Biomarkers

Sleep disturbances are associated with inflammatory and neurodegenerative processes, as well as greater Alzheimer's disease (AD) risk. However, the relationship between sleep duration and blood-based biomarkers (BBMs) of neurodegeneration and AD pathology remains understudied, particularly across sexes. In a community-based cohort, we examined the cross-sectional associations between self-reported sleep duration and BBMs of neurodegeneration (t-tau, NfL), AD pathology (p-tau-181), and astrocyte activation (GFAP), and explored potential sex-specific effects.

Our sample included 2,254 participants (mean age 69.9±8.4 years, 44.9% male) from the Framingham Heart Study Offspring (exam 9) and Omni 1 (exam 4) cohorts who completed their visits between 2011-2014. Sleep duration was assessed by a technician-administered questionnaire. Plasma t-tau, p-tau-181, GFAP, and serum NfL were measured using Single Molecule Array (SIMOA) assays. All BBMs were log-transformed and standardized for analysis. Model 1 adjusted for age, sex-at-birth, and cohort; Model 2 additionally adjusted for sleep apnea, cardiovascular risk factors, depression (defined as Center for Epidemiologic Studies Depression Scale ≥16 or antidepressant use), eGFR, APOE4, stroke, and all-cause-dementia.

Table 1 depicts sample characteristics. Longer sleep duration was associated with higher t-tau levels across all models (Model 1: β±SE=0.05±0.02, p <0.001; Model 2: 0.03±0.02, p = 0.036) (Table 2). There were statistically significant sex interactions for t-tau (p = 0.019), NfL (p = 0.006), and GFAP (p = 0.007). In sex-stratified analyses (Table 3), we observed positive associations between longer sleep duration and higher biomarker levels (t-tau: 0.10±0.02, NfL: 0.11±0.02, GFAP: 0.010±0.02; all p <0.001) in males but not in females. No significant associations or interactions were observed for p-tau-181.

Sex significantly modified associations between sleep duration and BBMs. Specifically, only males showed significant positive associations between longer sleep and elevated t-tau, NfL, and GFAP levels. Relative to females, males have heightened inflammatory responses to sleep disturbances, which may partially explain our sex-specific findings with BBMs of neuronal injury and astrogliosis. Although limited by cross-sectional design and self-reported sleep measures, the findings underscore the relevance of sex-specific analyses in sleep and neurodegeneration research. We will next explore non-linear associations as short and long sleep may relate to BBMs level differently. Future longitudinal studies with objective sleep measurements are needed.