Association between tooth brushing habits and incident disability in community-dwelling older adults: a prospective...

Japan is experiencing a rapid demographic shift toward a super-aged society, making the extension of healthy life expectancy a national priority. Oral self-care, particularly tooth brushing habits, may play a critical role in preventing functional decline among older adults, yet evidence in this population remains limited. This study aimed to examine the association between specific tooth brushing behaviors-frequency, timing (especially before bedtime), and use of interdental cleaning tools-and incident disability among community-dwelling older adults in Japan.

A three-year prospective cohort study was conducted in City A, Nara Prefecture, involving 7,785 older adults aged 65 and above who were free of disability at baseline. Tooth brushing habits were assessed via self-administered questionnaires, and incident disability was defined as new certification under Japan's Long-Term Care Insurance system. Covariates included age, gender, socio-economic status, lifestyle habits, physical and mental health, and oral health indicators. Multiple logistic regression analyses were used to estimate adjusted odds ratio (AOR) and 95% confidence interval (CI) for incident disability.

During the follow-up period, 713 participants (9.2%) developed incident disability. After adjustment for covariates and mutual adjustment for other tooth brushing habits, non-daily brushing before bedtime and non-use of interdental cleaning tools were associated with higher odds of incident disability (AOR for non-daily vs daily brushing before bedtime = 1.55, 95% CI: 1.19-2.01; AOR for non-use vs use of interdental cleaning tools = 1.34, 95% CI: 1.11-1.62). These associations were particularly evident among individuals aged 75 and older.

Daily brushing before bedtime and the use of interdental cleaning tools were independently associated with lower odds of incident disability among community-dwelling older adults. Nevertheless, causal inference is limited, and residual confounding cannot be ruled out.