Evaluating the impact of sugar warnings on beverage purchases in Hispanic populations - ABSTRACT Hispanic populations in the US have higher rates of obesity, diabetes, and nonalcoholic fatty liver disease compared to non-Hispanic white populations. These disparities are partially driven by higher intake of sugar- sweetened beverages (SSBs) among Hispanic adults. Warning labels on product packaging are a widely used public health strategy that could reduce SSB intake. Initial research on SSB warnings is promising. However, the effects of SSB warnings on behavior are unknown among Hispanic adults, a priority population for prevention of obesity, diabetes, and nonalcoholic fatty liver disease. Our overarching objective is to estimate the impact of sugar warnings on SSB purchases among Hispanic adults. In Aim 1, we will examine the impact of sugar warnings on SSB purchases in an RCT with Hispanic adults in North Carolina (n=550). The RCT will take place in the Mini Mart, a small convenience store designed to conduct research studies, allowing for experimental control in a realistic setting where participants spend actual money on real products. Participants will be randomly assigned to one of two trial arms: SSBs with sugar warning labels or SSBs with neutral control labels. At each of four visits over four weeks, participants will shop for beverages for their household. We will estimate the impact of sugar warnings on the primary outcome: total sugar purchased from SSBs as well as secondary outcomes, such as purchases of juice or drinks with non-caloric sweetener. Additionally, we will identify the psychological mechanisms explaining how sugar warnings influenced beverage purchases using in-depth qualitative interviews with a subset of RCT participants and longitudinal survey data from the RCT. In Aim 2, we will evaluate a sugar warning policy implemented in Mexico in 2020 to examine whether warnings were associated with reduced SSB purchases in the real world. We will use a longitudinal panel of 9,300 households over 5 years (2017-2021), with weekly beverage purchases linked to nutrition data. Using structural demand models, we will estimate the policy-linked changes in SSB purchases after Mexico's sugar warning policy was implemented. We will also estimate how much of the changes in SSB purchases were attributable to changes in consumer behavior (e.g., consumers switching products) or industry response (e.g., the food industry lowering the sugar content of beverages to avoid warning labels). This evaluation will allow us to move beyond the lab to understand how consumers react in the real world over a longer time, and also to measure the additional effects of warning policies on SSB purchases via changes to the products themselves. Together, the two studies will provide estimates of how sugar warnings could affect SSB purchases as well as mechanisms of the impact of sugar warning policies. To leverage the complementary data of both projects, we will bring together our Mexican collaborators and local Hispanic community partners to identify findings across aims that are most likely to have an immediate impact on obesity, diabetes, and non-alcoholic fatty liver disease in US Hispanic populations.
