You are on page 1of 17

Pediatric Exercise Science, 2010, 22, 135-151 © 2010 Human Kinetics, Inc.

A Multivariate Analysis of Performance in Young Swimmers
José M. Saavedra and Yolanda Escalante
University of Extremadura

Ferran A. Rodríguez
University of Barcelona
This study develops multivariate models to predict swimming performance based on multidimensional assessment. 66 male (age 13.6 ± 0.6 y) and 67 female (11.5 ± 0.6 y) swimmers undertook a test battery including a sports background and training questionnaire, anthropometry, general and specific fitness tests, and technique. Competitive performance (LEN scores in three best events) was the predicted variable. A multiple linear regression model explained 82.4% of performance variability in males (based on age, sitting height, 30-min test, 6 × 50 m at 1:30, and swimming index) and 84.5% in females (age, 30-min test, 6 × 50 m at 1:30, and velocity at 50 m). Discriminant analysis using a four-group split-sample approach correctly classified 94.1% of the best male swimmers (based on age, 30-min test, 6 × 50 m at 1:30, shoulder extension, arm span, and height), and 71.0% of the best females swimmers (30-min test, horizontal floating, velocity at 50 m, and age). Chronological age was the main predictor of performance in this age category. Main predictive variables pertained to the anthropometric (particularly in males), specific fitness (aerobic speed and endurance), and technical domains (particularly in females). In these ages competitions should be organized according to year of birth and not by age categories.

Sports performance is the result of a complex process involving many factors. Performance capacity has been studied using a multidimensional assessment approach in different sports such as ice hockey (9), rowing (35), orienteering (8), tennis (39), rugby (28), Australian football (19), soccer (25,29), handball (10), triathlon (41), weightlifting (11), and volleyball (12,13). If performance was taken as a continuous variable, multiple linear regression (MLR) is the most common statistical technique (10,12,25,39). If, however, performance is considered a discrete variable, usually categorized in two groups (i.e., selected/nonselected, elite/nonelite), discriminant analysis (DA) seems to be the preferred technique (1,8,10–12,19,28,29,35).

Saavedra and Escalante are with the University of Extremadura, Cáceres, Spain. Rodríguez is with the University of Barcelona, Barcelona, Spain. 135

More recently. have been able to explain between 20% and 93% of variability in performance.6 years). Therefore. and continuously modulated by the training process (13). to develop multivariate models explaining swimming performance from a multidimensional perspective).15) 100 breastroke (17) and 400 freestyle (4. and (c) to determine which group of variables actually predicted performance level at young age. Some studies predicting swimming performance using MLR.. based on individual genetic endowment. and (f) multidimensional evaluation. 66 males (13. such as neural network technology (23) and nonregression models (31) which have predicted performance on 99% and 60% of the cases. . All tests are widely used in swimmers’ assessment and were performed one week after the National summer championships. a considerable number of assessments were included as predictive variables to ensure comprehensive evaluation.17. the aims of this study were (a) to analyze swimming performance in young peripubertal athletes by developing multivariate predictive models (MLR and DA) based on a wide variety of assessments from a multidimensional perspective. One study has predicted performance in two groups (faster and slower) using DA. 100 freestyle (1.15. if these multivariate techniques are to be applied to talent selection in swimming. Swimming performance is determined by the interaction of morphological.18). with prediction percentages over 82% (21). (e) technical analysis. selected as the best national swimmers of their age category in Spain. especially if we bear in mind that only about forty out of the one hundred best swimmers in the 13–14 year old ranking will continue to rank among the one-hundred best at 17–18 years (38).6 ± 0.5 ± 0.6 years) and 67 females (11. which included assessment in the following domains: (a) sports background and training status. (b) to compare their predictive power with that of models developed in nonswimming related studies. According to the National swimming federation (RFEN) the “alevín” category includes females of 13–14 years and males of 11–12 years of age.e. psychological and technical factors. Methods Subjects Subjects were 133 young swimmers. it is of equal interest to identify the best individual young swimmers (MLR) as it is to predict the best group of swimmers (DA). physiological. The swimmers’ parents or legal tutors signed an informed written consent previously to their participation. (c) general fitness tests. (b) anthropometry. However. The study was approved by the Bioethics Committee of the University of Extremadura (Spain). both in adults (32. respectively. more novel statistical techniques have been used. Discrepancies may be related to differences in the performance criterion: 25 yd (32.34). (d) specific fitness tests. Assessment Procedures All subjects undertook a comprehensive battery of tests.18).4. In accordance with the aims of the study (i.136   Saavedra et al.34) and young swimmers (1. the most important competition of the season.

training volume per session (pool and dryland). subescapular. arm span. General Fitness Tests General fitness was assessed using the Eurofit test battery (7): shuttle run test assessed general aerobic endurance. In addition. and shuttle run test 10 × 5 m assessed agility-velocity. The following proportionality indices were calculated: body mass index. After exploratory analysis (i. and leg). skinfolds (triceps. lengths and widths (hand and foot). and pubic hair (males and females) development were rated in five stages by a medical expert by inspection. flamingo balance assessed general balance.e. chest girth/height index. breast (females). sitting height. abdominal. . biacromial breadth/height index. hand dynamometry assessed grip. biceps. gluteal. abdominals in 30 s assessed trunk power. and the three components (endomorphy. bitrochanterial/biacromial breadth index. etc. building specific MLR and DA models restricted to this specific domain). amount of specialized training. Somatotype was determined using the anthropometric method (16). breadths (biacromial. sit and reach assessed flexibility of the trunk and lower limbs. arm flexed. thigh. and training methods (circuit training. and wrist). Stages of genital (males). elbow. biiliac. plate tapping assessed segment velocity of the upper limbs.. and medial calf). bitrochanteric. Body composition was assessed using a two-compartment model (24).). mesomorphy. This questionnaire assessed the relationship between swimming performance and variables such as previous sports participation (both for swimmers and parents). Sexual maturation was assessed from the development of secondary sex characteristics according to Tanner (40). elastic bands. knee. we chose to use the linear regression equations based on the sum of six skinfolds used by Carter in Olympic athletes of any age and for each gender category (2). The age of menarche (females) was assessed by recall. front thigh. Measures included body dimensions (height. flexed arm hang assessed muscular resistance of the arms and shoulders. and weight). 6 on sports practice. arm span/height index. weights.Analysis of Performance in Young Swimmers 137 Sports Background and Training Status This domain was assessed by an ad hoc questionnaire including 26 items: 7 related to social background. number of weekly training sessions. and gluteal girth/height index. Anthropometry Anthropometric measurements were taken according to standardized procedures (30) by an ISAK (International Society for the Advancement of Kinanthropometry) certified anthropometrist. only training volume variables were used for further statistical analyses. horizontal jump assessed explosive strength of the lower limbs. Sum of six skinfolds was used as main adiposity index. number of meetings per season. although there is no wide agreement as to which regression equations should be used to estimate fat mass from skinfolds measures in athletic populations of children and adolescents. girths (chest. bitrochanterial breadth/height index. and 13 items on swimming training and competition. supraspinale. swimming practice (age of begin to swim and compete). and ectomorphy) were analyzed separately.

and hydrodynamic tests. Filming was carried out in a 25-m pool. within or at different events. the swimmer accelerated during the first 10 m (start) and maintained maximum speed during 10 m (finish).5 m).138   Saavedra et al. Specific Fitness Testing Specific fitness was assessed by means of flexibility. A 10-m test was used to assess maximum speed. a technical swimming index was calculated. Swimming tests were all performed at maximal effort. starting from an immersed position. anthropometry. Qualitative analysis consisted of technical assessment by an experienced coach with more than ten years of experience in underwater technical evaluation. on a separate trial from every plane. turn velocity (17.5–25 m). general fitness tests. The 30-min test (27) was used to assess aerobic endurance. 50-m personal best times were not . stroke rate. Subsequently. and ankle flexion and extension. and starting from the water. (b) to assess horizontal floating we measured the time lapsed between horizontal floating (initial position) and vertical floating in steady position. Semiquantitative kinematic analysis was performed during a 50-m all-out swim for the calculation of: start velocity (0–10 m).5–32. The 6 × 50 m with 1:30-min:s start at personal stroke was used to assess speed endurance (3). and (c) to assess vertical floating we rated the part of the body remaining out of the water from a zero score if the head was completely immersed up to seven if the water surface was at the level of the neck. Multidimensional Evaluation Combined analysis of variables from the different domains (sports background and training status. from video records on standard positions. both qualitative and semiquantitative was also undertaken. as the total number of errors made by each swimmer divided by the total number of possible errors (37). Hydrodynamic evaluation included glide and floatability tests (3): (a) to assess gliding hydrodynamic position and explosive strength of the lower limbs we measured by visual inspection the distance gained by gliding in prone position after turning. and after starting. and their technique compared with a standard technical pattern (23) to identify the most common technical errors (36). specific fitness tests. Flexibility testing included the measurement of the maximum range of movement of the shoulder. approach velocity (17. Swimmers were filmed from sagital and frontal planes while swimming in their fastest stroke.5 m). Performance Evaluation LEN table (22) was used to assess competitive performance level. stroke length. finish velocity (40–50 m). LEN scores measure how close a certain personal best time is from the World record in each competitive event. endurance swimming. Individual performance level was quantified as the sum of LEN scores in the three best personal events during the season. and swimming index (6). pull-out velocity (25–32. allowing also comparing two time records. and technical analysis) was made by developing multivariate models (see Statistical analysis for details). Technical Analysis Video analysis of swimming technique. sprint swimming.

e. backstroke.05). and 1.. p ≤ .. which measures the deviations within each group with respect to the total deviations.e. or breastroke) at any of four different race distances (i. freestyle.Analysis of Performance in Young Swimmers 139 included assuming that the swimmers were not specialized in sprint swimming at their young age. or until the maximum number of steps has been reached.e. Some typified variables (10-m all-out test. Pearson’s simple correlation coefficient (R). Performance was the predicted variable. This process continues until no variables in the equation can be removed and no variables not in the equation are eligible for entry.84 to enter).e. 800 –females–. provided the value of F was greater than a certain critical value (i.e.e..500 m –males). The sample-splitting method included initially the variable that most minimized the value of Wilks’s Lambda. F = 2.e. canonical correlation index. As a secondary analysis. Thus.. F = 3. and all parameters from technical quantitative analysis) were expressed as z values. and no further variables were selected in the process. glide distance after turn and start. D the worst) using a stepwise selection procedure.71 to remove). discriminant function) was Wilks’s Lambda.. not restricted to a single event (stroke and/or distance). and the number of candidate predictive variables in the model was limited to nv ≤ n / 5 (i. 6 × 50 m at 1:30 start. These criteria were chosen to obtain an overall assessment of competitive performance. Statistical Analysis Unless specified. subjects were classified by the sample-splitting method in four groups according to their performance level (i. 200. Before including a new variable. The next step was pairwise combination of the variables with one of them being the variable included in the first step. and simple and multiple linear coefficients of determination (R2 and Rm2) were calculated for all parameters in relation to performance level. Stepwise selection procedure consists in removing the variable with the largest probability of F if a certain preestablished value is exceeded (i. 50 m events are not included at the National championships program. In DA. butterfly.e.e. and the corresponding F-value was below a critical value (i.. All variables in the equation are again examined for removal. The criterion used to determine whether a variable entered the model (i. If this condition was not satisfied. 400. the process was halted. data are expressed as means ± SD (SD). 100.. partial correlation coefficients (Rp) with chronological are as a control variables were also computed. MLR models were developed using the stepwise selection procedure.. an attempt was made to eliminate some of those already selected if the increase in the value of Wilks’s Lambda was minimal. For each type of assessment. always with the condition that the F-value corresponding to the Wilks’s Lambda of the variable to select has to be greater than the aforementioned “entry” threshold. respectively. The normality and equal variance of the distributions were tested using the Kolmogorov-Smirnov and the Levene tests. Before MLR analysis. performance was evaluated as the result in fastest three competitive events swum in one of the fours strokes (i.10). Successive steps were performed in the same way. p ≤ . and percentage of subjects correctly classified . Wilks’s Lambda. the independent variable not in the equation with the smallest probability of F is entered if the value is smaller than a certain preestablished value (i. Then. in fact. The equation is recomputed without the variable and the process is repeated until no more independent variables can be removed. A the best group of swimmers. 13 variables) to prevent obtaining spurious relations among variables (26)..

99 ± 1.01 10.50 57.96 ± 7. respectively. The selected variables included in the MLR models explained 82. and Sexual Maturation Status of the Swimmers Age (years) Height (cm) Weight (kg) Pubic hair (stage)* Genital development (stage)* Breast development (stage)* Age of menarche (years) # Note. respectively) was . and age). 6 × 50 m at 1:30 start.12 ± 7. Table 2 shows the main results from the different assessments and variable correlates with competitive swimming performance.05.47 43.51 ± .. In the multidimensional evaluation..60 ± .56 171.18 3. The Statistical Package for Social Sciences (SPSS for Windows.61 ± 1. swimmers reached. specific endurance index. horizontal floating.63 ± 1. 30-min test. A P-value <.6% (females) of absolute World records in their three best events. Since power is 1-beta (1. Age.95 ± 8. Linear simple correlation (R) and partial correlations coefficients (Rp) calculated with age as control variable are shown for variables significantly correlated with performance (p ≤ ..8). Table 1  Gender. 30-min test.140   Saavedra et al. and sexual maturation status of the swimmers is presented in Table 1.2 = .5.3% of the sample. with only six variables for males (age. 40) # n = 19 Males (n = 66) 13. on average.17 2.4 ± 70.75 ± 7. Results A summary of chronological age. Multiple determination coefficients (Rm2) of MLR models for each sex are also shown.0 ± 76. and four variables for females (30-min test. Body Size. and swimming index).0 points (females) on average. Values are mean ± SD. 30-min test. about 50. was 508.0 .e. sitting height.92 ± . version 15.19 2.5% of the variance in competitive performance in males and females.4% and 84. 6 × 50 m at 1:30 personal stroke. predictive models correctly classified 72.80. Thus. 6 × 50 m at 1:30 start. Using a moderate effect size of . power was set at .158 3. Table 3 shows the DA of each assessment. for the whole sample and for each category group were computed as indicators of performance predictive capacity.05).. and in females (age.7 points (males) and 476. Swimming performance level.0) was used for all analyses.7% and 68. shoulder extension. assessed as the mean of the three best scores on the LEN table. velocity at 50 m personal stroke. and mean velocity at 50-m all-out test).57 - Females (n = 67) 11. according to Cohen (5). n = 66 and 67 for males and females. *According to Tanner (1962.82. and height). body size. beta was set at four times alpha (i. The variables selected to enter the models were five.05 was considered to be statistically significant.68 .55 154.20).8% (males) and 47. both in males (age. the calculated statistical power for the studied sample (i.e. Having set a two-tailed alpha value of .47 ± 1. arm span.

27 84.634 .12 9.396 .587 .56 171.03 ± 1.294 .414 .18 17.263 .27 .345 .15 9.80 ± 2.61 37.50 33.19 ± 5.507 .246 .398 .14 177.96 ± 7.48 ± 8.30 90.326 .22 ± 0.46 ± 2.50 80.06 ± 2.386 8.60 ± 0.12 ± 7.409 .87 ± 0.62 23.47 6.72 43.12 80.62 7.45 ± 0.55 ± 2.36 ± 2.318 .08 ± 0.82 ± 0.05 28.93 25.375 .380 .34 24.99 8.47 77.57 22.89 4410 ± 1136 .92 ± 0.281 Training status Training volume (h/wk) Training volume per session (m) Somatic evaluation Chronological age (years) Height (cm) Sitting height (cm) Arm span (cm) Weight (kg) Hand length (cm) Hand width (cm) Foot lenght (cm) Foot width (cm) Biacromial breadth (cm) Biiliac breadth (cm) Bitrochanteric breadth (cm) Knee breadth (cm) Elbow breadth (cm) Wrist breadth (cm) Chest girth (cm) Arm flexed girth (cm) Gluteal girth (cm) Thigh girth (cm) 13.465 R Test domain Test variables .86 ± 8.527 .77 57.Table 2  Main Results From the Different Assessments and Correlates With Competitive Swimming Performance Males (n = 66) Females (n = 67) Rp .10 158.77 ± 1.75 51.23 47.20 ± 5.401 .42 5.273 .265 .75 29.34 ± 7.287 .401 .263 .252 .34 4052 ± 929 Rm2 Mean ± SD R Rp Rm2 Mean ± SD 9.309 .85 ± 2.500 .314 .459 .504 .24 ± 4.82 27.256 .42 8.66 ± 0.25 82.46 ± 4.62 24.32 4.30 8.74 ± 0.243 .354 .263 N/A N/A .767* .95 11.287 141 (continued) .392 .07 ± 7.301 .441 .50 ± 1.55 154.20 ± 1.79 ± 1.78 ± 6.41 5.17 15.741* .92 ± 7.330 .304 .89 ± 1.12 ± 1.75 ± 7.51 ± 0.40 ± 0.430 .95 ± 8.350 .46 ± 0.579* .43 ± 0.

48 10.6 ± 14.571 –.360 .66 ± 1.86 13.08 ± 2.92 ± 4.53 ± 8.448 .09 ± 3.302 (continued) .49 3.91 78.279 .14 ± 2.68 ± 13.37 ± 11.40 ± 1.362 .08 3.18 51.33 28.67 ± 0.291 –.95 52.45 5.36 6.52 62.277 Mean ± SD 30.540 .12 ± .01 3.0 ± 16.03 2.03 –.99 53.87 61.92 4.84 ± 2.68 ± 4.52 20.39 22.92 ± 1.242 .89 ± 3.351 –.96 ± 0.287 .259 –.57 ± 2.04 Leg girth (cm) Arm span/height index Biacromial breadth/height index Chest girth/height index Gluteal girth/height index Body mass index (kg/m2) Sum of six skinfolds Fat mass (%) Endomorphy Mesomorphy Ectomorphy General fitness testing Shuttle run endurance (stage) Flamingo balance (attempts) Plate tapping (s) Sit-and-reach (cm) Horizontal jump (cm) Hand dynamometry (kg) Abdominals in 30 s (reps) Flexed arms hang (s) Shuttle run 10 × 5 m (s) Specific fitness testing Shoulder flexion (grades) Shoulder extension (grades) 9.04 20.310 Rp Rm2 R 142 Test domain Test variables –.65 ± 16.269 Rm2 R .312 .51 ± 1.32 ± 1.89 3.07 –.244 Mean ± SD 34.00 ± 2.57 ± 0.49 ± 3.91 11.74 ± 22.60 ± 13.58 ± 1.294 8.67 43.379 3.70 .275 .72 22.21 3.44 ± 1.59 18.93 9.26 49.26 15.550 .55 ± 14.76 ± 20.27 177.369 .78 ± 1.72 ± 2.65 ± 6.84 26.508 .38 12.64 ± 2.39 7.83 ± 1.20 2.46 9.03 65.21 21.49 19.385 .44 24.351 –.71 196.238 .346 .Table 2 (continued) Males (n = 66) Females (n = 67) Rp –.46 ± 6.268 –.278 .09 ± 0.48 ± 6.69 ± 2.365 .28 ± 1.72 ± 2.68 ± 1.91 ± 1.19 .

97 0.251 .415 .05).01 ± 0.00 ± 0.61 ± 6.76 ± 6.476 .458 .03 ± 0.613 Mean ± SD 165.07 0.Table 2 (continued) Males (n = 66) Females (n = 67) Rp Rm2 R Rp Rm2 R Test domain Test variables Mean ± SD 163.02 ± 0.46 0.00 ± 0.775* .24 ± 0.250 .97 0.00 ± 0.11 ± 0.824# .10 ± 0.674 .685* .19 Ankle flexion (grades) Ankle extension (grades) 10-m test maximal velocity (z value) 30-min test (m·s-1) 6×50 m at 1:30 start (z value) Gliding distance (prone.93 0.501 .317 .616 .96 .404 .58 112.630 .97 1.01 ± 0.97 0.93 0.484 .308 .654 .00 ± 0.97 0.97 0.00 ± 0.02 ± 0.00 ± 0.16 0. Pearson’s linear simple correlation (R) and partial correlations coefficients (Rp) calculated with age as control variable are indicated for variables significantly correlated with performance (P≤.539 .00 ± 0.552 .488 .36 ± 1.00 ± 0.488 0.845# 143 Note.96 0.97 .97 0.97 0.471 .01 ± 0.27 ± 9.286 .248 .528 .00 ± 0.490 .676* .91 7.466 .365 .93 0.10 ± 0.293 .700* . N/A: Not applicable (chronological age is the control variable for partial correlations) * Variables entered in MLR models # Multiple regression determination coefficients (Rm2) for all variables in MLR models .00 ± 0.01 ± 0.543 .97 .96 0.29 ± 1.448 .551 . m) Gliding after turn (z value) Gliding after start (z value) Technical analysis Start velocity (z value) Approach velocity (z value) Pull-out velocity (z value) Turn velocity (z value) Finish velocity (z value) Mean velocity (z value) Stroke rate (z value) Stroke length (z value) Swimming index (z value) MLR models 0.332 .94 .98 7.11 0.00 ± 0.526 .01 ± 0.97 0.97 0.411 .442 .840* .365 . Multiple determination coefficients (Rm2) of multiple linear regression (MLR) models for males and females are also shown at the bottom.02 ± 0.576* 0.94 0.599 .624 .00 ± 0.97 0.509 .50 108.98 1.02 ± 0.11 0.505 .9 ± 12.523 .459 .425 .01 ± 0.97 0.548 .

6% 75.397 . sexual maturation (pubic hair) 58.0% 64.3% .5% 68.4% 46.8% 46.597 .9% 31.2% 62.8% 35.2% .3% .8% 56.7% .5% 68.8% .3% 37.6% 56.3% 62.Table 3  Discriminant Analysis Models for the Different Assessments and the Overall Multidimensional Evaluation Evaluations Sports background and training status A Group (best swimmers) B Group C Group D Group (not-so-good swimmers) Whole sample Wilks’s L Canonical correlation index Variables in the model Somatic evaluation A Group (best swimmers) B Group C Group D Group (not-so-good swimmers) Whole sample Wilks’s L Canonical correlation index Variables in the model General fitness testing A Group (best swimmers) B Group C Group D Group (not-so-good swimmers) Whole sample Wilks’s L Canonical correlation index Variables in the model Specific fitness testing A Group (best swimmers) B Group C Group D Group (not-so-good swimmers) Whole sample Wilks’s L Canonical correlation index Variables in the model Males (n = 66) 70.2% 58.797 .7% 62.450 Training volume per session 88.722 Hand dynamometry. abdominals in 30 s 70.6% 52.3% .851 Age.4% 65.3% 35.1% .9% 0% 47.2% 43.8% 35.638 Hand dynamometry. horizontal floating (continued) 144 .5% 41.390 .724 .822 Height.789 30-min test.1% 68. shoulder extension Females (n = 67) None* 70.544 . 30-min test.5% 82. arm span.452 . shuttle run endurance 76.3% 34. age 37.3% 29.3% 31.807 6x50 m at 1:30 start.

523 . Data are percentage of subjects correctly classified in each of the four-category groups (A group. arm span.672 Turn velocity.Analysis of Performance in Young Swimmers 145 Table 3 (continued) Evaluations Technical analysis A Group (best swimmers) B Group C Group D Group (not-so-good swimmers) Whole sample Wilks’s L Canonical correlation index Variables in the model Multidimensional evaluation A Group (best swimmers) B Group C Group D Group (not-so-good swimmers) Whole sample Wilks’s L Canonical correlation index Variables in the model Males (n = 66) 82.4% 20.7% 62. best swimmers. as well as with DA models which categorized performance at four levels reaching predictive values in the fastest swimmers group of 94% in males and 71% in females.0% 76.. D group.4% 50.7% 68.6% 76.5% . velocity at 50 m personal stroke.605 . aerobic endurance. shoulder extension.427 .5% . 29% in soccer (25). or 21% in volleyball (12). Wilks’s Lambda. although it is worthy to note that that study was performed with recreational athletes (41). 6 × 50 m at 1:30 personal stroke. Comparison With Other Sports The results show a swimming performance prediction by MLR models of 82% (males) and 85% (females) based on only five predictive variables.5% 66.e.5% 64. Three of the variables selected (i.0% 62. and .691 Velocity 50 m individual stroke 71.893 Age.7% 43. age Note. age. These values are far higher than those reached in other sports: 46% in tennis (39). not-so-good swimmers) and within the whole sample. Discussion This study contributes with MLR models predicting performance as a continuous variable in young swimmers at the top national level with a high degree of predictive power (explaining over 82% of the variance in performance in both sexes).7% .2% 68.3% .5% 88. stroke rate 94. and variables entered in the model are indicated.5% 70.845 30-min test. height Females (n = 67) 65.405 .8% 62. * No discriminant predictive models were obtained. horizontal floating.1% 62. 30-min test. only in triathlon a higher prediction was achieved (98%).8% 72. canonical correlation index.

32) in our results. 28%. 20%) and 100 m freestyle (males. that most of these studies did not assess specific fitness. However. R2 = . 100 m breaststroke in Estonian swimmers (females.60).18. 17). it is worthy to note that our study is to our knowledge the only one in which subjects were categorized in four sample-splitted groups according to performance level. 69–70% in female handball (10). swimming competitions should be organized by year of birth rather than by age category (two years of age together). ³89%. females 85%).71). females. The variables selected in the model in both genders were age (males. In other studies the selected variables belonged to: physical fitness and anthropometric domain in ice hockey (9). which differs from other studies in which these were the most selected variables (1. or 100 m butterfly (males.25. 59%.32. Multiple Lineal Regression Models in Swimming As we have seen. 64% in orienteering (8). female 17%.55). and supports . and in the second the swimmers were adults. The models did not select anthropometric variables. and anthropometric domain (age) in orienteering (8). 100 m freestyle in Greek swimmers (males. aerobic endurance (males. This finding points at aerobic endurance as an essential factor for performance at this age. This is consistent with studies on young athletes (12. These values are higher than those obtained in performance prediction in 400 m freestyle in French swimmers (males and females. and the second strongest in males. and weightlifting (11). females. and 79% in volleyball (13). female 72%.59. 80% in Australian football (19). 94% in rugby (28). which showed similar or lower prediction values: 55% in ice hockey (9). 74%.4. 34) or higher than (males.15. R2 = .47. however. There are some studies.49. except sitting height in males (R2 = . R2 = . and speed endurance (males. prediction rates in sprint swimming (25 yd) are close to (males 75%. respectively. 82–84% in male handball (10). the results show high prediction performance values with a large percentage of variance explained by the MLR models developed (males 82%. females.39). These results are lower than those obtained in studies on other sports: 83% in rowing (35).146   Saavedra et al. an indicator of aerobic endurance which was validated with velocity at 4 mmol·L¯1 lactate as a reference criterion (27). which suggests that to minimize the effect of age on performance in the formative years. psychological domain in soccer (29). However. rugby (28). in which aerobic endurance was shown to be a strong predictive variable. With regard to DA. handball (10). With regard to the variables selected for our study. age and aerobic endurance (30-min test) entered the discriminant models in both sexes. 4). R2 = .34). 15). it should be remembered. rowing (35). R2 = . R2 = . variables discriminated among performance groups on 73% (males) and 68% (females) of all swimmers with six and four variables. however. 20%) in Australian swimmers of similar age and competitive level (1). was the strongest predictor of performance in females. Australian football (19). Despite the obvious difficulties this creates for categorization (26). speed endurance) entered the predictive models in both sexes. in agreement with a previous study (19). Age was the strongest predictor in males and the third in females. prediction of the best performance group was achieved at 94% in males and 71% in females. and 84% in weightlifters (11). technical domain in volleyball (13). The 30-min test. though in the first study the competitive level of the swimmers was lower.34). 18.

this has to be interpreted with caution since other factors.36–.001.918. These values are similar than those obtained in young U. females Rp=. In females.e. This argument has already been suggested as a possible explanation for different predictive models in males and females at this age (15).33) as in a previous report (R2=. 18). Rp=.Analysis of Performance in Young Swimmers 147 the notion that aerobic training should be one of the main training goals in young swimmers. are main predictors of performance as a continuous variable.50.54. In females though. This gives .61) were in fact highly correlated with swimming performance.898. faster and slower swimmers). although in those studies performance was categorized in only two groups (i. This may indicate that that training volume is only a key factor for performance in males. our results indicate that age and specific fitness. From an overall perspective. more even so in the females. The third common predictor for both sexes was the 6 × 50 m test swum at 1:30 at personal stroke. 46% of best swimmers correctly classified). Arm span has also been identified as a good predictor of performance in young swimmers using a MLR model (R2 = . females: 476 LEN points. particularly those related with aerobic and speed endurance. arm span. aerobic-anaerobic specific endurance also plays a role in determining performance. p < . in which technical variables reach high levels of correlation with performance even when controlling for age..12). except for females. partial correlations with chronological age as control variable (Table 2) showed that both aerobic endurance (males. With regard to the anthropometric assessment domain. this difference may be due to the better performance level of males in the sample as compared with the females (males: 508 LEN points. independently from age. the DA variables in males were height. Discriminant Analysis in Swimming The results show high performance prediction of the DA models both in males (73% of the whole sample and 91% of the best swimmers correctly classified) and in females (68% of the whole sample and 71% of the best swimmers correctly classified). 18). we need to take into account that both aerobic and speed endurance are also strongly correlated with age.. This test is considered an index of speed endurance (3) and suggests that mixed. the discriminant variables were age and sexual maturation (pubic hair.e. were not included in the model due to methodological reasons. and age (62% of best swimmers correctly classified).60). However.454 p < . larger stroke length). the swimming index was in males not as good as a predictor of performance (R2=. 21). most technical variables showed high partial correlations with performance (Rp=. which benefits swimming efficiency (i. since taller swimmers usually show a larger arm span. From the technical point of view. anthropometric (including sexual maturation) and technical variables are less relevant. and that general fitness. training volume per session (71% of best swimmers correctly classified) was the only variable entering the model. although at a somewhat lower level. However. females Rp=. we can see that in the sports background and training status domain. and speed endurance (males. whereas no variables were selected in females. swimmers of similar age (83%) and adult swimmers (88%. such as training intensity.55).01). Rp=. p = .S. On analyzing each of the assessments performed. The first two variables were also closely related in the current study (R = .

Finally. 6 × 50 m at 1:30 start at personal stroke. 30-min test. which suggests that the fastest females in 50 m are those who obtain the best performance also in longer distances at this age.. the relevance of turns at longer distances may be even higher. 50-m mean velocity.e. In females the discriminat variables were 30-min test. and 71% of best swimmers correctly classified). discriminant variables were hand dynamometry and abdominals in 30 s. all variables considered) the discriminant variables for males were age. is the main predictor of performance in this age category. In females. which informs on good body position while swimming. shoulder extension. The first of these variables underlines the importance of efficient turns. the three first variables were also selected by the MLR model. in the multidimensional evaluation (i. since quantitative analysis was made in a 50-m swim. p < .55. although with low discriminant power (35% of all swimmers correctly classified). the discriminant variables in males were turn velocity and stroke rate (63% of all swimmers correctly classified). Thus the relevance of this capacity in the formative years is to be acknowledged. Stroke rate points out the importance of swimming with the lowest cycle frequency to guarantee efficient swimming (20). and height (73% of all swimmers. Also body linear dimensions and shoulder flexibility seem to discriminate among best and not-do-good male swimmers. This last variable indicates general aerobic endurance which. In the technical analysis domain. several variables are correlated with age. 30-min performance also highlights the importance of aerobic endurance in swimming (14) at the formative stages. horizontal floating. As to specific fitness testing. Limitations Firstly. which appears to be the main single predictor of swimming . and shoulder extension (68% of all swimmers correctly classified).148   Saavedra et al. peripubertal age (24). an indication of the relevance of sexual maturation in swimming performance at young. together with horizontal floating (performance prediction: 66% of all swimmers correctly classified). Again. and (b) aerobic endurance and speed endurance are key factors for performance among young swimmers of this age. In females. the discriminant variables in males were 6 × 50 m at 1:30 start at personal stroke. was also closely related to specific aerobic endurance (R = . In general fitness testing in males. this is a strong indication that (a) chronological age..001). It is worthy to note that. and 94% of best swimmers correctly classified). 30-min performance were once more selected. A study carried out on swimmers of similar ages also selected this variable to predict performance (21). mean velocity in 50 m) was selected in the semiquantitative analysis (77% of best swimmers correctly classified). together with swimming speed and horizontal floatability. only second to swimming speed in 50 m.e. in spite of the relative homogeneity of the samples. and age (68% of all swimmers. given that drag is directly proportional to the cross sectional area of body (23). 30-min test. as discussed before. In females. also hand dynamometry and shuttle run endurance were selected (performance prediction: 46% of all swimmers correctly classified). aerobic endurance and age were selected in the DA model. only one variable (i. The first variable informs on speed endurance. a determining capacity in swimming performance in 100 and 200 m (14). in our study. arm span. and is in complete agreement with MLR results.

Secondly. The authors wish to thank James McCue for revising the English text and each and every one of the participants in the study. Therefore. however. considering the best three events would have given a better picture of the overall performance capacity of the swimmers. Thirdly. JC2007-00316). . Two practical implications may be derived: (a) in these ages competitions should be organized according to year of birth and not by age categories. reached similar or higher predictive power. and by identifying the best group of swimmers (DA: 94% males. outstanding chronological age and aerobic and speed endurance. We would like to acknowledge the work of the two anonymous reviewers who made the quality of the work being improved. The computation of partial correlations coefficients.. performance was assessed as the best score in each swimmers’ three fastest events.M. in which full specialization is not normally attained. to a certain extent. During completion of this paper J. while most swimmers were of peripubertal age. 24 (17 males. and (b) aerobic and speed endurance should be considered not only as capacities which guarantee long-term athletic development. allowed weighing this effect. the models have shown to be efficient in predicting performance. Finally. Likewise. the results show that this was not so. Perhaps using a different criterion would have changed our results. Most selected variables pertain to the anthropometric (particularly in males). both by detecting the swimmer’s individual potential (MLR: 82% of males’. Altogether. indirect estimations of physical and technical capacities and abilities. and not just in one. we must consider that the large number of subjects evaluated (n = 133) and the large number of tests performed (ninety). in most cases. ergospirometric measurements. and technical domains (particularly in females). Conclusions This study has developed multivariate models to predict swimming performance which. 7 females) were already at the adult sexual stage. but also as predictors of performance itself. in overall comparison with other sports. these results support the notion of the multidimensional nature of this sport. 71% females correctly classified). Acknowledgments This study was partially granted by the Spanish National Swimming Federation (Real Federación Española de Natación) and the Spanish Higher Sports Council (Consejo Superior de Deportes). the physical and technical assessments made are. however. Escalante was also a visiting researcher at the same University with partial grants awarded by the Autonomic Government of Extremadura (Junta de Extremadura) and Social European Funds (GRU07109-GRU08118). although we initially believed that this fact would influence performance. we took into consideration that the swimmers were in their formative years. and 84% females’ of variability in performance was explained by the model).e. biomechanical analysis) impractical. specific fitness. would have made more sophisticated measurements (i. Saavedra was a visiting researcher at the University of Ulster with a grant awarded by the Ministry of Education and Science of Spain (Ministerio de Educación y Ciencia. Y.Analysis of Performance in Young Swimmers 149 performance at this age category in both sexes.

Blanksby.. Sports Med. Hillsdale. Domrow. 1982. and J. 31:725–741. Kirwan. Gastin. NJ: Lawrence Erlbaum Associates. Lacour. Energy expenditure during front crawl swimming: predicting success in middle-distance events. Fry. 2007. San Diego: Karger. King. Physiological and anthropometric characteristics of Australian junior national. M. Deshaies. 20:861–866. Int. 2004. K. 2005. J. G. Fielding. In: Physical structure of Olympic athletes (part I).E. 10. Sci.L. 2007. P. and C. G. Costill. Georgieff. 35:59–66. and skill data to predict selection in a talent-identified junior volleyball squad.D. 9.D.K.R. 19. Ackland. 1979.. 36–50. Reilly.. and T. and novice volleyball players.D. R. Pavlicevic.J. 17. B. Modelo de estudio de la estructura condicional a través de un análisis multivariante enfocado a la detección de talentos en jugadores de balonmano. 1985. 12:175–191. 1986. Phys. B. J. J. 1999. Chiu. Fry. Exerc. 27:57–74. Haljaste. B. anthropometric. The use of physical fitness scores and anthropometric data to predict selection in an elite under 18 Australian rules football team. Heath. Physiques Sport. 6:266–270. pp. J. Newell (Eds. Am. B. and B. Keogh. 1985. J Sports Med Phys Fitness. 11. Jürimäe. 3. Anthropol. 8:23–28.. 7. Int. D. 16. Carter.). 1988. J.C. Anthropometric and performance variables discriminating elite American junior men weightlifters. 63:213–219.L. Roberts and K. Ciroslan. Porter. Body composition of Montreal Olympic athletes. Cicchella.E.). Eur. Hum. D. A. The relationship between anatomical characteristics and swimming performance in state age-groups championship competitors.Z. 2:125–133.150   Saavedra et al. J. and T. and biomechanical parameters in young swimmers.E. J Sports Sci.L. J. 2001. Res.N.C. 1993. D. 1967.A. Sci. and M. B. A psycho-biological profile of individual performance in junior hockey players. J. Strength Cond. physiological. Nassis.P. .H.L. J. L. 14. J. Analysis of swimming performance from physical. and D. G. 12. Bloomfield. Res.A.. Geladas. 2005. and N. Tech. Tests spécifiques d´évaluation du nageur.R. A multivariate analysis of kinanthropometric profiles of elite female orienteers. Act. Sports Med. A modified somatotype method.. 2. 18. J. Cestas: Fédération Française de Natation. Georgieff. 8. LeRoux. 2:30–36. Strength Cond. D. J. J. Padilla.. Swimming Res. Mov. Cazorla. In: Psychology of motor behavior and sport – 1978. Chatard. Schilling and L. Carter (Ed. Pediatr. 2006. Rodríguez. 2007. U. Anthropometric and physical traits affecting sprint swimming performance in young swimmers. Thiffault. Jürimäe. Sci. Influence de la morphologie et de l’entraînement sur la performance en natation. 25:1337–1144.. N. F. J. Kovaleski. C. 6. Handbook for the Eurofit Test of Physical Fitness. 13. Sport. The influence of anthropometrical and flexibility parameters on the results of breaststroke swimming. Creagh. 107–116. et al. Roma: CONI. 1987.. T. A. Energy system interaction and relative contribution during maximal exercise. Pargman. Cohen. Carter. Sports Med. 26:139–144.. G. and J. Council of Europe.A. Gabbett. and S. Champaign. 1988..C. G. Cazorla. pp. 21:902–908. Anz. and F.. Jagomägi. state. Ponchard. References 1. S. Vila. Fernández. M. J. The use of physiological. 4.. J.. Statistical power analysis for the behavioural sciences. 5. Med. Gabbett. and T.H. 19:70–81. 15. J. Anthropol. T. IL: Human Kinetics.

Pediatr. Tanner. 6:59–68. Liesen. 21. and M. 41. and W.B. Characteristics of youth soccer players aged 13-15 years classified by skill level. and A. and N.. W. 35. Vilas-Boas. and F. and D. Biostatistics.V.M. Scand J Med Sci Sports.L. G. 1994. 2007. Solanellas. 2000.P. Zeni´ c. A. 33. Sports Sci. J Sports Med Phys Fitness.L. Swimming fastest: a comprehensive guide to the science of swimming. 345-349. Rodríguez. Costill. J. MacDougall.). 1982. 36. 25. A. and F. Port. Ontario: BC-Decker. Hollmann. and H. 16:691–699. 1982. J Sports Sci. Identifying and developing rugby ta among 10-year-old boys: a practical model.J. Komi. Soares. 26.J.. 39. good and elite rowers. Green (Eds.P. L.J. 1991. R. and H. Franks. K. L. 1996. Marfell-Jones. F. R. 91:94–99. Apreciação de consistência dos resultados produzidos por avaliadores com experiência e formação similiares.. force production and anthropometric variables. Oliveira.G.. 11:87–92.. Non linear anthropometric predictors in swimming. Exerc. Malina. Zubcevi´ c. Physiol. R. Sports Med. Fernandes. Hughes. Maglischo.V. 31. Bouchard. Sokolovas.L. Physiological determinants of sprint swimming performance in children and young adults. The use of neural network technology to model swimming performance. and P.. Relationship between swimming velocity and lactic concentration during continuous and intermittent training exercises. Munich: Ligue Européene de Natation and German Swimming Federation. 29..Analysis of Performance in Young Swimmers 151 20. Wenger.W.E. 22. 24.M. and physical activity. Champaign.M. Keskinen.M. 1978. Carmo. J. Sci. 1998. Eur. Williams.J.L. Costa. pp. Ottawa: Canadian Association of Sports Sciences. Piennar. R. pp. 1962.M. pp. In Book of Abstracts. IL: Human Kinetics. Sports Med. S. and arm and leg speed performance of male and female swimmers as predictors of swim speed. 31:803–809. Madsen. R. 40. 18:695–702. Kilka. Relationship between power and sprint freestyle swimming. J. Malina..P.261-267. and J.. C.. Sci. Sekuli´ c. J. Int. A multidisciplinary approach to talent identification in soccer.). 2004. 32. H. Aroso. and J. and W. 1995. 2007.L. Eynde. J Sports Sci. 28. Porto. J. B. B. Streiner. Fernandes. First Annual Congress of the European College of Sport Science.J. In: Physiological testing of the elite athlete. J. M. 1:22–32. 1989...V. Maximum velocity swimming: Interrelationship of stroking characteristics.R. Analysis of USA swimming’s all time top times. Growth. Sharp. LEN. Troup. Soares. Comparative performance tables for swimming (1997-2000).G. Coll. Nevill. Spamer. Sports Exerc.. G. 13:377–385. and P. R. Smith. Silva. 27.D. Nice. Steyn. R. P. Antropol.. Mader. Br. 38.. A. et al.. Discriminant analysis of biomechanical differences between novice. 34. Multidisciplinary evaluation and performance prediction of tennis players of different age and sex categories. J. Silva. . and S. 75–115. 6:315–317. In Proceedings of the IV Congress of Notational Analysis of Sports. 1985. 2006. 30. J. J.A.S. J. Tavares (Eds. and W. Kinanthropometry. Champaign. 23. T. 2000. 2003. 2001. Oxford: Blackwell. Vilas-Boas. Santos-Silva. Growth and adolescence. J. 14:53–56. 37. Avaliação qualitative da técnica em Natação. 41:290–295. S. Port. Reilly. 6:74–77. Thorland. O. E. maturation. D. A. Cumming.M. and C. A. Prediction of sprint triathlon performance from laboratory tests. Ribeiro. Anthropometric measurements.A.M. and D. Norman. France: ECSS. N. Smith.D. Sports Sci. IL: Human Kinetics.J. J Sports Sci Med. Appl. 6:117–125.. J Sports Sci. Portugal: Center for Team Sports Studies (Faculty of Sports Sciences and Physical Education).M. 1998. M. Van Schuylenbergh. A. Hespel. H. Ross.J. Med. 2007. Tilli. The importance of technical evaluation on the training control in swimming – Subjective analysis of the swimming technique of a juvenile group of swimmers from the Association of the North of Portugal. Olbrecht. Spinks.P. 1996.M.J. R. J. 18:153–168.