Causal relationships between obesity-related anthropometric indicators and sepsis risk: a Mendelian-randomization study.

Background: Previous studies have reported an association between obesity and risk of sepsis. However, the results have been inconsistent, and no causal inference can be drawn from them. Therefore, we conducted a Mendelian-randomization (MR) study to investigate causal relationships between available obesity-related anthropometric indicators and sepsis risk.

Methods: We performed MR analyses using genome-wide association study (GWAS) summary statistics on 14 anthropometric indicators [namely body mass index (BMI), waist and hip circumferences (WC, HC), basal metabolic rate (BMR), whole-body fat mass (WBFM), trunk fat mass (TFM), leg fat mass (LFM), arm fat mass (AFM), body fat percentage (BFP), whole-body fat-free mass (WBFFM), trunk fat-free mass (TFFM), leg fat-free mass (LFFM), arm fat-free mass (AFFM), and whole-body water mass (WBWM)], sepsis, critical care sepsis, and 28-day death due to sepsis from the UK Biobank and FinnGen cohort. The primary method of MR analysis was inverse variance-weighted average method. Sensitivity analyses, including heterogeneity and horizontal-pleiotropy tests, were conducted to assess the stability of the MR results. Additionally, we applied multiple-variable MR (MVMR) to evaluate the effect of BMI on the relationship between each anthropometric indicator and sepsis risk.

Results: Our MR analysis demonstrated causal relationships between 14 anthropometric indicators and sepsis of different severities. After we adjusted for BMI, MVMR analyses indicated that WC, BMR, LFM, WBFFM, TFFM, AFFM, and WBWM remained significantly associated with the presence of sepsis (all  < 0.05). A sensitivity analysis confirmed the reliability of our MR results, and no significant horizontal pleiotropy was detected.

Conclusion: This MR study revealed that increases in obesity-related anthropometric indicators had causal associations with a higher risk of sepsis, which might provide important insights for the identification of individuals at risk for sepsis in community and hospital settings.