Deep water running has recently become a popular mode of aerobic training. When performing water running, the athlete “runs” in the deep end of a swimming pool with the aid of a commercial flotation belt that helps support the head above water. Anecdotal accounts have described the benefits of water running for injured athletes attempting to maintain cardiovascular conditioning during rehabilitation(28,29). In addition, noninjured athletes have found water running to be an attractive training alternative since it eliminates the orthopedic stress associated with land running(28,29).
Physiological responses to an acute submaximal or maximal bout of water running have been reported by several authors(2,23,25). However, the effect of chronic water run training on the maintenance of cardiorespiratory conditioning has not been investigated extensively, particularly among trained individuals(8). It has been reported that well-trained aerobic athletes experience a significant decline in cardiovascular fitness after approximately 2-3 wk of inactivity (6,27), a dilemma that many athletes face as a result of injury. In light of this, it seems important to evaluate alternate modes of exercise that will allow athletes to maintain aerobic conditioning during rehabilitation. Thus, the purpose of the present investigation was to determine whether water running could serve as an effective training alternative to land-based running for the purpose of maintaining aerobic performance in trained endurance athletes.
The subjects for this study were 16 aerobically trained male distance runners who ranged in age from 20 to 40 yr. Participation in this investigation was based on a) prior experience in running for durations of 60 min or longer, b) regular aerobic training in the range of 5-7 d·wk-1, and c) a minimum ˙VO2max of 55 ml·kg-1·min-1. Subjects were medically cleared and fully informed of the risks and stresses associated with the project before giving their written consent. In addition, subjects received a monetary stipend for their participation. Subjects were 28 ± 6 yr of age, weighed 69.8 ± 7.0 kg, were 176 ± 8 cm tall, and had a˙VO2max of 58.6 ± 3.4 ml·kg-1·min-1.
Following preliminary screening, subjects were assigned to one of two training groups matched by ˙VO2max, either treadmill run (R) or water run (WR). Both groups were required to follow the same workout schedule, which consisted of alternate high-intensity interval sessions (30 min at 90-100% ˙VO2max) and moderate-intensity recovery sessions (60 min at 70-75% ˙VO2max) performed 5 d·wk-1 for 6 wk. These workouts were consistent with average workouts done by the subjects prior to participation in this study. Subjects were evaluated at the beginning (day 0), halfway (day 21), and at the end (day 42) of the 6-wk experimental period.
Individual workouts for the 6-wk exercise period were based on preexperimental graded exercise tests (GXT). Due to the fact that heart rate and maximal oxygen uptake may be depressed by 10-15% during water run exercise(2,23), water and treadmill GXT were used for the WR and R groups, respectively, to equilibrate workout intensity.
Following a 2-wk orientation period in which instruction was given on proper water running technique, WR subjects performed a GXT in a swimming pool(≈27°C) while wearing an Aqua Jogger® flotation belt (Excel Sports Science, Eugene, OR) secured by a tether to the side of the pool. The water GXT was conducted using a modified Wilder protocol(28,29); exhaustion was defined as either volitional exhaustion or a drop in stride frequency greater than 5%. An acceptable stride was one in which a) the lead leg maintained at least 90° of knee flexion through the swing phase with the succeeding footplant extending a minimum of 0.5 foot anterior to the coronal plane and b) the trail leg extended at least 0.5 foot posterior to the coronal plane.
An open-circuit spirometry system was used to obtain expired respiratory gas during the water GXT. Subjects breathed through a Hans Rudolph two-way breathing valve secured to the head by a lightweight plastic head frame. One-minute samples of expired gas were collected into Douglas bags during the final 7 min of the water GXT. Gas samples were subsequently measured for total volume (STPD) using a Parkinson-Cowen dry gas flow meter and analyzed for O2 and CO2 concentrations using electronic gas analyzers(Applied Electrochemistry). Absolute ˙VO2 was calculated using the Haldane transformation. Heart rate was monitored continuously using a Polar Favor® heart rate monitor (Polar CIC, Port Washington, NY). Assessment of a valid water ˙VO2max was based on the plateau criterion, i.e., oxygen consumption rises by less than 2.0 ml·kg-1·min-1 or 150 ml·min-1 over two consecutive workloads (24).
Subjects in the R group performed a treadmill GXT, which was conducted using a modified Costill and Fox (4) protocol in a laboratory maintained at 20-22°C. Subjects began the treadmill test with a warmup consisting of a 3-min walk at 107 m·min-1 (4.0 mph) and 0% grade followed by a 3-min jog at 161 m·min-1 (6.0 mph) and 0% grade. Subjects then proceeded directly into a series of 3-min workloads. The initial workload was run at 241 m·min-1 (9.0 mph) and 0% grade followed by a workload of 255 m·min-1 (9.5 mph) and 0% grade. The speed of subsequent workloads was maintained at 255 m·min-1 while the treadmill was elevated by 1% with each successive workload until the subject reached exhaustion. Consistent with the water GXT, open-circuit spirometry was used to obtain expired respiratory gas during the treadmill GXT. All other procedures used during the preexperimental treadmill GXT were identical to those utilized during the preexperimental water GXT and have been previously outlined.
Results of the preexperimental water and treadmill GXT were used solely for the purpose of establishing individual workouts for the WR and R groups, respectively. This was done by identifying each subject's heart rate at 70, 75, and 90% of water (49.4 ± 7.3 ml·kg-1·min-1) or treadmill (59.5 ± 4.9 ml·kg-1·min-1) ˙VO2max. Subjects subsequently used heart rate to ensure that workouts during the 6-wk experimental period were performed at the prescribed exercise intensities of 70-75% and 90-100% ˙VO2max. Subjects in both groups were required to complete a minimum of two follow-up workouts for the purpose of verifying that the prescribed workout heart rates accurately matched the prescribed workout intensities.
Following preexperimental testing, both groups began a 6-wk workout period. Subjects in the WR group trained in a swimming pool maintained at≈27°C. In an effort to control for the effect of environmental heat and humidity on cardiovascular response, subjects in the R group exercised on motorized treadmills located in an air-conditioned environment(20-22°C).
During the 6-wk exercise period, both groups followed a prescribed workout schedule consisting of sessions of a) 30 min at 90-100% ˙VO2max and b) 60 min at 70-75% ˙VO2max alternated daily 5 d·wk-1. Workout intensity was regulated by heart rate using a Polar Favor® heart rate monitor. Average exercise heart rate was recorded for each workout.
On days 0, 21, and 42, subjects reported to the exercise physiology lab to perform a maximal treadmill test for the purpose of comparing preliminary and outcome data. Subjects fasted for 12 h and refrained from exercise for 48 h prior to this session. Subjects performed the treadmill test in an air-conditioned environment (20-22°C) at approximately the same time of day on each of the 3 testing days.
The maximal treadmill test was conducted using a modified Costill and Fox(4) protocol previously described. A computer-driven open-circuit spirometry system was used to analyze respiratory gas samples. Hard copy of ˙VE STPD, ˙VCO2, and ˙VO2 was provided at 30-s intervals and was used to calculate ˙VO2max, ventilatory threshold, and running economy. Assessment of a valid˙VO2max was based on criteria previously outlined(24). Ventilatory threshold was determined by plotting the ventilatory equivalents (˙VE·˙VO2-1,˙VE·˙VCO2-1) using a computer software package and was defined as an exponential increase in˙VE·˙VO2-1 without a similar increase in˙VE·˙VCO2-1(21). Ventilatory threshold data were analyzed by two independent evaluators who were provided the data double blind. Running economy was determined from the relative oxygen cost (ml·kg-1·min-1) of running at five submaximal workloads (3-min workloads at 161 m·min-1, 0% grade; 241, 0; 255, 0; 255, 1; and 255, 2%). Heart rate was monitored continuously via heart rate telemetry. Upon exhaustion, total run time to fatigue was recorded to the nearest second.
Additionally, on days 0 and 42 subjects were measured for body composition, which was determined from eight to 12 hydrostatic weighing trials with the subject in the prone position. The average of the final three underwater trials was used to represent the true underwater weight. Vital capacity was measured using a Pneumoscan KTC® spirometer. Residual volume was estimated using the conversion factor 0.24 (30). Body density was calculated according to the equation of Goldman and Buskirk(12) and percent body fat was calculated using the Siri equation (20). All pre- and postexperimental measurements of body composition were made by the same investigator.
Blood Sampling and Analysis
Blood samples were taken from an antecubital vein 2-4 min postexercise with the subject seated upright. Glucose and lactate were analyzed using a YSI 2300 Stat Plus Glucose and L-Lactate Analyzer that had previously been calibrated using standards of known concentrations. The intraassay variability of replicate glucose and lactate samples was 2%.
For the determination of norepinephrine, whole blood was collected into a 7.0-ml collection tube containing 7.0 mg reduced glutathione and 7.0 mg ethylene glycol tetraacetic acid 20 (EGTA) and immediately centrifuged at 3000 rpm for 25 min. Serum was harvested and frozen at -70°C for subsequent analysis using a high-performance liquid chromatography (HPLC) procedure based on the method of Moerman and De Schaepdryver (18). The intraassay variability for replicate samples of plasma norepinephrine was 2%.
Data were analyzed using a two-factor (group × time) analysis of variance (ANOVA) with repeated measures. Significant differences revealed by the ANOVA were identified using Tukey (honest significant difference [HSD]) post hoc analysis. Additionally, running economy data were analyzed using linear regression analysis. For all statistical measures, significance was established at the 0.05 level of probability.
Of the 18 subjects who began the study, two were forced to withdraw approximately 1 wk into the 6-wk training period. One subject dropped out of the R group due to personal reasons, while another withdrew from the WR group due to major surgery. On average, subjects trained 5.3 d·wk-1. Adherence among the R group was 98%, with injury being the only reason for missed workouts. Adherence among the WR group was 96%, with adverse weather and medical attention being the main reasons for missed workouts. Thirteen of 16 subjects a) completed every workout or b) missed no more than one workout over the duration of the study. On average, daily workout heart rate was 14% lower in WR versus R.
Preexperimental treadmill ˙VO2max(Fig. 1) was 58.4 ± 2.3 and 58.7 ± 4.7 ml·kg-1·min-1 for R and WR, respectively. Following 6 wk of workouts, no significant intra- or intergroup differences were observed for R (pre = 58.4 ± 2.3, post = 60.1 ± 3.6 ml·kg-1·min-1) and WR (pre = 58.7 ± 4.7, post = 59.6 ± 5.4 ml·kg-1·min-1). Similarly, ventilatory threshold (Fig. 2) was unaltered in R (pre = 47.5 ± 1.8, post = 48.2 ± 3.3 ml·kg-1·min-1) and WR (pre = 46.5 ± 6.4, post = 47.4 ± 6.7 ml·kg-1·min-1), nor were there any changes in running economy (Fig. 3) in R (pre = 48.4 ± 2.3, post = 48.9 ± 2.0 ml·kg-1·min-1 at 255 m·min-1) and WR(pre = 51.8 ± 2.0, post = 48.9 ± 2.0 ml·kg-1·min-1 at 255 m·min-1).
No significant changes were observed within or between groups following 6 wk of workouts for maximum heart rate (HRmax), ˙VEmax, or treadmill run time to exhaustion (Table 1).
Metabolic responses to maximal treadmill exercise are provided inTable 1. Glucose concentration following maximal treadmill exercise on day 0 was similar for R (7.9 ± 1.1 mM) and WR(7.4 ± 0.8 mM) and was essentially unchanged (P > 0.05) following the 6 wk of workouts. Likewise, maximal lactate levels were similar among R (7.9 ± 1.8 mM) and WR (7.8 ± 2.2 mM) at day 0. Over the 6-wk training period, lactate levels moved slightly higher for both groups, but these differences were not significant.
Plasma norepinephrine levels were similar among R (7.3 ± 0.5 ng·ml-1) and WR (7.2 ± 0.5 ng·ml-1) on day 0. These values were essentially unchanged (P > 0.05) after 3 and 6 weeks of workouts (Table 1).
Body composition data are summarized in Table 2. On day 0, treadmill runners and water runners were similar in total body weight(R = 69.9 ± 7.6, WR = 69.7 ± 6.8 kg), lean body mass (R = 60.2± 3.2, WR = 60.1 ± 5.9 kg), fat weight (R = 9.7 ± 5.3, WR= 9.6 ± 5.6 kg), and body fat percent (R = 13.9 ± 6.1, WR = 13.8± 4.5%). No changes were observed within or between groups after 6 wk of workouts.
The purpose of the present study was to determine whether deep water running could serve as an effective training alternative to land-based running for the purpose of maintaining physiological determinants of aerobic performance for trained endurance athletes.
The main finding of the present investigation was that ˙VO2max was not significantly altered within or between groups following 6 wk of workouts, which suggests that water running was an effective training alternative to land-based running for the maintenance of maximal aerobic capacity among trained aerobic athletes. This finding is in agreement with Eyestone et al. (8), who investigated trained collegiate males (˙VO2max = 57.4 ± 1.7 ml·kg-1·min-1) and reported that treadmill˙VO2max for water runners was not different compared to land-based runners after 6 wk of workouts, although both groups produced a decrease in˙VO2max of approximately 4% compared to day 0. The data of the present investigation showed an approximate 2% decline in ˙VO2max by day 21 followed by a 3% increase at day 42; however, these nonsignificant changes probably reflect normal day-to-day variation in maximal aerobic capacity (13).
Additional studies have examined the effect of chronic water run training among untrained individuals. Quinn et al. (19) reported that 4 wk (4 d·wk-1 for 30 min·d-1) of deep water running was ineffective in maintaining treadmill ˙VO2max in untrained collegiate females (˙VO2max = 39.9 ± 3.6 ml·kg-1·min-1). Conversely, Michaud et al.(17) demonstrated that an 8-wk progressive aerobic-interval deep water running program produced a significant 11% improvement in treadmill ˙VO2max among untrained subjects(˙VO2max = 2.25 ± 0.57 l·min-1).
For cross training to be effective in maintaining ˙VO2max, it must employ a training pattern that is equivalent in intensity and duration to the original exercise mode (11,14). In addition, the cross-training exercise modality must replicate the range of motion and skeletal muscle recruitment pattern of the original exercise mode(11,14). In the present study, water run workouts were equivalent in intensity and duration to treadmill workouts. Although not supported by electromyogram-based data, deep water running appears to recruit less lower extremity musculature compared to land-based running due to the fact that water running does not involve plantarflexion from a solid surface. Conversely, water running may recruit more upper-body musculature than land-based running due to hydrostatic resistance. Nevertheless, the motion of water running closely replicates land-based running and, in general, involves similar skeletal musculature and range of motion.
As with ˙VO2max, ventilatory threshold is considered a valid predictor of aerobic performance (9), particularly among athletic groups that are homogeneous for ˙VO2max. In the present study, both groups exhibited relatively high ventilatory threshold values throughout the 6-wk experimental period. Expressed as a percentage of˙VO2max, ventilatory threshold for the treadmill runners was 81.4, 81.7, and 80.1% on days 0, 21, and 42, respectively, while ventilatory threshold values for the water runners were 78.6, 80.0, and 79.6%˙VO2max. These values are representative of well-trained aerobic athletes, who typically produce ventilatory thresholds in the range of 75-85%˙VO2max(9). A relatively high ventilatory threshold is reflective of the effects of training on lactate production and clearance. During incremental treadmill exercise, blood lactate increases curvilinearly and is dependent on the number of skeletal muscle fibers recruited as well as the intensity of exercise (22). Well-trained individuals are able to produce and tolerate relatively high levels of blood lactate during intense exercise. In turn, this capacity to tolerate high lactic acidosis allows the athlete to exercise longer and at a greater intensity before exhausting. In the present study, relatively high blood lactate levels were evident for both groups on day 0, and no significant change in lactate response was observed within or between groups following 6 wk of workouts. Essentially, these data suggest that lactate response to maximal treadmill exercise was not affected by mode of exercise. The fact that lactate production and treadmill run time to exhaustion were not different between the treadmill and water runners indicated that lactate tolerance was similar between the two groups.
It has been demonstrated that in addition to ˙VO2max and ventilatory threshold, physiological economy is an important predictor of aerobic performance (3,9), especially among aerobic athletes who are similar in ˙VO2max. The present study demonstrated that running economy was similar within and between the two experimental groups when evaluated after 6 wk. This supports the contention that water running provided a sufficient exercise stress to maintain running economy at a well-trained level equivalent to that of the land-based runners. Several factors have been shown to influence running economy, including stride rate and frequency (15), body weight(7), and training (3). It is possible that hydrostatic resistance encountered during water run exercise favorably modified the water runners' stride mechanics, resulting in a more efficient stride (i.e., reduced overstriding). In turn, improvements in stride biomechanics may have contributed to the maintenance of running economy among the water runners.
In addition to physiological predictors of aerobic performance, run time to exhaustion was considered as a performance parameter in the present investigation. As with other experimental variables, treadmill run time to exhaustion was not significantly different within or between groups following 6 wk of workouts. Throughout the experimental period, the treadmill runners ran approximately 3 min longer during the maximal treadmill tests compared to the water runners despite being similar in ˙VO2max, ventilatory threshold, and running economy. This discrepancy may be explained in part by differences in anaerobic capacity and by psychological factors. (The fact that this difference was not statistically significant is attributed to the effect of relatively wide intragroup variations on the statistical analysis). The findings of the present investigation are in agreement with Eyestone et al.(8), who reported no significant decrement in 2-mile run performance after 6 wk of water run workouts.
Whole-blood glucose response to maximal treadmill exercise was similar for the treadmill and water runners throughout the experimental period. Blood glucose production during short, intense exercise is mediated primarily by hepatic glycogenolysis (26); hepatic glucose production increases in parallel with glucose utilization as exercise intensity increases(26). Hepatic gluconeogenesis, on the other hand, appears to contribute only about 10% to total glucose production during maximal exercise (1), despite increased levels of arterial lactate. In the present investigation, blood glucose concentration during maximal treadmill exercise was similar within and between the two groups throughout the experimental period and provides indirect evidence that hepatic glycogenolysis was not altered after 6 wk of water run training.
Norepinephrine levels at ˙VO2max were similar within and between the treadmill and water run groups throughout the experimental period and were comparable to values reported for well-trained runners and cyclists during incremental treadmill exercise (16). At present, there are no reported data regarding catecholamine activity as it relates to water running. Previous studies (11,14) investigating the effects of swim training on treadmill ˙VO2max have suggested that central training adaptations can be maintained by alternate nonspecific forms of aerobic training. Thus, these studies(11,14) provided indirect evidence that norepinephrine's influence on cardiac acceleration and contractility can be maintained through alternate forms of exercise. It seems logical that water running, which appears to be more transferable to treadmill running than swimming, would provide a sufficient stimulus to maintain catecholamine response during maximal treadmill exercise. This appears to have been true in the present investigation as plasma norepinephrine and heart rate were similar within and between the two groups during maximal treadmill exercise following 6 wk of workouts. If loss of conditioning had occurred among the water runners, a decrease in plasma norepinephrine concentration would have been expected (5).
In light of the important effect of body weight on the energy cost of running, changes in body composition were evaluated in the present study. As with several experimental variables, there were no intra- or intergroup differences in body composition following 6 wk of workouts. In general, percent body fat for both groups throughout the study was greater than values reported for elite long-distance runners (10).
Upon completion of the 6-wk workout period, subjects in the water run group were asked to respond to a written questionnaire. The purpose of the questionnaire was to subjectively evaluate physical and psychological effects experienced by the water run subjects a) during the treadmill˙VO2max tests on days 21 and 42 and b) upon resumption of their regular land-based training following completion of the study. In general, the water run subjects indicated that they experienced few adverse physical or psychological effects during the maximal treadmill runs with the exception of some expected delayed-onset muscle soreness. Similarly, the water runners responded that they were able to comfortably readapt to land-based running provided they resumed their preexperimental training regimen in a controlled manner.
In conclusion, the data of the present investigation demonstrated that water run exercise served as an effective training alternative to land-based running for the maintenance of physiological determinants of aerobic performance over a period of 6 wk among trained aerobic athletes. These results suggest that an athlete can perform effectively in land-based running, as determined by treadmill ˙VO2max, following 6 wk of water run workouts. As such, these findings have important implications for injured athletes attempting to maintain high levels of aerobic conditioning during rehabilitation and for noninjured athletes attempting to recover from musculoskeletal fatigue brought on by land running.
1. Ahlborg, G., P. Felig, L. Hagenfeldt, R. Hendler, and J. Wahren. Substrate turnover during prolonged exercise in man. J. Clin. Invest.
2. Butts, N. K., M. Tucker, and M. Greening. Physiologic responses to maximal treadmill and deep water running in men and women.Am. J. Sports Med.
3. Conley, D. L. and G. S. Krahenbuhl. Running economy
and distance running performance of highly trained athletes. Med. Sci. Sports Exerc.
4. Costill, D. L. and E. L. Fox. Energetics of marathon running. Med. Sci. Sports
5. Coyle, E. F., W. H. Martin, S. A. Bloomfield, O. H. Lowry, and J. O. Holloszy. Effects of detraining on responses to submaximal exercise. J. Appl. Physiol.
6. Coyle, E. F., W. H. Martin, D. R. Sinacore, M. J. Joyner, J. M. Hagberg, and J. O. Holloszy. Time course of loss of adaptations after stopping prolonged intense endurance training. J. Appl. Physiol. Respirat. Environ. Exerc. Physiol.
7. Cureton, K. and P. Sparling. Distance running performance and metabolic responses to running in men and women with excess weight experimentally equated. Med. Sci. Sports
8. Eyestone, E. D., G. Fellingham, J. George, and A. G. Fisher. Effect of water running and cycling on maximum oxygen consumption and 2-mile run performance. Am. J. Sports Med.
9. Farrell, P. A., J. A. Wilmore, E. F. Coyle, J. E. Billing, and D. L. Costill. Plasma lactate accumulation and distance running performance. Med. Sci. Sports Exerc.
10. Fleck, S. J. Body composition of elite American athletes. Am. J. Sports Med.
11. Gergley, T. J., W. D. McArdle, P. DeJesus, M. M. Toner, S. Jacobowitz, and R. J. Spina. Specificity of arm training on aerobic power during swimming and running. Med. Sci. Sports Exerc.
12. Goldman, R. F. and E. R. Buskirk. Body volume measurement by underwater weighing: description of a method. In:Techniques for Measuring Body Composition
, J. Brozek and A. Henschel(Eds.). Washington, DC: National Academy of Science, 1961, pp. 78-89.
13. Katch, V. L., S. S. Sady, and P. Freedson. Biological variability in maximum aerobic power. Med. Sci. Sports Exerc.
14. Magel, J. R., G. F. Foglia, W. D. McArdle, B. Gutin, and G. S. Pechar. Specificity of swim training on maximum oxygen uptake.J. Appl. Physiol
. 38:151-155, 1975.
15. Martin, P. E. and D. W. Morgan. Biomechanical considerations for economical walking and running. Med. Sci. Sports Exerc.
16. Mazzeo, R. S. and P. Marshall. Influence of plasma catecholamines on lactate threshold during graded exercise. J. Appl. Physiol.
17. Michaud, T. J., D. K. Brennan, R. P. Wilder, and N. W. Sherman. Aquarunning and gains in cardiorespiratory fitness. J. Strength Conditioning Res.
18. Moerman, E. J. and A. F. De Schaepdryver. Quantitation of catecholamines in urine and in plasma. Clin. Chim. Acta
19. Quinn, T. J., D. R. Sedory, and B. S. Fisher. Physiological effects of deep water running following a land-based training program. Res. Q. Exerc. Sport
20. Siri, W. E. The gross composition of the body. In:Advances in Biological and Medical Physics
, Vol. 49, J. H. Lawrence and C. A. Tobias (Eds.). City: Publisher, 1956, pp. 239-280.
21. Skinner, J. S. and T. H. McLellan. The transition from aerobic to anaerobic metabolism. Res. Q. Exerc. Sport
22. Stainsby, W. N. and G. A. Brooks. Control of lactic acid metabolism in contracting muscles and during exercise. Exerc. Sport Sci. Rev.
23. Svedenhag, J. and J. Seger. Running on land and in water: comparative exercise physiology. Med. Sci. Sports Exerc.
24. Taylor, H., E. Buskirk, and A. Henschel. Maximal oxygen uptake as an objective measure of cardiorespiratory performance. J. Appl. Physiol.
25. Town, G. P. and S. S. Bradley. Maximal metabolic responses of deep and shallow water running in trained runners. Med. Sci. Sports Exerc.
26. Wahren, J., P. Fehlig, G. Ahlborg, and L. Jorfeldt. Glucose metabolism during leg exercise in man. J. Clin. Invest.
27. Wilber, R. L. and R. J. Moffatt. Physiological and biochemical consequences of detraining in aerobically trained individuals.J. Strength Conditioning Res.
28. Wilder, R. P. and D. K. Brennan. Physiological responses to deep water running in athletes. Sports Med.
29. Wilder, R. P., D. K. Brennan, and D. E. Schotte. A standard measure for exercise prescription for aqua running. Am. J. Sports Med.
30. Wilmore, J. H. The use of actual predicted and constant residual volumes in the assessment of body composition by underwater weighing.Med. Sci. Sports Exerc.