The Obesity-Bone Paradox in Postmenopausal Women: Divergent Associations of BMI with Bone Mineral Density and Trabecular Bone Score in a Real-World DXA Cohort.

Background and

Objectives

Body mass index (BMI) has a complex relationship with skeletal health. Higher BMI is often associated with higher bone mineral density (BMD), but this apparent densitometric advantage may not fully reflect bone microarchitecture. Trabecular bone score (TBS) may provide complementary information in overweight and obese postmenopausal women. This study evaluated the associations between BMI, BMD, and TBS in postmenopausal women and assessed whether increased BMI is associated with discordant skeletal phenotypes characterized by non-osteoporotic densitometric status but impaired trabecular microarchitecture. Materials and

Methods

This retrospective observational study screened 110 DXA/TBS reports from patients evaluated in a tertiary DXA center. Inclusion criteria were female sex, postmenopausal status, available BMI or sufficient anthropometric data for BMI calculation, valid lumbar spine DXA, at least one valid hip DXA assessment, and valid lumbar TBS assessment. Four cases were excluded: two male patients and two non-menopausal female patients. The final analytical cohort included 106 postmenopausal women. Lumbar spine and hip BMD, T-scores, TBS, Bone Resilience Index, anthropometric variables, and menopausal variables were extracted from DXA/TBS reports. BMI was analyzed both continuously and by category: under/normal-weight, overweight, and obesity. Correlation analyses, group comparisons, and linear regression models were used to evaluate the relationship between BMI and skeletal parameters. Discordant skeletal phenotypes were defined descriptively as non-osteoporotic DXA findings combined with degraded TBS.

Results

The final cohort had a mean age of 65.2 ± 7.6 years and a mean BMI of 27.3 ± 4.5 kg/m2. BMI was positively associated with lumbar spine BMD (r = 0.392, p < 0.001), minimum femoral neck BMD (r = 0.382, p < 0.001), and minimum total hip BMD (r = 0.508, p < 0.001), but negatively associated with TBS (r = -0.307, p = 0.001). Mean TBS decreased across BMI categories, from 1.283 ± 0.084 in the under/normal-weight group to 1.255 ± 0.088 in the overweight group and 1.199 ± 0.109 in the obesity group. In the fully adjusted regression model including BMI, age, years since menopause, lumbar spine BMD, and lowest hip T-score, BMI remained an independent negative predictor of TBS (β = -0.0126, p < 0.001). Discordant skeletal phenotypes combining non-osteoporotic DXA status with degraded TBS were more frequent in patients with obesity.

Conclusions

In postmenopausal women, higher BMI was associated with higher BMD but lower TBS, supporting a divergent relationship between densitometric findings and DXA-derived trabecular texture parameters. These findings suggest that TBS may complement BMD in describing skeletal imaging status in overweight and obese women. Because fracture outcomes and complete clinical risk data were not available, the present results should be interpreted as evidence of imaging discordance rather than direct evidence of clinical underdiagnosis or misclassification.