Bidirectional relationship between insomnia and frailty in older adults: A 2-year longitudinal study

Abstract

Background: Most previous studies that examined the association of insomnia with frailty used cross-sectional designs. The temporal relationship between these factors is therefore largely unknown. This study aimed to examine the bidirectional relationship between insomnia and frailty by sex.

Methods: A 2-year longitudinal study involving all community-dwelling older adults living in a rural area in Japan (n = 3844). Validated measures of insomnia and frailty were employed. Insomnia was assessed using the Athens insomnia scale, and frailty using the Kihon checklist. We performed a cross-lagged panel model, adjusted for age, sex, years of education, employment status, self-rated health, complications (hypertension, diabetes, stroke, or osteoarthritis), BMI, physical activity, alcohol consumption, and smoking status, and assessed differences by sex.

Results: Poor sleep predicted the onset and worsening of frailty during follow up (standardized coefficient [95% confidence interval]: 0.076 [0.045, 0.107]). Frailty also predicted severe insomnia symptoms (0.074 [0.044, 0.104]). However, the temporal association between these conditions varied by sex. In older men, the effect of frailty on insomnia was stronger than that of insomnia on frailty. However, in women, the impact of insomnia on frailty was stronger than that of frailty on insomnia.

Conclusions: The primary potential cause of the association between insomnia and frailty may vary by sex, being frailty in men and insomnia in women. Sex-specific interventions to improve sleep quality and duration, and maintain functional abilities in daily life may contribute to the prevention and management of both frailty and insomnia in older adults.

Keywords: Community-dwelling older adults; Cross-lagged panel model; Kihon checklist; Longitudinal study; Sleep disorder.