Bone mineral density and hip structural analysis in postmenopausal Thai women: a cross-sectional study

Abstract

Hip Structural Analysis (HSA) provides information on proximal femoral geometry beyond conventional bone mineral density (BMD) measurements. However, data regarding bilateral hip geometry and its clinical correlates in Southeast Asian postmenopausal women remain limited. This study evaluated bilateral differences in HSA-derived parameters and examined associations between clinical characteristics and hip structural geometry in postmenopausal Thai women. A retrospective cross-sectional study was conducted in 1,096 postmenopausal Thai women (aged 66.26 ± 9.06 years) who underwent bilateral hip DXA examinations. HSA-derived parameters, including cross-sectional area (CSA), average cortical thickness (ACT), section modulus (SM), and buckling ratio (BR), were obtained at the narrow neck (NN), intertrochanteric (IT), and femoral shaft (FS) regions. Bilateral differences were evaluated using paired comparisons. Multivariable linear regression analyses were performed to examine associations between age, body mass index (BMI), years since menopause, and HSA-derived parameters. Multicollinearity was assessed using correlation analyses and variance inflation factors (VIFs). Total hip BMD did not differ significantly between sides, whereas several regional BMD and HSA-derived parameters demonstrated significant bilateral asymmetry. In multivariable analyses, higher BMI was consistently associated with greater ACT, CSA, and SM and with lower BR across anatomical regions. Increasing age was associated with lower ACT, CSA, and SM, whereas longer duration since menopause was associated with reduced cortical thickness and higher buckling ratio. Model explanatory power ranged from 13.5% to 33.9%. Correlation analysis demonstrated a strong association between age and years since menopause ( r = 0.931), although VIF values remained below conventional thresholds for severe multicollinearity. Significant bilateral differences in proximal femoral geometry were observed among postmenopausal Thai women despite similar total hip BMD values. Higher BMI was associated with greater cortical thickness, larger cross-sectional area, higher section modulus, and lower buckling ratio, whereas increasing age and longer duration since menopause were generally associated with less favorable structural measures. HSA-derived parameters may provide complementary information regarding hip structural characteristics beyond conventional BMD assessment. Further longitudinal studies are warranted.