Development of a risk score for the prediction of incident dementia in older adults using a frailty index and health checkup data: The JAGES longitudinal study.

Abstract

In Japan, the prevalence of dementia is expected to reach 4.7 million by 2025. This study aimed to develop a risk score for the prediction of incident dementia in community-dwelling older adults. In this longitudinal observational study, we used data from the Japan Gerontological Evaluation Study (JAGES) conducted in K City. We performed Cox regression analyses to develop three risk score models for the prediction of incident dementia in older adults using a frailty index and health checkup data. Analyses of the area under the receiver operating characteristic curve were conducted to compare the models' predictive abilities. We identified 6656 (9.2%) individuals who developed incident dementia during the observation period. The C-statistics of the risk scores ranged from 0.733 to 0.790. The risk score models were able to predict incident dementia in older adults and may help non-medical professionals detect dementia risk at an early stage.