Abstract
BACKGROUND: The effect of nighttime behaviors on cognition has not been studied independently from other neuropsychiatric symptoms. OBJECTIVE: We evaluate the following hypotheses that sleep disturbances bring increased risk of earlier cognitive impairment, and more importantly that the effect of sleep disturbances is independent from other neuropsychiatric symptoms that may herald dementia. METHODS: We used the National Alzheimer's Coordinating Center database to evaluate the relationship between Neuropsychiatric Inventory Questionnaire (NPI-Q) determined nighttime behaviors which served as surrogate for sleep disturbances and cognitive impairment. Montreal Cognitive Assessment scores defined two groups: conversion from 1) normal to mild cognitive impairment (MCI) and 2) MCI to dementia. The effect of nighttime behaviors at initial visit and covariates of age, sex, education, race, and other neuropsychiatric symptoms (NPI-Q), on conversion risk were analyzed using Cox regression. RESULTS: Nighttime behaviors predicted earlier conversion time from normal cognition to MCI (hazard ratio (HR): 1.09; 95% CI: [1.00, 1.48], p = 0.048) but were not associated with MCI to dementia conversion (HR: 1.01; [0.92, 1.10], p = 0.856). In both groups, older age, female sex, lower education, and neuropsychiatric burden increased conversion risk. CONCLUSION: Our findings suggest that sleep disturbances predict earlier cognitive decline independently from other neuropsychiatric symptoms that may herald dementia.