BACKGROUND:The pandemic of cardiovascular disease (CVD) and type 2 diabetes (T2D) requires the identification of new predictor biomarkers. Biomarkers potentially modifiable with lifestyle changes deserve a special interest. Our aims were to analyze: (a) The associations of lysine, 2-aminoadipic acid (2-AAA) or pipecolic acid with the risk of T2D or CVD in the PREDIMED trial; (b) the effect of the dietary intervention on 1-year changes in these metabolites, and (c) whether the Mediterranean diet (MedDiet) interventions can modify the effects of these metabolites on CVD or T2D risk. METHODS:Two unstratified case-cohort studies nested within the PREDIMED trial were used. For CVD analyses, we selected 696 non-cases and 221 incident CVD cases; for T2D, we included 610 non-cases and 243 type 2 diabetes incident cases. Metabolites were quantified using liquid chromatography-tandem mass spectrometry, at baseline and after 1-year of intervention. RESULTS:In weighted Cox regression models, we found that baseline lysine (HR+1 SD increase = 1.26; 95% CI 1.06-1.51) and 2-AAA (HR+1 SD increase = 1.28; 95% CI 1.05-1.55) were both associated with a higher risk of T2D, but not with CVD. A significant interaction (p = 0.032) between baseline lysine and T2D on the risk of CVD was observed: subjects with prevalent T2D and high levels of lysine exhibited the highest risk of CVD. The intervention with MedDiet did not have a significant effect on 1-year changes of the metabolites. CONCLUSIONS:Our results provide an independent prospective replication of the association of 2-AAA with future risk of T2D. We show an association of lysine with subsequent CVD risk, which is apparently diabetes-dependent. No evidence of effects of MedDiet intervention on lysine, 2-AAA or pipecolic acid changes was found. Trial registration ISRCTN35739639; registration date: 05/10/2005; recruitment start date 01/10/2003.