Diabetes is a major global health concern, associated with significantly increased mortality and high incidence of co-morbidities. The lysine-derived metabolite α-aminoadipic acid (2-AAA) was identified as a novel predictor of diabetes development in Framingham Heart Study (FHS) participants and validation samples (N~2,000). In these subjects, increased plasma 2-AAA in healthy individuals was associated with increased future risk of diabetes (12-year follow-up), identifying at-risk individuals even after adjustment for known risk factors. Several subsequent studies have confirmed the association between 2-AAA and diabetes, but the mechanisms remain unknown. Preliminary data support a role for 2-AAA in insulin secretion and diabetes pathophysiology, and suggest genetic determinants of 2-AAA relate to diabetes and diabetic complications. However it is not yet clear whether 2-AAA is itself causal in diabetes development, or is a biomarker for altered metabolic processes. Many questions remain as to mechanisms of action. In this proposal, we will examine the determinants of 2-AAA, by studying lysine-2-AAA metabolism in subjects with extreme levels of 2-AAA before and after dietary intervention (Aim 1); identify the genetic predictors of 2-AAA (Aim 2); and examine the genetic architecture of 2-AAA and disease (Aim 3). These aims will advance our long-term research objective, to understand determinants of 2-AAA, and establish utility of 2-AAA and related pathways as a novel therapeutic target in diabetes.