Positive association of total protein intake and bone mineral density (BMD) in women from the Framingham Offspring Study

Abstract

ObjectiveThe effect of protein on bone is controversial; further calcium intake (CA) may modify protein's effect. We evaluated the cross‐sectional association of % total energy from protein intake (TP%) with bone mineral density (BMD) at femoral neck (FN), trochanter (TR) and L2‐L4 spine (LS) in the Framingham Offspring Study, also whether CA modified any association.Methods1639 women and 1280 men completed a food frequency questionnaire and had BMD measured in 1991–95 or 1995–98. We used multivariable linear regression to examine the association of TP% with each BMD, adjusting for covariates.ResultsMean age was 60 years (SD:9.3) in women and 61 years (SD:9.4) in men. In women, positive associations were seen between TP% and each BMD [FN: β=0.0012(SE=0.0009), P=0.04, TR: β=0.0002(SE=0.0008), P=0.01, LS: β=0.002(SE=0.001), P=0.04]. No associations were seen in men (P range: 0.28–0.72). We stratified (a priori) by CA: women with <800mg/d CA showed strong associations [FN: β=0.002(SE=0.001), P=0.07, TR: β =0.002(SE=0.0009), P=0.04; LS: β=0.003(SE=0.001), P=0.07]. In women with ≥ 800mg/d CA the positive associations were not significant (P range: 0.27–0.42). No associations were seen in men (low CA group: P range= 0.15–0.50; high CA group: P range= 0.43–0.91).ConclusionIn this study higher protein intake was positively associated with BMD at multiple sites in middle aged women but not men. Longitudinal studies are needed especially to examine the role of CA in these associations.Grant Funding Source: NIH R01 AR053205, NIH R01 AR/AG 41398; and N01‐HC‐25195