Purpose: This study analyzed prospective associations between distinct trajectories of objectively measured physical activity (PA) and late adolescent bone parameters and explored the mediating effects of lean soft tissue, a surrogate of muscle mass to associations. Methods: Physical activity was measured by accelerometry starting at age 5 yr and continuing at 8, 11, 13, 15, and 17 yr in approximately 524 participants from the Iowa Bone Development Study. Sex-specific group-based trajectory modeling was used to construct developmental trajectories of moderate- and vigorous-intensity PA (MVPA) from childhood to late adolescence. At age 17 yr, proximal femur bone mineral density (aBMD) was assessed by dual X-ray energy absorptiometry, and its distribution was calculated by aBMD ratios. Specific geometric measures of the proximal femur were assessed using hip structural analysis. Results: A significant portion of the total effect of MVPA from age 5 to 17 yr on bone parameters at age 17 yr was explained by an increase in leg lean soft tissue in both sexes. For males and females, indirect effects were observed on the total and all regional proximal femur aBMD, and on the ratio between the inferomedial and superolateral neck aBMD. The effect on the ratio between the trochanter and the total proximal femur was specific to females, whereas the effect on the hip axis length was specific to males. Direct effects of MVPA on aBMD were identified only in males. Conclusions: Using robust mediation analysis, this is the first study addressing the indirect effect (through muscle) of PA across childhood and adolescence on proximal femur bone parameters. To improve bone health at the proximal femur, the results suggest PA interventions during growth that increase muscle mass, particularly in females.
The purpose of this study was to analyse associations between lean soft tissue (LST), a surrogate of skeletal muscle mass and key fracture-related geometric characteristics of the proximal femur. Moreover, we examined the role that muscle played on the proximal femur geometry in response to physical activity (PA). Participants were 83 young adults. Leg LST (exposure) was assessed by dual energy X-ray absorptiometry (DXA). Proximal femur geometry was derived from a left hip DXA scan. Geometric variables (outcomes) included the femoral neck axis length (FNAL), the femoral neck width (FNW), the neck–shaft angle and FNW|FNAL (an index of robustness). PA was evaluated by accelerometry. Linear regression was used to analyse relationships. Additional exposure variables included body height and mass. In males, leg LST explained 17.4% of variation in FNAL (P < 0.001) and 15% in FNW (P = 0.015). In females, it explained 8.8% of the variance in FNAL (P = 0.020). Associations remained significant in males, but not in females, when vigorous PA was added to the models. These results suggest that public health approaches to promote PA may be particularly important in females since vigorous PA seems to convey advantages in femur geometry and consequently in bone strength.