Purpose and methods: Body weight is the most extensively studied correlate of bone mass and is widely used as a covariate in statistical evaluations of bone mineral parameters. Lean tissue mass (LTM) also correlates with bone mass. We evaluated the correlations linking each of these two parameters with bone mineral content and bone mineral density in 41 prepubertal girls, including ten swimmers, 18 gymnasts and 13 nonathletes. Lean tissue mass, bone mineral content and bone mineral density were measured using dual-energy X-ray absorptiometry (Hologic QDR-1000/W; Hologic Inc., Waltham, MA, USA). Forward stepwise multiple regression was used to evaluate correlations linking bone mineral content or density (the dependent variables) to body weight or lean tissue mass (the independent variables).
Results: Body weight and lean tissue mass showed strong correlations with all bone mineral content and density measurements in the simple linear regression analysis, with lean tissue mass yielding the highest Pearson's correlation coefficients. In the multiple regression model, lean tissue mass consistently explained the largest proportion of the variance, whereas body weight had little influence or was eliminated from the model. The slopes of the regression lines of bone mineral content or density on body weight were significantly steeper in the subgroup of gymnasts (P < 0.001), whereas the slopes of the regression lines of bone mineral content or density on lean tissue mass were significantly less steep in the swimmers (P < 0.05).
Conclusion: Our data indicate that lean tissue mass is a significant predictor of bone mass in prepubertal girls and explains a larger part of the variance of bone mineral content and density than body weight. Use of body weight as a covariate in studies of bone mineral density may lead to erroneous results in prepubertal girls.