Objective measures of sleep in adults and older adults with and without depression: a systematic review and meta-analysis

Sleep architecture is poorly defined in people with depression, especially in older adults. We investigated differences in sleep macro- and micro-architecture between adults and older adults (>50 years) with and without depression. A systematic review identified 2135 papers through PubMed, Scopus, Web of Science and Embase databases. Two reviewers excluded articles using PRISMA guidelines. Fifteen articles met inclusion criteria. A random effects model meta-analysis was performed. NICE case-control guidelines were used to assess risk of bias. In the fifteen articles, 838 participants underwent objective sleep measurement (406 depression and 432 control). All adults with depression had less total sleep time, delayed sleep onset latency, higher wake after sleep onset, shorter rapid eye movement latency (ROL) and greater rapid eye movement REM density than controls. Two of these studies examined sleep architecture in 62 older adults (31 depression, 31 control). Older adults with depression had more stage 1 sleep, less stage 2 sleep, shorter ROL than older controls with no between-group difference in sleep efficiency, REM sleep or REM density. Six studies had poor case-control selection, which when removed nullified the effects of REM density and REM latency in all adults. Sleep micro-architecture measures could not be meta-analysed due to the limited studies available. Adults with depression had moderately worse sleep initiation and continuity than controls. Alterations to REM sleep were observed in adults with depression, however these effects were not robust. Sleep macro-architecture differences were not consistent in younger adults and older adults with depression, compared to relative controls.