Age-related decline in BMD is associated with changes in lean mass and modulated by sex and race

Abstract

Objective Studies have shown that the association between weight and BMD is primarily due to lean mass. Our objective was to determine whether the age-related decline in BMD was associated with changes in body composition (lean/fat mass) and whether it is influenced by sex and race/ethnicity. Study design This is a cross-sectional study of 76,780 adult participants (age 20-85 years) from the National Health and Nutrition Examination Survey (NHANES) 1999-2006 who had whole body dual energy X-ray absorptiometry (DXA) that provided soft tissue measurements for the total body lean and fat mass and bone measurements for the lumbar spine (LS), total hip (TH) and femoral neck (FN). We performed a multivariate linear regressions stratified by sex with BMD of the 3 skeletal sites as outcome, age as predictor and total body lean and fat mass, and race categories [African American (AA), Caucasian American (CA) and Mexican American (MA)] as covariates. For all data presented p-values were Results In both sexes and across age groups BMD was higher in AA than in other groups. In males, LS BMD did not change significantly with age. TH BMD was stable until age 40 and declined thereafter at a rate of approximately 5% per five years. There was a positive correlation between decline in BMD and lean mass (3% decrease in BMD per 10 kg decrease in lean mass), while adding fat mass did not have significant effect. Findings were similar at the FN (3.3% decrease in BMD per 10 kg decrease in lean mass) but the decline started at age 20. In females, the rate of decline in BMD and lean mass did not differ between perimenopause (45-60 years) and post menopause (over 60 years). LS did not change significantly with age. TH BMD was stable until age 45 and declined thereafter at a rate of approximately 5.1% per five years. There was a positive correlation between the decline in BMD and lean mass (3.2% decrease in BMD per 10 kg decrease in lean mass), while adding fat mass did not have a significant effect. The observation was similar in FN with approximately 5% decrease in BMD per five years after age 45 and 3.5% decrease in BMD per 10 kg decrease in lean mass. Conclusion Decline in BMD with age was associated with the decrease in lean mass in both sexes and all 3 racial/ethnic groups. Based on this cross-sectional study we cannot determine whether the decline in lean mass causes decrease in BMD or whether the reduction in both BMD and lean mass are due to other factors associated with aging. Nevertheless, when designing interventions for optimizing bone health in the elderly it may be prudent to preferentially consider those that also improve total body lean mass.