Differences in Stroke Phases, Arm-Leg Coordination and Velocity Fluctuation due to Event, Gender and Performance Level in Breaststroke

Differences in Stroke Phases, Arm-Leg Coordination and Velocity Fluctuation due to Event, and Performance Level in Breaststroke HIDEKI TAKAGI 1, SEIJI SUGIMOTO 1, NAOHIKO NISHIJIMA 1, BARRY D.WILSON 2 1 Institute of Health and Sports Sciences, University of Tsukuba, Ibaraki, Japan 2 School of Physical Education, University of Otago, Dunedin, New Zealand ABSTRACT The purpose of this study was to analyze stroke phases, arm-leg coordination and trunk motion fluctuation during breaststroke in elite male and female 50, 100 and events at the 9th FINA World Swimming Championships, Fukuoka 2001. Four phases of the arm stroke and three phases of the leg kick as well as phases of simultaneous arm and leg propulsion and recovery were identified from video of swimmers motions below the surface. The duration of each phase was expressed as a proportion of the whole stroke cycle. Three measures of the arm-leg coordination, percent simultaneous arm-leg recovery time (%SRT), percent arm lag time (%ALT) and percent simultaneous arm-leg propulsion time (%SPT) were calculated. Mean mid-pool swimming hip velocity (V), stroke rate (SR) and stroke length (SL) were also calculated. In addition, the intra-cycle hip velocity of the swimmers was obtained by cinematographic analysis. The SR decreased and SL increased significantly as the event distance increased. For the arm-leg coordination the %ALT, %SPT and %SRT indicated significant differences between event, gender and performance level. In particular, for increasing event distance and for the higher performing swimmer the lower the %SPT and the higher the %SRT. In addition, the range of the intra-cycle hip velocity fluctuation in the lower performing group was greater than the higher performing group. The non-propulsive phase seems to be a key factor for better performance; the breaststroke swimmers must avoid rapid deceleration during the non-propulsive phase by adopting a low resistance posture and stroking technique. Key words: Biomechanics, Race analysis, Stroke phases, Arm-leg coordination, Intra-cycle motion, Breaststroke, Swimming INTRODUCTION Breaststroke is a very challenging stroke because of the discontinuous propulsive action of the arms and legs and its complex time synchronization (Soares et al., 1999). Several studies concerning the coordination of arm and leg motion, and the fluctuation of intra-stroke velocity in breaststroke have been conducted. Craig et al. (1988) investigated the temporal and velocity changes during the stroke cycle for a range of stroke rates, and suggested that the mechanics of swimming may be more critical in the breaststroke than in other competitive stroke styles. A successful breaststroke swimmer was reported to combine more effectively a high intra-cycle peak velocity with relatively long stroke periods (Manley and Atha, 1992). Larger fluctuations of intra-cycle velocity caused a greater energy expenditure (Vilas-Boas, 1996), indicating that the competitive breaststroker should increase their mean velocity without increasing velocity fluctuations. Chollet et al. (1999) suggested that the phases of the breaststroke cycle were not modified in a similar manner with increasing swimming velocity. A more expert breaststroker was reported to change the synchronization of arm and leg stroking patterns with different race

tempos more effectively than a lower ability swimmer to reduce the drag (Soares et al., 1999). These studies indicate that arm-leg coordination seems to be very important for breaststroke but few studies have described differences of the arm-leg coordination by event, gender and performance level using comprehensible measures. Coordination has been described for the arms only in front crawl and a simple index of coordination, the ratio of the lag time between the start of propulsion by one arm and the end of propulsion by the other was developed as a means of understanding arm stroking coordination (Chollet et al., 2000). Describing the complicated arm-leg coordination in breaststroke may form the basis for developing an index of coordination for breaststroke similar to that for front-crawl. Almost all data in the studies described above were obtained in experimental conditions rather than during racing. In view of the recent advances in breaststroker s race performances we believe it is necessary to investigate the technique of the world s current top-level breaststrokers during racing. Therefore, the purpose of this study was to compare differences in stroke phases, arm-leg coordination and intra-cycle hip velocity fluctuation in breaststroke due to event and performance level for both male and female swimmers at the FINA 2001 World Championship. METHOD Subjects The subjects studied were male and female participants in the preliminary, semi-final and final races of the 50, 100 and breaststroke events in the 9th FINA World Swimming Championships, Fukuoka 2001. Due to limitations of the underwater cameras field angle, only swimmers who swam in lanes 4 or 5 in each race were analyzed. The subjects were assigned to two groups, males and females. To investigate differences related to performance, the subjects were also assigned to either a group Event Performance level Female Male Total 4 4 4 9 11 6 13 15 10 8 6 9 6 5 9 14 11 18 Total 35 46 81 Table 1 Number of subjects in each group comprising those eliminated in the preliminaries () or a group comprising those who advanced to the semifinal (). When the same swimmer swam more than twice in lane 4 or 5 in the same events, the data from the best race performance was used. The number of subjects in each group is shown in Table 1. Definition of Terms The definitions of stroke phases and measures of arm-leg coordination during breaststroke are shown in Figure 1. The stroke phases were modified from those of Maglischo (1993). The number of phases was different from Maglischo s following consideration of practical limitations of the video analysis. The arm stroke was divided into four phases defined as follows, i) Recovery: the period from the maximum flexion of the elbows underneath the breast till the arms are completely stretched in front of the face. ii) Glide: the period from when the arms are completely stretched forward till the first observed lateral movement of the hands. iii) Out-sweep: the period from the first observed lateral movement of the hands till the first observed down and backward movement of the hands. iv) In-sweep: the period from the end of the out-sweep till the hands come together for the recovery action. Maglischo (1993) suggested that the out-sweep motion does not produce any propulsion. However, Schleihauf et al. (1988) and Thayer et al. (1986) suggested a possibility of propulsion during the out-sweep. Therefore, we included both the out-sweep and the insweep in the propulsive phase.

2 3 4 5 3 1 1 1-2: Recovery 1-2: Sweep 2-3: Glide 2-3: Lift & Glide 2 3-4: Out-sweep 3-4: Recovery 4-5: In-sweep Duration of arm stroke Duration of arm-propulsion 1 2 3 4 5 Arm Recovery 1 Glide 1 Out-sweep 1 In-sweep 1 Recovery 2 Glide 2 4 1 2 3 4 Leg Sweep 1 Lift & Glide 1 Recovery 1 Sweep 2 %SRT %ALT %SPT Duration of leg-propulsion Duration of leg stroke Fig.1 Definition of stroke phase and measures of arm-leg coordination in breaststroke. Curve lines labeled 1 to 4 and 1 to 5 in the figure are representative 2D motion of the toes and fingers during one stroke. The block diagrams describe the phases of the stroke with time increasing from left to right in the figure The leg kick was divided into three phases defined as follows, i) Sweep: the period from maximum dorsiflexion of the ankle to maximum extension of knee as the feet come together. ii) Lift & Glide: the period from the end of the sweep until the legs are in line with the body and just beneath the surface. iii) Recovery: the period from the first observed forward movement of feet by flexing at knees to the maximum flexion of the knees. During the leg kicking motion, only the sweep phase was defined as a propulsive phase. The durations of phases of arm and leg motion were expressed as a proportion of stroke duration (Figure 1). Phases of separate and simultaneous arm and leg motion were calculated as measures of armleg coordination: simultaneous recovery time (%SRT), indicating a non-propulsive phase of the stroke, the difference between the end of the arms propulsion and the start time of the legs sweep as a percentage of stroke time. Percent arm lag time (%ALT), the time from the start of leg propulsion to the beginning of arm propulsion divided by stroke duration expressed as a percentage. Simultaneous propulsion time (%SPT) indicating a phase of simultaneous propulsion of arms and legs, the difference between the end time of the legs sweep and the start time of the arms out-sweep as a percentage of stroke time. To evaluate the magnitude of the fluctuation in intra-cycle hip velocity, the minimum velocity of the hip was expressed as a percentage of mean swimming velocity (%MinV). Data Collection and Analysis The stroke phases and measures of arm-leg coordination and intra-cycle velocity fluctuation were analyzed from underwater views of three video cameras (Victor TK-1270T) recording at 60 Hz, with a 1/100s digital shutter. Two of the cameras were placed on the pool floor beneath the lane lines of 4 and 5 with the lens axis vertical approximately 15 m away from the wall viewing the swimmers in those lanes as they passed overhead. The other was positioned on the pool floor on lane 2 with

Table 2 Means and standard deviations and statistical comparisons for stroke parameters, arm-leg coordination and intra-velocity fluctuation by event, gender, and performance level Dependent variable V (m/s) SR (strokes/min) SL (m) %SRT %ALT# Factor Group N Mean SD Events Performance level Events Performance level Events Performance level 27 1.46 ± 0.13 26 1.46 ± 9 28 1.40 ± 9 Female 35 1.38 ± 8 Male 46 1.48 ± 0.11 38 1.39 ± 0.10 43 1.48 ± 9 27 58.08 ± 6.08 26 48.41 ± 4.50 28 39.58 ± 4.61 Female 35 47.75 ± 8.41 Male 46 49.22 ± 9.77 38 57 ± 8.15 43 46.82 ± 9.76 27 3 ± 0.20 26 1.82 ± 0.21 28 2.15 ± 0.30 Female 35 1.78 ± 0.31 Male 46 1.88 ± 0.38 38 1.69 ± 0.30 43 1.96 ± 0.35 27 17.85 ± 3.15 26 21.81 ± 8.98 28 32.93 ± 11.22 Main effect Male 35 21 ± 7.67 Female 46 26.48 ± 17 Performance 38 22.34 ± 9.48 level 43 26.09 ± 11.37 27 13.89 ± 6.66 Events 26 19.31 ± 10.14 28 25.86 ± 10.15 Male 35 19.60 ± 10.88 Female 46 19.89 ± 9.95 Performance 38 16.89 ± 8.75 level 43 22.30 ± 10.98 * 27 17.07 ± 11.17 Events 26 9.00 ± 13.13 28-3.00 ± 13.10 %SPT Male 35 8.80 ± 13.93 Female 46 6.59 ± 15.71 Performance 38 12.18 ± 13.10 level 43 3.44 ± 15.37 27 34.50 ± 20.38 Events 26 27.25 ± 15.12 28 33.86 ± 20.83 %MinV Male 35 33 ± 19.88 Female 46 33.41 ± 18.51 Performance 38 24.64 ± 14.75 level 43 38.41 ± 20.24 Significant level at * p <5, p<1, p<01, No Significant # Interaction: Events x the lens axis perpendicular to the swimming Events Tukey HSD * *

direction. All cameras viewed the swimmers at approximately the 25-m mark of the pool. A three-dimensional volume 4.8m long (in the swimming direction), 3.0 m wide, and 1.2 m deep was calibrated using dropped scale lines suspended from the lane cable. Custom written digitizing software was used for digitizing the centroid of the hip joints frame by frame (30 Hz) for each subject. The DLT procedure (Abdel-Aziz, 1971) was used to reconstruct three-dimensional coordinates of the hip joints centroid enabling calculation of intra-cycle hip velocity free of perspective errors. Random errors from the digitizing process were reduced using a recursive second-order Butterworth digital filter with a frequency cutoff of 6 Hz. The number of frames digitized was different for each subject. To time normalize the data of the whole stroke by the event and gender, the frame-series hip velocity data for each subject was fitted to a spline function and interpolated using PC- Mathematica, 4.2 software. After these procedures, the mean intra-cycle horizontal hip velocity of each frame was calculated by event, gender and performance level. To compare the stroke phases and arm-leg coordination among the events or between genders, data of a complete cycle near the 25m mark in the first length of the races were adopted as representative data regardless of events. Using data from the 25m stage of the races was considered to reduce the possible confounding effects of fatigue and race tactics on our measurements over the event distances. Mean mid-pool velocity (V), stroke length (SL) and stroke rate (SR) were also determined at the 25m mark. The values were calculated by using images recorded from above the pool by a total of five cameras repeating the methods of Wakayoshi et al. (1992). Statistical Analysis Three-way analysis of variance (ANOVA) and Tukey post hoc testing were used to analyze the effects of three factors of event (three levels), gender (two levels) and performance (two levels) on the stroke phases and the arm-leg coordination measures as response variables. The normality of all observed variables was tested by using the skewness and kurtosis statistically. The overall significance level was set at alpha = 5. Basic statistics of means and standard deviations (SD) of the stroke phases and the arm-leg coordination measures in each group were computed using SPSS for Windows Version 1J. SL (m) 2.8 2.6 2.4 2.2 1.8 Female y = 007x 2-982x + 4.9263 R 2 = 0.9094 SL (m). 2.8 2.6 2.4 2.2 1.8 Male y = 007x 2-0.1094x + 5.414 R 2 = 0.8957 1.6 1.6 1.4 1.4 1.2 30 35 40 45 50 55 60 65 70 SR (strokes/min) 1.2 30 35 40 45 50 55 60 65 70 SR (strokes/min) Fig.2 Relationship between the mean stroke rate (SR) and the mean stroke length (SL) by event, gender and performance level

Arm Recovery Glide Out-sweep In-sweep Leg Recovery Lift & Glide In-sweep Female Female 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Male Male 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Fig.3 Percentages of the each stroke phase to the whole stroke duration in arm and leg motion by event, gender and performance level RESULTS The mean and SD values for stroke phases, arm-leg coordination and hip velocity fluctuation by event, gender and performance level are shown in Table 2. To evaluate the stroke parameters SL and SR, the relationship between the means by event, gender and performance level are shown in Figure 2. A quadric regression curve for each male and female is also indicated in the respective graphs. There was a significant inverse relationship between SL and SR, (p< 01) for both male and female swimmers. For V and SL, there were significant differences between all groups (Table 2). The V for male or qualified was significantly faster than female or eliminated respectively, and the V for shorter events except for between and was significantly faster than for longer events. The SL for male or qualified was significantly longer than female or eliminated respectively, and the SL for shorter events was significantly shorter than for longer events. For the SR, there was a significant difference between events, with the SR for shorter events being significantly higher than for longer events. The duration of each stroke phase for arm and leg motion as a percentage of the whole stroke duration by event, gender and performance level are shown in Figure 3. The non-propulsive phase of the arm stroke, which consists of the recovery and glide, tended to be of longer duration as the event distance increased. The non-propulsive phase of leg action, which consists of lift & glide and recovery, increased with an increase of event distance, in particular, in the events. Comparing the mode of stroking between the different performance levels; the glide phase in the arm stroke of the group was significantly longer (p < 5) than that for the group in all events, and for leg motion, the percentage of the lift & glide phase tended to be longer in the group than in the group. The arms and legs simultaneous propulsion and recovery, %SPT and %SRT, and lag time from begin of leg-propulsion to arm-propulsion, %ALT by event, gender and performance level are shown in Figure 4

50% %SRT %ALT %SPT 40% 30% 20% 10% 0% -10% -20% Female Male Fig.4 Percentage of the simultaneous recovery time (%SRT), the arm lag time (%ALT) the simultaneous propulsion time (%SPT) and SD (shown as error bars) and Table 2. For the %ALT, there were significant differences between events and performance level groups. As the events became longer, the %ALT became higher, and %ALT in the group was significantly higher than that in the group (Table 2). The %SRT also increased significantly with an increase in event distance. In contrast, the %SPT decreased significantly with an increase in event distance. In particular in the events, the %SPT indicated negative values except for the female s group. These negative values mean that the arm stroking began a brief interval after finishing the leg sweep i.e. the propulsive phase of the arm did not overlap that of the leg. Fluctuation of horizontal intra-cycle hip velocity by event, gender and performance are shown in Figure 5 and Table 2. The X axis is time, normalized to mean time, and starts at begin of the recovery action of the arms. Comparing to within the same event and gender, the mean hip velocity during a stroke cycle was higher in the group than in the group. The difference between the two groups was most marked at or near the minimum hip velocity rather than at maximum hip velocity. This was consistent with the %MinV of the group being significantly lower than for the group. There was no difference in the %MinV between event and gender (Table2). DISCUSSION AND IMPLICATIO We adopted the dependent variables at the first length in the race as representative data

for each subject and event in order to reduce the confounding effects of fatigue and race tactics. Although the stroke characteristics in the or event naturally changed Female 0 10 20 30 40 50 60 70 80 90 100 %Time (normalized to mean time) Male Male 0 10 20 30 40 50 60 70 80 90 100 %Time (normalized to mean time) Female 0 10 20 30 40 50 60 70 80 90 100 110 120 %Time (normalized to mean time) Male 0 10 20 30 40 50 60 70 80 90 100 110 120 %Time (normalized to mean time) Female 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 %Time (normalized to mean time) Male 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 %Time (normalized to mean time) Fig.5 Fluctuation of intra-cycle horizontal velocity of hip joint during a stroke plotted against time (normalized to mean time)

as the event progressed, the values of V, SR and SL at the first length have been shown to be strongly correlated with the mean value of the variable in the whole race (Thompson et al., 2000). Therefore, to evaluate the stroke characteristics for different events by using data from only this stage of the races was considered reasonable. For the SR and SL, our findings are largely in agreement with those of previous studies (Kennedy et al., 1990; Chengalur and Brown, 1992; Wakayoshi et al., 1992; Thompson et al., 2000). For example, the SR and SL were inversely related to each other, better swimmers adopted a greater SL than less proficient swimmers. In addition the SR values were not significantly different by gender and performance levels. For the V, the results for top level performers were naturally faster than the previous results but the mean values of V were comparable to previously published results. To evaluate validity of the stroke coordination measures in this study, we compared them with those reported in the previous studies done by Chollet et al. (1999) and Soares et al. (1999). In these studies changes of stroke phases for both arms and legs motion were examined for swims at pre-determined speeds corresponding to the actual 50, 100 and competition. Chollet et al. (1999) indicated that the percentage of the glide phase increased and the recovery phase decreased as the event distance increased consistent with what we have shown in Figure 3. In contrast, Soares et al. (1999) showed that only the absolute duration of the arm recovery became significantly higher as the event distance increased. Compared to previous research, we present new measures to describe differences of the arm-leg coordination due to event, gender and performance level. These new measures showed that the %ALT became higher as the distance increased, and the %ALT of qualified swimmers was significantly higher than for the eliminated swimmers. For the %SPT, there were significant differences between gender and performance level. As the event distance increased, the %SPT value changed from positive to negative dramatically, a change indicative of decreasing overlap of propulsion. Moreover, the qualified swimmers adopted a significantly smaller %SPT than the eliminated swimmers. In contrast, the %SRT values tended to increase as the distance increased, with no difference between performance levels. The results are new evidence that the swimmers adjust to a change of event distance by altering the timing of the arm motion. Maglischo (1993) pointed out that there are three general styles of breaststroke timing, i.e. overlap, continuous and glide. The three styles correspond roughly speaking to - as overlap, the - as continuous, and the - as glide timing. The %SPT quantifies this timing difference for the different event distances. Another major finding concerned with the arm-leg coordination is that the group had significantly higher %ALT and lower %SPT than the group. These results reveal that the better swimmers delay the beginning of arms motion during the legs sweep, in other words they extend the arms glide phase. Since the better swimmers adopted a longer non propulsive glide phase, this implies that the non-propulsive phase is perhaps more important for higher performance than the propulsive phase. The reason why the non-propulsive phase is a key factor for performance can be established by examining the intra-cycle hip velocity fluctuation during a stroke. We proposed the %MinV as indicative of the magnitude of fluctuation of intra-cycle hip velocity. The %MinV of the group was significant lower than for the group, indicating that the intracycle hip velocity fluctuation was more significant in poor swimmers rather than in better swimmers. The hip velocity minimum during the arms glide phase tended to be also lower in poor swimmers than the better swimmers although there was no statistical

difference. For the maximal intra-cycle hip velocity, there was no remarkable difference between the two groups although there was a tendency to have a higher velocity at the end of the arms in-sweep in the better swimmers. It is, therefore, important not to decelerate rapidly during the recovery phase, and to keep a relatively higher hip velocity during the arms glide phase. High variations in swimming speed within a stroke cycle are well known to impose a high energy cost (Vilas-Boas, 1996). Therefore, the breaststroker should consider how to minimize the fluctuations of intra-cycle velocity. Colman et al. (1998) compared the intra-cycle velocity variation of flat and undulation breaststroke styles, and concluded that there was considerably less difference between the maximum and minimum velocity peaks in the most undulating style than in the lowest flat style. Different styles appeared to be homogeneously distributed between the group and the group but since swimmers styles were not explicitly studied this is a possible area for future study. Regardless of style, a characteristic of better performing swimmers was the lesser reduction in velocity during the non-propulsive phase. CONCLUSION In this paper, we have compared stroke phases, arm-leg coordination and intra-cycle hip velocity fluctuation during breaststroke by event, gender and performance level. We found i) that differences for event distances were mainly changes in arms recovery and glide phases while the legs kick timing was unchanged; ii) better performance seems to be identified by changes in technique to reduce the decrease in hip velocity during the non-propulsive phase. This may be the result of decreased simultaneous propulsion or an increase in the glide phase of the arms in the better swimmers. ACKNOWLEDGMENTS We would like to acknowledge the Fédération Internationale de Natation (FINA) and the Japanese Amateur Swimming Federation for their support of this study. The authors would like to extend a special thanks to the members of the race analysis project, Teruo Nomura, Koji Wakayoshi, Takeshi Matsui, Keisuke Okuno, Futoshi Ogita, Yuji Ohgi, Yasushi Ikuta, and Masanobu Tachi for all their support throughout this project. REFERENCES Abdel-Aziz, Y.I., and Karara, H.M. (1971). Direct linear transformation from comparator coordinates into object space coordinates in close-range photogrammetry. Proceedings of the Symposium on Close-Range Photogrammetry (pp. 1-18). Falls Church, VA: American Society of Photogrammetry. Chengalur, S. N. and Brown, P. L. (1992). An analysis of male and female Olympic swimmers in the 200- meter events. Canadian Journal of Sports Sciences, 17, 104-109. Chollet, D., Chalies, S. and Chatard, J. (2000). A new index of coordination for crawl: Description and usefulness. International Journal of Sports Medicine, 21(1), 54-59. Chollet, D., Chollet, C. T. and Gleizes, F. (1999). Evolution of co-ordination in flat breaststroke in relation to velocity. In K. L. Keskinen, P. V. Komi and A. P. Hollander (eds.), Biomechanics and Medicine in Swimming VIII (pp. 29-32). Jyväskylä: Gummerus Printing. Colman, V., Persyn, U., Daly, D. and Stijnen, V. (1998). A comparison of the intra-cyclic variation in breaststroke swimmers with flat and undulating styles. Journal of Sports Sciences, 16(7), 653-665. Craig, A. B., Boomer, W. L. and Skehan, P. L. (1988). Patterns of velocity in competitive breaststroke

swimming. In B. E. Ungerechts, K. Wilke and K. Reischle (eds.), Swimming Science V (pp. 73-77). Champaign, IL: Human Kinetics. Kent, M. R. and Atha, J. (1973). Intra-cycle retarding force fluctuations in breaststroke. The Journal of Sports Medicine and Physical fitness, 13(4), 274-280. Kent, M. R. and Atha, J. (1975). Intracycle kinematics and body configuration changes in the breaststroke. In L. Lewillie and J. P. Clarys (eds.), Swimming II (pp. 58-63). Baltimore: University Park Press. Kennedy, P., Brown, P., Chengalur, S. N. and Nelson, R. C. (1990). Analysis of male and female Olympic swimmers in the events. International Journal of Sports Biomechanics, 6, 187-197. Maglischo, E. W. (1993). Swimming Even Faster. Mountain View, Ca: Mayfield Publishing. Manley, P. K. and Atha, J. (1992). Intrastroke velocity fluctuations in paced breaststroke swimming. In D. Maclaren, T. Reilly and A. Lee (eds.), Swimming Science VI (pp. 151-159). London: E & FN Spon. Schleihauf, R. E., Higgins, J. R., Hinrichs, R., Luedtke, D., Maglischo, C., Maglischo, E. W. and Thayer, A. (1988). Propulsive techniques: front crawl stroke, butterfly, backstroke, and breaststroke. In B. E. Ungerechts, K. Reischle and K. Wilke (eds.), Swimming Science V (pp. 53-59). Champaign, IL: Human Kinetics. Soares, P. M., Sousa, F. and Vilas-Boas, J. P. (1999). Differences in breaststroke synchronization induced by different race velocities. In K. L. Keskinen, P. V. Komi and A. P. Hollander (eds.), Biomechanics and Medicine in Swimming VIII (pp. 53-57). Jyväskylä: Gummerus Printing. Thayer, A. L., Schleihauf, R. E., Higgins, J. R., Hinrichs, R. N., Luedtke, D. L., Maglischo, C. W. and Maglischo, E. W. (1986). A hydrodynamic analysis of breaststroke swimmers. In J. G. Hay (eds.), Starting, Stroking and Turning (pp. 131-143). Iowa: Department of Exercise Science, University of Iowa. Thompson, K. G., Haljand, R. and MacLaren, D. P. (2000). An analysis of selected kinematic variables in national and elite male and female 100-m and 200-m breaststroke swimmers. Journal of Sports Sciences, 18(6), 421-431. Vilas-Boas, J. P. (1996). Speed fluctuation and energy cost of different breaststroke techniques. In J. P. Troup, A. P. Hollander, D. Strasse, S. W. Trappe, J. M. Cappaert and T. A. Trappe (eds.), Biomechanics and Medicine in Swimming VII (pp. 167-171). London: E & FN Spon. Wakayoshi, K., Nomura, T., Takahashi, G., Mutoh, Y. and Miyashita, M. (1992). Analysis of swimming races in the 1989 Pan Pacific swimming championships and 1988 Japanese Olympic trials. In D. Maclaren, T. Reilly and A. Lee (eds.), Swimming Science VI (pp. 135-141). London: E & FN Spon.