Abstract
Abstract Introduction Impairments in the microstructure of the limbic tracts are seen in early stages of the Alzheimer’s disease (AD) spectrum. Disrupted sleep has been linked with AD. We examined associations between measures of limbic white matter tracts and objective sleep parameters in cognitively unimpaired older adults. Methods This cross-sectional study included 170 community-dwelling cognitively unimpaired older adults (mean±SD: age=67.2±5.2y) participating in NYU studies of sleep, aging, and memory. Subjects completed polysomnography (NPSG) and brain MRI. Sleep measures of interest included total sleep time (TST), NREM stages 2, 3 and REM sleep duration, sleep efficiency and fragmentation, and AHI4% and REM AHI4%. Microstructural properties of the cingulum, uncinate fasciculus (UF), and fornix were estimated using diffusional tensor imaging (DTI) metrics including radial kurtosis (RK), radial diffusivity (RD), and fractional anisotropy (FA). Linear mixed-effects regression models examined associations between sleep variables and DTI metrics. Models were adjusted for age, race, education, sex, time between NPSG and MRI. Results Participants were 71.2% female, 44.1% Black/African-American, and had 16.8±2.5y of education. In the right cingulum, higher AHI4% in REM was associated with decreased RK(β[IV]= β[SE}; β[AHI4%] =-0.0014[0.00054], p=0.01) and shorter stage 2 sleep duration was associated with decreased RK (β [stage 2 sleep duration] =0.00062[0.00026], p=0.02). In both hemispheres of the cingulum, increased sleep fragmentation in stage 3 was associated with increased RD (β [sleep fragmentation, right] = 0.0017[0.00080], p=0.03; β [sleep fragmentation, left] = 0.0026[0.00087], p=0.003), and higher AHI4% in REM was associated with reduced FA (β [AHI4% left] =-0.00039[0.00019], p=0.048; β [AHI4% right] =-0.00055[0.00017], p=0.02). In the right UF, with increasing age, reduced TST was associated with lower RK(β[TST]=0.0000047[0.000002], p=0.02). In the right UF, reduced time in REM sleep was associated with increased RD (β [REM duration] =-0.0022[0.00078], p=0.006). In the fornix, higher AHI4% was associated with lower FA(β[AHI4%] =-0.00061[0.00029], p=0.04). Conclusion In this sample of community-dwelling cognitively unimpaired older adults, reduced sleep duration and OSA severity (NREM and REM) were associated with limbic white matter tracts alterations. Longitudinal studies are needed to investigate the role of disrupted sleep in limbic white matter microstructure alterations in early stages of the AD spectrum. Support (if any) AASM-BS-231-20, NIAK23AG068534, NIAR01AG082278, R01AG083975, AARG-D-21-848397, BrightFocus-A2022033S