Variation in estimates of in‐home sleep architecture in diverse population with and without cognitive impairment

Abstract

AbstractBackgroundSleep disturbances are highly prevalent in patients with cognitive impairment and have been increasingly recognized as an important risk factor for dementia. However, in‐lab sleep studies are costly and often impractical at a population level, particularly with diverse populations. There are limited data on measures of sleep continuity and architecture in the home setting among Black, Mexican American (MA) and non‐Hispanic White (NHW) adults, with and without cognitive impairment.MethodTo date, the ongoing Health and Aging Brain Study‐Health Disparities (HABS‐HD)‐Dormir study has enrolled 732 participants using a community‐based participatory research approach. In‐home sleep testing was assessed using WatchPAT (WatchPAT‐200) and the zzzPAT software (5.2.79.7p), which provides estimates of sleep duration, sleep efficiency, and rapid eye movement (REM) and non‐REM (NREM) sleep duration using a proprietary algorithm. Dementia, mild cognitive impairment (MCI) or normal cognition (NC) were determined based on a comprehensive clinical evaluation.ResultOn average, participants were aged 66.7±8.4 (50‐92) years, including 64.2% women; 34.4% MA, 23.8% Black and 41.8% NHW adults; 22.1% classified as MCI or dementia. After adjustment for age, sex, race, body mass index and comorbidities, participants with MCI or dementia compared to those with NC had shorter adjusted mean sleep duration (6.3 vs. 6.7 hours, p = 0.002), lower sleep efficiency (77.2% vs. 82.1%, p<0.001), longer sleep latency (32.3 vs. 25.7 minutes, p = 0.006) and longer REM sleep latency (116.5 vs. 95.4 minutes, p = 0.002); had lower REM sleep percentage (20.1% vs. 23.1%, p<0.001), and higher NREM sleep percentage (80.0% vs. 76.9%, p<0.001). The association between sleep staging (REM or NREM sleep percentage) and cognitive diagnoses was only significant among Black participants (p for interaction = 0.02, Figure 1).ConclusionIndividuals with MCI/dementia differed from those with NC in key measures of sleep architecture assessed by a home sleep test, suggesting both the utility of in‐home assessments to estimate sleep in middle‐aged and older diverse samples and the need to understand the race differences underlying associations between sleep architecture and cognition in community‐based studies.