Two measures of physical activity as predictors of bone mass in a young cohort

Objective: To compare the association of two measures of physical activity with bone mass in healthy children and young adults, as part of a larger st udy on bone mineral acquisition in youth. Design: Cross-sectional observation study. Setting: General community, outpatient study. Participants: Subjects included 103 non-Hispanic white female (n = 54) and male (n = 49) healthy volunteers aged 9 to 25 years. Main Outcome Measures: Self-reported physical activity was measured by a 3- day activity diary of all activities and a questionnaire designed to captur e recreational activities throughout the year. Activity was expressed as ho urs per week of weight-bearing and non-weight-bearing activity. Bone mass a t the hip, spine, and whole body was measured by dual x-ray absorptiometry. Results: The activity measures were not well correlated with each other. In males, weight-bearing and non-weight-bearing activity reported in 3-day di aries was positively associated with bone mass at the hip, spine, and whole body (p < 0.05). Among females, only weight-bearing activity measured by t he yearly questionnaires was significantly positively associated with bone mass (p < 0.05). In males and females, weight-bearing activity was more hig hly correlated with bone mineral than was non-weight-bearing activity. In a ddition, the associations between activity and bone mass varied by skeletal site. Conclusions: The association between physical activity and bone mass varied both in direction and in significance depending on the physical activity i nstrument used. Gender differences were observed in the associations betwee n specific activity instruments, type of activity (weight-bearing and non-w eight-bearing), and bone mass at different skeletal sites. Variability asso ciated with the two physical activity measures may contribute to discrepant findings in this study and in the literature.