Do non‐pharmacological sleep interventions affect anxiety symptoms? A meta‐analysis

Abstract

Research indicates a bidirectional relationship between sleep and anxiety, with findings suggesting anxiety can precede poor sleep and vice versa. Evidence suggests sleep-related thought processes associated with anxiety are involved in the maintenance of insomnia. Previous meta-analyses provide some evidence to suggest cognitive behavioural therapy for insomnia moderately improves anxiety, yet little research has investigated the effect of other sleep interventions on anxiety symptoms. The aim of this meta-analysis was to review whether non-pharmacological sleep interventions have an impact on anxiety symptoms immediately post-intervention. A systematic search of electronic databases was conducted to identify all randomized control trials (RCTs) investigating non-pharmacological sleep interventions that included anxiety symptoms as an outcome. Forty-three RCTs (n = 5945) met full inclusion criteria and were included in a random-effects meta-analysis model. The combined effect size of non-pharmacological sleep interventions on anxiety symptoms was moderate (Hedges' g = -0.38), indicating a reduction in symptoms. Subgroup analyses found a moderate effect for those with additional physical health difficulties (g = -0.46), a moderate effect for those with additional mental health difficulties (g = -0.47) and a moderate effect for those with elevated levels of anxiety at baseline (g = -0.43). A secondary meta-analysis found a large effect of non-pharmacological sleep interventions on sleep-related thought processes (g = -0.92). These findings indicate non-pharmacological sleep interventions are effective in reducing anxiety and sleep-related thought processes, and these effects may be larger in patients with anxiety. This has clinical implications for considering sleep interventions in the treatment of anxiety.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 38 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.