Meta-Analysis of the Antidepressant Effects of Acute Sleep Deprivation.

Abstract

To provide a quantitative meta-analysis of the antidepressant effects of sleep deprivation to complement qualitative reviews addressing response rates. English-language studies from 1974 to 2016 using the keywords sleep deprivation and depression searched through PubMed and PsycINFO databases. A total of 66 independent studies met criteria for inclusion: conducted experimental sleep deprivation, reported the percentage of the sample that responded to sleep deprivation, provided a priori definition of antidepressant response, and did not seamlessly combine sleep deprivation with other therapies (eg, chronotherapeutics, repetitive transcranial magnetic stimulation). Data extracted included percentage of responders, type of sample (eg, bipolar, unipolar), type of sleep deprivation (eg, total, partial), demographics, medication use, type of outcome measure used, and definition of response (eg, 30% reduction in depression ratings). Data were analyzed with meta-analysis of proportions and a Poisson mixed-effects regression model. The overall response rate to sleep deprivation was 45% among studies that utilized a randomized control group and 50% among studies that did not. The response to sleep deprivation was not affected significantly by the type of sleep deprivation performed, the nature of the clinical sample, medication status, the definition of response used, or age and gender of the sample. These findings support a significant effect of sleep deprivation and suggest the need for future studies on the phenotypic nature of the antidepressant response to sleep deprivation, on the neurobiological mechanisms of action, and on moderators of the sleep deprivation treatment response in depression.