Abstract
BackgroundOsteoporosis and dementia are two common disorders mainly affecting ageing population, and heel estimated bone mineral density (eBMD) measured by quantitative ultrasound (QUS) has been demonstrated to be a reliable and non-invasive method for assessing skeletal health. The aim of this study is to determine the association between eBMD and incident dementia in older adults.MethodsThis retrospective cohort study employs UK Biobank data of 131,030 adults aged ≥ 60 years without dementia at baseline. Cox proportional-hazards models were used to investigate the association between eBMD and incident dementia, with the C-index evaluating the discriminative potential of eBMD.ResultsAmong participants (52.1% women, median [IQR] age was 64 [62-66] years), there were 4,572 cases (3.5%) of incident dementia. Minimal model showed that participants with low eBMD (< 0.467 g/cm2) had a 14% increase in the rate of dementia incidence (HR 1.14, 95% CI 1.06-1.23; P <.001), and each standard deviation (SD) decrease in eBMD was associated with a 49% increase in dementia risk (HR 1.49, 95% CI 1.19-1.86; P <.001). Such association remained significant after further adjustment for potential confounders. Stratified analyses revealed that lower eBMD increased dementia risk in male participants (HR 1.17, 95% CI 1.06-1.31; P =.003) and in participants with neutral (HR 1.18, 95% CI 1.05-1.33; P =.007) or low genetic risk (HR 1.36, 95% CI 1.01-1.83; P =.04). Sensitivity analyses showed similar results. However, discriminative analyses showed minimal improvement after adding eBMD to dementia prediction models.ConclusionLower heel eBMD is independently associated with increased dementia risk among older adults.