Social
Library Use and Functional Disability, Dementia, and Mortality Among Older Adults in Japan: An Instrumental Variable Analysis Qiuyi Liu* Qiuyi Liu Liu Liu Liu Liu Liu Liu Liu Department of Social Epidemiology, Graduate School of Medicine and School of Public Health, Kyoto University
Background
Cultural participation has been increasingly recognized as a determinant of health in later life, yet causal evidence remains relatively limited, in part due to self-selection and unmeasured confounding. Libraries represent widely accessible cultural facilities that support reading, learning, and social engagement, but their potential health effects have rarely been evaluated using causal inference methods.
Objective
To estimate the causal effect of library use on the incidence of dementia, functional disability, and mortality among older adults in Japan.
Methods
We conducted a prospective cohort study using data from the Japan Gerontological Evaluation Study (JAGES) 2019–2022, including101,815community-dwelling adults aged 65 years and older without baseline functional disability or dementia. Frequency of library use was self-reported at baseline and converted into an approximate monthly visit measure. Incident functional disability, dementia, and all-cause mortality were ascertained through linkage with national long-term care insurance and resident registries. To address endogeneity in library use, we applied an instrumental variable approach using network-based geographic distance from each participant’s residence to the nearest library as the instrument. We applied the two-stage residual inclusion (2SRI) instrumental-variable framework to Cox proportional hazards models and Aalen’s additive hazards models. All models were adjusted for age, gender, educational background, marital status, household income, employment status, and population density.
Results
We observed 11,412 (11.20%) functional disability, 5,164 (5.07%) dementia, and 5,540 (5.44%) deaths. In the IV Cox model, each additional monthly visit to the library was associated with a 0.56 times risk of dementia (HR = 0.56, 95%CI: 0.34-0.90). In the additive hazards model, the hazard difference was −0.015 per person-year (95% CI: −0.023 to −0.006), corresponding to 150 fewer cases per 10,000 person-years. Point estimates indicated the association of library use with decreased risks of functional disability and mortality, but their confidence intervals were wide.
Conclusions
Using geographic accessibility to libraries as an instrument, this study provides causal evidence that engagement with libraries may contribute to healthier aging among older adults, particularly by reducing dementia risk. Libraries may function as community-level cultural resources with potential relevance for population aging and health policy.