[Untitled]

To investigate whether body morphology, obesity and its long time evolution were associated with lumbar and femoral bone mineral density (BMD) in premenopausal women of the same age. Design: Cross-sectional study. Subjects: 72 healthy premenopausal women born in 1950 (42 years) with a regular physical activity. Measurements: BMD measured by dual-X-ray absorptiometry (DEXA) at lumbar spine and proximal femur; body weight, body mass index (BMI), BMI at 20 years (BMI-20), increase in BMI since age of 20 (BMI->20), body circumferences (breast, waist, hip) and their ratios (WHR, BHR, WBR), smoking and alcohol intake. Results: Lumbar spine BMD did not correlate with any anthropometric measurement. Femoral BMDs correlated positively with weight, BMI, BMI-20, breast, waist, WHR and BHR. The BMI-20 explained the 5% and the current BMI the 13% of variance of total femur BMD. After adjustment for weight or BMI, breast circumference and BHR remained significantly correlated with all femoral BMDs sites except neck. Weight was the best predictor for neck BMD (R2 = 0.08; p < 0.02), and BHR for Ward's triangle (R2 = 0.12; p < 0.01) and trochanter (R2 = 0.10; p < 0.001). Alcohol intake, cigarette smoking, and age of menarche were not related to BMDs. Conclusion: In premenopausal women of the same age, lumbar spine BMD was not associated with any anthropometric measurement. Greater BHR and its long time of evolution may be determinants of greater femoral BMD (trabecular), whereas body weight may be determinant of femoral neck BMD (cortical). Further studies are needed to determine whether large breast to hip ratio may be considered as a protective factor for femoral osteoporosis.