ObjectivesTo elucidate potential pathways of dietary risk of type 2 diabetes (T2D) by defining a model that describes how identified patterns of habitual beverage consumption associate with identified networks of circulating cardiometabolic plasma biomarkers. MethodsThis study included 1,231 cases and 1,560 controls from a nested case-control study of T2D within the Nurses’ Health Study I. Participants completed validated food frequency questionnaires assessing their habitual beverage intake and provided plasma samples to assess 27 different plasma biomarkers of cardiometabolic risk. Common exploratory factor analysis (EFA) was used to identify common factors that separately described beverage consumption patterns and biomarker networks. False discovery rate corrected multivariable-adjusted regression elucidated the relationships between beverage and biomarker factors, and their associations with T2D risk. ResultsEFA revealed five factors describing beverage consumption patterns and seven factors explaining underlying biomarker networks. Preferential consumption of alcoholic beverages was associated with reduced risk of T2D (P < 0.0001) and lower concentrations of detrimental biomarkers of endothelial dysfunction (P = 0.014). Also, a preferential consumption of low-calorie sweetened beverages (LCSBs) was associated with increased risk of T2D (P < 0.0001), and lower concentrations of insulin-like growth factor binding protein 1 and 2, and soluble leptin receptor (P = 0.010). ConclusionsThese analyses provide new mechanistic insight into how different patterns of beverage consumption may relate to T2D risk, and indicate that preferential consumption of LCSB may exert diabetogenic effects through mechanisms involving the disruption of insulin-like growth factor and leptin signalling. Funding SourcesNational Institutes of Health.
Read full abstract