Is Poor Sleep Quality Associated with Lower Levels of and Faster Declines in Cognitive Function? Results from the Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) Study

Abstract

AbstractBackgroundPoor sleep quality has been linked to worse cognitive function, but most research has been conducted in predominantly White samples. We investigated whether poorer sleep quality is associated with lower levels of and faster declines in cognition and whether these associations differ by race/ethnicity.MethodWe used data from KHANDLE (N = 1,438; mean age: 75.7 years), a multi‐ethnic cohort of Kaiser Permanente Northern California members. We created a global sleep score by summing scores on six sleep components assessed by a modified Pittsburgh Sleep Quality Index: subjective sleep quality, latency, duration, disturbances, sleep medication use, and daytime dysfunction. Executive function (EF) and verbal memory (VM) were assessed by the Spanish and English Neuropsychological Assessment Scale (3 waves each separated by 18 months). Linear regression models assessed the associations between sleep quality and baseline cognition. To understand associations with cognitive decline, we fit linear mixed‐effect models with random intercepts accounting for practice effects and interview mode. Covariates included baseline age, gender, income, education, depression, physical activity, and central adiposity. We included interactions between time (years since baseline) and sleep and time and covariates to determine associations between sleep and cognitive decline. To assess racial/ethnic differences, we included interactions (race*sleep in linear regression models and time*sleep*race in mixed‐effect models for each racial/ethnic category) and performed likelihood ratio tests to determine the presence of interactions.ResultAfter adjusting for covariates, sleep quality was not associated with baseline EF or VM and the associations did not differ by race/ethnicity (Figure 1). Overall, poorer sleep quality was associated with faster declines in EF (β time*sleep = ‐0.02, 95% CI ‐0.03, ‐0.004: Figure 2). The associations between sleep and declines in EF were modified by race/ethnicity (β time*sleep*Black = ‐0.06, 95% CI ‐0.09, ‐0.03), such that poorer sleep quality was associated with faster declines in EF among Black adults only (β time*sleep = ‐0.06, 95% CI ‐0.08, ‐0.04; Figure 3).ConclusionPoorer sleep quality was associated with faster cognitive decline, and the association was particularly strong among Black older adults. Further research is needed to elucidate the mechanisms through which sleep influences cognitive decline among Black populations.