Causal Relationships Between Soluble ST2, Heart Failure, and Sepsis: Bidirectional and Multivariable Mendelian Randomization Analyses
Preprint
Abstract
BACKGROUND: Soluble ST2 (sST2) predicts poor outcomes in heart failure (HF) and sepsis, but does it actually drive these conditions or just tag along for the ride? We used bidirectional Mendelian randomization (MR) to find out, testing six directional causal pathways among sST2, HF, and sepsis. METHODS AND RESULTS: Our approach involved bidirectional two-sample Mendelian Randomization(MR) analyses drawing on genome-wide association study (GWAS) summary statistics: sST2 data came from deCODE Genetics (n=30,931), HF data from the Heart Failure Molecular Epidemiology for Therapeutic Targets(HERMES) Consortium (n=977,323), and sepsis data from FinnGen R12 (n=500,348) . We looked at six directions. IVW was the main method, with (Mendelian Randomization–Egger regression)MR-Egger, weighted median, (Mendelian Randomization Pleiotropy RESidual Sum and Outlier)MR-PRESSO sensitivity tests, and multivariable MR (MVMR). We did a cis-SNP analysis too, using only IL1RL1 variants. None of the six pathways showed a causal effect. Genetically predicted sST2 had no link to sepsis (OR: 1.01; 95% CI: 0.94–1.08; P=0.869) or HF (OR: 0.99; 95% CI: 0.92–1.07; P=0.867).cis-SNP analysis gave the same answer for sST2→sepsis (OR: 1.03; 95% CI: 0.98–1.09; P=0.223). MVMR adjusting for HF: still nothing (OR: 1.01; 95% CI: 0.93–1.09; P=0.862). HF and sepsis did not cause each other either - HF→sepsis (OR: 0.98; 95% CI: 0.79–1.23; P=0.882), sepsis→HF (OR: 1.07; 95% CI: 0.96–1.20; P=0.236). Going the other way, genetic risk for HF or sepsis did not affect sST2 levels. CONCLUSIONS: We found no evidence that sST2 causes HF or sepsis. The picture that emerges is one where sST2 goes up because patients are sick-not the other way around. This makes sST2 a useful prognostic signal but probably not something worth targeting with drugs.We also found no direct causal link between HF and sepsis. Clinical Perspective: What Is New?: This bidirectional Mendelian randomization study is the first to comprehensively examine potential causal relationships among soluble ST2, heart failure, and sepsis using genetic instruments. We found no evidence that genetically predicted sST2 levels causally influence the risk of heart failure or sepsis, despite strong observational associations reported in clinical studies. Similarly, genetic liability to heart failure or sepsis does not appear to causally affect circulating sST2 levels, suggesting sST2 elevation is a consequence rather than a cause of these conditions. What Are the Clinical Implications?: These findings suggest that sST2 functions primarily as a prognostic biomarker reflecting disease severity rather than as a driver of pathophysiology, which has implications for its use in clinical decision-making. The strong observational associations between sST2 and adverse cardiovascular outcomes likely reflect confounding or reverse causation rather than direct causal effects. Drug development efforts targeting the ST2/IL-33 signaling pathway should consider that modulating sST2 levels may not directly prevent heart failure or improve sepsis outcomes.Keywords
This publication has 18 references indexed in Scilit:
- Research Progress of Biomarkers for Sepsis and Precision MedicineEmergency Medicine International, 2025
- Strengthening the reporting of observational studies in epidemiology using mendelian randomisation (STROBE-MR): explanation and elaborationBMJ, 2021
- Genome-wide association and Mendelian randomisation analysis provide insights into the pathogenesis of heart failureNature Communications, 2020
- Clinical and Prognostic Significance of sST2 in Heart FailureJACC, 2019
- sST2 Predicts Outcome in Chronic Heart Failure Beyond NT−proBNP and High-Sensitivity Troponin TJACC, 2018
- Genomic atlas of the human plasma proteomeNature, 2018
- The Biology of ST2: The International ST2 Consensus PanelThe American Journal of Cardiology, 2015
- Mendelian randomization: genetic anchors for causal inference in epidemiological studiesHuman Molecular Genetics, 2014
- Avoiding bias from weak instruments in Mendelian randomization studiesInternational Journal of Epidemiology, 2011
- Soluble ST2 plasma concentrations predict mortality in severe sepsisIntensive Care Medicine, 2010