Bone mineral density in adolescent female athletes: relationship to exercise type and muscle strength.

PURPOSE This study investigated the influence of different exercise types and differences in anatomical distribution of mechanical loading patterns on bone mineral density (BMD) in elite female cyclists, runners, swimmers, triathletes, and controls (N = 15 per group). Associations between leg strength and BMD were also examined. METHODS Areal BMD (g x cm(-2)) was assessed by duel-energy x-ray absorptiometry (DXA) (total body (TB), lumbar spine (LS), femoral neck (FN), legs, and arms). Right knee flexion and extension strength was measured using a Cybex Norm isokinetic dynamometer at 60 degrees x s(-1). RESULTS Runners had significantly higher unadjusted TB, LS, FN, and leg BMD than controls (P < 0.05); higher TB, FN, and leg BMD than swimmers (P < 0.05); and greater leg BMD than cyclists (P < 0.05). Absolute knee extension strength was significantly (P < 0.01) correlated (0.33 < or = r < or = 0.44) with TB, FN, LS, and leg BMD for all groups combined. Weaker but still significant correlations (0.28 < or = r < or = 0.33) existed for normalized (per leg lean tissue mass) knee extension strength and all BMD sites, except FN BMD. There were no significant correlations between absolute or normalized knee flexion strength and any of the BMD variables. Absolute knee extension strength was entered as the second independent predictor for LS and leg BMD in stepwise multiple linear regression analysis (MLRA), accounting for increments of 4% and 12%, respectively, in total explained variation. CONCLUSION We conclude that running, a weight bearing exercise, is associated with larger site-specific BMD than swimming or cycling, that the generalized anatomical distribution of loads in triathlon appears not to significantly enhance total body BMD status, and that knee extension strength is only a weak correlate and independent predictor of BMD in adolescent females.