Plasma proteome and incident myocardial infarction: sex-specific differences

Abstract

Background and aims: Few population-based cohort studies, including both men and women, have explored circulating proteins associated with incident myocardial infarction (MI). This study investigated the relationships between circulating cardiometabolic-related proteins and MI risk using cohort-based and Mendelian randomization (MR) analyses and explored potential sex-specific differences.

Methods: The discovery cohort included 11 751 Swedish adults (55-93 years). Data on 259 proteins assessed with Olink proximity extension assays, biochemical, and questionnaire-based information were used. Participants were followed up for incident MI and death over 8 years through linkage to Swedish registers. Replication analyses were conducted on the UK Biobank sample (n = 51 613). In MR analyses, index cis-genetic variants strongly related to the proteins were used as instrumental variables. Genetic association summary statistic data for MI were obtained from the CARDIoGRAMplusC4D consortium and FinnGen.

Results: Forty-five proteins were associated with incident MI in discovery and replication samples following adjustment for potential confounders and multiple testing. In the secondary analysis, 13 of the protein associations were sex-specific, with most associations identified among women. In MR analysis, genetically predicted higher levels of renin, follistatin, and retinoic acid receptor responder protein 2 were linked to an increased risk of MI. Tissue factor pathway inhibitor, tumor necrosis factor receptors 1 and 2, placenta growth factor had an inverse association with MI.

Conclusions: This study identified both new and confirmed previously established associations between circulating proteins and incident MI and, for the first time, suggested sex-specific patterns in multiple protein-MI associations.

Keywords: Cohort; Mendelian randomization; Myocardial infarction; Proteomics.

Structured Graphical Abstract

Summary of the main results of the cohort-based and Mendelian randomization analyses on the association between circulating proteins and incident myocardial infarction. FDR, false discovery rate; MI, myocardial infarction.

Figure 1

Association of 45 replicated circulating proteins with incident myocardial infarction in SIMPLER cohorts (n = 11 751). Hazard ratios (HR) with 95% confidence intervals (CI) are expressed per standard deviation unit change in protein measurements. Proteins were identified and passed FDR adjustment in the discovery sample, were validated in a replication sample (P < 0.05), and had P < 0.05 in the multivariable analysis. The models were adjusted for age (as the time scale), sex, education attainment, project, baseline cigarette smoking, alcohol consumption, walking/bicycling, exercise, adherence to the mDASH diet, body mass index, systolic blood pressure, and blood levels of fasting glucose, low-density lipoprotein cholesterol, and high-density lipoprotein cholesterol. * Protein METRNL was not available in the UK Biobank data and, therefore, was not replicated

Figure 2

Sex-specific association between replicated circulating proteins and incident myocardial infarction in SIMPLER cohorts. Results for women are shown in the panel A, and for men – panel B. Only proteins that had a statistically significant interaction term of protein × sex in relation to MI in SIMPLER and were replicated in the UK Biobank were included. Hazard ratios (HR) with 95% confidence intervals are expressed per standard deviation unit change in protein measurements. The models were adjusted for age (as the time scale), education attainment, project, baseline cigarette smoking, alcohol consumption, walking/bicycling, exercise, adherence to the mDASH diet, body mass index, systolic blood pressure, and blood levels of fasting glucose, low-density lipoprotein cholesterol, and high-density lipoprotein cholesterol. Proteins identified in the primary analysis based on the entire cohort are highlighted in bold. * Protein MMP-2 was not available in the UK Biobank data and, therefore, was not replicated

Figure 3

Potential causal associations of circulating proteins with the risk of myocardial infarction, Mendelian randomization analysis. OR, odds ratio; CI, confidence interval.