0085 Associations Between Light Exposure and Cognitive Performance in the Multi-Ethnic Study of Atherosclerosis

Abstract Introduction Mis-timed or suboptimal light exposure patterns can disrupt sleep-wake rhythms and have been linked to adverse health outcomes. Prior work links circadian health to cognitive function, but the direct effect of light on cognition remains unclear. We investigated associations between actigraphy-derived light measures and cognitive performance in a diverse cohort of older adults, hypothesizing that higher daytime light exposure would be associated with better cognition. Methods We conducted a cross-sectional analysis using multivariable linear regression among 1,710 participants (mean age 68.4 years; 54.4% female) with valid wrist actigraphy from the MESA Sleep Ancillary Study and cognitive testing data from MESA Exam 5. White light exposure (lux) was recorded in 30-second epochs by Actiwatch Spectrum devices (Philips Respironics) and log-10 transformed prior to analysis. The primary exposure was average daytime illuminance; secondary exposures included average duration of time in ≥ 1,000 lux (TALT1000), average nighttime illuminance, and average 24-hour illuminance. “Nighttime” was defined as the 8-hour window centered around each participant’s average sleep midpoint, and “daytime” as the remaining 16-hour window. Cognitive performance was assessed using the Digit Symbol Coding (DSC) test. Covariates included sociodemographics, health, lifestyle, and sleep factors, as well as study site, and season. Results Daytime light exposure averaged 1.29 (SD 0.44) log₁₀ lux and varied across sites (mean range 1.14–1.39 log₁₀ lux) and seasons, with higher levels in summer. Higher average daytime illuminance was significantly associated with better DSC test scores in the unadjusted model (β = 6.26; 95% CI: 4.26, 8.27). The association persisted after sequential covariate adjustment (β range 1.92 - 3.12; all p < 0.05). Average 24-hour light showed similar associations. Average nighttime light (mean 0.25 log₁₀ lux, SD 0.24) was negatively associated with DSC in the unadjusted model but with adjustment the association attenuated (β = -3.10; 95% CI: -6.76, 0.55). TALT1000 was positively associated with DSC in crude models but attenuated to non-significance after adjustment. Conclusion Average daytime light was positively associated with cognitive performance, even after adjusting for multiple demographic, health, and sleep variables. These findings support the potential of optimizing daytime light environments to support healthy cognitive aging. Support (if any)