX, Tao DL, Xu K Y, Earnest CP, Mc Naughton LR. Heart rate variability threshold values for early-warning nonfunctional overreaching in elite female wrestlers. J Strength Cond Res 27: 1511–1519, 2013.
41. Toichi M, Sugiura T, Murai T, Sengoku A. A new method of assessing cardiac autonomic function and its comparison with spectral analysis and coefficient of variation of RR interval. J Auton Nerv Syst 62: 79–84, 1997.
42. Umeda T, Suzukawa K, Takahashi I, Yamamoto Y, Tanabe M, Kojima A, Katagiri T, Matsuzaka M, Totsuka M, Nakaji S, Sugawara N. Effects of intense exercise on the physiological and mental condition of female university judoists during a training camp. J Sports Sci 26: 897–904, 2008.
43. Yeh RG, Chen GY, Shieh JS, Kuo CD. Parameter investigation of detrended fluctuation analysis for short-term human heart rate variability. J Med Biol Eng 30: 277–282, 2010.
44. Zatsiorsky VM, Kraemer WJ. Science and Practice of Strength Training. Champaign, IL: Human Kinetics, 2006

stress recovery; HRV analysis; training loads; sport performance

Copyright © 2014 by the National Strength & Conditioning Association.