Anthropometry fails in classifying bone mineral status in postmenopausal women

This study tested two hypotheses: (1) that simple anthropometric parameters can be used to identify patients at risk of decreased bone mineral content and (2) that an inverse relationship exists between waist:hip ratio (WHR) and bone mineral density (BMD). Bone mineral content (BMC) and BMD were eva luated by dual-energy X-ray absorptiometry in 1873 free-living women. Of th ese, 1819 (97%) were post-menopausal. One thousand and thirteen women (54%) had normal BMD, 705 (38%) osleopenia and 155 (8%) osteoporosis. Body weigh t (Wt), body mass index and arm muscle and fat areas were significantly low er in osteoporotics than osteopenics (p < 0.0001) and in these latter than controls (p < 0.0001). However, values of WHR were similar in all groups (p = ns). Body weight was the anthropometric parameter better correlated with BMC (rho = 0.650, p < 0.0001) and only Wt and age were identified as signi ficant predictors of bone mineral status (normal-BMD/osteopenic/osteoporoti c) at polytomous logistic regression (p = 0.0001 for each). However, Wt cou ld not be employed as an indicator of bone mineral status at the individual level because of high variations in BMC for the same lever of Wt. Under- ( <5th percentile) and normal-Wt (5th-95th percentile) women had the same fre quency of osteopenia (39%) while it was lower in over-Wt (> 95th) women (13 %). The frequency of osteoporosis was higher in under- than normal-Wt women (37 vs 7%) and none of the over-Wt women had osteoporosis. This study show s that: (1) simple anthropometric measurements cannot be used to select sub jects at risk of decreased BMC and, (2) BMD does not vary with WHR.