Abstract

Plenary Lecture: Integrating Metabolite Profiles for Less Impersonal Medicine

Robert Gerszten, Principal Investigator, Massachusetts General Hospital

Insulin resistance and endothelial dysfunction typically precede type 2 diabetes mellitus (DM) by decades. Thus, a substantially elevated risk of cardiovascular events can be observed well before overt hyperglycemia is present. This highlights the importance of early identification of individuals at elevated cardiometabolic risk, particularly since the delay or prevention of cardiometabolic disease is possible via both behavioral and pharmacological approaches. An emerging set of technologies, based on mass spectrometry, enables the monitoring of hundreds of metabolites from biological samples (“metabolomics”). Combining biomarker and phenotypic data in human populations provides a rich opportunity to identify metabolic signatures of cardiometabolic diseases, which may have implications for both understanding biology and clinical prevention. In prior studies of the Framingham Heart Study (FHS) Offspring Cohort, we have identified and validated metabolite profiles of those destined to develop overt diabetes. The strongest individual predictors of future DM included branched chain amino acids, aromatic amino acids, specific triacylglyerol species, and 2-aminoadipic acid (a putative lysine degradation product). These metabolites predict DM above and beyond clinical risk factors and biochemical markers. By integrating metabolite data with genome wide arrays for common variants, we have identified 23 novel genetic determinants of human metabolism, including 8 loci previously implicated in human diseases. All of our data have been made publicly available. In addition to highlighting recently completed and ongoing studies, this presentation will also review limitations of prior work and future directions.