A multivariate analysis of performance in young swimmers

Area of Science:

• Sports Science
• Human Performance
• Biomechanical Analysis

Background:

• Predicting athletic performance in youth swimmers is crucial for talent identification and training optimization.
• Multidimensional assessments offer a comprehensive approach to understanding factors influencing swimming success.

Purpose of the Study:

• To develop and validate multivariate models for predicting competitive swimming performance in male and female youth swimmers.
• To identify key anthropometric, fitness, and technical variables that predict swimming performance.

Main Methods:

• A test battery including questionnaires, anthropometry, general and specific fitness tests, and technique analysis was administered to 66 male and 67 female swimmers.
• Multiple linear regression and discriminant analysis were employed to build predictive models for swimming performance (LEN scores).

Main Results:

• Multiple linear regression models explained 82.4% of performance variability in males and 84.5% in females, highlighting age, anthropometry, and specific fitness tests.
• Discriminant analysis achieved high classification accuracy (94.1% for males, 71.0% for females) using distinct sets of predictive variables.
• Chronological age emerged as the primary predictor of performance in this age group.

Conclusions:

• Multidimensional assessments effectively predict swimming performance in young athletes.
• Predictive variables differ between genders, emphasizing anthropometrics for males and technique/aerobic capacity for females.
• Competitions for these age groups should be based on year of birth rather than broad age categories to ensure fair competition.