Abstract
Physical activity (PA) is critical for maximizing bone development during growth. However, there is no consensus on how well existing PA measurement tools predict bone strength. PURPOSE: The purposes of this study were to compare four methods of quantifying PA (pedometer, 3-d PA recall (3DPAR), bone-specific PA questionnaire (BPAQ), and past year PA questionnaire (PYPAQ)), in young girls and to evaluate their ability to predict indices of bone strength. METHODS: A total of 329 girls aged 8-13 yr completed a pedometer assessment, the 3DPAR, the BPAQ, and a modified PYPAQ. Peripheral quantitative computed tomography was used to assess bone strength index (BSI) at metaphyseal (4% distal femur and tibia) sites and strength-strain index (SSI) at diaphyseal (femur = 20%, tibia = 66%) sites of the nondominant leg. Correlations and hierarchical multiple regression were used to assess relationships among PA measures and indices of bone strength. RESULTS: After adjusting for maturity, correlations between PA measures and indices of bone strength were positive, although low (r = 0.01-0.20). Regression models that included covariates (maturity, body mass, leg length, and ethnicity) and PA variables showed that PYPAQ score was significantly (P < 0.05) associated with BSI and SSI at all sites and explained more variance in BSI and SSI than any other PA measure. Pedometer steps were significantly (P < 0.05) associated with metaphyseal femur and tibia BSI, and 3DPAR score was significantly (P < 0.05) associated with metaphyseal femur BSI. BPAQ score was not significantly (P > 0.05) associated with BSI or SSI at any sites. CONCLUSIONS: A modified PYPAQ that accounts for the duration, frequency, and load of PA predicted indices of bone strength better than other PA measures.