This review aimed to assess the impact of behavioural therapy for insomnia administered alone (BT-I) or in combination with cognitive techniques (cognitive-behavioural therapy for insomnia, CBT-I) on depressive and fatigue symptoms using network meta-analysis. PubMed, Scopus and Web of Science were searched from 1986 to May 2015. Studies were included if they incorporated sleep restriction, a core technique of BT-I treatment, and an adult insomnia sample, a control group and a standardised measure of depressive and/or fatigue symptoms. Face-to-face, group, self-help and internet therapies were all considered. Forty-seven studies were included in the meta-analysis. Eleven classes of treatment or control conditions were identified in the network. Cohen's d at 95% confidence interval (CI) was calculated to assess the effect sizes of each treatment class as compared with placebo. Results showed significant effects for individual face-to-face CBT-I on depressive (d = 0.34, 95% CI: 0.06-0.63) but not on fatigue symptoms, with high heterogeneity between studies. The source of heterogeneity was not identified even after including sex, age, comorbidity and risk of bias in sensitivity analyses. Findings highlight the need to reduce variability between study methodologies and suggest potential effects of individual face-to-face CBT-I on daytime symptoms.
Keywords: Cognitive behavioural therapy; Depression; Fatigue; Insomnia; Network meta-analysis.
Copyright © 2017 Elsevier Ltd. All rights reserved.