Objective
This study aimed to examine the association between self-reported sleep problems and cognitive decline in community-dwelling older people. We hypothesized that daytime somnolence predicts subsequent cognitive decline.
Methods
This is a longitudinal study in a 3.2-year follow-up, with 18-month intervals. The setting is the Washington Heights-Inwood Community Aging Project. There were 1098 participants, who were over 65 years old and recruited from the community.
Sleep problems were estimated using five sleep categories derived from the RAND Medical Outcome Study Sleep Scale: sleep disturbance, snoring, awaken short of breath/with a headache, sleep adequacy, and daytime somnolence. Four distinct cognitive composite scores were calculated: memory, language, speed of processing, and executive functioning. We used generalized estimating equations analyses with cognitive scores as the outcome, and time, sleep categories and their interactions as the main predictors. Models were initially unadjusted and then adjusted for age, gender, education, ethnicity, depression, and apolipoprotein E-ε4 genotype.
Results
Increased daytime somnolence (including feeling drowsy/sleepy, having trouble staying awake, and taking naps during the day) was linked to slower speed of processing both cross-sectionally (B = −0.143, p = 0.047) and longitudinally (B = −0.003, p = 0.027). After excluding the demented participants at baseline, the results remained significant (B = −0.003, p = 0.021).
Conclusions
Our findings suggest that daytime somnolence may be an early sign of cognitive decline in the older population. Copyright # 2015 John Wiley & Sons, Ltd.