Objectives: To examine whether self-reported symptoms of insomnia independently increase risk of cognitive decline in older adults.
Design: Longitudinal cohort study.
Setting: The four sites of the Established Populations for Epidemiologic Studies of the Elderly.
Participants: Six thousand four hundred forty-four community-dwelling men and women age 65 and older who had no more than one error on the Short Portable Mental Status Questionnaire (SPMSQ) at baseline and an in-person interview at the third annual follow-up (FU3).
Measurements: Insomnia was defined as report of trouble falling asleep or waking up too early most of the time. Cognitive decline was defined as two or more errors on the SPMSQ at FU3. Logistic regression was used to determine risk of cognitive decline associated with insomnia, controlling for demographic, behavioral, and health-related factors. Analyses were stratified by sex and depressed mood.
Results: Among nondepressed men, those reporting symptoms of insomnia at both baseline and FU3 had an adjusted odds ratio (OR) of 1.49 (95% CI = 1.03-2.14) for cognitive decline, relative to those with no insomnia at FU3. Men with insomnia at FU3 only were not at increased risk (OR = 1.16, 95% CI = 0.82-1.65). These relationships were not found in women. Men and women with depressive symptoms at FU3 were at increased risk for cognitive decline independent of insomnia.
Conclusion: Chronic insomnia independently predicts incident cognitive decline in older men. More sensitive measures of cognitive performance may identify more subtle declines and may confirm whether insomnia is associated with cognitive decline in women.