Mendelian randomization of plasma lipidome, inflammatory proteome and heart failure

Abstract

Aims: Heart failure (HF) is a global health issue, with lipid metabolism and inflammation critically implicated in its progression. This study harnesses cutting-edge, expanded genetic information for lipid and inflammatory protein profiles, employing Mendelian randomization (MR) to uncover genetic risk factors for HF.

Methods: We assessed genetic susceptibility to HF across 179 lipidomes and 91 inflammatory proteins using instrumental variables (IVs) from recent genome-wide association studies (GWASs) and proteome-wide quantitative trait loci (pQTL) studies. GWASs involving 47 309 HF cases and 930 014 controls were obtained from the Heart Failure Molecular Epidemiology for Therapeutic Targets (HERMES) Consortium. Data on 179 lipids from 7174 individuals in a Finnish cohort and 91 inflammatory proteins from a European pQTL study involving 14 824 individuals are available in the HGRI-EBI catalogue. A two-sample MR approach evaluated the associations, and a two-step mediation analysis explored the mediation role of inflammatory proteins in the lipid-HF pathway. Sensitivity analyses, including MR-RAPS (robust adjusted profile score) and MR-Egger, ensured result robustness.

Results: Genetic IVs for 162 lipids and 74 inflammatory proteins were successfully identified. MR analysis revealed a genetic association between HF and 31 lipids. Among them, 18 lipids, including sterol ester (27:1/18:0), cholesterol, 9 phosphatidylcholines, phosphatidylinositol (16:0_20:4) and 6 triacylglycerols, were identified as HF risk factors [odds ratio (OR) = 1.037-1.368]. Cholesterol exhibited the most significant association with elevated HF risk [OR = 1.368, 95% confidence interval (CI) = 1.044-1.794, P = 0.023]. In the inflammatory proteome, leukaemia inhibitory factor receptor (OR = 0.841, 95% CI = 0.789-0.897, P = 1.08E-07), fibroblast growth factor 19 (OR = 0.905, 95% CI = 0.830-0.988, P = 0.025) and urokinase-type plasminogen activator (OR = 0.938, 95% CI = 0.886-0.994, P = 0.030) were causally negatively correlated with HF, whereas interleukin-20 receptor subunit alpha (OR = 1.333, 95% CI = 1.094-1.625, P = 0.004) was causally positively correlated with HF. Mediation analysis revealed leukaemia inhibitory factor receptor (mediation proportion: 23.5%-25.2%) and urokinase-type plasminogen activator (mediation proportion: 9.5%-10.7%) as intermediaries in the lipid-inflammation-HF pathway. No evidence of directional horizontal pleiotropy was observed (P > 0.05).

Conclusions: This study identifies a genetic connection between certain lipids, particularly cholesterol, and HF, highlighting inflammatory proteins that influence HF risk and mediate this relationship, suggesting new therapeutic targets and insights into genetic drivers in HF.

Keywords: Mendelian randomization; causality; heart failure; inflammatory proteome; lipidome; mediator.

Figure 1

Study design schematic. Two‐sample Mendelian randomization (MR) was first used to assess the causal association between 179 lipids and heart failure (HF) and to derive the total causal effect (β_total) to obtain the associated lipids. Subsequently, two‐step MR was used to assess the mediation effect of inflammatory factors, with the first step assessing the causal effect of liposomes causally associated with HF on the inflammatory proteome (β_1) and the second step assessing 91 inflammatory proteomes causally associated with HF to derive the causal effect (β_2), which thus allowed for the calculation of the mediation effect (β_1 × β_2). A–C denote the order of MR estimation performed. HERMES, Heart Failure Molecular Epidemiology for Therapeutic Targets; pQTL, proteome‐wide quantitative trait loci.

Figure 2

Mendelian randomization (MR) estimates for lipidome and heart failure (HF). (A) The volcano plot reveals that 36 lipid species exhibited significant causal associations with HF (red dots). (B) The bubble plot indicates that the causal relationship between 31 lipids and HF was confirmed by the robust MR method, of which 18 were risk factors and 13 were protective factors. OR, odds ratio.

Figure 3

Heatmap of the Mendelian randomization results for the inflammatory proteome and heart failure. OR, odds ratio. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 4

Significantly mediated Mendelian randomization estimates of lipidome and inflammatory protein levels. nsnp, number of single nucleotide polymorphism.

Figure 5

Leukaemia inhibitory factor receptor and urokinase‐type plasminogen activator identified as inflammatory mediators in the lipid–heart failure (HF) causal pathways. The arrows represent the direction of the risk effect, with red for increased levels or high HF risk and green for decreased levels or low HF risk. For example, when phosphatidylcholine (16:0_22:6) levels are elevated, leukaemia inhibitory factor receptor levels are reduced, and HF risk is increased.