Summary
Depression is associated with disturbances to sleep and the 24-hour sleep-wake pattern (known as the rest-activity rhythm or RAR). However, there remains a need to identify the specific sleep/RAR correlates of depression symptom severity in zpopulation sub-groups, such as strained dementia caregivers, who are at elevated risk for major depressive disorder. We assessed the cross-sectional associations of sleep/RARs with non-sleep depression symptom severity among n=57 (mean age=74, standard deviation=7.4) strained dementia caregivers who were currently without clinical depression. We derived sleep measures from polysomnography and actigraphy, modeled RARs using a sigmoidally transformed cosine curve, and measured non-sleep depression symptom severity using the Hamilton Depression Rating Scale (HRDS) with sleep items removed. The following sleep-wake measures were associated with greater depression symptom severity (absolute Spearman correlations ranged from 0.23–0.32): more time awake after sleep onset (WASO), higher RAR middle level (mesor), relatively shorter active periods (alpha), earlier evening settling time (down-mesor), and less steep RARs (beta). In multivariable analysis, high WASO and low RAR beta were independently associated with depression symptom severity. Predicted non-sleep HDRS means (95% confidence intervals) in caregivers with and without these characteristics were: normal WASO/beta=3.7 (2.3–5.0), high WASO/normal beta=5.5 (3.5–7.6), normal WASO/low beta=6.3 (3.6–8.9), high WASO/low beta=8.1 (5.3–10.9). Thus, in our sample of strained caregivers, greater sleep fragmentation (WASO) and less sustained/sharply segregated resting and active periods (low RAR beta) uniquely correlate with depression symptom severity. Longitudinal studies are needed to establish whether these independent sleep-wake correlates of depression symptoms explain heightened depression risk in dementia caregivers.