Comparison of the effect of physical activity on the relationship of somatotype components and BMI to body composition.

Introduction: The endo- and mesomorphy components of somatotype and the BMI are assumed to relate to body composition constituents e.g. adiposity and lean body mass. Although physical activity may change body shape and composition, it is not clear what the influence of physical activity may be on the relationship of somatotype and BMI to these constituents. Purpose: To compare the effects of self-reported physical activity levels on the relationships of somatotype components and of BMI to body composition constituents. Methods: A total of 54 women and 77 men were enrolled in this study. Both women and men were divided in two activity level groups: sedentary and sportive. Anthropometric measures allowing for the calculation of somatotype according to the Heath-Carter method were taken. Body composition was also assessed by using dual energy X-ray absorptiometry. Pearson correlation coefficients and linear regression analysis were used to calculate the variance explained by the components of somatotype and BMI in body composition constituents. The significance of the difference between correlation coefficients of the sedentary and the sportive groups was calculated using Fisher r to z transformations. Results: In the sedentary group BMI explained 67% and 56% of the variance in relative adiposity (p<0.001), while endomorphy explained 70% to 77% of the variance in relative adiposity (p<0.001) in women and men respectively. In the sportive group BMI explained 24% and 29% of the variance in relative adiposity (p<0.01), while endomorphy explained 60% to 75% of the variance in relative adiposity (p<0.001) in women and men respectively. Conclusion The results of this study suggest that endomorphy is a better predictor of body composition constituents than BMI, in the physically sportive population in particular. The use of BMI remains a dubious indicator of body composition constituents and should therefore be avoided in the assessment of nutrition and/or health status in physically sportive adults.