Sports Medicine

, Volume 35, Issue 7, pp 597–617

Monitoring of Performance and Training in Rowing


  • Jarek Mäestu
    • Institute of Sport Pedagogy and Coaching SciencesUniversity of Tartu
    • Institute of Sport Pedagogy and Coaching SciencesUniversity of Tartu
  • Toivo Jürimäe
    • Institute of Sport Pedagogy and Coaching SciencesUniversity of Tartu
Review Article

DOI: 10.2165/00007256-200535070-00005

Cite this article as:
Mäestu, J., Jürimäe, J. & Jürimäe, T. Sports Med (2005) 35: 597. doi:10.2165/00007256-200535070-00005


Rowing is a strength-endurance type of sport and competition performance depends on factors such as aerobic and anaerobic power, physical power, rowing technique and tactics. Therefore, a rower has to develop several capacities in order to be successful and a valid testing battery of a rower has to include parameters that are highly related to rowing performance. Endurance training is the mainstay in rowing. For the 2000m race, power training at high velocities should be preferred to resistance training at low velocities in order to train more specifically during the off-season. The specific training of the international rower has to be approximately 70% of the whole training time. Several studies have reported different biochemical parameters for monitoring the training of rowers. There is some evidence that plasma leptin is more sensitive to training volume changes than specific stress hormones (e.g. cortisol, testosterone, growth hormone). In rowing, the stress hormone reactions to training volume and/or intensity changes are controversial. The Recovery-Stress Questionnaire for Athletes measures both stress and recovery, and may therefore be more effective than the previously used Borg ratio scale or the Profile of Mood States, which both focus mainly on the stress component. In the future, probably the most effective way to evaluate the training of rowers is to monitor both stress and recovery components at the same time, using both psychometric data together with the biochemical and performance parameters.

In the world of sport, athletes and coaches push themselves harder and harder in order to achieve the best results in competition. However, by increasing either the frequency, duration or intensity of training, they risk creating excessive fatigue that may lead to functional impairment, described as ‘overtraining syndrome’, ‘staleness’ or ‘burnout’.[1]The aim of sport training is to accustom the human body through different training loads and competitions, at the same time minimising the risk of illness, injury and fatigue in the period leading up to the competition. Thus, an athlete’s body must become accustomed to training loads that are intense enough to displace the homeostasis of an athlete. Once the adaptation to a certain training load has occurred, a greater load must be applied to get further improvement and stressful high-intensity training periods are necessary to obtain high performance in sports.[2]It has been shown that systematic recovery periods in the training process are necessary to prevent an overtraining syndrome and/or staleness.[3]Training should be organised in periods of stressful heavy training to induce sufficient training response followed by a period of reduced load to allow recovery and an increase in performance.[4,5]A problem for coaches is that athletes respond differently to the same training loads. A load that is too high for one athlete may have no training effect at all to another. It is evident, however, that underestimation or overestimation of performance level, trainability and insufficient recovery will lead to: (i) inappropriate training response of the athlete; or (ii) overreaching and eventually staleness, burnout syndrome or overtraining.[68]Hoffman et al.[9]pointed out that peak athletic performance depends on the proper manipulation of training volume and intensity as well as providing adequate rest and recovery between practice sessions.

Overtraining is a state of overexertion, which may be attained when the training load and the associated disturbances in homeostasis are not matched by recovery.[5]The first phase of overtraining is quickly reversible and is referred as ‘overreaching’.[10,11]Overreaching is characterised by underperformance, which is reversible within a short-term recovery period of 1–2 weeks and can be rewarded by a state of supercompensation (an increase in performance ability allowing 1–2 weeks of regeneration after a short-term phase of overtraining).[10,11]The term ‘staleness’, is the unwanted end result of overtraining,[11]and ‘overtraining syndrome’ should have the same meaning.[10]

Repetition of a training stimulus will result in a weaker response and, after a time, responses to training will asymptotically reach a limit. Further increases in performance can be reached by increasing intensity and/or duration of training.[12]For rowers, it has been demonstrated that training kilometres rowed are positively related to success in championships.[13,14]However, the risk of overtraining increases with training volume, particularly with monotonous training.[6,8,15]

A problem in studying training effects is the complexity of the goals of training, because different capacities (e.g. aerobic, anaerobic, strength) of the athlete have to be developed and improved,[5,16]so it also makes the monitoring process complicated. As a typical power endurance sport, rowers need physical strength to achieve high power per stroke, endurance to sustain this power for 2000m, as well as special motoric and tactical skills.[13,1618]Mood state also seems closely related to performance.[4,17,19]

Evaluation of the actual trainability and diagnosis of possible overload and overtraining are already among the most complicated tasks in sports medicine and sports psychology.[6,7,20,21]The monitoring process is effective if it has profound scientific foundations. The aim of the current review article is to give an overview of the methods used to monitor training in rowing and to provide some future suggestions to make the training monitoring process more effective and achieve better performance results during competitions.

1. Characteristics of Rowing Performance

Rowing is primarily a strength-endurance type of sport. The typical rowing competition takes place on a 2000m course and lasts, depending on a boat type and weather conditions, 5.5–7.0 minutes. During a race, muscle contraction is relatively slow and about 32–38 duty cycles per minute are used. In international rowers, the power per stroke may reach as high as 1200W and the average power per race is about 450–550W.[14]During competition, a rower depends mainly on his or her aerobic metabolism because energy stores and glycolysis are limited to cover the energy demand only for approximately 1.5–2.0 minutes.[14]Aerobic power can be defined as the maximal oxygen consumption (V̇O2max) as estimated during a performance that lasts from 2 to 10 minutes.[22]Aerobic and anaerobic energy contributions on a 2000m rowing race in different studies are presented in table I. According to Roth et al.,[23]the energy of the 2000m race was provided 67% aerobically and 33% anaerobically, 21% alactic and 12% lactic. Secher et al.[24]found that the aerobic energy contribution may be up to 86%.
Table I

Mean contribution of aerobic and anaerobic energy during rowing in different studies using elite heavyweight male rowers

Many factors affect physical performance during rowing. Power depends on aerobic and anaerobic energy supplies balanced by efficiency or technique.[22]Efficiency is expressed as the relationship between energy expenditure and boat velocity and it depends on the technical skill of the rower. Efficiency varies from as much as 16% to 21%, even during ergometer and tank rowing.[29,30]Differences in efficiency between rowers and non-rowers have been demonstrated, while no differences were observed between elite lightweights selected for the World Championships team compared with those who did not make the team.[30]This indicates that efficiency on an ergometer is only a rough estimate of technique in the boat.[22]

Testing an athlete is an attempt to evaluate his or her sport-specific performance. The easiest way of doing this is to measure the shortest time needed to cover a particular rowing distance. However, this is rather complicated, because external factors such as wind, water currents and temperature may influence the result. Furthermore, a need may exist to evaluate the individual contribution to a boat, including as many as eight rowers.[22]Accordingly, rowing ergometers are commonly used to measure individual performance parameters in rowers and training changes. Although rowing an ergometer does not require the same skills as on-water rowing, it has been observed that the ergometer simulates the biomechanical and metabolic demands of on-water rowing.[31]The dynamic rowing ergometer Rowperfect seems to more closely reflect the on-water rowing technique than the commonly used Concept II rowing ergometer.[32]Stroke angle/length, which did not vary with rate, was similar for both forms of rowing. The mean trunk, thigh and lower leg angles at the catch and finish of the stroke were also similar across the stroke rates.[32]However, rowing technique is a very complex task and consists of components such as balance, efficiency, and maintaining the boat speed during the recovery phase, which cannot be measured on an ergometer. Thus, estimating rowing technique as a complex on a rowing ergometer is of little value in analysing on-water technique.

Rowing ergometers should be considered valuable tools in testing, but they should be used with care when developing endurance during the preparation period because they may seriously affect the technique of on-water rowing.

Many researchers (see table II) have found performance-predictive parameters for rowers to predict 2000m rowing ergometer performance[28,3339]and two of them[35,36]also developed performance-predictive parameters for 2000m single scull distance. These studies used rowers of different levels, classification (i.e. scullers, sweep rowers) and sex. This may be a reason why each study had different equations of performance-predictive parameters. However, all these studies reported either V̇O2max in L/min or maximal aerobic power (Pmax) in W obtained on incremental test to exhaustion,[3339]or during standardised 2000m all-out rowing ergometer test,[28]to be an important parameter in predicting performance over a 2000m rowing ergometer distance. For example, Ingham et al.[34]and Womack et al.[39]found that V̇O2max, Pmax, and oxygen consumption (V̇O2) at a power eliciting a blood lactate concentration of 4 mmol/L were closely related to the 2000m ergometer performance time. In contrast, Riechman et al.[38]found that 2000m rowing ergometer performance time is best characterised by the mean power of an all-out 30-second ergometer test (i.e. a measure of anaerobic lactic power) and V̇O2max. Similar results were obtained by Jürimäe et al.,[36]who reported Pmax and mean power of 40 seconds of work (i.e. a measure of anaerobic lactic power) to be the best predictive parameters of 2000m rowing ergometer performance.
Table II

Performance predictive parameters for rowers on 2000m ergometer distance in different studies

It is now possible to determine aerobic capacity on the water due to solid-state portable measurement equipment, but to our knowledge there are no studies of the use of this type of apparatus in rowing. These results, in a rower’s natural environment can be interesting, e.g. how high can V̇O2max be on water compared with ergometer testing? However, these tests are logistically difficult to organise and they are probably more time consuming than laboratory testing. Presently, the testing procedures of rowers in their natural environment are conducted by our laboratory and are focused mostly on anaerobic threshold and V̇O2max differences between ergometer and on-water rowing.

Jürimäe et al.[36]compared ergometer rowing with on-water rowing and found that from different anthropometric and body composition characteristics only muscle mass was correlated to 2000m single scull distance time, while almost every anthropometric variable was related to 2000m rowing ergometer distance time. Similarly, Russell et al.[28]reported that anthropometric parameters predicted the 2000m rowing ergometer distance best when compared with metabolic parameters and a combination of categories. Thus, care should be taken when interpreting the rowing ergometer results to predict on-water rowing performance, because anthropometric variables may have too big an influence on the result. In smaller and lighter rowers, on-water rowing speed is usually compensated by higher physiological parameters, which is clearly indicated by the fact that in international regattas some lightweights may easily compete with their heavier peers.

The accurate analysis and assessment of various components of performance within the training context is an important process for coaches and sport scientists to include as an integral aspect of the training and competition programme of a rower. Determinants of competitive success include various psychological attributes such as self-motivation,[40]technical skills (including balance),[41]coordination with other crew members,[42] in addition to the physiological characteristics of muscular endurance, aerobic power, anaerobic power and strength.[43]It is difficult to rate these parameters, although everything depends on appropriate physiological characteristics. Some of these parameters can be compensated with others, for example, good technique may compensate for lower physiological parameters. If a rower has no coordination with team members he/she can still be a good single sculler. These parameters must be viewed together to allow better understanding of a rower’s state. Therefore, a good testing battery for a rower needs to consist of several parameters to determinate his or her performance and to make the selection process more effective. A key aspect to consider regarding a physiological test is the extent to which it is actually correlated with rowing performance.[44]

Changes in performance capacity can be analysed during all-out rowing tests in a rowing boat over various distances or on a rowing ergometer. Pmax during a standardised test (e.g. 2-, 6- and 7-minute all-out, 500, 2000, 2500 and 6000m all-out tests) can be used for evaluation of the exercise capacity.[22,25,26,35,36,39,4450]However, Steinacker et al.[20]argued that Pmax is subject to motivation of the rower tested and, therefore, may not be sensitive enough to monitor a complete rowing season. A question was then raised by the authors that more reliable test programmes, such as fast ramp tests could be used for measuring rowing performance, because they may fit into a training programme more easily. However, our experience (unpublished data) has showed that 1 minute of maximal rowing on a rowing ergometer has no relationship with changes in rowing ergometer performance over 2000m after heavy training periods. This is supported by the study of Smith,[44]who found no changes in 500m rowing ergometer time nor power after 3 weeks of overload training with a 33% increase in the frequency and a 30% increase in training volume and the following tapering week in international male and female rowers. This may be explained by the fact that anaerobic energy production may have too big an influence on the test results and it is well known that in successful rowers anaerobic capacity trainings are <10% of the whole training time.[14]

Endurance capacity is an important result of training and regeneration.[16,27]Higher performance at a fixed or individual lactate threshold means higher maximum performance, but there is a wide scattering of individual data of rowers. In successful rowers, the 4 mmol/L lactate threshold is in the range of 75–85% of their Pmax.[14,17]It should be taken into account that different testing protocols and different types of ergometers used may contribute different levels of 4 mmol/L blood lactate levels. For example, Lormes et al.[51]found higher lactate levels for a given heart rate on the Gjessing than on the Concept II rowing ergometer, possibly because of power losses in the transmission system of the Gjessing rowing ergometer. Maximal lactate values decrease with higher lactate threshold, as an indicator of increased endurance capacity.[14]However, using fixed or individual lactate threshold values measured on a rowing ergometer as guidelines for training intensity must be viewed with caution because they do not exactly mirror the blood lactate at steady-state for rowers.[52]Lactate threshold and maximum lactate values are also influenced by preceding exercise and muscle glycogen stores.[14]Thus, all variables must be standardised before testing the athlete to avoid difficulties in interpretation of the test results. In a glycogen-deficient state, maximum lactate and performance are depressed and lactate threshold virtually increased, but in the state of overreaching or overtraining without glycogen deficit, maximum lactate and performance capacity, and lactate threshold are decreased or lactate threshold unchanged.[16,20,53]

Nowadays, the blood lactate response to exercise is a commonly accepted tool for performance assessment and training prescription.[14,53]The blood lactate response has been thoroughly investigated and described using variety of terms and definitions.[54,55]The anaerobic threshold has been one of the most commonly used terms for describing the blood lactate response. Anaerobic threshold can be defined as the workload that can be performed by the oxidative metabolism and at which blood lactate production and release are balanced during continuous exercise.[14,5456]

To determine anaerobic threshold, numerous concepts and definitions have been published in the last decades.[5763]However, although a number of competing models exist to fit blood lactate concentration data during incremental exercise, there has been little comparison between different concepts in rowers.

Steinacker[14]and Wolf and Roth[64]have reported that the submaximal aerobic capacity measured as the power that elicits a blood lactate level of 4.0 mmol/L is the most predictive parameter of competition performance in trained rowers, especially in small boats such as singles and doubles. However, some authors have questioned the physiological significance of a fixed blood lactate value of 4.0 mmol/L, which does not take into account the individual kinetics of the lactate concentration curve.[65,66]Moreover, power at a blood lactate level of 4.0 mmol/L has not been reported to represent a steady-state workload in rowing.[52,67]Different anaerobic threshold concepts (fixed blood lactate values of 2, 3 and 4 mmol/L, Dmax,1 modified Dmax, LTLOG,2 increases of 1 mmol/L and over the resting level [for further explanation see Jürimäe et al.[70]]) and their relationships to rowing performance were studied by Jürimäe et al.[70]using 21 male national standard rowers. The results of this study indicated the anaerobic threshold value to be 3.7 mmol/L determined by the LTLOG method, which was lower than the suggested 4 mmol/L value by different authors[14,52,64]and may therefore be a better indicator for selecting training intensities without accumulation of blood lactate. The authors concluded that LTLOG represented the 2000m all-out rowing ergometer performance best. Moreover, a deflection point using the LTLOG method was very easily detected, allowing more accuracy for each athlete. Thus, the LTLOG value, which detects the anaerobic threshold with less subjectivity, may be a more appropriate measure of training modality in rowing; however, it must be still confirmed in future studies.

Coen et al.[71]described an on-water graded exercise test battery to measure individual anaerobic threshold and sport-specific performance on coxless pairs. The results of the study indicated that heart rates and performance on individual anaerobic threshold were significantly related between ergometer and on-water rowing. Furthermore, a graded incremental exercise test on water was strongly related to competition results over 2000m, especially when seat positions were considered,[71]which seems expected. However, performing on-water graded incremental exercise was extremely time consuming and the weather conditions may have too big an influence to make proper conclusions. Therefore, for reliable longitudinal data, the rowing ergometer is still the best apparatus for testing the performance parameters of rowers.

In summary, aerobic and anaerobic capacities both seem to be important parameters in competitive rowing, despite factors such as the level, sex, and body mass. Also, both capacities are important to measure to better understand the performance of a rower. For anaerobic threshold level, the LTLOG method may represent the threshold intensity better than the commonly used 4 mmol/L blood lactate concentration. It should be considered that a 2000m rowing ergometer performance is more suitable for rowers who compete in big boats such as fours and/or eights, to ensure a similar performance time. When performance of rowers in small boats is measured, a 2500m ergometer distance appears to more closely reflect the metabolic effort of on-water rowing on singles and doubles.[22,26,36,72]

2. Characteristics of Rowing Training

During rowing competition, anaerobic alactic and lactic as well as aerobic capacities are stressed to their maximum.[16]Therefore, the training of successful rowers has to be built up with a focus on aerobic training with the proper relationship of strength training and anaerobic training.

Endurance training (training at blood lactate concentration of 2–4 mmol/L) is the mainstay of success in rowing.[14,17,47,7375]Training of successful athletes is characterised by extensive as well as intensive endurance training with approximately 70–80% of the time spent on the water.[7577]Intense endurance training above the anaerobic threshold may be important for improvement of V̇O2max during the competitive season, but should not amount to >10% of the training volume.[14]Over a year, the percentage of specific rowing training time on the water is approximately 52–55% for an 18-year-old, 55–60% for a 21-year-old, and up to 65% for the older athlete.[78]However, the amount of time of specific rowing training must increase together with the qualification of a rower. Strength training is approximately 20% for an 18-year-old and 16% for an adult athlete, and general athletic training is in the range of 26–33%.[78]It is important to increase specific rowing training with increased training experience.[16]

The preparation period of rowers usually starts in October, where the main goal of training is to build up a base through extensive aerobic endurance training (90% of the total training time).[79]It appears that during the preparation period, rowers should de-emphasise resistance training at low velocities and emphasise power development at higher velocities (i.e. train more specifically for the types and velocities of movements used in the rowing technique and at speeds necessary to mimic the competitive pace).[46]The main period for developing strength-endurance is from January to March. The competitive period usually starts in March or April and culminates for elite rowers in late August or early in September with the World Championships. During the competition period, aerobic training is still the most important (about 70% of total training). Approximately 25% of the training during the competitive season is aerobic-anaerobic (blood lactate concentration 4–8 mmol/L) training and the rest is purely anaerobic (blood lactate concentration >8 mmol/L) training.[79]Steinacker et al.[53]investigated the time course of rowing velocity and ergometer results of a coxed eight during the training camp before the Junior World Championships in 1995. Such a preparatory training programme typically has a duration of approximately 4 weeks: 2 weeks high-intensity/high-volume training, tapering for 1 week and in the last week special preparation for the finals. The slowest boat speed of the 2000m distance was observed during the high-volume/high-intensity training, and the fastest boat speed was observed after the tapering period and at the World Championships, indicating that the rowers had overcome the over-reaching state after the high-intensity period and realised their performance capacity, indicating no overtraining.

In rowing, the problem with studying training effects is the complexity of the goals of training because different capacities (aerobic, anaerobic, power, strength, tactical skills) have to be improved.[16]This causes timing problems because several capacities cannot be developed at the same time. For example, endurance and sprint training are not appropriate to develop in the same training session. For rowers, one of the most important tasks is the maintenance of strength gains while training to enhance aerobic endurance simultaneously,[80]as resistance training is one physical performance factor of a complete annual training programme.[13,17]Research has shown that a sequence of resistance training prior to endurance training may be preferred for the off-season of rowing.[81,82]Bell et al.[80]found significant strength gains with a training frequency of three times per week for 10 weeks and was maintained for at least 6 weeks where the main goal was to develop aerobic endurance, and strength training was conducted only once or twice per week.[80]Whether strength gains can be maintained beyond 6 weeks while performing endurance training is not known[80]and needs further research. Bell et al.[81]also showed that organising strength and endurance training into sequential programmes can influence the physiological adaptation to training in rowers.

Most rowers also use unspecific and cross training to increase training tolerance and to avoid overtraining. During cross training, different muscle groups are recruited, which may allow partial recovery of other muscle groups and, therefore, the advantages of cross training seems to be ‘peripheral’ effects, enhancing or maintaining strength in power training and ‘central’ effects by decreasing the monotony of training.[16]

3. Selected Biochemical Indices of Monitoring Training in Rowing

In the world of training and coaching, different biochemical indices in blood are measured to prevent and to diagnose overtraining syndrome as an unwanted result of an athlete’s training regimen. However, to date there are still no valid diagnostic tools that would help us to prevent overtraining. Different hormonal responses were often proposed for monitoring overreaching and overtraining situations and also for the recovery period.[7,8,20,49,53,8387]Endogenous hormones are essentially involved in exercise-induced short- or long-term adaptations and influence the regeneration phase through the modulation of anabolic and catabolic processes after exercise.[88]Hormonal mechanisms that help mediate both short-term homeostatic control and long-term cellular adaptations to any type of stress are imposed on humans. For example, cortisol and growth hormone exert an essential role both in short-term (control of utilisation of energy substrates, mobilisation of protein resources) and prolonged stable (amplification of the translation process, supply of protein synthesis by ‘building materials’) adaptation to exercises.[89]

When planning the hormone sampling, the timing of sampling must always be considered because most of the hormones exhibit circadian rhythms and different aspects may have an influence on the result.[9092]Numerous investigations have studied the effects that different types of prolonged physical stress have on the hormones of hypothalamus-pituitary-adrenocortical axis.[5,83,93]Prolonged heavy endurance training has been found to cause the increase and decrease in the morning basal levels of cortisol and testosterone, respectively.[93]While resting levels of cortisol have reported to be unchanged[94]or decreased[95]after endurance training in male athletes.

In rowers, the further improvement of performance capacity was associated with increased growth hormone and cortisol levels and elite rowers have higher values of cortisol and growth hormone compared with national and medium performance level rowers.[49]These results are somewhat in contrast to Steinacker et al.,[96]who reported higher cortisol values for athletes who were not selected to the National Junior team of Germany, indicating higher catabolic activity. At an early stage during the training camp in 1996 in the German Junior Team when the training load was highest, basal cortisol levels increased by 18% and decreased slightly afterwards.[53]Elevated basal cortisol levels are often seen as a normal stress response to high-intensity training.[16,96]For example, the increase in cortisol level was found in junior rowers after anaerobic training during the last days before blood sampling.[96]There were no changes in plasma cortisol and testosterone concentrations after 2 hours of rowing at an intensity of 75% of anaerobic threshold.[97]

Steinacker et al.[5]found a 10% increase in human growth hormone from baseline during a high-load training phase (training load approximately 180 min/day for 2 weeks), and a decrease of 30% during the tapering phase where training volume was reduced for about 30%, while the intensity was maintained. The fact that not only the training intensity but also the volume may alter testosterone and growth hormone levels was supported in a study by Mäestu et al.[84]This study showed that almost doubling the training volume caused significant increases in fasting testosterone and growth hormone concentrations, which returned to the pre-stress level after the second tapering week, indicating a lag period. In summary, the results of these studies may suggest that the hormone responses of the hypothalamus-pituitary-adrenocortical axis are not specific and do not exactly mirror the amount of physical stress in rowing.

Adlercreutz et al.[98]and Härkonen et al.[99]stated that a condition of overstrain might exist in an athlete if at least one of the following two criteria are fulfilled: (i) free testosterone/cortisol ratio lower than 0.35 × 10-3; and/or (ii) a decrease in the free testosterone/cortisol ratio of ≥30%. During a rowing season, no significant relationships were found between the free testosterone/cortisol ratio and rowing ergometer performance parameters in male rowers.[93]However, a decrease in the free testosterone/cortisol ratio was observed after the training camp (range 4–40%), but these changes were not significantly related to performance parameters.[93]The authors concluded that the criterion of a decrease in the free testosterone/cortisol ratio of ≥30% or the free testosterone/cortisol ratio lower than 0.35 × 10-3 cannot be regarded as a first sign of overtraining.[93]In contrast, a significant decrease in the free testosterone/cortisol ratio as a result of an intensive training period and a slight increase after reduction of training intensity was observed in international-level junior rowers.[96]Meanwhile, the rowing ergometer performance was improved. Therefore, the free testosterone/cortisol ratio seems to be more useful as an indicator for a status of insufficient time to recover from training.[93]

Catecholamines stimulate cardiovascular and metabolic reactions and indicate physical and psychological stress.[100103]All-out rowing is associated with extremely high plasma catecholamine levels: 19 nmol/L and 74 nmol/L for adrenaline and noradrenaline, respectively.[104,105]Higher noradrenaline concentrations were noted during endurance training at similar heart rate on the Gjessing ergometer compared with rowing in the boat,[106]but adrenaline values were not statistically different. The authors concluded that, because of the higher sympathoadrenergic activation when exercising on the ergometer, the intensity of ergometer rowing should be set carefully.[106]

The plasma leptin level is identified as an adipocyte-derived hormone and its receptor has highlighted the regulation of appetite, thermogenesis and metabolism.[107]For a review see Friedman and Halaas,[107]and Flier.[108]However, it has become evident that leptin does not act only as an ‘adipostatic hormone’,[109]but it also profoundly affects hypothalamic neuroendocrine function as a result of negative energy flux.[85]Several studies showed that endurance exercise sessions decrease the plasma leptin concentration after 48 hours, in association with a preceding decrease in insulin,[110]while short-term exhaustive exercise has no immediate or delayed effect on circulating leptin concentration.[111]However, it has been recently demonstrated that a 30-minute all-out rowing ergometer test caused an immediate decrease in plasma leptin levels, probably by the involvement of all major muscle groups of the human body.[112]In the literature, the responses of plasma leptin to training are controversial. It has been suggested that fasting plasma leptin is not regulated in a dose-response manner in competitive male athletes,[113]while it has shown to be decreasing in heavy training in highly trained rowers[85,114]and increasing when training load is decreased. These differences could be explained by the differences in physical stress, as the training regimen of the swimmers in the Noland et al.[113]study was intensive interval training, while in Mäestu et al.[114]it was low-intensity high-volume training. Accordingly, it could be suggested that fasting plasma leptin response during prolonged heavy training stress may depend on the total amount of the physical stress.[114]

Jürimäe et al.[115]demonstrated a significant relationship between plasma leptin and training volume without changes in body mass. Furthermore, Simsch et al.[85]demonstrated a positive relationship between leptin levels and rowing performance. These results are in contrast to those of Petibois et al.[116]who argued that plasma leptin levels in top-level athletes parallel the changes in body fat and are not sensitive to an increase in training volume for trained individuals. Desgorces et al.[117]showed no changes in fasting plasma leptin concentration after 36 weeks of training. However, training volume might have been too low (i.e. 45 min/day in session 1 and 77 min/day in session 2) to induce a change in fasting leptin concentration. However, they showed the changes in post-exercise leptin concentrations, suggesting an improved regulation of energetic status after training (i.e. the better lipid availability). The hypothesis exists that leptin expression in the adipocyte is related to energy flux and triglyceride loss.[118]Further studies are needed to investigate the role of plasma leptin in highly trained athletes as a marker of training stress or overtraining. In addition to leptin, we have recently demonstrated that adiponectin could also be another adipocytokine to use as a marker of training stress in rowers.[119,120]

Recently, Urhausen and Kindermann[121]suggested that instead of resting hormone concentrations, maximal exercise-induced hormonal responses during and after a period of training overload should be studied to assess the adaptivity of the athletes. The elevation of hormone levels after rowing exercises performed at the maximal possible rate may be an overall expression of the dependence of hormone changes on the exercise intensity.[49]This means that there should be a wide reserve to increase the hormone responses to exercise. Only a sufficient performance capacity has to be achieved to evoke a correspondingly large rise in hormone concentrations.[49]

Despite the episodic character of growth hormone secretion, its response during exercise is characterised by a continuous increase of blood level during the exercise.[122]Growth hormone response to submaximal exercise has been found to decrease or disappear as a result of training.[123,124]There is also a possibility that fatigue may modulate growth hormone response.[122]For example, Urhausen et al.[86]demonstrated a decrease in exercise-induced rise of growth hormone in cyclists, while Lehmann et al.[16]and Mäestu et al.[114,125]found no change in resting nor exercise-induced values of growth hormone in runners and rowers, respectively. Therefore, the exercise-induced changes in growth hormone concentrations may have practically less value of training monitoring in rowers. However, in these studies, different exercise protocols and subjects were used and so they are not exactly comparable.

In the state of overtraining and overreaching, an intra-individually decreased maximum rise of cortisol[115,126]and insulin[126]has been found after a standardised exhaustive exercise test. Moreover, Jürimäe et al.[115]also demonstrated a significant decrease of exercise-induced (2000m all-out rowing ergometer test) leptin concentrations after a high-load training cycle in rowers. These studies may indicate that if hypothalamic downregulation exists during overreaching, analysis of exercise-induced cortisol and leptin concentrations may provide valuable information for detecting signs of overreaching, because in a state of fatigue the hypothalamic downregulation may be more sensitively detected. However, this is an area for future, well controlled research.

For several years, serum creatine kinase activity has been measured as a parameter of muscular stress in training-associated studies. A particularly important consideration relating to the use and interpretation of creatine kinase activity values in the sports sector is the dependence of this parameter on the nature of the stress.[127]Creatine kinase activity reflects training intensity and muscular strain only at the beginning of a training phase, decreasing creatine kinase activity levels suggest a muscular adaptation to training.[14,20,86]Morning levels of creatine kinase activity mainly represent its release during the previous day and are also influenced by creatine kinase activity clearance (50–80% per day).[96]Steinacker et al.[96]also found higher creatine kinase activity values in the junior rowers who were not selected for the national team despite their lower physical power.

In summary, it seems that there are a lack of valid biochemical markers of training stress in rowing. However, some recent studies[85,114,115]have reported the decrease of leptin in high-load training phases. It could be suggested that the maximal exercise-induced changes in leptin values may represent the more sensitive marker of overreaching for rowers.[115]

4. Selected Psychometric Instruments of Monitoring Training in Rowing

In addition to clinical findings, metabolic and hormonal values, the level of psychologically related stress and recovery seems to reflect well the clinical state of athletes.[2,5]Furthermore, mood state (e.g. motivation and striving for success) seems to be closely related to actual performance.[2,17,128130]To date, the most common psychometric instruments used are: the one-item Borg ratio scale,[129]which was developed to subjectively measure the intensity of the exercise; the Profile of Mood States (POMS),[130]which measures only current stress; and the Recovery-Stress Questionnaire for Athletes (RESTQ-Sport),[131]which allows measuring both subjectively perceived stress and recovery.

Marriot and Lamb[132]found a highly consistent relationship between Borg ratio scale perceptions of exertion on a rowing ergometer and heart rate. When the rating of perceived exertion was used as a means of producing an appropriate training heart rate, it was satisfactory but only at the higher intensities of effort (ratings 15 and above). Urhausen et al.[133]found significantly higher ratings of subjective exertion during the ‘stress test’ at the intensity of 110% of individual anaerobic threshold until exhaustion in overtrained endurance athletes. However, the one-item construction of the Borg ratio scale cannot assess different aspects of recovery and stress.[10]Moreover, it is difficult to interpret what causes the change of the scale after standardised exercise and, therefore, proper intervention is complicated. Thus, the Borg ratio scale is not suitable for monitoring training in highly trained athletes.

There is convincing evidence that athletes can be distinguished on the basis of psychological skills and emotional competencies.[134]POMS was initially developed as an economical method of identifying and assessing transient, fluctuating affective state.[130]POMS consists of 65 items and it yields a global measure of mood, consisting of: tension, depression, anger, vigour, confusion and fatigue. An overall score is computed by summarising the five negative mood states and subtracting the positive mood state (vigour). POMS has been used to measure the mood state of rowers,[2,40]swimmers[128]and runners.[135]During a training programme, mood state (the POMS was used three times during the season) did not differ between those who adhered to the training programme and the dropouts; however, those who remained in training had higher self-motivation.[40]During training, mood state increased, but surprisingly remained elevated in those who did not make the team and decreased in those who were successful.[40]Verde et al.[135]concluded that resting heart rate, sleep patterns and hormonal changes do not provide a useful early warning that the peak of the performance has been passed, but the POMS does show a consistent pattern of loss of vigour and fatigue during heavy training for 3 weeks in runners. Morgan et al.[128]measured mood states of swimmers throughout the season. At the beginning, the swimmers exhibited the ‘iceberg profile’, an indicator of a mentally healthy state. In the high-load phase, mood disturbances increased and a profile reflected poor mental health. After reducing the intensity, the swimmers demonstrated the ‘iceberg profile’ again. However, it should be stated that five of the six scales of POMS measure the negative mood characteristics of tension, anger, fatigue, depression and confusion and, therefore, POMS only vaguely reflects recovery processes. Berger and Motl[136]argued that a decrease in a negative mood state may not necessarily indicate mood benefits. Thus, when using POMS as the psychometric tool to monitor the mood states of athletes, the main focus is on stress-related behaviour and it might be not appropriate to evaluate the recovery of athletes.

Restricting the analysis to the stress dimension alone is insufficient, especially in high-performance areas, since the management of training intensity and volume is tightly linked to outstanding performance.[5]A psychometrically based instrument to assess the recovery-stress state is the RESTQ-Sport.[131]The recovery-stress state indicates the extent to which persons are physically and/or mentally stressed, whether or not they are capable of using individual strategies for recovery as well as which strategies are used.[4]This questionnaire was created to get distinct answers to the question “How are you?”[10]and addresses physical, subjective, behavioural and social aspects using a self-report approach.[2]The theory behind the questionnaire is that an accretion of stress in everyday life, coupled with weak recovery potential, will cause a variation of the psychophysical general state.[10]A Likert-type scale is used with values ranging from 0 (never) to 6 (always), indicating how often the respondent participated in various activities during the past 3 days/nights. The mean of each scale can range from 0 to 6, with high scores in the stress-associated activity scales reflecting intense subjective strain, whereas high scores in the recovery-oriented scales mirror plenty of recovery activities.[4]However, it has to be considered that the RESTQ-Sport and the POMS are not direct measures of physiological states of the organism, both instruments reflect the subjective representation of these states.[2]

Several studies with rowers have showed that the POMS scales of depression, anger and fatigue are negatively correlated with recovery-associated scales of RESTQ-Sport, at the same time vigour is positively correlated and vice versa. A positive relationship exists between the stress-related scales of RESTQ-Sport and depression, anger and fatigue, while vigour appears to be negatively correlated with stress scales.[4,19,137]Longitudinal studies in athletes have shown that the RESTQ-Sport can sensitively monitor stress and recovery processes in training camps and throughout the season, and the 19 scores of the RESTQ-Sport have acceptable levels of reliability.[2,4,5,19,138]Moreover, a dose-response relationship was demonstrated between training volume (daily rowed kilometres) and the somatic components of stress and recovery.[4,5]These results together indicate that RESTQ-Sport could be a potential tool for monitoring the training of elite athletes using subjective stress and recovery indices. However, so far, RESTQ-Sport has mainly been used in German and American athletes, thus it would be beneficial if this questionnaire could be translated and modified into other languages to take into account the quiddity of the nation. The Estonian version of RESTQ-Sport has successfully been used on rowers during different training periods.[138140]

In summary, through utilisation of the RESTQ-Sport, coaches and athletes can be informed of the importance of daily activities and how these activities are related to the recovery-stress state of an athlete compared with the frequently used one-item Borg scale or POMS, which generally measure the stress-related behaviour and, therefore, could not be sufficient in high-performance areas. Previous studies[4,19,53,85,138,139]utilising RESTQ-Sport suggest that it is appropriate to measure recovery-stress state changes during rowing training in highly trained male athletes.

5. Studies on Monitoring Training in Rowing

In the literature, there are not very many longitudinal studies that deal with training monitoring of rowers (table III). Most of them are 3–5 weeks in duration, i.e. one microcycle. Only some of them[33,49,50,93]are longer than one microcycle. However, these longer studies tend to monitor only one specific pattern or the time interval between two different testing batteries is too long, making the analysis more difficult.
Table III

Studies involving training monitoring in rowers studies

Recently, a study reporting the changes in training and performance of top Norwegian rowers over the last 30 years was published.[141]They reported a 20% increase in training volume and nowadays their international medal winners train 1100–1200 hours/year.

There is also a lack of studies that deal with rowing performance and resistance training. Bell et al.[8082]have investigated high-velocity resistance training (HVRT) in relation to rowing training. However, they have compared HVRT to anaerobic power and reported no significant changes in anaerobic power after different training programmes.[142]High-intensity resistance training was also used in the study of Simsch et al.[85]and the stagnation in performance of an incremental ergometer test was detected; however, the performance increased significantly after endurance training. It is known that building up the strength-endurance capacity during the off-season is one part of a rower’s winter training programme. The impact of low-velocity resistance training (e.g. frequency, intensity) on rowing performance during the off-season needs some further research.

Achieving the maximal result during competitions is the major purpose of athletic training. A rower is not always in his or her best shape during major competitions, therefore, valuable information could be obtained from studies that deal with specific preparation for competitions at different levels of rowers. However, it has to be stated that these studies must not be case studies, but controlled ones (control group would benefit), which are difficult to prepare because most athletes do not want to experiment with their training prior to competition. Furthermore, the competition results are always difficult to analyse because they depend on factors such as the effects of counterparts and weather conditions.

6. Multi-Level Approach of Training Monitoring

The evaluation of the clinical state of an athlete, i.e. of current trainability and of the diagnosis of overload and overtraining, is already one of the most complicated tasks in sports medicine.[7,8,53,85,143]There is a cascade of various responses to prolonged training that can be used in training monitoring. It is also evident that only some of the parameters are reliable and specific enough. One should also distinguish between parameters in which the inter-individual response differs, such as creatine kinase activity, hormonal parameters or mood state, and parameters that are hard to tolerate, such as physical performance.[53]The hypothalamus acts as the central integrator of all afferent signals to the brain and has an important role in the regulation of the central responses to stress and training.[108]Such integration involves information from autonomic nerve system afferents, direct metabolic effects, hormones and also different information from different brain centres.

There is experimental evidence that all metabolic hormones have hypothalamic receptors.[144]Leptin and insulin depress the activity of excitatory neurons in the lateral hypothalamus[145,146]and have effects on energy expenditure, body mass control and sympathetic activity.[108]High levels of leptin have been found to inhibit activation of the hypothalamus-pituitary-adenocortical axis and inhibit cortisol release.[147]Therefore, studying the effects of leptin may have the advantage of knowing the amount of stress affecting the organism.

In the high-load training phases, which are essential to achieve improvements in performance through overreaching, decreases of steroid hormones could be observed,[16,91,93,96,98,148,149]suggesting that hypothalamic downregulation will occur in a state of overreaching. This was also demonstrated in experimental training by Lehmann et al.[150]and Barron et al.[151]At present, the mechanism by which the hypothalamus senses metabolic imbalance and fatigue in athletes is speculative.

Using 11 elite rowers during their preparation for the 1996 Atlanta Olympics, Kellmann and Günther[4]found that the alteration of extensive endurance training was well reflected in psychological measures using RESTQ-Sport. High duration was indicated by elevated levels of stress and simultaneously lowered levels of recovery. Moreover, the scales ‘somatic complaints’, ‘lack of energy’, ‘fitness/injury’ and ‘fitness/being in shape’ described the dose-response relationship with the training load. However, the different trends in the RESTQ-Sport scales may be explained by the different time courses of hormones and corresponding scales.[53]For example, ‘somatic complaints’ were highest with the highest training load and elevated cortisol concentrations as well as creatine kinase activity.

Steinacker et al.[5]found disturbance of the homeostasis after high-intensity, high-volume training for 3.2 hours/day after 18 days in male junior rowers. This disturbance was also reflected in psychometric scales, in performance, metabolic and hormonal parameters (depression of peripheral and steroid hormones), and was restored and supercompensated after tapering as indicated by the boat speed.[5]Therefore, psychometric scales, in which changes have known to be related to blood hormone concentration changes, may be an alternative measure of an athlete’s current state.

Monitoring the current levels of both stress and recovery has the possible advantage that problems may be detected before symptoms of overtraining are likely to appear.[2]Steinacker et al.[5]found that both performance and hormonal indices of training were reflected by the scores of the RESTQ-Sport. One week of a heavily increased training volume (100% compared with previous week) indicated increased levels of stress and decreased level of recovery-associated activities with a significant changes in fatigue and social relaxation in male junior rowers.[139]

For monitoring, it is also important that mood is correlated to physical performance ability, hormonal parameters and metabolic data.[53,140]Naessens et al.[152]demonstrated a U-shaped relationship between subjective fatigue ratings and sympathetic tone (basal noradrenaline excretion). RESTQ-Sport allows monitoring of mood state in athletes, but different scales of RESTQ-Sport have different time courses that have to be taken into account.[2]Physical complaints, fatigue, and general stress[5,138]as well as conflicts/pressure[139]were highest with highest training load, elevated cortisol concentrations and high creatine kinase activity.[53]Sleep quality, personal accomplishment, self-efficacy and general well-being were lowest with the highest training volume.[138,139]Moreover, cortisol was found to be related to all of the stress scales (except disturbed breaks) of RESTQ-Sport after 3 weeks of high-volume training in male rowers.[138]Fatigue has also been found to peak together with sympathetic activation (noradrenaline secretion).[53]From the perspective of a biopsychological stress model,[53,153]recovery and stress should be treated using a multi-level approach dealing with psychological, emotional, cognitive behavioural/performance, and social aspects of the problem, considering these aspects both separately and together.[5]

In summary, it is suggested that investigating physiological and psychological aspects of rowers are an advantage of more effective training monitoring in highly trained rowers.

7. Future Investigations and Recommendations

To date, most of the studies on training monitoring in rowers have investigated only one microcycle and the following recovery period. To our knowledge, only the study of Simsch et al.[85]has used two consecutive microcycles and the recovery period between them. However, these two microcycles were different from each other. During the first training cycle the focus was mainly on resistance training and during the other microcycle the focus was on endurance training. In the future, the studies with two or more training cycles and recovery periods between them, could possibly provide the advantage of learning more about sustainable training load and different markers of monitoring in rowing training. However, these studies are difficult to control and are usually costly. Recent studies also confirm that recovery and stress should be monitored simultaneously in high performance areas to prevent athletes from overtraining.[4,5,19,53,137139]

In addition, studies with individual analysis may have an advantage of so far used cluster analyses because athletes respond differently to similar training loads. For example, if after a high-load training cycle, some of the subjects show no change in performance, some of them improve and the rest have worsened performance. In this situation, the results of athletes with worsened performance, who may experience overreaching or even overtraining syndrome, disappear into the whole cluster, making the final analysis difficult and more speculative.

The main principle of testing is minimum testing – maximum reliable information. The more direct the link between a parameter and a specific performance, the more value the testing has.[122]As mentioned in section 1, testing a rower is a complex task. As it is known from the literature, rowers with similar V̇O2max may have different rowing performance.[14]It appears that the V̇O2max is not the best predictive parameter for rowers, but an ability to sustain high V̇O2 during competition (i.e. endurance capacity). Therefore, it is advisable to determine the V̇O2 during the simulated rowing ergometer race, either on 2000 or 2500m all-out rowing, and the average V̇O2 should be recorded.

8. Conclusions

Training monitoring studies in rowers should become more specific in their nature. There appears to be no single marker of training monitoring and possible overtraining in rowers. In the future, studies should focus on different fasting blood biochemical markers during different training periods. For example, some studies[84,85]indicate that the plasma leptin could be used as a marker of training stress during low-intensity training in rowers, while maximal exercise-induced stress hormone changes may indicate early signs of overreaching.[121]There are also supportive studies[3,4,11,19,121,139,142]that indicate psychometric data for training monitoring.


Dmax is the point on the blood lactate concentration curve that yields the maximal perpendicular distance to the straight line formed by the two end datapoints.[68]


LTLOGis the power output at which the blood lactate begins to increase when the log (blood lactate) is plotted against the log (power output).[69]



No sources of funding were used to assist in the preparation of this review. The authors have no conflicts of interest that are directly relevant to the content of this review.

Copyright information

© Adis Data Information BV 2005