Archives of Gerontology and Geriatrics Plus · Apr 16, 2026

Frailty risk stratification using a senior volunteer-driven frailty check-up and subsequent long-term care and healthcare expenditures: A cohort study

By Tanaka T, Yoshizawa Y, Lyu W et al.

Background The frailty check-up (FC) program in Japan, conducted by trained senior volunteers, provides multidimensional assessment data that have been shown to predict disability and mortality. This study examined the associations between an FC-derived frailty risk score and subsequent long-term care (LTC) costs, medical expenses, and total healthcare expenditures. Methods This longitudinal cohort study examined community-dwelling older adults in Kashiwa City, Japan, who underwent FC during 2017–2019. Participants were categorized into three risk groups using an FC-derived weighted frailty risk score based on eight predictors. Annual LTC and medical costs were obtained from municipal LTC insurance receipts and the national Kokuho databases for fiscal years 2019–2022. Means and cost ratios, adjusted for age, sex, height, chronic disease, and lifestyle, were estimated using generalized linear mixed models with a negative binomial distribution. Results This study analyzed 1882 community-dwelling older adults (mean age: 77.0 ± 6.0 years; female, 74.3%). Expenditures increased over time in all risk groups. In 2022, the adjusted mean LTC costs were 30.7, 76.7, and 132.7 × 1000 Japanese yen (JPY) in the low-, moderate-, and high-risk groups, respectively, with corresponding total healthcare costs of 551.0, 779.8, and 982.3 × 1000 JPY. The high-risk group incurred 3.75-fold higher LTC costs and 1.51-fold higher healthcare costs than the low-risk group. Conclusions An FC-derived frailty risk score was associated with subsequent increases in LTC and healthcare expenditures among community-dwelling older adults. These findings suggest that frailty risk stratification using community-based FC data may inform resource planning in super-aged societies.

doi.org/10.1016/j.aggp.2026.100277Open accessBiomarkers

AgingResearch.News