Abstract MP95: Association between Artificially Sweetened Beverages and the Incidence of Type 2 Diabetes in Postmenopausal Women
Background. Evidence demonstrating a relationship between artificially sweetened beverages (ASB) and the incidence of type 2 diabetes (T2D) is limited with mixed results. Among postmenopausal women and independent of their baseline adiposity, sugar sweetened beverages (SSB), and physical activity, ASB consumption is hypothesized to be associated with an increase in T2D incidence.
Methods. Participants were 62,082 postmenopausal women, aged 50-79 at enrollment, from the Women’s Health Initiative Observational Study with plausible dietary intake calculated from the FFQ and free of T2D before the annual visit 3 (1996-2001). ASB was self-reported as intake of diet drinks (e.g. 12 oz of Diet Coke or diet fruit drinks) during the past 3 months. Servings of ASB were condensed into 5 groups: never or <3/month (reference), 1-6/week, 1/day, 2-3/day, or >3/day. Incident T2D cases were defined as reported use of any hypoglycemic medication anytime after AV3 up to September 2010. Total energy was calculated from the FFQ. Multivariable logistic regression assessed associations of ASB intake and T2D incidence. To control for residual confounding by adiposity, multivariate models were stratified by BMI categories and adjusted for BMI as a continuous variable.
Results. There were 4,649 incident T2D cases. Compared to the reference, the odds of developing T2D at higher levels of ASB intake were: OR=1.06, 95% CI 0.98-1.14; OR=1.30, 95% CI 1.17-1.44; OR=1.34, 95% CI 1.18-1.52; OR=1.41, 95% CI 0.85-2.35; respectively in fully adjusted models (p trend = 0.08). Fully adjusted models, stratified by category of BMI, are given in the table below.
Conclusion. In a dose-response fashion, women with higher ASB consumption had a greater risk of incident T2D independent of BMI, physical activity, SSB, and total energy. This association was strongest among obese women. Caution in interpreting these results is recommended due to potential underreporting of energy since bio-marker calibrated energy intake was not utilized in the present analysis.
- © 2013 by American Heart Association, Inc.