A systematic review and meta-analysis of placebo versus no treatment for insomnia symptoms

Abstract

This systematic review and meta-analysis aimed to determine the size of the placebo effect for insomnia symptoms when comparing placebo treatment with no treatment. PsycINFO, MEDLINE, and CINAHL databases were systematically searched for studies allocating participants with insomnia symptoms (diagnosed or self-reported) to receive a placebo that they were led to believe was an active treatment or to a no treatment control group. Thirteen independent studies (n = 566) met inclusion criteria. Meta-analysis indicated a reliable placebo effect whereby placebo treatment led to improved perceived sleep onset latency (SOL; Hedges g = 0.272), total sleep time (TST; Hedges g = 0.322), and global sleep quality (GSQ; Hedges' g = 0.581), when compared with no treatment. There was no effect on objective assessment of SOL, however only a few studies reported this outcome and there were insufficient sample sizes to meta-analyse other objective outcomes. Moderator analysis indicated that the placebo effect for perceived insomnia symptoms was quite consistent across different variables. The present findings provide strong evidence for placebo effects for perceived insomnia symptoms, but not on the only objective measurement with sufficient sample size to meta-analyse, namely objective SOL. This has important implications for the treatment of insomnia symptoms and the design and interpretation of clinical trials for insomnia symptoms. [Abstract copyright: Copyright © 2017 Elsevier Ltd. All rights reserved.]