Flatwater kayaking is a sport that places exceptional demands on the upper body and trunk musculature (Tesch, 1983). Previous research papers (Bishop, 2000; Fry and Morton, 1991; Gray et al., 1995; Tesch, 1983) suggest that flatwater kayak paddlers possess high values for maximal aerobic and anaerobic capacities and upper-body muscle strength.

Kayak paddlers spend the majority of their race at or around peak VO₂ (Bishop, 2000) and obtain the majority of the required energy from the aerobic system (Bishop, 2000; Fernandez et al., 1995). Zamparo et al., 1999 concluded that the fraction of average power provided by oxidative processes increased with the distance covered, whereas that provided by anaerobic sources decreased. In brief, the aerobic contribution, expressed as a fraction of VO₂ max, was shown to be 73% for the 500m and 85% for the 1000m (lasting approximately 1min 45 and 3min 45 respectively). These data support that of an earlier study performed in six elite kayak paddlers (Tesch et al., 1976). The importance of the anaerobic contribution, however, cannot be discounted. Studies such as Bishop, 2000 and Fernandez et al., 1995 suggest that Olympic kayak paddlers not only need a high aerobic power, but the anaerobic contribution is also very important for successful performance.

Anthropometric data available for male and female, elite sprint canoe/kayak paddlers suggest a homogenous shape and size (Ackland et al., 2003). Ackland et al., 2003, assessed 50 male and 20 female sprint canoe/kayakers who competed at the Sydney Olympic Games (2000) representing 9 countries (Table 1). Sydney Olympic paddlers compared to paddlers represented at the Montreal Olympics in 1976, were approximately five kilograms heavier on average. However, with comparable skin fold values for the two groups, it was suggested by Ackland et al., 2003 that the subjects in the present sample have a higher proportion of lean body mass. It was therefore speculated by Ackland et al., 2003 that the morphology of elite paddlers have altered during the past 25 years and shifted toward a heavier but more lean physique. The winning times for the k1 500m and k1 1000m events from the two Olympic games (1976 vs. 2000) show a vast improvement and give some indication of the developmental trends and support the suggestion of improved physical capacity over the two distances. Over the 500m distance, the times went from winning in 1:46.41 in 1976 to the present winning time of 1:37. 919, and over the 1000m event, from 3:48.20 to 3:25.898. Arguably, the technological advancements in boat and paddle design should also be considered as it was suggested by Robinson et al., 2002 that the time differences presented have been closely related to these technological advancements, as well as changes in the paddlers themselves.

Ackland et al., 2003 noted that sprint kayak paddlers possess unique characteristics not commonly observed in the general population. These include a lean body composition with proportionately large upper body girths and narrow hips (for males). The mean somatotype recorded by Ackland et al., 2003 (1.6 - 5.7 - 2.2 for males, 2.4 - 4.7 - 2.0 for females) demonstrated that kayak paddlers are best described as mesomorphs.

Despite the suggestion put forward by Ackland et al., 2003 that elite sprint kayak paddlers physical characteristics are homogeneous, Fry and Morton, 1991 and van Someren et al., 1999 argue that although success is not correlated with greater body mass, the kayaker can be of considerable size without compromising performance. Casual inspection of the data in Table 1 suggests a range of morphological characteristics of this population. It is interesting to note that a comparable on water sport, rowing, relies heavily on the morphological characteristics of their athletes for talent identification as there are different weight classes. However, when analysing the morphological characteristics of elite rowers, especially where size restrictions apply, there is a relatively small variation. Previous studies (Secher, 1990; 1992; 1993; Shephard, 1998) suggest that successful rowing competitors are very tall, with a large lean body mass and aerobic power. However, a slight variance may be accepted and incorporated in a variety of positions in sports where the athletes size and shape are a product of the nature of the sport (Ackland et al., 2003). Kayaking is one such sport, where although paddlers possess unique characteristics not commonly observed in the general population, there is no single trait that distinguishes an elite kayak paddler.

Considering the potential role of physical characteristics, when examining paddlers using the ergometry system, all subjects overcome the same resistance in order to perform work, irrespective of body mass (Bishop, 2000; van Someren et al., 1999). Thus, an increased body mass may not only not compromise ergometry performance, but might enhance it. Kayak racing is performed on water and as mentioned above in the study by Ackland et al., 2003, kayakers are now heavier and have a greater percent lean body mass compared to 25 years ago. While a larger individual may have a larger absolute peak VO₂, potentially, a too large body mass of the kayaker may negatively affect the relative peak VO₂ attainable and cause the kayak to sit deeper in the water, increasing wetted area of the kayak (Jackson, 1995). This increased wetted area will increase the frictional and wave drag (Jackson, 1995) thereby increasing the resistance that must be overcome by the kayaker to propel the kayak forward (Pendergast et al., 2005). Thus, the question then arises: how are power output and kayak drag affected with an increased lean body mass; do they cancel each other out, or is there an advantage in being light? Further research is essential to examine the full potential of the advantages and disadvantages between the range of body types mentioned above and examine the energy cost associated.

The energy cost of kayaking is determined by the drag of the kayak and the efficiency of the kayaker to overcome that drag. As such, the importance of both drag and efficiency in determining the metabolic requirement of kayaking, which is highly variable, is critical. The energy cost of paddling a given distance increases with increasing velocities according to a power function (Pendergast et al., 1989; 2003; 2005; Zamparo et al., 1999). Peak kayak paddling performance is therefore dependant upon maximal metabolic power (aerobic and anaerobic) complimented with superior locomotion economy.

Elite male kayakers appear homogeneous in shape and physical size, being differentiated from the general population by their greater upper body girth and narrow, hips (Ackland et al., 2003) and demonstrate superior aerobic and anaerobic qualities (Hahn et al., 1988; Tesch et al., 1976; Tesch, 1983; Pendergast et al., 1989; Zamparo et al., 1999). Kayakers have reported VO₂peak values of around 58 ml·kg^-1·min^-1 (4.7 L·min^-1) and lactate values of around 12 mM during laboratory and on water testing. For kayaking, a sport that relies on high maximal aerobic power, the anaerobic energy system also seems to be important for successful performance. van Someren and Oliver, 2001 reported that the mean lactate threshold occurred at a blood lactate concentration of 2.7 mmol·L^-1, at a HR of 170 beats·min^-1 and a VO₂ of 44.2 ml·kg^-1·min^-1. The lactate threshold presented corresponded to a percentage of 89.6% of the maximum heart rate and 82.4% of the VO₂ peak.

Although the absolute values of peak VO₂for kayaking have been described to be quite high (Tesch, 1983), they are not quite as high as other sporting events such as road cycling, rowing or running. Billat et al., 1996 showed on the kayak ergometer, the power output of kayak paddlers at VO₂peak was only 57% of the power output produced by cyclists on their respective ergometer. In other sports, such as road cycling and distance running where the lower body is the main source of energy output, values tend to be around 73 ml·kg^-1·min^-1 (Billat et al., 1996, Lee et al., 2002) and 74 ml·kg^-1·min^-1 (Lucia et al., 1999) respectively, much higher than those reported for kayaking. When comparing the results from bicycle ergometry in Tesch et al., 1976 to the results presented for kayak paddling in the same study, it was reported that the oxygen uptake for the 500m races corresponded to 77% of the individual VO₂peak during the leg exercise. In the 1000m, the oxygen uptake corresponded to 87% of VO₂ peak of the leg exercise. It was also reported by Hahn et al., 1988 that the peak oxygen uptake recorded on the kayak ergometer was 89.1% of the maximal oxygen uptake achieved on the arm/leg ergometer. These studies suggest that physiological differences exist between upper and lower body aerobic exercise. It was therefore speculated that if the kayakers VO₂ were to be normalised for arm mass and the cyclists for leg mass for example, the differences observed in VO₂ may not be quite as large as those presented and thus compare favourably with other endurance sporting events.

It was the purpose of this review to summarise published physiological data relating to men’s and women’s kayaking. It can be concluded that flatwater kayaking is characterised by exceptional demands on upper body performance. A successful kayaker not only requires high aerobic power, but a high anaerobic energy yield and great upper body muscle strength is also of great importance. Consequently the peak relative VO₂ value attainable is negatively affected.

Exercise physiologists have utilised a variety of testing methods to estimate VO₂ peak in kayakers (Table 2). The proportion of whole body musculature required during exercise in studies testing kayak paddlers varies. Physiological tests of kayak paddlers have been carried out during on water analysis (Gray et al., 1995; Pendergast et al., 1989; Tesch, 1983; van Someren et al., 1999), kayak ergometry (Billat et al., 1996; Bishop et al., 2002; Fry and Morton, 1991; Hahn et al., 1988; Pendergast et al., 1979; Tesh et al., 1976), arm cranking (Bergh et al., 1976; Pendergast et al., 1979; Tesch, 1976; 1983; Tesch and Lindeberg, 1984), bicycle ergometry (Bergh et al., 1976; Pendergast et al., 1979; Pyke et al., 1973; Tesch et al., 1976), treadmill running (Sidney and Shephard, 1973; Tesch, 1983; Tesch et al., 1976) and combined activity of the legs and arms (Bergh et al., 1976; Hahn et al., 1988; Pendergast et al., 1979). The effects of proportion of muscle used must be taken into account when discussing kayak paddler performance.

Although a number of testing methods have been preformed, the ideal method of testing is to measure oxygen consumption on the water. Six male Swedish sprint kayakers of Olympic standard were reported to reach a peak oxygen uptake of 4.67 L·min^-1 during an on water 1000m race (Tesch, 1983). In another study by van Someren et al., 1999 9 well trained kayak paddlers produced an average peak value of 4.27 L·min^-1 for VO₂ for the same race distance at maximal effort, lower than the 4.71 L·min^-1 and 4. 67 L·min^-1 reported when measuring elite Swedish kayak paddlers (Tesch et al., 1976 and Tesch, 1983 respectively). Considering that all studies examined paddlers during maximal efforts, it was speculated that the differences observed were a result of the subject characteristics in the study by van Someren et al., 1999. The subject population were not of elite standard and one could argue they were at a lower skill and conditioning level.

Findings by Fry and Morton, 1991 further support van Someren et al., 1999. It can therefore be assumed from these results presented that the more skilled paddlers are more likely to obtain a greater peak rate of oxygen consumption. Studying 38 kayak paddlers from the Western Australian championships, Fry and Morton, 1991 classified the paddlers as either state team paddlers and non-state team members based on an objective selection policy, including performance time and position. Using a Monarch mechanically braked bicycle ergometer mounted on a kayak frame, the VO₂peak values for the subjects were determined using a progressive test to exhaustion. The mean VO₂peak values for the state team kayakers reached 4.78 L·min^-1, a value significantly higher than the mean VO₂peak for the non-state team paddlers (3.87 L·min^-1). However, when maximal oxygen consumption was expressed in ml·kg^-1·min^-1, although remaining higher, there was no significant difference between the state and non-state team paddlers. Bishop, 2000 explains that while a large aerobic power is very important, anthropometric characteristics can also influence performance. Paddlers of the state team were found to be slightly heavier and taller than the less successful paddlers. Significant strength differences and reduced, but not significant skinfold measurements, were also found to be associated with the state team members. To account for the difference in oxygen consumption, Fry and Morton, 1991 suggested that the aerobic power to weight ratio is not as important for kayak success as absolute aerobic power. This implies that the kayaker can afford to be large without detriment to success provided they can produce high levels of aerobic power.

A potential confounder to the suggestion from Fry and Morton, 1991 above is that no measures of technique or skill, other than time to complete the task, were recorded. One could argue that on the basis of rowing studies by Smith and Loschner, 2000 skilled kayak paddlers are better able to minimise any excess body movements within the kayak to provide a more powerful and efficient stroke compared to their sub-elite counterparts. Rowing studies, such as Loschner et al., 2000 and Smith and Loschner, 2000 have analysed the movement of a rowing scull and found the amount of yaw (sideways deviation), pitch (bobbing movement) and roll (roll of the boat from side to side) induced in the boat by a rower affected the efficiency of boat propulsion and hence influenced the velocity of the rowing scull. Considering the complex nature of kayaking, this could potentially manifest as a change in oxygen consumption of the athlete if differences in technique are assumed between performance level and would be a fertile area of future investigation.

Maximal measures of oxygen uptake are typically determined in a laboratory during treadmill running or pedalling a cycle ergometer. However, athletes who are predominantly involved in upper-body work may not be accustomed to this form of exercise. Consequently, testing the lower body is potentially inappropriate as it is not specific to the trained task of the athlete and as a result the performance may not elicit optimal results of VO₂peak. Stromme et al., 1977 determined the VO₂peak of cross-country skiers, rowers and cyclists during uphill running on a treadmill and during maximal performance on their specific sport activity. All athletes attained higher levels of VO₂peak during their specific sport activity than during treadmill running (male and female skiers: 2.9% and 3.1%, respectively, p < 0.001, rowers: 4.2%, cyclists: 5.6%, respectively, p < 0.01).

Although the absolute values of peak VO₂ for kayaking have been described to be quite high (Tesch, 1983), they are not quite as high as other sporting events such as road cycling, rowing or running (Table 3). An apparent difference is also evident when examining VO₂peak in relative units among the different sporting disciplines. A number of researchers (Fry and Morton, 1991; Hahn et al., 1988; Tesch, 1983; van Someren and Oliver, 2001) have suggested values of the relative VO₂peak for kayaking (58 ml·kg^-1·min^-1) compare favourably with water sports such as swimming (58.4 ml·kg^-1·min^-1; Roels et al., 2005). However in sporting activities where the lower body is the main source of energy output, such as road cycling and distance running, values tend to be around 73 ml·kg^-1·min^-1 (Billat et al., 1996, Lee et al., 2002) and 74 ml·kg^-1·min^-1 (Lucia et al., 1999), much higher than those reported for kayaking. Similarly in rowing, although the peak absolute VO₂ of rowers compared favourably with the results obtained for road cycling and middle distance runners (~5-6 L·min^-1), when the VO₂ was expressed in relative units (~64 ml·kg^-1·min^-1) these were not quite as high as the mean values obtained for the other endurance type athletes mentioned above. The differences may be explained by the fact that generally, long distance runners are small, thin and lightweight compared to kayakers or rowers, who have a larger body mass. One body type and training style will favour better performance in each sporting discipline. Another important concept is what is the active muscle mass involved in generating the VO₂max. For example, if the VO₂max of a kayaker is divided by the mass of only the upper body (discounting the legs that are not used extensively in kayaking) their relative VO₂max may be even high than that of distance runners.

To describe the differences present Billat et al., 1996 examined the power output at VO₂peak of different sporting athletes. Billat et al., 1996 studied 41 elite national class sportsmen; road cyclists (n = 9), flatwater kayak paddlers (1000m) (n = 9), middle distance swimmers (400m) (n = 9) and long distance runners (3000-10 000m) (n = 14). The athletes were required to perform two tests; a maximal VO₂ test and an all out exercise bout performed at the power or velocity that elicited VO₂peak (see Table 3). Each subject performed the exercise tests on their specific ergometer, using a Cosmed device (a telemetric oxygen measuring system) to measure VO₂peak. Notably, on the kayak ergometer, the power output of kayak paddlers at VO₂peak was only 57% of the power output produced by cyclists on their respective ergometer (Table 3). Billat et al., 1996 suggested it was the result of the smaller muscle mass involved with paddling a kayak. It can be speculated that if the kayakers VO₂ were to be normalised for arm mass and the cyclists for leg mass for example, the differences observed in VO₂ may not be quite as large as those presented and thus compare favourably with other endurance sporting events. Interestingly, in the study by Billat et al., 1996 when examining the time to exhaustion for kayak paddlers, a significantly higher time was reported when comparing the results of cyclists (p < 0.05). Furthermore, it was also reported by Billat et al., 1996 that the heart rates and blood lactate levels for the kayak paddlers were similar among all sports except for swimmers, who generated lower HR and blood lactate levels at the end of the incremental tests.

van Someren and Oliver, 2001 reported the mean lactate threshold of kayak paddlers at a blood lactate concentration of 2.7 mmol·L^-1, at a HR of 170 beats/min and a VO₂ of 44.2 ml·kg^-1·min^-1. The lactate threshold presented, corresponded to a percentage of 89.6% of the maximum heart rate and 82. 4% of the VO₂ peak. These values presented indicate the extreme nature of kayaking and the undue demands it places upon the anaerobic system. Considering that, it was noted by Bishop, 2000 earlier, that kayak paddlers spend the majority of their race at or around VO₂peak. In another study by DalMonte and Leonardi, 1976, it was reported that the onset of blood lactate accumulation of international kayakers occurred in a time frame between 79 and 87% of VO₂peak. Furthermore Bunc and Heller, 1991, who defined the ventilatory threshold as the level of maximal exercise at which the subject is still capable of working close to steady state, reported that the values of the ventilatory threshold in international male canoe/kayaker paddlers corresponded to an exercise intensity of between 83 and 85% of VO₂peak. At intensities greater than the anaerobic threshold, the rates of anaerobic glycolysis and subsequent lactate production are very high (Bishop et al., 2001).

When the rate rise of blood lactate is plotted as a function of work load for sedentary subjects and white water kayakers (Pendergast et al., 1979), it is clear that the anaerobic threshold for sedentary subjects occurred at a lower work load (75W) requiring approximately 70% of their arm VO₂peak compared to the kayakers. The anaerobic threshold for kayakers increased to approximately 125W, which corresponded to approximately 80% of their arm VO₂peak. This notable difference observed between the sedentary subjects and their kayaking counterparts demonstrates the kayak paddlers’ ability to withstand high levels of arm exercise before fatigue sets in.

Tesch, 1983 suggested that the elite paddlers examined in his study exhibited a well developed anaerobic capacity for upper-body exercise. The relatively high peak blood lactate concentration values observed following maximal kayak racing (13 mM; Tesch, 1983) indicate a significant anaerobic contribution to kayak performance (Bishop, 2000) (Table 4). It was suggested by Pendergast et al., 1979 that the kayaker paddlers were capable of developing at least twice the anaerobic power of sedentary subjects for arm cranking exercises. In light of this however, Pyke et al. (1973) suggested that to assess the physiological responses to work by kayakers, the circular movement of hand/arm cranking on an ergometer does not closely simulate the movements required on the water. It is essential to utilise specific ergometers to allow for an accurate comparative analysis.

Associated with the results presented above, were the findings presented by Davis et al., 1976 and Tesch and Lindeberg, 1984 who reported kayakers to have lower blood lactate levels during similar absolute exercise intensities or work loads relative to maximal oxygen uptake when compared to sedentary subjects performing arm exercise. Therefore, as previously mentioned by Tesch, 1983, in kayaking and numerous other sporting events that require high aerobic energy supply, the anaerobic energy system seems to be an important factor for successful performance.

To compare the lactate levels among a number of different sporting athletes Tesch and Lindeberg, 1984 studied 7 elite male kayak paddlers, 6 national calibre weight/power lifters, 8 local and national calibre body builders and 6 physically active non athletes. All subjects performed upper body exercise seated on an arm crank ergometer. Results suggested that the kayakers exhibited a significantly lower blood lactate concentration (p < 0.05) at all power outputs tested. The authors suggested that factors other than muscle volume determined the rate of blood lactate accumulation during progressive arm exercise. This was most evident when comparing the lactate response in the kayak paddlers (endurance trained athletes) with that of the strength trained athletes, who exhibited significantly greater upper body mass. Similarly, lower blood lactate levels were reported by kayak paddlers than their sedentary subject counterparts during comparable absolute or relative work loads. It was suggested by Pendergast et al., 1979 that low values recorded when measuring the lactate of kayakers was a result of a relatively high early lactate threshold and thus a decreased release of lactate (relative to sedentary subjects) from trained muscle performing the submaximal work. As the work intensity increased, the kayakers were more able to perform aerobically for a longer period of time, therefore delaying the onset of blood lactate accumulation.

When five moderately active subjects underwent a month of kayak training, Ridge et al., 1976 reported a significant reduction (up to a 6%) in the level of blood lactate accumulated at the same relative work loads on the kayak ergometer compared to the pre-test. Furthermore, training has also shown to decrease the arterial lactate for a given exercise load in the study by Klassen et al., 1970. The data presented by Klassen et al., 1970 reflects a training-induced decrease in overall release of lactate from tissues to blood as well as an increase in clearance of lactate from plasma during exercise. The muscular lactate clearance reflects a reduction in glycogenolysis and enhanced muscle enzyme capacity (Bergman et al., 1999; Stallknecht et al., 1998). An enhanced lactate transport capacity in the muscle has also been suggested to contribute to the high clearance following training (Brooks et al., 1999).