Abstract Cardiovascular diseases are the leading cause of death globally. Consequently, metabolomics studies have in recent years aimed at identifying relevant biomarkers, yet no studies have been performed in twin populations, which reduce confounding due to genetic and environmental factors. We included 11,217 twins (age at intake, 47–94 years (mean = 65)) from the Swedish Twin Registry, holding data on 173 nuclear magnetic resonance metabolomic biomarkers and nationwide register-based data on diagnoses of ischemic stroke, ischemic heart disease, myocardial infarction, angina pectoris, and coronary artery disease. Only incident cases (i.e., exclusively holding diagnosis after blood sampling) were included for statistical analyses, which were performed at the individual level and the twin-pair level by Cox regression analyses. Lastly, biomarkers significant in both analyses were inspected, hence conditioning on the twin pair design. Fifty-one biomarkers were found to be associated with myocardial infarction after correction for multiple testing, all showing a hazard ratio above 1. LASSO regression analysis of these biomarkers identified four biomarkers, all related to ApoB lipoprotein biology, potentially reflecting a pro-atherogenic effect. Investigation of biomarkers with p -values < 0.05 identified 20 biomarkers for ischemic heart disease, all showing hazard ratios below 1 and primarily related to ApoA1 lipoprotein biology, potentially reflecting an anti-atherogenic effect. Lastly, three biomarkers, i.e., acetoacetate, bOHbutyrate, and isoleucine, were found for ischemic stroke, all showing a hazard ratio above 1. Taken together, different biomarkers associated with the disease phenotypes indicate that their molecular profiles are different, despite their common basis. The ApoB stands out as a promising biomarker for myocardial infarction. Graphical abstract
Skovgaard et al. (Fri,) studied this question.