A high-throughput NMR metabolomics platform8 was used for the quantification of 68 lipid and abundant metabolite measures from baseline serum samples of the FINRISK, SABRE, and BWHHS cohorts. All metabolites were measured in a single experimental setup, which allows for the simultaneous quantification of both routine lipids, total lipid concentrations of 14 lipoprotein subclasses, fatty acid composition such as monounsaturated (MUFA) and polyunsaturated fatty acids (PUFA), various glycolysis precursors, ketone bodies and amino acids in absolute concentration units (Supplemental Table 1).8 (link) The targeted metabolite profiling therefore includes both known metabolic risk factors and metabolites from multiple physiological pathways, which have not previously been examined in relation to CVD risk in large population studies. The 68 metabolite measures were assessed for association with incident CVD events using a hypothesis-generating biomarker discovery approach with subsequent replication in two independent cohorts. Spearman’s correlations of the metabolites are shown in Supplemental Figure 1. The NMR metabolomics platform has previously been used in various epidemiological studies9 ,10 (link),16 (link),17 (link),20 (link)–22 (link),31 (link),32 (link), details of the experimentation have been described9 ,24 (link), and the method has recently been reviewed.8 (link),19 (link)A subset of 679 serum samples from the FINRISK study were additionally profiled with liquid-chromatography mass spectrometry (LC-MS) using the Metabolon platform33 (link) in a case-cohort design for comparison of biomarker associations with incident CVD (expanded methods online). The biomarker associations were further compared with those obtained by LC-MS-based profiling of the Framingham Offspring Study (fifth examination cycle, n=2289 fasting plasma samples), as described in detail previously.13 (link),14 (link) Since several fatty acid biomarkers were not measured by LC-MS, the quantification was analytically confirmed by comparing NMR and gas chromatography in the Cardiovascular Risk in Young Finns Study (YFS, n=2193 fasting serum samples).34 (link) Metabolite profiling data collected at two-time points in YFS9 was further used to examine associations of dietary intake with the circulating biomarkers, and tracking of concentrations within the same individuals over 6 years.