Data are from 1378 women aged 75 years and older recruited in the Toulouse EPIDémiologie de l'OStéoporose cohort. Body weight and composition (dual energy X-ray absorptiometry), knee extension strength, and physical activity habits were assessed. The INDIRECT macro for SPSS (SPSS Inc, Chicago, IL) was used to determine total and specific indirect effects of physical activity on UGS through body weight (BW) and composition as well as muscle mass and strength.
Knee extension strength [95% confidence interval (CI) 0.0091, 0.0231] and BW (95% CI 0.0037, 0.0412), but not fat mass or muscle mass, were mediators of the relationship between physical activity and gait speed. However, muscle mass significantly mediated the relationship between physical activity and muscle strength (95% CI −3.0722, −0.6265).
Results suggest that muscle strength and BW are the most significant mediators of the relationship between physical activity and UGS. Although the specific effect of muscle mass appears negligible at first sight, its role should not be overlooked given its impact on muscle strength.