Abstract 044: Differential Effects of Sugar-sweetened, Artificially Sweetened, and Unsweetened Beverages on Taste Preference but Not CVD Risk Factors in a 12-Month RCT

Abstract

Introduction: Given the adverse health effects of added sugars, particularly in liquid form, dietary guidelines recommend replacing sugar-sweetened beverages (SSB) with non-caloric options. However, data are limited regarding the effects of artificially sweetened (ASB) compared to unsweetened (USB) beverages. Hypothesis: We tested the hypothesis that, as replacements for SSB, USB vs. ASB would produce greater reductions in CVD risk factors (USB<ASB<SSB), body weight and adiposity (USB<ASB=SSB), and preference for sweet taste (USB<ASB=SSB). Methods: Participants were adults aged 18 to 40 y who habitually consumed SSB. We randomly assigned 203 participants (121 males, 82 females) to three groups, and delivered SSB, ASB, or USB to their homes for 12 months. Outcomes included ratio of serum triglyceride to HDL-cholesterol (TG:HDLC, primary outcome), insulin sensitivity (HOMA-S), body weight, and fat mass (DXA). To assess preference for sweet taste, we obtained sweetness ratings for sucrose solutions ranging in concentration (mass/volume) from 0% to 18% using a 10-cm visual analog scale (0=not at all sweet; 10=extremely sweet). We fitted a logistic curve to each participant’s ratings, and defined sweetness threshold as the concentration corresponding to 5 cm on the scale. Results: Participant retention was 92% (186 of 203). Consumption of assigned beverages increased by 1 to 2 servings/d (12 fl oz per serving) for each group (P<0.001), and consumption of SSB declined to almost 0 for those receiving ASB or USB. Neither TG:HDLC nor HOMA-S changed significantly for any group (P>0.20). Change in body weight did not differ between groups (P=0.65) despite observed weight gain with consumption of SSB (mean±SE 1.2±0.6 kg, P=0.03) but not with ASB or USB (both 0.6±0.6 kg, P>0.20). Consistent with an a priori hypothesis, we found effect modification by a measure of baseline adiposity (P=0.01). Body weight increased by 0.48±0.17 kg more for each kg of trunk fat with consumption of SSB vs. USB (P=0.005). The increase was 0.39±0.17 kg/kg with SSB vs. ASB (P=0.02). A similar pattern of effect modification was observed for change in fat mass. Change in sweetness threshold differed between groups (P=0.03), due to a decrease with consumption of USB (median -0.5%, interquartile range 1.3, P=0.005) and no change with SSB or ASB. Conclusions: Replacing SSB with non-caloric beverages had no effect on pre-specified CVD risk factors. Among individuals with central adiposity, replacing SSB with either ASB or USB had a favorable effect on body weight, consistent with prior findings. However, USB were a better replacement than ASB for altering taste preference, a finding with plausible implications for promoting adherence to prescribed low-sugar diets. Consistent with epidemiologic data, the benefits of eliminating consumption of SSB on CVD risk factors may require longer periods of study for the general population.