Insomnia and multiple health outcomes: umbrella review of meta-analyses of prospective cohort studies

T T Wu,Y L Zou,K D Xu,X R Jiang,M M Zhou,S B Zhang,C H Song
DOI: https://doi.org/10.1016/j.puhe.2022.11.021
IF: 4.984
Public Health
Abstract:Objectives: To evaluate existing evidence of prospective cohort studies on associations between insomnia and multiple health outcomes. Study design: An umbrella review of meta-analyses of prospective cohort studies. Methods: A systematic search was undertaken in Pubmed, Embase, Cochrane, and Web of Science from inception to October 2021 to find meta-analyses of prospective cohort studies investigating the association of insomnia with any health outcome. The summary relative risk (SRR) for each meta-analysis was recalculated with random-effects model. The methodological quality and the quality of evidence were assessed by the A Measurement Tool to Assess Systematic Reviews and Grading of Recommendations, Assessment, Development and Evaluation, respectively. Results: A total of 25 published meta-analyses of prospective cohort studies, reporting 63 SRRs for 29 unique outcomes were included. Insomnia was mainly related to cardiovascular outcomes and mental disorders. The former comprised atrial fibrillation (SRR: 1.30, 95% confidence interval: 1.26 to 1.35), cardiovascular diseases (1.45, 1.29 to 1.64), coronary heart disease (1.28, 1.10 to 1.50), myocardial infarction (1.42, 1.17 to 1.72), and stroke (1.55, 1.39 to 1.72). The latter involved alcohol abuse (1.35, 1.08 to 1.67), all mental disorders (2.16, 1.70 to 3.97), anxiety (3.23, 1.52 to 6.85), depression (2.31, 1.90 to 2.81), suicidal ideation (2.26, 1.79 to 2.86), suicidal attempt (1.99, 1.31 to 3.02), and suicidal death (1.72, 1.42 to 2.08). Besides, insomnia enhanced the risk of Alzheimer's disease (1.51, 1.06 to 2.14) and hyperlipidemia (1.64, 1.53 to 1.76). Conclusion: Insomnia exhibits considerable adverse outcomes, primarily comprises cardiovascular outcomes and mental disorders, but further studies with robustly designed trials are needed to draw firmer conclusions.
What problem does this paper attempt to address?