Higher-level functional capacity and incident long-term care insurance certification: a competing risks analysis

Higher-level functional capacity (HLFC) may capture functional vulnerability beyond basic activities of daily living in older adults. However, the association of baseline HLFC with subsequent incident long-term care insurance (LTCI) certification remains unclear when death is considered a competing event.

This prospective cohort study included 688 community-dwelling Japanese adults aged 65 years or older who participated in the 2017 wave of the Community Empowerment and Care for Well-being and Healthy Longevity Study. HLFC was assessed once at baseline using the Tokyo Metropolitan Institute of Gerontology Index of Competence (TMIG-IC), including the total score and the formal subdomain scores for Instrumental Self-Maintenance, Intellectual Activity, and Social Role. Participants were followed for 8 years. Incident LTCI certification at Support Level 1 or higher was analyzed using Fine-Gray competing-risk regression, with death before certification treated as a competing event. The fully adjusted Fine-Gray model included age group, sex, living arrangement, physical activity, smoking status, alcohol consumption, and disease status.

Among 688 complete-case participants, 145 developed incident LTCI certification and 79 died before certification. In the fully adjusted model, higher HLFC total score was associated with a lower cumulative incidence of incident LTCI certification (sHR per 1-point increase, 0.941; 95% CI, 0.891-0.994; p = 0.029). Higher Social Role score was also associated with a lower cumulative incidence (sHR, 0.803; 95% CI, 0.702-0.918; p = 0.001), whereas the continuous per-point Instrumental Self-Maintenance and Intellectual Activity estimates included the null after full adjustment. Multiple-imputation and lagged-event analyses generally retained the direction and broad magnitude of the complete-case estimates. Exposure-by-log(time) diagnostics indicated that the Fine-Gray associations varied over follow-up. Estimates also varied under alternative age specifications, with particularly marked attenuation of the total-score association.

Higher baseline HLFC total and Social Role scores were associated with a lower subsequent cumulative incidence of incident LTCI certification in the prespecified age-group-adjusted analysis. The total-score association was substantially attenuated under alternative age adjustments, while the magnitude of the Social Role association also varied.