0078 Eye of the Beholder: Response Surface Analysis of Subjective-Objective Cognition Across Actigraphic Sleep in Older Adults

Abstract Introduction A discrepancy between subjective and objective measures of cognition may indicate those at risk for missed cognitive impairment diagnosis in older adults. Past work of ours suggests aspects of sleep may impact associations between subjective and objective cognition in cognitively healthy older adults. A critical next step is to distinguish how the direction (e.g., how the relationship differs if objective cognition is low vs. high) and degree (e.g., magnitude) of the discrepancy is associated with sleep, to identify those at risk for missed cognitive impairment diagnosis. Therefore, the present study extended past work by investigating how the direction and degree of the subjective-objective cognition discrepancy is related to actigraphic sleep in older adults. Methods Cognitively healthy older adults (N=37, Mage=69.4±4.9 years, 35 women) completed cognitive tasks [Stroop Color Word (inhibition), Posner Cueing (exogenous and endogenous orienting attention), Sternberg (working memory)], Cognitive Failures Questionnaire, and wore a GENEActiv actiwatch for 2 weeks [sleep onset latency (SOL), total sleep time (TST), sleep efficiency]. Polynomial regressions were used to test if the interaction between subjective and objective cognition was associated with actigraphic sleep, controlling for depressive symptoms and number of medications. If the interaction was significant, response surface analyses were used to probe the interaction to test the direction or degree of concordance and discrepancy between cognition measures across varying levels of sleep. Results The interaction between cognitive complaints and exogenous orienting attention was significantly associated with SOL (β=-.78, p=.02). Response surface analysis contrasts revealed a significant degree of discrepancy (contrast-estimate=1.86, p=.008), suggesting SOL increases as the subjective-objective cognition discrepancy gets larger in either direction. Conclusion Greater subjective-objective cognition discrepancy, regardless of direction, is associated with longer objective SOL, while a smaller discrepancy is associated with shorter Interestingly, there were no significant associations for TST or sleep efficiency. Findings may reflect a shared hyperarousal mechanism (e.g., insula or amygdala neural activity), which may interfere with prefrontal-mediated self-appraisal of cognitive functioning. Future studies should investigate SOL as a potential treatment target for reducing the discrepancy between cognition measures, with the aim of reducing missed cognitive impairment diagnosis. Support (if any) Sleep Research Society Foundation (PI: Costa); T32HL086210