Geographic heterogeneity in the association between educational attainment and QUS-defined bone status among adults in Southwest China: a cross-sectional study

Although education is commonly regarded as a favorable factor for bone health, its association with bone status may vary across different socioeconomic and ecological contexts. We examined the association between educational attainment and QUS (quantitative ultrasound)-defined bone status across five provinces in Southwest China and explored province-specific indirect statistical associations.

This cross-sectional study included 73,600 adults aged ≥ 30 years from Sichuan, Chongqing, Guizhou, Yunnan, and Tibet, recruited through a multistage, stratified cluster sampling design within the China Multi-Ethnic Cohort. QUS-defined bone status was classified as normal, reduced, and markedly reduced using calcaneal quantitative ultrasound-derived T-scores. Geographic effect modification was evaluated using cumulative logit ordinal regression. Province-specific path analysis was further conducted to explore potential indirect associations involving health behaviors, nutritional factors, and socioeconomic status. Partial proportional-odds models were applied in sensitivity analyses.

Participants had a mean age of 51.54 years (SD, 11.43), and 62.1% were women. Overall, 60.6%, 35.0%, and 4.4% of participants were classified into the normal, reduced, and markedly reduced QUS-defined bone-status categories. Education showed marked regional heterogeneity in its association with bone status (P < 0.001): higher educational attainment was associated with poorer bone status in Sichuan (β=-0.088, P < 0.05), Guizhou (β=-0.047, P < 0.05), and Tibet (β=-0.067, P < 0.05), better bone status in Chongqing (β = 0.045, P < 0.05), and no association in Yunnan (β = 0.015, P = 0.155). Path analyses showed distinct provincial patterns: higher alcohol consumption and higher dairy intake in Sichuan; higher income, higher dairy intake, and reduced smoking in Chongqing; lower physical activity in Guizhou and Yunnan; and no statistically significant indirect association was identified in Tibet. In PPOM sensitivity analyses, high (vs. low) educational attainment was associated with lower odds of worse bone status in Chongqing at both cumulative thresholds (OR = 0.738, 95% CI, 0.652-0.837; OR = 0.688, 95% CI, 0.541-0.875), higher odds of worse bone status in Guizhou (OR = 1.174, 95% CI, 1.018-1.352), and no clear association in Yunnan (OR = 0.873, 95% CI, 0.696-1.094).

Educational attainment showed province-specific associations with QUS-defined bone status in Southwest China. Indirect statistical associations involving socioeconomic, behavioral, and dietary factors require confirmation in longitudinal studies.