Traffic-light labels and financial incentives to reduce sugar-sweetened beverage purchases by low-income Latino families: a randomized controlled trial

Abstract

Objective: The objective of this study was to test the effectiveness of financial incentives and traffic-light labels to reduce purchases of sugar-sweetened beverages in a community supermarket. Design: In this randomized controlled trial, after a 2-month baseline period (February-March 2014), in-store traffic-light labels were posted to indicate healthy (green), less healthy (yellow), or unhealthy (red) beverages. During the subsequent 5 months (April-August 2014), subjects in the intervention arm were eligible to earn a $25 in-store gift card each month they refrained from purchasing red-labeled beverages. Setting: Urban supermarket in Chelsea, Massachusetts, a low-income Latino community. Subjects: Participants were customers of this supermarket who had at least one child living at home. A total of 148 customers (n=77 in the intervention group and n=71 in the control group) were included in the final analyses. Results: Outcomes were monthly in-store purchases tracked using a store loyalty card and self-reported consumption of red-labeled beverages. Compared to control subjects, the proportion of intervention subjects who purchased any red-labeled beverages decreased by 9% more per month (p=0.002). More intervention than control subjects reduced their consumption of red-labeled beverages (−23% vs. −2% for consuming ≥1 red beverage/week, p=0.01). Conclusions: Overall, financial incentives paired with in-store traffic-light labels modestly reduced purchase and consumption of sugar-sweetened beverages by customers of a community supermarket.

Impact of Sugar-Sweetened Beverage Warning Labels on Consumer Behaviors: A Systematic Review and Meta-Analysis

“…38 Eleven studies were conducted in the U. S. 28,31,32,34,35,38,41,47,49,50,52 ; 3 in Canada 27,29,45 ; 2 in Australia 36,47 ; 2 in New Zealand 30,39 ; and 1 each in Brazil, 33 the United Kingdom, 48 Uruguay, 37 Germany, 51 and the Netherlands. 40 Thirteen studies recruited adults, 28,29,31,32,36,39,40,[45][46][47][48][49]52 4 recruited children, 34,37,38,50 4 recruited both children and adults, 27,30,33,51 and the remaining 2 did not report the age of the participants. 35,41 Five studies recruited >70% women, 29,39,47,…”

“…40 Thirteen studies recruited adults, 28,29,31,32,36,39,40,[45][46][47][48][49]52 4 recruited children, 34,37,38,50 4 recruited both children and adults, 27,30,33,51 and the remaining 2 did not report the age of the participants. 35,41 Five studies recruited >70% women, 29,39,47,49,52 12 studies had a roughly equal split of men and women, 27,28,30,31,33,37,38,45,46,48,50,51 and the remaining 6 did not report gender distribution. 32,[34][35][36]40,41 Data collection settings varied among the 22 experimental studies: 9 were conducted online, [28][29]…”

“…35,41 Five studies recruited >70% women, 29,39,47,49,52 12 studies had a roughly equal split of men and women, 27,28,30,31,33,37,38,45,46,48,50,51 and the remaining 6 did not report gender distribution. 32,[34][35][36]40,41 Data collection settings varied among the 22 experimental studies: 9 were conducted online, [28][29][30][31]46,[48][49][50]52 5 in supermarkets, 27,34,39,45,47 4 in school or college, 32,37,40,51 2 in hospital cafeterias, 35,41 1 in both supermarkets and schools, 33 and 1 at home. 36 However, only 8 studies measured real purchases in these settings, 27,[33][34][35]39,42,45,…”

“…32,[34][35][36]40,41 Data collection settings varied among the 22 experimental studies: 9 were conducted online, [28][29][30][31]46,[48][49][50]52 5 in supermarkets, 27,34,39,45,47 4 in school or college, 32,37,40,51 2 in hospital cafeterias, 35,41 1 in both supermarkets and schools, 33 and 1 at home. 36 However, only 8 studies measured real purchases in these settings, 27,[33][34][35]39,42,45,47 whereas the remaining 15 studies asked participants to answer questions regarding their purchase intentions or choices shown on an electronic device. 28−32,36,38,40,46,48−52 Sample sizes varied substantially across studies, ranging from 148 to 32,452 participants; 4 studies used the number of beverages purchased within a specific period (rather than individuals) as the level of analysis.…”

“…Labels are classified into 6 different forms: (1) SNP, which displays a circle, octagon, or rectangle within which sugar content information is provided (e.g., "high in sugar"); (2) SHE, which displays a circle, octagon, or rectangle within which the adverse health impact of SSB consumption is provided (e.g., "Drinking beverages with added sugar contributes to obesity, diabetes, and tooth decay"); (3) TNP, which describes sugar content information in a sentence (e.g., "This product is high in sugar"); (4) THE, which describes the adverse health impact of SSB consumption (e.g., "WARNING: Drinking drinks with added sugar contributes to obesity, type 2 diabetes, and tooth decay"); (5) GHE, which displays a picture of an obese belly or decayed teeth with relevant descriptions; and (6) GNP, which displays a picture of sugar added in the drinks with corresponding descriptions. Eight studies adopted THE, 32,36,38,41,46,49,50,52 10 adopted SNP, [27][28][29][30][31]33,37,39,45,51 3 adopted GHE, 30,46,48 5 adopted SHE, 28,35,40,45,47 2 adopted GNP, 46,48 and 2 adopted TNP. 29,34 Seven studies tested multiple forms of SSB warning labels in different arms.…”