Hip BMD Change in Aging Men: A Non-Linear Association with Lean Mass Change Explains More Variation than Does Weight Loss or Fat Mass Change
Osteoporosis – Epidemiology (Clinical)
Aging and Other Disorders of Bone and Mineral Metabolism (Clinical)
Oral Presentations, Presentation Number: 1091
Session: Concurrent Oral Session 16: Osteoporosis Epidemiology II
Sunday, September 13, 2009 10:00 AM - 10:15 AM, Colorado Convention Center, Four Seasons Ballroom 1-3
* , UNITED STATES, , US, , UNITED STATES, , UNITED STATES
Purpose: Associations between weight loss and loss of bone mineral density (BMD) have been reported. However, the relationships between longitudinal changes in BMD and changes in soft tissue composition have received little attention.
Methods: In the MrOS study, 2487 community-dwelling men aged ≥65 yrs had baseline and repeat assessment (mean 4.7 yrs later) of height; weight; and DXA-based measures of total hip BMD, lean mass and fat mass. We considered four measures: total mass and the mutually exclusive factors that comprise total mass (lean, fat, and bone mineral content). Appendicular measures were the sum of arm and leg masses. Changes in total hip BMD, appendicular lean mass (ALM), appendicular fat mass (AFM), total body lean mass (TBL), and total body fat mass (TBF) and their correlations were calculated. Standardized coefficients and partial R2 from multiple linear regression were used to assess the contribution that change in each mass variable made to the variation in BMD loss. All models were adjusted for age and clinical site. Potential confounders included race, alcohol use, smoking, medical history, medication use, self-reported health, and physical activity.
Results: Proportionately, the largest declines were seen for ALM, with negligible changes in TBL (Table). Both AFM and TBF increased; weight and BMD had similar declines. BMD change correlated positively with ALM change (age- and weight-partial r = 0.18) and negatively with TBF change (age- and weight-partial r = -0.13). The best-fitting multivariate model with total hip BMD change as the dependent variable included age, clinical site, ALM change, and TBF change. The relationships between a) ALM and BMD change and b) TBF and BMD change were nonlinear: loss in mass correlated with loss in BMD but gain in mass was not associated with gain in BMD. This multivariable model accounted for 20.2% of the variation in total hip BMD change, with ALM change accounting for more variation than TBF change. Estimates were unaltered by potential confounders.
Conclusions: Decline in lean mass and increase in fat mass were each associated with loss of BMD, but change in BMD was more strongly associated with change in lean mass, potentially as a result of shared biological processes and/or biomechanical interactions. Gains in lean or fat mass were not associated with gain in total hip BMD in this aging group of men.
* Presenting Authors(s):
, UNITED STATES