Interactive associations of nocturnal sleep disturbance and vascular risk with prospective cognitive decline in clinically normal elderly individuals: Findings from the National Alzheimer's Coordinating Center Uniform Data Set

Abstract

AbstractBackgroundWe determined whether Nocturnal Sleep Disturbance (NSD) and vascular risk act together to promote prospective cognitive‐decline in clinically normal older adults; and, evaluated the unique influence of their combined risk on prospective cognitive decline beyond that of commonly used Alzheimer’s disease (AD) biomarkers.MethodLongitudinal study utilizing data from the National Alzheimer's Coordinating Center (NACC) Uniform Data set (UDS). Participants (N=361) were cognitively normal at baseline and had baseline medical data to quantify vascular risk, using an adaptation of the Framingham Heart Study general cardiovascular disease (aFHS‐CVD) risk‐score and CSF‐Aβ, CSF P‐tau, CSF T‐tau and MRI‐imaging data with at least one UDS follow‐up visit. The Neuropsychiatric Inventory Questionnaire characterized NSD and incident mild cognitive impairment (MCI) diagnosis during UDS follow‐up characterized prospective cognitive decline. Logistic mixed‐effects models with random intercept and slope, controlling for age, sex, education, APOE‐ε4 and their interactions with time examined associations between the NSD/ FHS‐CVD risk score and longitudinal cognitive‐decline.ResultOf the 361 participants, 223 (62%) were women and 35 (9.7%) had NSD. The proportion of males versus females with sleep problems was 10.9% vs. 9.3% respectively. For participants with NSD and no NSD, the mean (SD) age was 71 (7.3) and 70 (5.7) years and average follow‐up time was 5.2 (2.6) and 4.9 (2.7) years, respectively. Both NSD (OR: 1.42, P < .003) and higher aFHS‐CVD risk score (OR; 1.63, P < .001) were significantly associated with increased/faster likelihood to develop incident MCI. The interaction of NSD and the aFHS‐CVD risk‐score with time was significant (P < .001) suggesting an increase in the likelihood of conversion to MCI increased over time. Stratifying aFHS‐CVD risk score into tertiles, NSD participants in the highest (OR: 2.82, P < .003) and middle tertile (OR: 2.38, P < .001) were significantly more likely to develop incident MCI, compared with participants without NSD in the lowest aFHS‐CVD risk score tertile, suggesting a synergistic effect. This effect remained robustly associated with incident MCI even after adjustment for AD biomarkers.ConclusionIn elderly cognitive‐normal individuals, NSD and vascular risk may be alternate and non‐invasive measures of assessing risk of prospective cognitive‐decline in preclinical AD.