Insomnia in Community-Living Persons with Advanced Age

To evaluate the epidemiology of insomnia, including demographic and clinical correlates, in older adults.

Cross-sectional.

Community.

Yale Precipitating Events Project participants (N=379; mean age 84.3; 67.8% female; 11.9% African American).

Insomnia Severity Index (ISI), with scores of 8 and higher indicating insomnia, which was further stratified according to ISI score as mild (8-14), moderate (15-21), or severe (22-28). Baseline characteristics included age, sex, race, education, smoking, obesity, medical conditions, depressive symptoms (Center for Epidemiologic Studies Depression score ≥16), cognitive impairment (Mini-Mental State Examination score <24), restless legs syndrome (RLS), self-reported sleep-disordered breathing (SDB), medications, and daytime sleepiness (Epworth Sleepiness Scale (ESS), range 0-24).

Insomnia was established in 163 (43.0%) participants (average ISI score 12.3 (mild)). For the entire sample, average baseline characteristics were as follows: 30.1% did not complete high school, 5% were current smokers, 19.2% were obese, 28.2% had cardiovascular disease, 19.3% had chronic lung disease, 27.2% had depressive symptoms, 16.1% had cognitive impairment, 36.8% had RLS, and 3.4% had self-reported SDB; mean number of medications was 9.2, and mean ESS was 6.4. In multivariable regression models, only depressive symptoms (adjusted odds ratio (aOR)=8.34, 95% confidence interval (CI)=4.49, 15.47) and RLS (aOR=2.49, 95% CI=1.48, 4.21) were significantly associated with insomnia.

In a sample of older adults with high medical burden and polypharmacy, insomnia was highly prevalent but unexpectedly mild and associated only with depressive symptoms and RLS. The discordance of high prevalence but mild severity of insomnia in the oldest adults highlights the need for diagnostic confirmation with objective measures of sleep disturbances, whereas the strong associations with depressive symptoms and RLS inform priorities in managing insomnia.