Designing warnings for sugary drinks: A randomized experiment with Latino parents and non-Latino parents.

Sugary drink warnings are a promising policy for reducing sugary drink consumption, but it remains unknown how to design warnings to maximize their impact overall and among diverse population groups, including parents of Latino ethnicity and parents with low English use. In 2019, we randomized US parents of children ages 2-12 (n = 1078, 48% Latino ethnicity, 13% low English use) to one topic (one of four warnings, or a neutral control), which they viewed on three designs (text-only, icon, and graphic) to assess reactions to the various warnings on sugary drinks. All warning topics were perceived as more effective than the control (average differential effect [ADE] ranged from 1.77 to 1.84 [5-point Likert scale], all p < .001). All warning topics also led to greater thinking about harms of sugary drinks (all p < .001) and lower purchase intentions (all p < .01). Compared to text-only warnings, icon (ADE = 0.18) and graphic warnings (ADE = 0.30) elicited higher perceived message effectiveness, as well as greater thinking about the harms of sugary drinks, lower perceived healthfulness, and lower purchase intentions (all p < .001). The impact of icon warnings (vs. text warnings) was stronger for parents with low English use, compared to those with high English use (p = .024). Similarly, the impact of icon (vs. text warnings) was stronger for Latino parents than non-Latino parents (p = .034). This experimental study indicates that many warning topics hold promise for behavior change and that including images with warnings could increase warning efficacy, particularly among Latino parents and parents with low English use. Clinical Trial Registration: NCT04382599.