Eur J Appl Physiol (1981) 47: 385-391 European Journal of Applied Physiology and Occupational Physiology Springer-Verlag 1981 V0 2 Peak During Free Swimming Using the Backward Extrapolation of the 0 2 Recovery Curve Richard R. Montpetit, Luc A. Léger, Jean-Marc Lavoie, and Georges Cazorla Département d'éducation physique, Université de Montréal, CEPSUM, 2100, boul. Edouard-Montpetit, Montréal, P.Q., Canada, H3C 3J7 Institut National du Sport et de!'education Physique, Paris, France Summary. The purpose of this study was to eval.~.rnte the accuracy and feasibility of estimating oxygen consumption (V0 2 ) dur~ng maximal swimming by using the backward extrapolation (BE) of the V0 2 recovery curve to time zero. Two series of experiments were conducted. In. the first, the validity of the BE method was ascertained by comparing the V0 2 peak values obtained during free swimming (Douglas bag technique) with those estimated by the BE method during recovery after the same tests. These results were also compared with V0 2 peak measured during uphill treatmill running. V0 2 peak measured during maximal free swimming and estimated by the BE method during the recovery period of that test, were not significantly different. No significant difference was found between V0 2 peak uphill treadmill running and free swimming. In the second series of experiments, V0 2 peak was measured in each of 28 swimmers during uphill running ( conventional method) and during unimpeded swimming using the BE method. The mean value was significantly higher during swimming (X= 3.37) than during running (X= 3.23). Seventy-five percent of the subjects had higher V0 2 peak during swimming. The swimming speed during maximal swimming effort was 10% higher when the BE method was used then when the same subjects were tested by the conventional Douglas bag technique. The BE method is a reliable and valid way to measure V0 2 peak in maximal swimming and allows the swimmer to use bis specifically trained musculature fully. Key words: Swimming - V0 2 peak - Backward extrapolation - Method - T esting specificity
386 R. R. Montpetit et al. Introduction lt is well known that the maximal V0 2 of swimmers is habitually higher during treadmill running than during free swimming (Magel and Faulkner 1967; Holmer 1974). Although the lack of specificity of the free swimming test has never been questioned, our experience with the testing of international caliber swimmers has led us to doubt somewhat the task validity of the free swimming test, using the conventional mouthpiece and valve. equipment along with the tubing and supporting paraphernalia, to measure V0 2 peak. In the conventional test the mechanics of swimming is influenced, since the head of the subject remains straight, the breathing pattern is altered and more importantly the overall technique is modified i.e., reduced rotation, no flip turn. Judging from the comments of the swimmers, it is felt that they cannot fully exploit their swimming capacity and technique. The results of previous studies (Magel et al. 1975; McArdle et al. 1978; Stn>mme et al. 1977) show that the maximum V0 2 of which a subject is capable is highly specific to the musculature employed during maximal exercise and that it is specific to the training mode. This implies that a test, involving large muscle mass and optional use of the specifically trained muscles should elicit the highest value for V0 2 (referred to hereafter as V0 2 peak), when comparing the results obtained by different test procedures (Stn;)mme et al. 1977). The present study was designed to elucidate this problem. Using a novel technique to measure V0 2 in maximal effort first used by di Prampero et al. (1976) and shown to be valid and reproduci~le in a recent study from our laboratory (Léger et al. 1980), we measured V0 2 peak during maximal free crawl swimming using either conventional tech~iques or the backward extrapolation method (BE), and compared it to V0 2 peak ij?. running. To ascertain the validity of the BE technique during swimming, V0 2 peak was measured with both the conventional method and the backward extrapolation technique during the same test. From a practical point of view, it was felt important to assess the usefulness of the BE method for swimming. To date, no V0 2 peak value has ever been collected in field conditions in which the swimmer is allowed to perform without being hindered by the equipment and/or testing conditions. Methods Two series of experiments were conducted. In series A comparisons were made between the V0 2 peak measured during treadmill running, free swimming using the Douglas bag method and free swimming using the backward extrapolation method immediately following test 2. In series B, V0 2 peak directly measured during treadmill running was compared to V0 2 peak measured after a swimming test using the BE method. Series A. Eleven swimmers participated in these experiments (Table 1). All subjects were trained competitive swimmers and had performed the test procedure at least once before the experiments reported here. Only one maximal test was performed per day. All the tests were completed within a 2-week oeriod. The order of testin11: was randomized.
V0 2 Peak During Free Swimming 387 Table 1. Characteristics of subjects Experiment Age Height Weight V elocity maxa (year) (cm) (kg) (m. çl) Series A 15.6 173.3 65.3 1.29 n = 11 (13-19) (161-182) (47.2-74.9) (1.15-1.39) Series B 14.6 165.8 57.1 1.39 n = 28 (12-18) (152-182) (41-74.9) (1.23-1.49) Values are means with range given in parentheses a Velocity at which V0 2 peak free swimming was reached Maximum running was performed at variable speeds and grades in a multistage continuous test. Initially, the treadmill was set at 4.8 km h- 1, 10% grade, and the subject ran for 3 min. Following a 5-min rest, the subject then ran at 9.7 km h- 1, 0% grade, for 5 min. Thereafter the speed was set at 11. 3 km h- 1 for the duration of the test but the slope of the treadmill bed was raised every 3 min by 2.. 5% increments until exhaustion of the subject. Oxygen uptake determinations were made with the standard Douglas bag open circuit technique. AH swimming experiments were performed in a 50-meter pool in which the water temperature was 26-27 C. The maximal test consisted of swimming at progressively increasing velocities (increments of 0.2 m s- 1 every 4 min) until the subject was unable to maintain the pace for at least two lengths of the pool. To assure a constant velocity for each level, pylons were placed at every 10 m at the bottom of the pool. The subject's passing over each pylon had to coïncide with an underwater audio signal. The starting speed was 0.8 m s- 1 The middle 40 m were timed to the nearest 0.1 s during the last level to ascertain precisely the velocity at the time of exhaustion. Duplicate collections were made during the last 2 min of each level. Expiratory volumes were collected in meteorological balloons placed in a free-wheeling cart that followed the swimmer along the side of the pool. The tubing was reduced to a minimum (1.6 m) and extended from the breathing apparatus to a wooden pole that projected 90 from the cart and directly over the swimmer. In addition to the collections during the test, after the cessation of the exercise four consecutive collections were made to permit the calculation of V0 2 using the BE method. The collection period was carefully timed from the beginning of expiration and the end of expiration some 20 s later (range 18.9-21.3). The BE method has been described elsewhere (Léger et al. 1980) and has been shown to be valid for the measurement of V0 2 peak. The same equipment and breathing valve (Daniels) were used for all tests in both series. Volumes of expired gas were measured in a calibrated Tissot spirometer. Gas samples were drawn from the bags and analyzed for 0 2 and C0 2 (Beckman E-2, and LB-1). Both analysers were calibrated before and after each experiment with gases of known concentrations (micro-scholander technique). Series B. Twenty-eigh.t young competitive swimmers participated in these experiments (Table 1). They were tested for V0 2 peak both during treadmill running (same procedure as in series A) and after a maximal swimming test using the BE method only. The loads for swimming were imposed exactly the same way as in series A. Differences between means were tested for significance with a one-way analysis of variance for repeated measurements in series A and with a t-test for paired observations in series B. Pearson's correlation coefficients were computed to examine the relationship between the V0 2 peak values measured during the different experimental conditions. Results Series A. V0 2 peak, minute ventilation at V0 2 peak, respiratory exchange ratio and Pearson's correlation coefficients for the two experimental conditions are
Table 2. Comparison of V0 2 peak, VE> and R during treadmill running and swimming Series A (n = 11) Series B (n = 28)!.;.) OO OO Treadmill running (Douglas bag method) Swimming (Douglas bag method) (BE method)a V0 2 peak, 1 min- 1 3.91 ± 0.55 3.81 ± 0.59 3.79 ± 0.66 Max. pulmonary ventilation 127 ± 21 111±19 101 ± 17b l min- 1 BTPS Respiratory exchange ratio 1.08 ± 0.04 1.02 ± 0.04 i.11 ± o.o5b r 0.99 0.94 Values are means ± SD a V0 2 peak measured by the backward extrapolation method (BE) b VE and R for the first collection immediately following the end of exercise Treadmill running (Douglas bag method) 3.23 ± 0.71 106 ± 18 1.07 ± 0.05 0.95 Swimming (BE method) 3.37 ± 0.77 96 ± 27b 1.10 ± 0.14b :;o :;o a: 0 ::i... "Cl ('1) ::t... ('1)... e:..
V0 2 Peak During Free Swimming 389 5.0 y= 0.885x + o.m r = 0.95 SEE = :!: 0.197 u =~ 0 ::: l.. 0 E.~ 3.5 E E ~ 3.0 0 Cl- "' N = ::> 2.5 2.5 3.0 15 V02 peak treadmill running { l min- 1 ) Fig. 1. Comparison of V0 2 peak measured uphill treadmill running and during maximal swimming using the BE of the 0 2 recovery curve method 5 0 shown in Table 2. There was no significant difference between the V0 2 peak determined during treadmill running and the V0 2 peak measured either by the conventional (direct) method or the BE method during free maximal swimming. Three subjects had equal or higher V0 2 peak swimming than running. The correlation between the V0 2 peak values running and swimming was almost perfect (r = 0.99). V0 2 peak measured during maximal free swimming with the conventional breathing equipment and estimated by the BE method during the recovery of that test, were not significantly different (Table 2). Prediction of V0 2 peak in swimming from the BE V0 2 peak provided a good estimate of the true maximal V0 2 during swimming, as the standard error of the estimate (SEE) expressed in % of the mean was small ( ± 3. 7 % ). Series B. The V0 2 peak for maximal unimpeded swimming (BE method) was significantly higher (p < 0.01) than the V0 2 peak running (Fig. 1). Twenty-one out of twenty-eight subjects had equal or higher V0 2 peak swimming than running. The mean difference between V0 2 peak swimming and running was large (X= +5.2%, range -10 to + 14%) and was well above the expected intra-individual variation. Based on seven test-retest determinations the BE method for the evaluation of V0 2 peak after free swimming was found to be highly reliable (r = 0.92). The lack of a significant difference with a paired t-test analysis between the two tests indicated good reproducibility.
390 R. R. Montpetit et al. Discussion The major finding of this study is that V0 2 peak free swimming unimpeded is higher than V0 2 peak running (Series B). Seventy-five percent of the swimmers had higher V0 2 swimming than running. This finding concurs with ~he data of Str0mme et al. (1977) who have shown that athletes reached higher V0 2 max in their specific sport activities than in uphill treadmill running. In series A, V0 2 peak at time zero from the backward extrapolation method was not significantly different from that of the V0 2 peak directly measured during the free swimming test. This is in accord with the data already published from our laboratory (Léger et al. 1980). The standard error of the estimate was only 0.141 min- 1 or 3.7%. This is lower than the 5.1 % found in our previous study and is probably due to the fact that great care was taken not to eut respira tory cycles during the gas collections ( see methodology). It was postulated in a previous paper (Léger.et al. 1980) that such a procedure could reduce SEE. The identical results for V0 2 during maximal swimming found by direct measurement and. by the BE method, indicate that the latter technique may be used to measure V0 2 peak in swimming, as it may be in treadmill running (di Prampero et al. 1976), as long as certain test conditions are respected. These conditions are: (1) that the exercise is progressive and continuous and leads to exhaustion in more than 4-5 min; (2) that no delay exists between the end of the exercise and the beginning of the gas collections; (3) that the gas collection is started at the beginning of expiration and stopped approximately 20 s la ter at the end of expiration and ( 4) that the exercise is not of supramaximal intensity or of short duration (<5 min). As opposed to supramaximal exercise (di Prampero et al. 1973; Roberts and Morton 1978), there is no delay to the onset of 0 2 recovery after a maximal multistage continuous test.. For all swimmers, the mean speed at which V0 2 peak was reached was higher swimming without equipment than during the conventional test involving mouthpiece and tubing (Table 1). For the nine swimmers who participated in both series, the speed corresponding to V0 2 peak was 1.25 m s- 1 and 1.39 m ç 1 ( + 10%) for the conventional and the BE method respectively. The mean speed for a 400 m race in these subjects (1.46 m ç 1 ) was just slightly higher than that attained during the progressive test without equipment. The BE method presents many advantages over the conventional method in that no respira tory equipment is needed during the test. In addition, it allows the athlete to perform without any modification of his technique and permits him to push himself to maximal exertion without any restriction whatsoever. Thus, the work task imposed on the swimmers in these experiments gave full credit to the specifically trained muscle mass of these individuals and allowed them to perform in a technically correct manner. In summary, the results of this study indicate that V0 2 peak determinations during swimming can be made in ideal field conditions with good accuracy and without equipment. This allows sport scientists and coaches to test swimmers during conditions that reproduce very closely the competitive situation (in addition to using their preferred stroke) without altering in any way the mechanics of swimming.
V0 2 Peak During Free Swimming 391 References Astrand PO, Eriksson B, Nylander I, Engstrom P, Karlsberg B, Saltin B, Thoren C (1963) Girl swimmers. Acta Paediatr [Suppl] 147: 5-75 di Prampero PE, Peeters L, Margaria R (1973) Alactic 0 2 debt and lactic acid production after exhausting exercise in man. J Appl Physiol 34: 628-632 di Prampero PE;Cortili G, Magnani P, Saibene F (1976) Energy cost of speed skating and efficiency of work against air resistance. J Appl Physiol 40: 584-491 Holmer I, Lundin A, Eriksson B (1974) Maximum oxygen uptake during swimming and running by elite swimmers. J Appl Physiol 36:711-714 Holmer I (1974) Physiology of swimming in man. Acta Physiol Scand [Suppl] 407 Léger LA, Seliger V, Brassard L (1980) Backward extrapolation of V0 2 max from the 0 2 recovery curve. Med Sei Sports Exercise 12: 24-27 Magel JR (1966) Maximum oxygen uptakes of college swimmers. Ph.D. dissertation, The University of Michigan, Ann Arbor, Michigan Magel JR, Faulkner JA (1967) Maximum oxygen uptakes of college swimmers. J Appl Physiol 22: 929 Magel JR, Foglia GF, McArdle WD, Gutin B, Pechar GS, Katch FI (1975) Specificity of swim training on maximum oxygen uptake. J Appl Physiol 38: 151-155 McArdle WD, Glaser RM, Magel JR (1971) Metabolic and cardiorespiratory response during free swimming and treadmill walking. J Appl Physiol 30: 733-738 McArdle WD, Magel JR, Delio DJ, Toner M, Chase JM (1978) Specificity of run training on V0 2 max and heart rate changes during running and swimming. Med Sei Sports 10: 16-20 Roberts AD, Morton AR (1978) Total and alactic 0 2 debts after supramaximal work. Eur J Appl Physiol 38: 281-289 Str0mme SB, Ingjer F, Meen HD (1977) Assessment of maximal aerobic power in specifically trained athletes. J Appl Physiol 42: 833-837 Accepted May 18, 1981