Orexin, Sleep, and Cognition in Alzheimer Disease: Non-REM Oscillatory Activity and Neural Resilience.

Abstract (English)

BACKGROUND AND OBJECTIVES: Sleep-wake dysregulation and elevated CSF orexin have been implicated in Alzheimer disease (AD). Sleep spindles (SPs) and slow oscillations (SOs) are linked to cognition and neurodegeneration; however, their relationship with CSF orexin concentrations in symptomatic AD has not been characterized. We investigated whether nonrapid eye movement (NREM) SP-SO activity is associated with CSF orexin and whether these oscillatory features moderate associations between orexin, cognition, neuropsychiatric symptom severity, and AD biomarkers. METHODS: This prospective observational cohort study was conducted at a tertiary memory clinic in Lleida, Spain. Individuals aged ≥60 years with biomarker-confirmed mild-to-moderate AD (National Institute on Aging-Alzheimer's Association criteria) underwent overnight polysomnography and morning CSF sampling. SP and SO were detected using validated automated algorithms with independent verification and visual quality control. CSF was assayed for orexin-A, amyloid-β42 (Aβ42), phosphorylated tau181 (pTau181), total tau, and YKL-40. Cognitive performance Alzheimer's Disease Assessment Scale-Cognitive Subscale ([ADAS-Cog], Mini-Mental State Examination (MMSE), California verbal learning test, ROCF) and neuropsychiatric symptoms (NPI) were assessed longitudinally over 36 months. Associations were examined using generalized linear models with robust estimators adjusted for age, sex, Aβ42, and apnea-hypopnea index. Multiple comparisons were controlled using false discovery rate correction. Interaction terms assessed moderation effects. RESULTS: Sixty participants (30 women; mean age 74.7 years) were included. Longer SO duration and higher SP density and power were associated with lower CSF orexin concentrations (SP density: β = -187.37 pg/mL, 95% CI -344.93 to -29.80). Orexin was not associated with global sleep continuity metrics. Higher CSF orexin concentrations were associated with worse global cognition (ADAS-Cog: β = 0.014, 95% CI 0.003-0.024; MMSE: β = -0.01, 95% CI -0.011 to -0.004) and greater neuropsychiatric symptom severity (NPI at: β = 0.03, 95% CI 0.011-0.041). Higher orexin was also associated with higher pTau181 (β = 0.11, 95% CI 0.04-0.19), total tau, and YKL-40 (β = 0.37, 95% CI 0.17-0.57). Significant orexin × SP-SO interactions were observed, such that greater oscillatory activity attenuated the adverse associations between orexin and cognitive outcomes, independent of Aβ42 and tau. DISCUSSION: In biomarker-confirmed AD, NREM SP and SO activity are associated with CSF orexin concentrations and moderate associations between orexin and longitudinal cognitive and neuropsychiatric outcomes. Limitations include the observational design and absence of a comparator group, precluding causal inference and limiting contextualization relative to normal aging. NREM oscillatory metrics and orexin concentrations may represent complementary physiologic markers for disease monitoring and therapeutic targeting. TRIAL REGISTRATION INFORMATION: Role of Hypoxia and Sleep Fragmentation in AD; ClinicalTrials.gov Identifier: NCT02814045.