Diet, intermediate risk markers and risk of type 2 diabetes

This thesis aimed to study the relation of diet with risk of type 2 diabetes and intermediate risk markers of diabetes. We investigated the effect of cis9, trans11 conjugated linoleic acid (c9,t11 CLA) supplementation on pulse wave velocity and cardiovascular risk factors in a randomized, controlled trial among 401 adults. After 6 months of supplementation, the mean ( SE) pulse wave velocity did not change in the CLA group (delta 0.00 0.07) compared with the placebo group (delta 0.09 0.06). No effects were found on cardiovascular risk factors. This study does not support an anti-atherosclerotic effect of c9,t11 CLA. Furthermore, we studied the association of dietary carotenoids intake with prevalence of metabolic syndrome in 374 men. We found that higher dietary intakes of total carotenoids, lycopene and beta-carotene related to a reduced prevalence of metabolic syndrome with 58%, 45% and 42%, respectively. Higher carotenoids intake may thus reduce the risk for metabolic syndrome. Moreover, we verified diabetes cases that were ascertained in the EPIC-NL cohort, which consists of 40,011 adults aged 20-70 years. Diabetes, ascertained via self report, linkage to hospital discharge diagnoses registers and a urinary glucose strip test, was verified against general practitioner information. In total, 2289 diabetes cases were ascertained. After verification, 67% were defined as having diabetes, 22% as non-diabetic individuals, and 15% as uncertain. Macronutrient intakes in relation to diabetes risk were studied in the EPIC-NL cohort. We found that diets high in total and animal protein may increase diabetes risk, whereas vegetable protein did not relate to diabetes risk. We also found that high glycemic index (GI) and glycemic load (GL) diets increased diabetes risk, with 8% and 27% per SD increase, respectively. Conversely to the findings in EPIC-NL, dietary GI and GL were not statistically significantly related to increased risk of diabetes in the InterAct study, which includes 12,403 diabetes cases and 16,835 sub-cohort members from eight European countries. These differences may be due to discrepancies in GI value assignments to foods reported in the FFQ. Therefore, it is difficult to draw a firm conclusion on the relation of GI and GL with diabetes risk. We studied the relation of dairy products with diabetes risk in the InterAct study. Total dairy products, milk and yogurt were not related to diabetes risk. Higher cheese intake related to a reduced diabetes risk (HR 0.84, 95%CI: 0.70, 1.01, P-value for linear trend 0.02). This study does not support a role for total dairy products in diabetes prevention, but inverse associations with specific dairy sub-types, cheese in particular, may be present. In a case-cohort study, nested in the EPIC-NL cohort, we found that higher plasma uric acid related to higher risk of diabetes, with a HR of 4.61, 95%CI: 3.38, 6.29 in the highest versus lowest quartile. Adjustment for diabetes risk factors led to attenuation (HR 1.51, 95%CI: 1.02, 2.25). A uric acid related dietary pattern did not influence the relation. This study confirms that uric acid may be an independent risk factor for diabetes.