Insomnia, sleep duration, and risk of anxiety: A two-sample Mendelian randomization study

Abstract

Background: The effect of insomnia and sleep duration on risk of anxiety has been assessed based on traditional epidemiological studies. However, the inconsistent conclusions do not establish causal associations. This study aimed to explore the causal associations of insomnia, short sleep, and long sleep with anxiety.

Methods: We used summary statistics from three large-scale genome-wide association studies (GWAS) of European ancestry to perform a two-sample Mendelian randomization (MR) study. MR analyses were mainly conducted with the inverse-variance-weighted (IVW) method. To evaluate the robustness of our findings, we performed the weighted-median approach, the MR-Egger method, and the MR-robust adjusted profile score (MR-RAPS) method for sensitivity analyses.

Results: There was a statistically significant association of genetically predicted insomnia with anxiety using the IVW method (OR = 1.36, 95% CI = 1.23-1.51, P < 0.001). Genetically predicted short sleep was potentially associated with anxiety using IVW method (OR = 1.67, 95% CI = 1.08-2.60, P = 0.022). However, sensitivity analyses did not find the causal association of short sleep with anxiety (all P > 0.053). We did not observe a statistically significant causal association of genetically predicted long sleep with anxiety (OR = 0.91, 95% CI = 0.48-1.74, P = 0.775).

Conclusions: We found strong evidence that insomnia but not short sleep and long sleep has a causal effect on anxiety. The characteristics of insomnia should be incorporated into anxiety prevention and intervention strategies, which have important public health significance.

Keywords: Anxiety; Causal association; Insomnia; Mendelian randomization; Single nucleotide polymorphism; Sleep duration.