1165 Insulin Resistance Moderates REM, But Not NREM, Sleep Stage Durations and Episodic Memory Performance Among Older Black American Adults

Abstract Introduction Sleep architecture - comprising rapid eye movement (REM) sleep, light sleep (N1/N2), and slow-wave sleep (SWS) - is essential for metabolic regulation and memory consolidation. Sleep disturbances adversely affect insulin resistance (IR), which can lead to hippocampal impairment. Therefore, disturbed sleep and metabolic dysfunction, together, may negatively influence episodic memory. This relationship needs further elucidation among Black Americans, who experience disproportionally higher sleep disruption, IR, and cognitive decline, likely due to increased psychosocial/environmental stressors relative to non-Blacks. This study examined whether associations between sleep architecture and episodic memory performance differed by IR levels in a sample of older Black adults. Methods Black adults (N=83) ages ≥55 years (mean age 61.8±5.2, 71.1% female) residing in South Florida (Monroe, Miami-Dade, Broward, and Palm Beach counties), were drawn from NIH-funded studies (MOSAIC, n=51; PRAISE, n=32), yielding baseline demographic and clinical data for the proposed analyses. Participants completed a medical history questionnaire, provided blood samples, wore a Fitbit (Inspire 2) device that captured sleep stage durations. All sleep data were averaged, providing one data point for each sleep metric. IR was assessed using fasting glucose and insulin to calculate HOMA-IR [(glucose (mg/dL) × insulin (U/mL))/405]. Episodic memory was measured using the NIH Toolbox Picture Sequence Memory Test, where age- and education-adjusted T-scores were derived. We used moderation regression analysis to determine whether insulin resistance moderated associations between total sleep duration or sleep stage durations and episodic memory. Results HOMA-IR significantly moderated the relationship between REM duration and episodic memory performance (ΔR²=0.11, ΔF=9.75, p=0.0025), controlling for sex and diabetes diagnosis. Among insulin-sensitive participants, greater REM duration predicted better episodic memory. Whereas higher insulin resistance was associated with poorer memory performance as REM duration increased. Moderation regression analysis found no significant associations between light sleep, SWS, or total sleep duration and episodic memory. Conclusion Findings indicate that insulin resistance modifies REM sleep and episodic memory interactions among older Black adults. Future studies should investigate metabolic health as a pathway linking sleep architecture to cognition and inform interventions targeting both sleep and metabolic risk. Longitudinal research may clarify how improving these factors could influence cognitive aging. Support (if any)