The association between body fat distribution and bone mineral density: evidence from the US population

Objective To investigate the association between different body fat distribution and different sites of BMD in male and female populations. Methods Use the National Health and Nutrition Examination Survey (NHANES) datasets to select participants. The weighted linear regression model investigated the difference in body fat and Bone Mineral Density (BMD) in different gender. Multivariate adjusted smoothing curve-fitting and multiple linear regression models were used to explore whether an association existed between body fat distribution and BMD. Last, a subgroup analysis was performed according to age and gender group. Results Overall, 2881 participants were included in this study. Compared to males, female participants had lower BMD (P < 0.05) and higher Gynoid fat mass (P < 0.00001), while there was no difference between Android fat mass (P = 0.91). Android fat mass was positively associated with Total femur BMD (Males, beta = 0.044, 95% CI = 0.037, 0.051, P < 0.00001; Females, beta = 0.044, 95% CI = 0.039, 0.049, P < 0.00001), Femoral neck BMD (Males, beta = 0.034, 95% CI = 0.027, 0.041, P < 0.00001; Females, beta = 0.032, 95% CI = 0.027, 0.037, P < 0.00001), and Total spine BMD (Males, beta = 0.036, 95% CI = 0.029, 0.044, P < 0.00001; Females, beta = 0.025, 95% CI = 0.019, 0.031, P < 0.00001). The Gynoid fat mass, subgroup analysis of age and ethnicity reached similar results. Conclusion Body fat in different regions was positively associated with BMD in different sites, and this association persisted in subgroup analyses across age and race in different gender.