Dissecting causal networks of inflammatory factors and metabolites in heart failure: A mediation Mendelian randomization study

Abstract

Heart failure (HF) represents a major global health burden, with complex pathophysiology involving inflammatory and metabolic pathways. However, the causal relationships among inflammatory factors, metabolites, and HF risk remain unclear. This study aims to identify inflammatory factors causally associated with HF risk, characterize metabolic alterations causally linked to HF development, and explore potential mediating roles of metabolites in inflammation-related HF pathogenesis. We conducted a comprehensive Mendelian randomization study using genetic data from FinnGen biobank (29,218 HF cases and 381,838 controls), combined with genome-wide association studies data for 91 inflammatory factors and 1400 metabolites. We performed bidirectional and mediational Mendelian randomization analyses to investigate causal relationships and potential mediating effects of metabolites in inflammation-related HF pathogenesis. We identified 9 inflammatory factors (out of 91) causally associated with HF risk, including 3 risk-promoting factors (interferon gamma: OR = 1.080, matrix metalloproteinase-1: OR = 1.081, tumor necrosis factor-beta: OR = 1.064) and 6 protective factors (CD40L receptor: OR = 0.954, DNER: OR = 0.943, IL-10: OR = 0.950, LIFR: OR = 0.911, TNFSF12: OR = 0.933, uPA: OR = 0.927). More than 13 metabolites (out of 1400) showed robust associations with HF risk, with N-methyl-2-pyridone-5-carboxamide demonstrating the strongest evidence (OR = 1.058). Further analysis revealed 23 significant inflammatory factor-metabolite pairs (out of 1000s of possible combinations), suggesting potential mechanistic pathways through which inflammatory factors influence HF development. This study establishes a comprehensive causal framework linking inflammation to HF through specific metabolic alterations, identifying novel biomarkers and potential therapeutic targets. The findings suggest that combined interventions targeting both inflammatory and metabolic pathways might offer improved strategies for HF prevention and treatment.

Keywords: Mendelian randomization; blood metabolites; heart failure; inflammatory factors; mediation analysis; precision medicine.

Figure 1.

Causal associations between 9 inflammatory factors and heart failure risk.

Figure 2.

Sensitivity analyses for the association between leukemia inhibitory factor receptor and heart failure (HF). (A) Scatter plot showing the relationship between single nucleotide polymorphisms (SNP) effects on leukemia inhibitory factor receptor (x-axis) and HF (y-axis), with regression lines representing different Mendelian randomization (MR) methods. (B) Funnel plot displaying the relationship between the precision and magnitude of causal estimates for individual SNPs. (C) Leave-one-out analysis showing the stability of causal estimates when excluding each SNP in turn. (D) Forest plot showing individual SNP effects and their 95% confidence intervals.

Figure 3.

Sensitivity analyses for the association between tumor necrosis factor ligand superfamily member 12 and heart failure (HF). (A) Scatter plot showing the relationship between single nucleotide polymorphisms (SNP) effects on leukemia inhibitory factor receptor (x-axis) and HF (y-axis), with regression lines representing different Mendelian randomization (MR) methods. (B) Funnel plot displaying the relationship between the precision and magnitude of causal estimates for individual SNPs. (C) Leave-one-out analysis showing the stability of causal estimates when excluding each SNP in turn. (D) Forest plot showing individual SNP effects and their 95% confidence intervals.

Figure 4.

Reverse Mendelian randomization analysis results.

Figure 5.

Forest plot of Mendelian randomization (MR) analysis for metabolites with strong evidence and heart failure (HF).

Figure 6.

Forest plot of Mendelian randomization (MR) analysis for the causal effect of inflammatory factors on metabolites.