Sleep–Wake Disturbances and Frailty in Community-Living Older Persons

OBJECTIVES: To evaluate the association between sleep–wake disturbances and frailty.
DESIGN: Cross-sectional.
SETTING: New Haven, Connecticut.
PARTICIPANTS: Three hundred seventy-four community-living persons aged 78 and older.
MEASUREMENTS: Frailty was based on the Fried phenotype, and sleep–wake disturbances were defined as daytime drowsiness, based on an Epworth Sleepiness Scale (ESS) score of 10 or greater, and as subthreshold and clinical insomnia, based on Insomnia Severity Index (ISI) scores of 8 to 14 and greater than 14, respectively.
RESULTS: Mean age was 84.3; 87 (23.8%) participants were drowsy, 122 (32.8%) had subthreshold insomnia, 38 (10.2%) had clinical insomnia, and 154 (41.2%) were frail. There was a significant association between drowsiness and frailty, with unadjusted and adjusted odds ratios (ORs) of 3.79 (95% confidence interval (CI)=2.29–6.29) and 3.67 (95% CI=2.03–6.61), respectively. In contrast, clinical insomnia was significantly associated with frailty in the unadjusted analysis (OR=2.77, 95% CI=1.36–5.67) but not the adjusted analysis (OR=1.93, 95% CI=0.81–4.61)), and subthreshold insomnia was not associated with frailty in the unadjusted or adjusted analysis.
CONCLUSION: In older persons, sleep–wake disturbances that present with daytime drowsiness, but not insomnia, are independently associated with frailty. Because drowsiness is potentially remediable, future studies should determine whether there is a temporal relationship between drowsiness and frailty, with the ultimate goal of informing interventions to reverse or prevent the progression of frailty.