Abstract P293: The Effect of Diet Beverage Intake on Measures of Diabetes Control: A Pilot Study

Abstract

Introduction: Diet beverages are calorie free beverages sweetened with non-nutritive sweeteners (NNS). People with diabetes are at high macro and microvascular risk and are the highest per capita consumers of diet beverages as they tend to consume them directly in place of sugar sweetened beverages. This behavior has been endorsed by dietetic and scientific organizations; and diet beverages are marketed synonymously with better health, suitable for weight loss, and thus advantageous for diabetes control. The underlying public health concern is the lack of data to support or refute this concept. To begin addressing this evidence gap we carried out a randomized, two period crossover trial testing the effect of habitual diet beverage intake compared to habitual bottled water intake (avoidance of all NNS) on clinical markers of diabetes control and potential mechanistic pathways. Hypothesis: Habitual diet beverage intake will raise glycemia, lower kidney function, and lower diet quality compared to water intake. Methods: We carried out a randomized, two period crossover trial of 12 adults with type 2 diabetes who were being treated with oral medications and were free of other major chronic disease. In two separate 4 week periods with a washout period between them, participants were randomized to consume 24 oz. a day of a diet beverage (DB) of choice in 1 period and 24 oz. a day of bottled water (H2O) in the other while maintaining usual lifestyle habits and diabetes treatment regimens across periods. The primary outcome was change in glycemia (fructosamine). Secondary outcomes were kidney function (eGFR via serum creatinine) and diet quality (healthy eating index score, HEI-2010). All clinical measures were collected by standardized protocol and laboratory measurement. Diet was assessed by 3 random 24 hour recalls each period (1 weekend, 2 weekday) using a web based automated self-administered platform and diet quality was calculated by estimating the HEI-2010. Mixed-model repeated measures linear regression was used to analyze the continuous outcomes data. Results: 10 women and 2 men, mean (SD) age 58.3 (7.9) years with mean (SD) HbA1c % of 7.1 (1.4) at baseline were randomized and completed the study. Glycemia (fructosamine) was reduced, mean (SE) of -3.9 (3.2) umol/L during the H2O period and increased 0.6 (3.2) umol/L during the DB period, p=0.36. The results for kidney function (eGFR) were: H2O: -0.4 (3.7) v. DB: -2.3 (-3.7) mL/min/1.73m 2 , p=0.72. Diet quality was lower during the DB period relative to the H2O period (58.0 v. 61.5, p=0.25). Conclusion: Although the magnitude and precision of the treatment effects limits the inference, these results suggest the need for larger and longer trials to more definitively test this hypothesis and inform the role diet beverages may have in diabetes control.