Menu Close

Sugar-Sweetened Beverage Health Warnings and Purchases: A Randomized Controlled Trial


Grummon, Anna H.; Taillie, Lindsey Smith; Golden, Shelley D.; Hall, Marissa G.; Ranney, Leah M.; & Brewer, Noel T. (2019). Sugar-Sweetened Beverage Health Warnings and Purchases: A Randomized Controlled Trial. American Journal of Preventive Medicine, 57(5), 601-610. PMCID: PMC6803129


INTRODUCTION: Five U.S. states have proposed policies to require health warnings on sugar-sweetened beverages, but warnings' effects on actual purchase behavior remain uncertain. This study evaluated the impact of sugar-sweetened beverage health warnings on sugar-sweetened beverage purchases.
STUDY DESIGN: Participants completed one study visit to a life-sized replica of a convenience store in North Carolina. Participants chose six items (two beverages, two foods, and two household products). One item was randomly selected for them to purchase and take home. Participants also completed a questionnaire. Researchers collected data in 2018 and conducted analyses in 2019.
SETTING/PARTICIPANTS: Participants were a demographically diverse convenience sample of 400 adult sugar-sweetened beverage consumers (usual consumption ≥12 ounces/week).
INTERVENTION: Research staff randomly assigned participants to a health warning arm (sugar-sweetened beverages in the store displayed a front-of-package health warning) or a control arm (sugar-sweetened beverages displayed a control label).
MAIN OUTCOME MEASURES: The primary trial outcome was sugar-sweetened beverage calories purchased. Secondary outcomes included reactions to trial labels (e.g., negative emotions) and sugar-sweetened beverage perceptions and attitudes (e.g., healthfulness).
RESULTS: All 400 participants completed the trial and were included in analyses. Health warning arm participants were less likely to be Hispanic and to have overweight/obesity than control arm participants. In intent-to-treat analyses adjusting for Hispanic ethnicity and overweight/obesity, health warnings led to lower sugar-sweetened beverage purchases (adjusted difference, -31.4 calories; 95% CI= -57.9, -5.0). Unadjusted analyses yielded similar results (difference, -32.9 calories; 95% CI= -58.9, -7.0). Compared with the control label, sugar-sweetened beverage health warnings also led to higher intentions to limit sugar-sweetened beverage consumption and elicited more attention, negative emotions, thinking about the harms of sugar-sweetened beverage consumption, and anticipated social interactions. Trial arms did not differ on perceptions of sugar-sweetened beverages' added sugar content, healthfulness, appeal/coolness, or disease risk.
CONCLUSIONS: Brief exposure to health warnings reduced sugar-sweetened beverage purchases in this naturalistic RCT. Sugar-sweetened beverage health warning policies could discourage sugar-sweetened beverage consumption.


Reference Type

Journal Article

Year Published


Journal Title

American Journal of Preventive Medicine


Grummon, Anna H.
Taillie, Lindsey Smith
Golden, Shelley D.
Hall, Marissa G.
Ranney, Leah M.
Brewer, Noel T.

Article Type





United States of America


North Carolina


Grummon - 0000-0002-8705-038X
Taillie - 0000-0002-4555-2525