LAUSR

Purpose: Bone and lean mass loss and cognitive impairment are prevalent in elder adults and have been hypothesized to share a potential link. Methods: This nationwide cross-sectional study systemically sampled elder adults aged ≥65 years and conducted the door-to-door survey. The causal diagrams help to decide which covariates were included in the generalized linear mixed models (GLMMs). The structural equation modeling (SEM) was performed for the validation. Results: A total of 535 participants were enrolled and categorized into the normal (67.3%), mild cognitive impairment (18.3%), and dementia groups (14.4%). With increasing in the severity of cognitive impairment, the bone marrow density and lean mass consistently showed the trend of decreasing values. In the GLMMs, a significant association existed between the decrease of the bone mineral density (BMD) and the Mini-Mental State Examination (MMSE) (β = 5.819 scores per g/cm2 decrease, p = 0.0305) with adjustment of the age, sex, and physical activity. The SEM models confirmed that the MMSE was significantly and directly predicted by the age (β = 0.1363, p = 0.0003) and BMD (β = 0.1251, p = 0.0006) independently and indirectly predicted by lean mass (β = 0.1138, p = 0.0003) through the bone density path. Conclusion: In conclusion, an independent association between bone loss and cognitive impairment was existed rather than the confounding effect and the decrease of lean mass indirectly contributed to cognitive impairment by influencing the bone density.