Mendelian Randomization Study Results Add to Knowns About Insomnia and Sepsis Link

Bruce G. Thorkildsen, MD

Recently published findings from a large-scale Mendelian randomization (MR) study support the idea of a potential causal association between genetically predicted insomnia and the risk of sepsis. These data were in line with previous research and results from observational data, which reported that insomnia increases the risk of altered immune response and bloodstream infection.

The analysis included 593,724 individuals of European ancestry with insomnia from the UK Biobank and 23andMe, and 10,154 cases of sepsis, also from the UK Biobank. Led by Bruce G. Thorkildsen, MD, assistant professor of medicine at St. Luke’s School of Medicine, a genetically predicted doubling in the prevalence of insomnia was associated with an odds ratio (OR) for sepsis of 1.37 (95% CI, 1.19-1.57; P = 7.6 x 10-6). The association between insomnia and sepsis was more pronounced among women compared with men (women: OR, 1.44; 95% CI, 1.24-1.68; men: OR, 1.10; 95% CI, 0.86-1.40).

Analyzed from March 2022 to March 2023, the trial used data from a genome-wide association study of 2.4 million individuals from those 2 registries that identified 55 independent genetic variants (R² <.01) strongly associated with insomnia (P <5 x 10^-8). In total, the genetic variants used in the main analysis explained 3.9% of the variance of insomnia.

Investigators first conducted a main analysis using the inverse-variance weighted (IVW) method, calculating the combined association across the Wald ratios for all single-nucleotide variants (SNVs), putting more emphasis to the SNVs with the lowest variance. To eliminate biases, the group also completed sensitivity analyses using weighted median, weighted mode, and MR-Egger regression. The proportion of the association between insomnia and sepsis was mediated through 4 strong cardiometabolic risk factors of sepsis: body mass index (BMI), type 2 diabetes (T2D), smoking, and cardiovascular disease (CVD).

READ MORE: Significantly Larger Hypothalamic Volume in Narcolepsy Type 1 Following H1N1 Infection

At the conclusion of the study, findings on the main analysis were also supported by the analysis using sepsis data from FinnGen. In addition, sensitivity analyses using weighted median, weighted mode, and MR-Egger regression also supported the main analysis. One-third of the association between genetically predicted insomnia and the risk of sepsis was mediated through BMI, T2D, smoking, or CVD, but with a likely direct association independent of these factors (OR, 1.23; 95% CI, 1.03-.145; P = .02).

When separating each cardiovascular risk factor, only BMI significantly mediated the association between genetically predicted insomnia and risk of sepsis. Despite this, investigators concluded that the large portion of the association mediated by these factors supports previous studies that these factors are potentially caused or worsened by insomnia and are also potential causes of sepsis. "However, most of the association between insomnia and risk of sepsis was not explained by these factors, indicating that insomnia may have a substantial direct influence on sepsis risk," they wrote.

In terms of strengths and limitations, Thorkldsen et al noted that, "Our findings are strengthened by the consistent results across sensitivity analyses robust to pleiotropy and the evaluation in a separate outcome cohort with no overlap with the exposure GWAS. An important limitation to our study is that it only included individuals of European ancestry, and we encourage future studies to evaluate whether our findings replicate in other ancestry groups."

REFERENCE
1. Thorkildsen MS, Gustad LT, Mohus RM, et al. Association of genetically predicted insomnia with risk of sepsis: a Mendelian randomization study. JAMA Psychiatry. Published online August 9, 2023. doi:10.1001/jamapsychiatry.2023.2717