International Journal of Sports Physiology and Performance, 2014, 9, 959-965 http://dx.doi.Org/10.1123/ijspp.2013-0419 ©2014 Human Kinetics, Ino. Assessment of Fatigue Thresholds in 50-m All-Out Swimming Susana M. Soares, Ricardo J. Fernandes, J. Leandro Machado, Jose A. Maia, Daniel J. Daly, and Joao P. Vilas-Boas Context: It is essential to determine swimmers’ anaerobic potential and better plan training, understanding physiological effects of the fatigue. Purpose: To study changes in the characteristics of the intracyclic velocity variation during an all-out 50-m swim and to observe differences in speed and stroking parameters between these changes. Methods: 28 competitive swimmers per formed a 50-m front-crawl all-out test while attached to a speedometer. The velocity-time (v[fj) curve off all stroke cycles was analyzed per individual using a routine that included a wavelet procedure, allowing the determination of the fatigue thresholds that divide effort in time intervals. Results: One or 2 fatigue thresholds were observed at individual level on the v(t) curve. In males, when 1 fatigue threshold was identified, the mean velocity and the stroke index dropped (P < .05) in the second time interval (1.7 ± 0.0 vs 1.6 ± 0.0 m/s and 3.0 ± 0.2 vs 2.8 ± 0.3 m/s, respectively). When 2 fatigue thresholds were identified, the mean velocity of the first time interval was higher than that of the third time interval (P < .05), for both male (1.7 ± 0.0 vs 1.6 ± 0.1 m/s) and female (1.5 ± 0.1 vs 1.3 ± 0.1 m/s) swimmers. Conclusion: One or 2 fatigue thresholds were found in the intracyclic velocity-variation patterns. Concurrently, changes in velocity and stroke parameters were also observed between time intervals. This information could allow coaches to obtain new insights into delaying the degenerative effects of fatigue and maintain stable stroke-cycle characteristics over a 50-m event. Keywords: aquatic exercise, metabolism, speed decline, anaerobic transition zone, stroke parameters The importance of the ATP-PC and glycolytic energy pathways during competitive swimming is generally accepted. In fact, as the majority of swimming events last less than 2 minutes, success largely depends on anaerobic energy production.1 With this in mind, a number of tests have been developed to evaluate swimmers’ anaerobic potential, but none have fully satisfied either researchers or coaches, as they were mainly indirect. The most direct measure ments—muscle biopsy2 and magnetic nuclear resonance3—are expensive, invasive, and/or limited in providing information on total anaerobic energy production in exercises that involve more than a single muscle or muscle group.2 The results obtained with indirect tests commonly represent a mechanical expression of jumping,4 run ning,5 or cycling effort,6 but information on physiological changes, fatigue, and changes in stroke pattern is not provided. Coaches still cannot obtain specific information to help determine training and competitive strategy. The gold standard for anaerobic potential evaluation is the Wingate Anaerobic Test (WAT).7 The WAT has several limitations and is very unspecific for swimming even when performed using arm-cranking ergometers,8 since rotational arm-cranking movement is quite different from swimming patterns.9 In-water swimming tests attempting to mimic the WAT (30-s all-out tests) were developed using tethered swimming (in both adolescent10 and international- level swimmers11), but once more they were not able to provide information on fatigue onset and consequent technical changes. Soares and Maia are with the Faculty of Sport, Center for Research, Edu cation, Innovation and Intervention in Sport (CIFI2D), and Fernandes, Machado, and Vilas-Boas, the Faculty of Sport, Center for Research, Educa tion, Innovation and Intervention in Sport (CIFI2D), Porto Biomechanics Laboratory (LABIOMEP), University of Porto, Porto, Portugal. Daly is with the Dept of Kinesiology, KU Leuven, Leuven, Belgium. Address author correspondence to Susana Soares at susana@fade.up.pt. Although all energy systems are active at the start of an effort, the domination of the ATP-PC in all-out short anaerobic efforts is well accepted. It is not clear when ATP-PC domination ends and glycolysis starts. Various authors indicate different limits, particu larly 1 to 5 seconds,12 7 to 10 seconds,13 5 to 15 seconds,14-16 and 10 to 20 seconds.17 When this bioenergetic approach is applied to training, it is assumed that from 8 to 12 seconds a gradual rise in the contribution of glycolysis occurs,12 indicating the depletion of the ATP and PC affecting swimming technique. Some evidences exists that stroke frequency, stroke length, and stroke index change concurrently with a drop in swimming velocity,18 but it is not clear when exactly these changes occur. This knowledge is fundamental to allow coaches to design more efficient training and competitive strategies. With this in mind, Soares et al19 used wavelets to analyze the intra-stroke-cycle force-to-time (F[fj) curve in a 30-second tethered swimming test, concluding that changes occurred in the frequency content of the F(?) curve might be associated with anaerobic fatigue thresholds. However, tethered swimming is different from free swimming, and some bias may exist when using this testing condition. Free-swimming velocity patterns have recently been analyzed using nonchronometric approaches during in-water tests similar to the WAT. Smolka and Ochmann20 developed an anaerobic efficiency test (using 100-m freestyle swimming at maximal intensity), pur portedly based on the classic WAT. Nevertheless, the test had some limitations, as it was conducted in a 25-m swimming pool involving turns, disturbing the analysis of the time progression of speed; it was much longer than the WAT; and the swimming speed was collected by 5 cameras for manual video digitizing (highly time consuming). Potentially, the analysis of fatigue via the decrease in veloc ity (or power) over time during an all-out exercise could provide important information concerning not only the anaerobic potential of a swimmer but also the dynamics of the ATP-PC cycle to lactic 959