Nonlinear Associations Between Sleep Duration and Incident LTCI Certification in Older Adults: Findings From the...

To examine the association between self-reported sleep duration and incidence of long-term care insurance (LTCI) certification in older Japanese adults.

Prospective cohort study. Nonlinear associations were explored using restricted cubic spline (RCS) modeling.

We prospectively followed 9576 community-dwelling adults (3267 men; 6309 women; mean age, 71.1 years) for a mean of 10.3 years.

Sleep duration was categorized into six 1-hour increments. The incidence of LTCI certification was identified via municipal records. Cox proportional hazards models and RCS analysis were adjusted for sociodemographic, lifestyle, and clinical factors. Same analyses were conducted by sex and age group (65-74 vs ≥75 years).

In multivariable-adjusted models, using 7 to <8 hours/d as the reference category, long sleep duration (≥9 hours/d) was consistently linked to a higher risk of incident LTCI certification (hazard ratio [HR], 1.17; 95% CI, 1.03-1.32 for 9 to <10 hours/d; HR, 1.26; 95% CI, 1.11-1.42 for ≥10 hours/d). Although short sleep duration (<6 hours/d) was associated with elevated LTCI risk in partially adjusted models, the association was attenuated and no longer statistically significant in the fully adjusted model (HR, 1.21; 95% CI, 0.99-1.48). RCS analysis indicated a nonlinear association between sleep duration and LTCI risk (P for nonlinearity <.001), although the association was more pronounced for long sleep duration. This nonlinear pattern was more evident among women and adults aged 65-74 years but less pronounced in men and those 75 years or older. A similar tendency was also observed for severe cases (care need level ≥2).

A nonlinear pattern suggested increased LTCI risk at the extremes of sleep duration, with a more pronounced risk at longer sleep durations. These findings suggest that sleep patterns represent modifiable risk factors for functional decline and long-term care needs in older adults.