Network meta-analysis examining efficacy of components of cognitive behavioural therapy for insomnia

Abstract

Cognitive behavioural therapy for insomnia (CBTI) is recommended as first-line treatment for insomnia. CBT-I is a multi-component intervention comprising psychoeducation, sleep restriction, stimulus control, cognitive, and relaxation therapy. The relative efficacy of its components has yet to be investigated with state-of-the-art meta-analytic methods. PubMed, MEDLINE, PsycINFO, PsycARTICLES, and CINAHL were searched according to a pre-registered protocol using search terms indicative of insomnia and CBTI. Baseline-to-post-treatment effect sizes (Cohen's d) were calculated in a component network meta-analysis. Eighty studies representing 15,351 participants (mean age 44.9 years, 70.1 % female) were included. For the primary outcome insomnia severity, a significant positive effect for sleep restriction therapy (d = -0.45; 95 % CI: [-0.63; -0.36]) was found. Overall, the results suggest that sleep restriction therapy improves self-reported sleep continuity and sleep quality, and stimulus control therapy improves self-reported and objective total sleep time. No significant effects of psychoeducation, relaxation therapy, and cognitive therapy, and no further significant effects of any CBT-I component on objective sleep parameters were found. The most common sources of bias were a lack of blinding, missing outcome data, and the absence of study protocols. The current results suggest that sleep restriction therapy and stimulus control therapy are the most effective components of CBT-I.

Keywords: CBT-I; Cognitive behavioural therapy; Component network meta-analysis; Insomnia disorder.