Listen to your heart…But what do you really want to know?
Written by Peter Joffe
For many years, measuring heart rate in sport is one of the most popular ways to monitor and evaluate workload. It is easy to implement, it is non-invasive, and it is based on scientific knowledge.
In the last decades, the wide availability of different types of commercial heart rate (HR) monitors and software made it possible for every athlete to collect continuously his/her HR data and even perform relatively complicated analysis. Thanks to the help of HR monitoring, it looks like training today can be much more effective and scientific. It is not so straightforward, however.
Though heart function is simple – to pump blood, it is a very complexly regulated machine, and its connection with physical conditioning is not as simple as it may look.
In this article, I would like to discuss different HR measures, their possible interpretations, and suitability for evaluating the athlete’s fitness, training load, current conditions, and preparedness.
Possible HR measures
For many decades coaches have been counting pulses in their training sessions and beyond. One of the significant advantages of HR monitoring is that you can derive variable data basically from the same source: rate of the heartbeats. I am sure that most of the readers are familiar with different HR measures. However, there are some relatively new methods, and traditional approaches are now understood better. So, probably, it is worth talking about this. I will give average/normal values where possible; however, keep in mind that this is just for reference.
Resting HR (HRrest).
HRrest is a minimum rate of heart beats when there are no physical workload, minimum emotional and psychological influences. It is better to measure in the standard conditions (e.g. immediately after wake up in the morning). Normal values : 60-90, however endurance athletes can have lower than 40.
HR maximum (HRmax)
HRmax is a necessary reference point for the evaluation of intensity. The only problem with it is that to achieve truly HRmax subject should give” all and out,” which is not always possible for elderly and non-active people. The most used theoretical calculation is 220-age. In my opinion, another formula is closer to reality: 208-0.7x age (Tanaka, Monahan, & Seals, 2001).
HR-reserve
HRreserve, is calculated from HRmax and HRrest:
HRreserv= HRmax-HRrest.
Since the athletic heart has lower HRrest and maximum HR is generally not different between athletes and non-athletes, athletes have more heartbeats in reserve when exercising. For instance, if we assume maximum HR 200 for two given individuals, one of them, who is an athlete, can increase his/her HR from 40 to 200, which is 160 beats, whereas the non-athlete has only 130 (200-70).
HR during exercise (HRexercise)
HRexercise is usually expressed as a percentage of HRmax. Some researchers argue that intensity expressed as a fraction of HRreserve during exercise more precisely reflects a fraction of maximum oxygen uptake than it does intensity expressed as per cent of HRmax.
The fraction of HRreserv= [(HRexcercise – HRrest)/ HRreserve]
Of course, there are no average values for HRexercise, but there is a conventional separation of HRexercise on different intensity zones.
Usually, it follows guidance (% of HRmax): 50-60 very easy; 60-70 easy; 70-80 moderate; 80-90 hard; 90-100 maximum.
Perhaps, more scientific (though much more complicated) way is to define HR intensity zones with reference to physiological thresholds such as aerobic/ventilatory and anaerobic /respiratory compensation. These training zones should be established individually during the test, where the lactate accumulation or gas exchange curve rates are plotted against HR (Sparks, Coetzee, & Gabbett, 2017).
Recovery HR (HRrecovery).
It shows how fast the heart slows down after exercise. It can be expressed as a difference between HR achieved in the last 15-30 sec of exercise and the first minute of recovery. More extended periods can be used as well (usually no longer than five minutes). It is generally accepted that first rapid decrease in HR is connected with parasympathetic reactivation, whereas later reduction is more work-dependent. However, some authors pointed out that sympathetic influence, which depends on work intensity and duration, is present even in the first seconds of recovery. Especially, accumulated anaerobic by-products and acidosis can play major roles. That is, possibly, the reason why children, with their minor anaerobic contribution to exercise, have faster HRrecovery than adults (Martin Buchheit, Duche, Laursen, & Ratel, 2010).
For avoiding intensity bias, a sub-maximal HRrecovery test may be preferable. I will talk about test later. Average values for HRrecovery after 1 min that gives me my TomTom watch are the following:1-20 poor; 20-30 fair; 30-40 good; >40 excellent. Honestly, I don’t know how they came to this guidance and workload is not taken into account.
There are highly cited researches which connected HRrecovery with the risk of mortality. Though they are based on an ageing population with already exiting heart problems, these data are used as “cutting points”: HRrecovery <12 after one minute and < 22 after two minutes is associated with a higher risk of death (Cole , Blackstone , Pashkow , Snader , & Lauer 1999; Shetler et al., 2001).
Heart rate variability (HRV).
HRV becomes increasingly popular nowadays. It measures variation in HR. If, for example, a person has an average HR of 60 beats per minute, this doesn’t mean that time between every two beats is exactly 1 sec. Sometimes it maybe 1.1 sec and sometimes 0.9. HRV may be important because it reflects a balance between parasympathetic (generally slow down HR) and sympathetic (generally increases HR) systems. Generally, higher HRV is considered as a positive sign in physical conditioning. HR fluctuates in different frequencies: very low, low, and high. High frequencies oscillations presumably reflect parasympathetic influences whereas very low—sympathetic and low — both.
There are many methods to analyse HRV, which are divided into two main groups: time-domain and frequency-domain methods (for review see (“Heart rate variability: standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology,” 1996)). Coach and/or sport scientist needs HR monitor, which can measure beat-to-beat intervals and computer software which can produce analysis.
One of the most simple and, as some researchers argue, most reliable for sports purposes method is RMSSD (root mean square of the successive differences). It is a time-domain method for analysing high-frequency osculations. RMSSD analysis does not demand a lot of time for collecting data ( 1 min may be enough, though usually data is collected for 3-5 min). It can be easily done in Excel (Plews, Laursen, Stanley, Kilding, & Buchheit, 2013). Further, in this article, I will always keep in mind RMSSD when I talk about HRV. HRV can be measured during sleep, after wake up, during exercise, and after exercise. Average RMSSD values (mean-SD) are 42-15 msec (Nunan, Sandercock, & Brodie, 2010). Athletes usually have higher values.
HR acceleration.
That is a relatively new index. It reflects how fast the heart can achieve the necessary rate after the start of the exercise. It is hypothesised that the faster HR acceleration is beneficial for better oxygen supply at the first stages of the exercise. Thus fitter athletes have faster HR acceleration. Though it is probably true, it is not easy to measure HR acceleration practically because such a test demands a highly standardised workload and, more importantly, the same base-line HR values, which would be difficult to control in reality. For these reasons and due to a lack of research on this measure, I will not discuss it further in this article.
Long-term heart adaptations to exercise (assessing fitness).
During exercise, body requirements in oxygen, fuel, and the removal of metabolic by-products increase dramatically. For coping with that demand, blood flow rises to 7 times (Joyner & Casey, 2015); thus, the heart needs to pump much more blood than in rest. As a consequence, the athlete’s heart has to undergo some necessary adaptations. It is bigger in mass and volume, it is less stiff (that allows better filling), and in a wide range of intensities, it beats slower. To understand why the latter is the case, we need to consider the cardiac output and stroke volume.
The cardiac output is a product of HR and stroke volume (the amount of blood ejected in one beat). At rest, cardiac output is the same for athletes and non-athletes, but athletes can achieve much higher values during exercise. This is due to a larger stroke volume. As it was mentioned, the athletic heart is bigger, more powerful and has better filling properties than the ordinary heart. Also, endurance athletes have higher blood volume and better venous return during exercise. All these facilitate stroke volume. Elite athletes have 1.7 times at rest and up to 2 times during exercise higher stroke volume than sedentary men (Joyner & Casey, 2015).
Thus at rest or during sub-maximal exercise, when there is no need for maximal cardiac output, the athlete’s heart beats slower. Endurance athletes may have HR at rest lower than 40, whereas, for the normal population, it is around 70. In its turn, lower HR allows more time for heart-filling, thus eventually increasing stroke volume. During the same sub-maximal exercise, better-adapted subjects generally have lower HR than a person who adapted less. It is necessary to note that such exercise should be sport-specific; thus, you cannot compare rower and runner by giving them sub-maximal running tests.
Our heart has autonomic neural regulation which is provided by sympathetic and parasympathetic (vagal) systems. Parasympathetic impulses down-regulate HR whereas sympathetic influence up-regulate it. Since athletic heart normally beats slower than non-athletic, it is generally accepted that athletes have more profound parasympathetic influence than ordinary people. Thus, when doing different HR measures, we are expecting following adaptations for fitter subjects (Borresen & Lambert, 2008):
1. Lower HRrest due to better parasympathetic regulation and higher stroke volume.
2. Higher HRV at rest due to a better balance between sympathetic and parasympathetic systems.
3. Lower HR for the same exercise because the relative intensity is lower, and stroke volume is higher.
4. Faster HRrecovery after the same exercise due to better parasympathetic influence.
5. Faster HRV recovery after the same exercise due to better autonomic balance.
HR for assessing intensity and workload
Intensity
HR is a good estimation of aerobic intensity in continuous exercise. The idea is simple: if the intensity is higher, muscles need more blood/oxygen; therefore, the heart pumps at higher rates. It is not so simple in intermittent exercises, like for example, sports games, as well as for assessing anaerobic workload and resistance/plyometric activities. HR has some inertia; hence it does not always come up with rapid changes of intensity. Sometimes maximal bursts of activity are relatively short, and HR does not have enough time to achieve its highest level.
Nevertheless, HR measures may provide useful information for sports game’s coaches. The Spanish football team’s head coach was really surprised when I showed him his player’s HR graphs during the training session. He found that he is talking too much between exercise, and players’ HRs are significantly falling during these periods. Or HR measured during 4 min of small-side games showed huge variability between players because some worked really hard during these games whereas others did not.
During conditioning training sessions, players perform exercises where HR measures give useful information about the aerobic strain. Commercially available HR monitors provide coaches with the graphical representation of the HR, which can reflect the relative intensity of the game periods and a percentage of time spent in different HR zones. It can be very helpful in planning and analysing training load. Some researchers argue that to gain improvements in aerobic fitness, players have to spend at least 7% of training time in a high-intensity zone (>90% HR max) (Manzi, Bovenzi, Impellizzeri, Carminati, & Castagna, 2013). And finally, unusually low or high HR combined with the high perceived difficulty of exercise (RPE) can warn the coach that the player is tired or unwell.
Quantifying internal workload
As in opposite to external work, which can be measured as the distances, speeds, and accelerations, internal work means physiological strain imposed on an athlete during an exercise. This can be different for the same external work for different athletes. There are suggestions to use HR for quantifying internal workload during training sessions. It is the easiest way to collect physiological data, and there is a logical connection between HR and exercise’s difficulty.
Nevertheless, to establish a scientifically based relationship between HR and internal workload is not a simple task. It is clear that exercises’ relative internal difficulty is not strictly proportional to HR ( e.g., 1 min exercise with HR 200 probably more than two times harder than 1 min of HR 100 exercise). So there is a need to find more reasonable proportions.
Sally Edwards suggested to divide HR on five intensity zones separated by 10 per cents of HR max and to give “ weighting coefficient” for each zone: 50-60% – 1 : 60-70 % – 2 : 70-80% – 3 : 80-90%-4; 90-100% – 5 (Foster et al., 2001). So now exercise with HR 200 beats/min would be 5 times harder than 100 beats/min (assuming that 200,in this case, is a HRmax). That is probably much closer to reality. Nevertheless, this method looks like oversimplification because there is no scientific ground for it, neither for weighting coefficients nor for HR zones definition.
Eric Banister proposed Training Impulse (TRIM) for quantification of internal workload. This impulse is based on exercise average HR and its duration. Sum of TRIMs at specific HRs may be used if training session has multiple exercises and intensities. Banister proposed special formula for dealing with the non-linear relationship between HR and exercise physiological strain:
TRIMP(men) = sum of (D x HRr x 0.64e 1.92xHRr)
TRIM(women) sum of ( D x HRr x 0.86e1.67xHRr)
Where:
D is the duration in minutes at particular HR
HRr is the particular HR as a fraction of HRreseve
This formula was developed based on the experimentally observed relationship between heart rate and lactate accumulation rate in the incremental test. It was established for the average athlete. For improving this method’s precision, HR – blood lactate function can be calculated individually for every athlete (Manzi, Iellamo, Impellizzeri, D’Ottavio, & Castagna, 2009). Also, HR zones may be specified with reference to lactate thresholds that makes them more scientifically grounded (Stagno, Thatcher, & van Someren, 2007).
However, in my opinion, there is a fundamental flaw in this method. That is the relationship between blood lactate, HR and internal load in intermittent activities. In continuous exercise, though blood lactate itself is not the reason for fatigue, it, like an alarming signal, can indicate that fatigue is coming. The course of blood lactate accumulation coincides with a build-up in metabolic by-products and homoeostasis’s disturbances. All these processes influence an increase in HR; hence some logical construction can be built.
It is not so in intermittent activities. As it was mentioned before, HR in such exercises not always reflects their real physiological strain in the same manner as it does in continuous running, thus relationship between HR and internal workload based on lactate curve, derived from continuous incremental test, may be not applicable (or, at least, not precise) in sport games (Impellizzeri, Rampinini, & Marcora, 2005).
There is another method which is suggested for calculation of TRIM. Though it based on psychological measurement (Borg’s RPE scale) and probably doesn’t look as scientific as complicated Banister’s formula, nevertheless it is probably not less reliable. At least, both methods showed pretty good correlation between them (Alexiou & Coutts, 2008; Foster, et al., 2001). This calculation is really simple: internal training load can be defined as product of RPE (from 1 to 10) and session duration in minutes. The advantage of RPE is that it reflects overall difficulty of training, including resistance, anaerobic and psychological strains.
HR for assessing preparedness, fatigue and exercise prescription
One of the most difficult and important tasks in coaching practice is evaluating athletes’ current conditions and preparedness for competitions. This is necessary for planning and monitoring a workload. Many physiological, psychological and training factors should be taken into account. One of the most challenging problems is giving athletes optimal workload and, at the same time, don’t over-train him/her.
Fatigue is an inseparable part of the training process. It may be in the form of functional overreaching and chronic overtraining. Former is considered as part of positive adaptation, which is followed, after tapering, by super-compensation and improvement in performance. The latter is a negative condition and is followed by sustainable performance impairment, psychological apathy and even drop-out. Distinguish between these two conditions before overtraining already happens is not an easy task. The idea is that HR measures may provide some additional information on this matter.
From different physiological methods, HR is probably the most simple way which is available for every coach. However, the disadvantage of HR measures is that they may fluctuate for many reasons independently from what we are interested in most – athlete’s physical conditions. Besides, even connecting HR indexes with an athlete’s current physical state is not straightforward and poorly understood.
Essentially, assessing physical form through HR indices, we are looking at autonomic balance. We hope that disturbance in this balance, reflected in HR fluctuations, warn us about fatigue and/or overtraining; thus, we can adjust training load. We expect that HR indices show enhanced or maintained autonomic balance if the athlete is in good shape and an opposite tendency if his/her conditions deteriorate. However, as we see further, it is not always the case. My analysis of the literature revealed very contradicting results. Just give you a few examples.
In their review, Stanley et al. gave examples of how training prescriptions based on HR autonomic indices were superior to traditional planning (Stanley, Peake, & Buchheit, 2013). Participants in these studies were moderately trained subjects; thus, the conclusions to the high-level performance are questionable.
Lamberts et al. trained 14 good-level cyclists for four weeks. During the training, those who had faster HRrecovery later showed a tendency to perform better in 40 km trial. The authors concluded that HRrecovery may be a valuable tool to assess training status (Lamberts, Swart, Capostagno, Noakes, & Lambert, 2010).
In their literature review, L.Bosquet et al. concluded that fluctuations in HR indices related to training interventions are small and fell within possible day-to-day variations. Thus their correct interpretations require comparison with other signs and symptoms (Bosquet, Merkari, Arvisais, & Aubert, 2008).
Bellenger and colleagues made the same conclusion. In their meta-analysis, they found that HR indices may change in the same manner in overreaching and non-fatigued states thus, additional measures are needed to understand athlete’s conditions (Bellenger et al., 2016).
Thorpe et al. investigated whether HR measures may help track morning fatigue in elite football players during the competitive season. They concluded that other methods such as the perceived rating of fatigue, delayed onset of muscle soreness (DOMS), and sleeping quality are superior and more helpful. (Thorpe et al., 2016)
So, why literature is so inconclusive?
Firstly, mentioned above, significant HR variations may be partly responsible. The coefficient of variation for HRrest and HRV rest is 10 and 12 % respectively. It is even much higher for exercise and post-exercise measures (e.g. 60% for HRV exercise)(M. Buchheit, 2014). For dealing with such variations, data should be collected more often. Some authors suggested using seven days of rolling average HRV and HRrest data instead of single day values for exercise prescription (Le Meur, Pichon, et al., 2013). Although this may reduce variations, planning today’s training based on data collected a few days ago may put in question the main goal of these methods – online monitoring and planning. Besides, not all athletes can and willing to collect morning HR data consistently.
The second reason may be that HR induces autonomic balance responds differently in already highly trained individuals compared to the average athletic population (Borresen & Lambert, 2008). High-level athletes can already have “fine-tuned” heart autonomic regulation, and interactions between different influences during different training and competition periods is much more complex than it can be for fitness enthusiasts.
Thirdly, methodological issues in HR measuring are very important. When and how they were collected, what type of analysis was used and what changes should be considered as meaningful can make huge differences in their interpretations.
And finally, studies, which claimed the superiority of HR measures for exercise prescription, often used an individual approach for HR participants, whereas others were prescribed collectively. The individual approach is clearly better, and that might bias the results.
The very idea that HR should be “better” if an athlete is in good form is questionable. For example, an increase in HRexercise and a decrease in HRrecovery do not necessarily indicate impair in the form (M Buchheit, Simpson, Al Haddad, Bourdon, & Mendez-Villanueva, 2012). Sometimes when an athlete is on the peak of his preparedness, some healthy level of increasing sympathetic influence is present-“sympathetic mobilisation” (M. Buchheit, 2014)) , and that may make some or all HR indexes look worse. On the opposite, when an athlete is overtrained parasympathetic system can “put brakes” on the heart and to limit its ability to pump at higher rates thus, the sportsman may be fatigued despite having relatively good HR characteristics.
For example, though faster HRrecovery means positive adaptations in the long-term, this does not necessarily mean that athlete is in good form. Sometimes in the opposite, faster HRrecovery combined with higher RPE may signal overreaching (Le Meur, Buchheit, Aubry, Coutts, & Hausswirth, 2016). In another study, Le Meur et al. found that overreached triathletes had significantly lower HRexercise than controls (Le Meur, Hausswirth, et al., 2013). Both positive and negative adaptations may influence a decrease in HRV.
For overcoming uncertainty, whether a decrease in HRV may be a consequence of positive, sympathetic mobilisation rather than a negative decrease in vagal modulation, a new index was suggested: Ln RMSSD/R-R interval ratio (Plews, Laursen, Kilding, & Buchheit, 2012). When sympathetic mobilisation is present, HRrest is higher; thus R-R interval (average time between two beats) is smaller, and Ln HRV/R-R ratio will be bigger. This may smooth HRV decrease or even overturn it. However, even this new index remains an individual for every athlete and should be considered carefully (Plews, et al., 2013).
In summary:
From all HR indices, morning HR and HRV, HRexercise and HRrecovery are most suitable for analysis of physical form and preparedness. They should be implemented together with other methods.
HRV after exercise (HRV recovery) and, especially, HRV during exercise have high variability and multiple influences such as blood pressure, baroreflex activity, and metabolic disturbance. This, probably, makes their application assessed for autonomic balance impractical (M. Buchheit, 2014).
It is difficult to draw a conclusion about an athlete’s current condition based exclusively on HR indexes.
Practice
Test
Test design is a very important issue when it comes to HR indices. It should be valid, reliable, not time-consuming and not physically demanding. Recently, 5 to 5 test, which can measure HRexercise, HRrecovery, and HRV recovery in one sub-maximal exercise, was suggested (M. Buchheit, 2014). This test consists of 5 min sub-maximal running (cycling or rowing) below the aerobic threshold and 5 min recovery. Last 30 seconds of exercise gives us HRexercise ,
HRrecovery can be estimated from 1-st minute of the recovery, and HRV recovery can be calculate during last 3 min of rest. Such a test may be easily incorporated into warm-up, it is less intensity-dependent and psychologically non-demanding. HRrecovery measured in the earlier recovery (during 1-st minute) better reflects parasympathetic influence than later periods that are more intensity-dependent (Martin Buchheit, Papelier, Laursen, & Ahmaidi, 2007). Due to the limited usefulness and demand for extra time, collecting HRV recovery may be excluded. That will make the test shorter (6 min).
Decision making
When deciding whether to use a particular test or measure in real life, the practitioner always has to weigh its usefulness and necessity against the players’ and coaches’ comfort. Training and recovery processes must not be distracted by scientific toys. In my opinion, despite some promising results, the practical application of HR measures for understanding athlete current form and exercise prescription remains difficult. This is especially true for team sports.
For being meaningful, HR measures should be collected often in highly standardised conditions and constantly be compared against other physiological and psychological measures. The same nature of the sports games with their very variable and unpredictable activities, multi-intensity training, complicated psychological influences, and busy players’ schedules make these methods problematic in reality. Maybe it is better to allow players a nice sleep instead of bothering them with HR measures in every morning.
This doesn’t mean that I am sceptical about HR measures. Actually, I use them a lot. HR may provide valuable information that can add to the overall picture. However, I want to emphasise that coach should understand what he/she wants to derive from HR data and all limitations of its application. Otherwise, instead of useful scientific help, counting pulse on every occasion turns out to be annoying and distracting. I hope that the following conclusions may be helpful for coaches and athletes.
Conclusion
1. In the long term, adaptation to exercise may decrease HRrest, HRexercise, and increasing in HRV and speed of HRecovery. That is due to a decrease in the relative difficulty of exercise and adaptation of the cardiovascular system itself. A coach may consider such changes as a positive effect of the training intervention.
2. Though HR may provide useful information about aerobic strain, evaluation of training workload; it does not comprehensively reflect all training intensity particularities(especially in intermittent exercise).
3. Calculating TRIM through HR is relatively complicated and possibly provides no advantages compare to TRIM calculating through RPE.
4. Usage HR indices for assessing athlete current form, preparedness, recovery, and exercise prescription is possible only with a combination of other methods such as RPE, sleeping quality, mood questioner, DOMS assessment, physical tests, etc. HR measures do not comprehensively reflect such aspects of fatigue/recovery as energy restoration and muscle damage.
5. When using for mentioned above aims, HR indices should be measured consistently and in standard conditions.
6. Meaningful changes should be compared to natural day-to-day variability.
7. Significant deviation from average values rather than particular tendency (e.g. higher HRrest or slower HRrecovery) should be considered as possible signs of impairment in form for high-level athletes. These average values are specific for every athlete. Training period and intensity have to be taken into account.
Reference
Alexiou, H., & Coutts, A. J. (2008). A comparison of methods used for quantifying internal training load in women soccer players. International Journal of Sports Physiology and Performance, 3(3), 320-330.
Bellenger, C. R., Fuller, J. T., Thomson, R. L., Davison, K., Robertson, E. Y., & Buckley, J. D. (2016). Monitoring athletic training status through autonomic heart rate regulation: a systematic review and meta-analysis. Sports Medicine, 46(10), 1461-1486.
Borresen, J., & Lambert, M. I. (2008). Autonomic control of heart rate during and after exercise : measurements and implications for monitoring training status. Sports Med, 38(8), 633-646.
Bosquet, L., Merkari, S., Arvisais, D., & Aubert, A. E. (2008). Is heart rate a convenient tool to monitor over-reaching? A systematic review of the literature. Br J Sports Med, 42(9), 709-714.
Buchheit, M. (2014). Monitoring training status with HR measures: do all roads lead to Rome? Front Physiol, 5(73).
Buchheit, M., Duche, P., Laursen, P. B., & Ratel, S. (2010). Postexercise heart rate recovery in children: relationship with power output, blood pH, and lactate. Applied Physiology, Nutrition, and Metabolism, 35(2), 142-150.
Buchheit, M., Papelier, Y., Laursen, P. B., & Ahmaidi, S. (2007). Noninvasive assessment of cardiac parasympathetic function: postexercise heart rate recovery or heart rate variability? American Journal of Physiology-Heart and Circulatory Physiology, 62(1), H8.
Buchheit, M., Simpson, M., Al Haddad, H., Bourdon, P., & Mendez-Villanueva, A. (2012). Monitoring changes in physical performance with heart rate measures in young soccer players. European Journal of Applied Physiology, 112(2), 711-723.
Cole , C. R., Blackstone , E. H., Pashkow , F. J., Snader , C. E., & Lauer , M. S. (1999). Heart-Rate Recovery Immediately after Exercise as a Predictor of Mortality. New England Journal of Medicine, 341(18), 1351-1357. doi: 10.1056/nejm199910283411804
Foster, C., Florhaug, J. A., Franklin, J., Gottschall, L., Hrovatin, L. A., Parker, S., . . . Dodge, C. (2001). A new approach to monitoring exercise training. The Journal of Strength & Conditioning Research, 15(1), 109-115.
Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. (1996). Circulation, 93(5), 1043-1065.
Impellizzeri, F. M., Rampinini, E., & Marcora, S. M. (2005). Physiological assessment of aerobic training in soccer. Journal of Sports Sciences, 23(6), 583-592.
Joyner, M. J., & Casey, D. P. (2015). Regulation of increased blood flow (hyperemia) to muscles during exercise: a hierarchy of competing physiological needs. Physiological reviews, 95(2), 549-601.
Lamberts, R. P., Swart, J., Capostagno, B., Noakes, T. D., & Lambert, M. I. (2010). Heart rate recovery as a guide to monitor fatigue and predict changes in performance parameters. Scand J Med Sci Sports, 20(3), 449-457.
Le Meur, Y., Buchheit, M., Aubry, A., Coutts, A. J., & Hausswirth, C. (2016). Assessing Overreaching With HRR: What is the Minimal Exercise Intensity Required? International Journal of Sports Physiology and Performance, 1-14.
Le Meur, Y., Hausswirth, C., Natta, F., Couturier, A., Bignet, F., & Vidal, P. P. (2013). A multidisciplinary approach to overreaching detection in endurance trained athletes. Journal of Applied Physiology, 114(3), 411-420.
Le Meur, Y., Pichon, A., Schaal, K., Schmitt, L., Louis, J., Gueneron, J., . . . Hausswirth, C. (2013). Evidence of parasympathetic hyperactivity in functionally overreached athletes. Med Sci Sports Exerc, 45(11), 2061-2071.
Manzi, V., Bovenzi, A., Impellizzeri, M. F., Carminati, I., & Castagna, C. (2013). Individual training-load and aerobic-fitness variables in premiership soccer players during the precompetitive season. The Journal of Strength & Conditioning Research, 27(3), 631-636.
Manzi, V., Iellamo, F., Impellizzeri, F., D’Ottavio, S., & Castagna, C. (2009). Relation between individualized training impulses and performance in distance runners. Med Sci Sports Exerc, 41(11), 2090-2096.
Nunan, D., Sandercock, G. R., & Brodie, D. A. (2010). A quantitative systematic review of normal values for short?term heart rate variability in healthy adults. Pacing and Clinical Electrophysiology, 33(11), 1407-1417.
Plews, D. J., Laursen, P. B., Kilding, A. E., & Buchheit, M. (2012). Heart rate variability in elite triathletes, is variation in variability the key to effective training? A case comparison. European Journal of Applied Physiology, 112(11), 3729-3741.
Plews, D. J., Laursen, P. B., Stanley, J., Kilding, A. E., & Buchheit, M. (2013). Training adaptation and heart rate variability in elite endurance athletes: opening the door to effective monitoring. Sports Medicine, 43(9), 773-781.
Shetler, K., Marcus, R., Froelicher, V. F., Vora, S., Kalisetti, D., Prakash, M., . . . Myers, J. (2001). Heart rate recovery: validation and methodologic issues. Journal of the American College of Cardiology, 38(7), 1980-1987.
Sparks, M., Coetzee, B., & Gabbett, T. J. (2017). Internal and External Match Loads of University-Level Soccer Players: A Comparison Between Methods. J Strength Cond Res, 31(4), 1072-1077.
Stagno, K. M., Thatcher, R., & van Someren, K. A. (2007). A modified TRIMP to quantify the in-season training load of team sport players. J Sports Sci, 25(6), 629-634. Stanley, J., Peake, J. M., & Buchheit, M. (2013). Cardiac parasympathetic reactivation following exercise: implications for training prescription. Sports Medicine, 43(12), 1259-1277.
Tanaka, H., Monahan, K. D., & Seals, D. R. (2001). Age-predicted maximal heart rate revisited. J Am Coll Cardiol, 37(1), 153-156.
Thorpe, R. T., Strudwick, A. J., Buchheit, M., Atkinson, G., Drust, B., & Gregson, W. (2016). Tracking Morning Fatigue Status Across In-Season Training Weeks in Elite Soccer Players. International Journal of Sports Physiology and Performance, 11(7), 947-952. doi: 10.1123/ijspp.2015-0490