Abstract

Evidence suggests neutral or moderately beneficial effects of dairy intake on type 2 diabetes mellitus risk. Nevertheless, evidence on associations with early phases of type 2 diabetes remains inconsistent. We aimed to examine associations between dairy-type intake with prediabetes risk and longitudinal insulin resistance. The analytic sample consisted of 6770 participants (aged 62 +/- 4 years, 59% female) free of (pre-)diabetes at baseline from the prospective population-based Rotterdam Study. Dairy intake was measured at baseline using food frequency questionnaires. Data on prediabetes (fasting blood glucose 6.1-6.9 mmol/L or non-fasting 7.7-11.1 mmol/L) and the longitudinal homeostatic model assessment of insulin resistance (HOMA-IR) were available from 1993-2015. Associations with these outcomes were analyzed with dairy intake in quartiles (Q4 vs. Q1) and continuous using multivariable Cox proportional hazard models and linear mixed models. During a mean follow-up of 11.3 +/- 4.8 years, 1139 incident prediabetes cases were documented (18.8%). In models adjusting for sociodemographic, lifestyle and dietary factors, a higher intake of high-fat yogurt was associated with lower prediabetes risk (HRQ4vsQ1 0.70, 95% CI 0.54-0.91 and HRserving/day 0.67, 0.51-0.89). In addition, a higher intake of high-fat milk was associated with lower prediabetes risk (HRQ4vsQ1 0.81, 0.67-0.97, HRserving/day 0.88, 0.79-0.99). Associations were found for low-fat dairy, low-fat milk and total cheese with a higher prediabetes risk (HRserving/day ranging from 1.05-1.07, not significant in quartiles). Associations with longitudinal HOMA-IR were similar to prediabetes for high-fat yogurt, low-fat dairy and low-fat milk. Fermented dairy, low-fat yogurt, high-fat cheese, cream and ice cream were not associated with the outcomes. In conclusion, a higher intake of high-fat yogurt was associated with a lower prediabetes risk and lower longitudinal insulin resistance. Additionally, high-fat milk was associated with a lower prediabetes risk. Some low-fat dairy types were inconsistently associated with these outcomes. Studies are needed to confirm associations and to examine the influence of confounding by population characteristics.