539 Interactions Between Cardiorespiratory Fitness and Sleep Apnea in Predicting Risk of Alzheimer’s Disease

Abstract Introduction Several studies suggest a link between obstructive sleep apnea (OSA) and Alzheimer’s disease (AD). Additionally, frequent exercise is associated with more favorable AD biomarker profiles, and emerging evidence suggests greater cardiorespiratory fitness levels may be associated with lower risk of cognitive decline. We investigated whether cardiorespiratory fitness modifies the association of OSA and risk of AD. Methods A subset of the Wisconsin Sleep Cohort study participants with study visits starting in 2000 (n=1182, 46% female, mean [range] age at baseline=57 [37–82] years) completed multiple [range, 1–5] in-laboratory protocols that included overnight polysomnography, anthropometric measurements, and questionnaires. Additionally, the National Death Index was searched to determine cause of death among decedents. Cox proportional hazards models estimated relative hazards of AD (self-reported physician diagnosis or indication of AD on the death certificate) associated with the joint effects and the interaction of OSA – characterized by the base 10 logarithm of the apnea-hypopnea index (log10(AHI+1)) – and cardiorespiratory fitness (an index based on age, sex, BMI, self-reported physical activity, and resting heart rate). Additionally, the sample was stratified by fitness level at the 3rd quartile (>75th percentile compared to <75th percentile) and the hazard ratio for log10(AHI+1) was estimated for the lower and higher fitness-level groups. Results were adjusted for age, sex, BMI, and education. Results There were 10 incident cases of AD. The mean [range] fitness level was 7.1 [0–12.3]. 28% of the sample had moderate OSA (AHI 5–15); and 26% had severe OSA (AHI>15). Higher log10(AHI+1) was associated with greater hazards (p=0.03) of AD and there was a significant interaction between log10(AHI) and fitness (p=0.04), such that at greater fitness levels, the effect of log10(AHI) on AD was mitigated. In stratified analysis, among the less fit, the hazard ratio for an increment of 1 in log10(AHI) was 12.8 (95% CI, 1.1–153.8, p=0.04); among those who were more fit, the hazard ratio was not significant. Conclusion More severe OSA is associated with higher risk of AD, and this risk is greater among those with lower levels of cardiorespiratory fitness. Support (if any) This work was supported by US NIH grants R01AG058680, R01AG062167, R01HL62252, 1R01AG036838 and 1UL1RR025011.