Variability in Sleep is Associated with Inflammatory Biomarkers: Results from the Einstein Aging Study

Abstract Background Prior research links poor subjective sleep and higher systemic inflammation, both of which are risk factors for cognitive decline in older adults. To expand upon previous findings, we investigated associations between multiple objective aspects of sleep health (including sleep variability) and inflammation (including stimulated inflammatory markers) among a diverse sample of older adults. Method Participants free of dementia (N = 218, mean age = 77.5 years, 66% females; 46% White, 38% Black, 14% Hispanic/other, MCI positive = 61) enrolled in The Einstein Aging Study were included. Wrist actigraphy (Spectrum Plus) over 16 days yielded measures of sleep disturbance by wake after sleep onset (WASO), sleep duration as total night sleep time (TST), sleep timing (midpoint of night sleep), and variability of each sleep measure. Inflammatory markers included circulating CRP and interleukin (IL)‐1β, IL‐4, IL‐6, IL‐8, IL‐10, and tumor necrosis factor (TNF)‐𝛼 plus lipopolysaccharide (LPS)‐stimulated levels of these same cytokines. Separate linear regression models evaluated the association of each inflammatory marker and the average level and variability of each sleep measure. Models controlled for BMI, gender, age, education, and race/ethnicity. Result Circulating inflammatory markers were generally not associated with objective sleep measures. However, after controlling for covariates, associations were observed between shorter average WASO and higher IL‐4, later average night sleep midpoint and higher IL‐6, and more variability in TST with higher IL‐1β. For stimulated cytokines, more WASO variability was associated with higher levels of all stimulated cytokines except IL‐6 and IL‐1β. More variability in sleep duration (TST) was associated with higher IL‐8 (but not other cytokines). Mean and variability of sleep timing, and mean WASO and TST were not significantly associated with any of the stimulated cytokines. Conclusion The present findings add substantially to research suggesting that sleep patterns are related to inflammation in older adults. Stimulated cytokines [IL‐4, IL‐10, IL‐8, TNF𝛼] were more strongly related to variability measures than mean sleep measures. Since stimulated cytokine levels reflect inflammatory reactivity to immune challenge, the present findings suggest future work should explore underlying relationships between poor sleep and central disruption, and possibly neurodegeneration, in older adults. Specifically, poor sleep (greater sleep variability) might contribute to higher reactivity of neuroinflammation pathways.