Tracking within-athlete changes in whole body fat percentage in wheelchair athletes

Purpose: To evaluate tracking of within-athlete changes in criterion measures of whole-body fat percentage (dual energy X-ray absorptiometry; DXA) with skinfold thickness measures (Σ 4, 6, or 8) in wheelchair basketball players. Methods: This longitudinal study tracked body composition of sixteen international wheelchair basketball players at 5 time points over a 15-month training/competition period. The primary outcome was DXA-derived whole-body fat percentage (BF%), with Σ 4, 6, or 8 skinfolds (mm) as the predictor variable. Data were analysed using a linear mixed model with restricted maximum likelihood (random intercept, with identity covariance structure) to derive the within-athlete prediction error for predicting criterion BF% from Σ skinfolds. This prediction error allowed us evaluate how well a simple measure of the Σ skinfolds could track criterion changes in BF%; that is, we derived the change in Σ skinfolds that would have to be observed in an individual athlete to conclude that a substantial change in criterion BF% had occurred. All data were log-transformed prior to analysis. Results: Σ 8 skinfolds were the most precise practical measure for tracking changes in BF%. For the monitoring of an individual male wheelchair basketball player, a change in Σ 8 skinfolds by a factor of greater than 1.28 (multiply or divide by 1.28) is associated with a practically meaningful change in BF% (≥1 percentage point). Conclusions: Σ 8 skinfolds can track changes in BF% within individual wheelchair athletes with reasonable precision, providing a useful field monitoring tool in the absence of often impractical criterion measures.