Abstract 3079: The Traditional Mediterranean Diet and Incidence of Diabetes Mellitus Following Myocardial Infarction: The GISSI-Prevenzione Study
Background While risk factors to prevent CHD in diabetics are well-established, very little is known about risk factors to prevent diabetes in patients with CHD. We tested the hypothesis that a traditional Mediterranean diet would be associated with lower incidence of diabetes among patients post-MI.
Methods We used prospectively ascertained information among 8,291 patients with recent (≤3 months) MI and free of diabetes at baseline (diabetes medication use, physician diagnosis, or fasting glucose [FG] ≥126 mg/dl). Usual dietary habits were assessed at baseline and updated at 0.5 and 1.5 yrs. A Mediterranean Dietary Score was defined according to intakes of cooked vegetables, raw vegetables, fruits, fish, and olive oil. Incident diabetes was diagnosed by either new diabetic medication use or elevated FG (≥126 mg/dl) assessed at 0.5, 1, 1.5, 2.5, and 3.5 yrs. We also evaluated incidence of diabetes plus impaired glucose tolerance (IGT; FG ≥100 and <126 mg/dl), after excluding individuals with IGT at baseline. Risk was assessed using Cox proportional hazards with time-varying covariates.
Results During 26,795 person-yrs, 998 individuals (12%) developed diabetes. After multivariable adjustment, consumption of a traditional Mediterranean diet was associated with significantly lower incidence of diabetes and of diabetes+IGT (Table⇓). Differences in serum triglycerides, HDL cholesterol, systemic inflammation as assessed by fibrinogen and leukocyte count, and dietary intakes of butter/other oils (consumed as replacements for olive oil) appeared to mediate ~25% of the observed lower risk (not shown; to be presented).
Conclusions Incidence of both diabetes and IGT is very high in the years following MI (not due to undiagnosed disease, given our use of fasting glucose to define prevalent diabetes or IGT). Greater adherence to a traditional Mediterranean diet may substantially lower this risk, having implications for dietary counseling in patients post-MI.