Predictive metabolites for incident myocardial infarction: a two-step meta-analysis of individual patient data from six cohorts comprising 7,897 individuals from the the COnsortium of METabolomic Studies

AIMS: Myocardial infarction (MI) is a major cause of death and disability worldwide. Most metabolomics studies investigating metabolites predicting MI are limited by the participant number and/or the demographic diversity. We sought to identify biomarkers of incident MI in the Consortium of Metabolomics Studies (COMETS).

METHODS AND RESULTS: We included 7,897 individuals aged on average 66 years from six intercontinental cohorts with blood metabolomic profiling (n = 1,428 metabolites, of which 168 were present in at least 3 cohorts with over 80% prevalence) and MI information (1,373 cases). We performed a two-stage Individual Patients Data meta-analysis. We first assessed the associations between circulating metabolites and incident MI for each cohort adjusting for traditional risk factors, and then performed a fixed effect inverse-variance meta-analysis to pull the results together. Finally, we conducted a pathway enrichment analysis to identify potential pathways linked to MI.On meta-analysis, 56 metabolites including 21 lipids and 17 amino acids were associated with incident MI after adjusting for multiple testing (false discovery rate, FDR < 0.05), and 10 were novel. The largest increased risk was observed for the carbohydrate mannitol/sorbitol (HR [95% CI] = 1.40[1.26-1.56], p-value < 0.001), whereas the largest decrease in risk was found for glutamine (HR [95% CI] = 0.74[0.67-0.82], p-value < 0.001). Moreover, the identified metabolites were significantly enriched (corrected p-value < 0.05) in pathways previously linked with cardiovascular diseases, including aminoacyl-tRNA biosynthesis.

CONCLUSIONS: In the most comprehensive metabolomics study of incident MI to date, 10 novel metabolites were associated with MI. Metabolite profiles might help to identify high-risk individuals before disease onset. Further research is needed to fully understand the mechanisms of action and elaborate pathway findings.

TRANSLATIONAL PERSPECTIVE: In the largest meta-analyses covering six international cohorts, we identify 10 novel and 46 known metabolites associated with incident MI, that can be used to identify at-risk individuals before disease onset. Our results improve our understanding of the molecular changes that take place in MI development and provide potential novel targets for clinical prediction and a deeper understanding of causal mechanisms.