AI-generated research brief — verify at source
Added-Sugar Warning Labels on US Restaurant Menus Significantly Reduce Sugar Ordered
Both tested added-sugar warning label designs meaningfully reduced the amount of added sugar participants ordered from restaurant menus, with the icon-plus-text label achieving an 11% reduction at dinner — equivalent to 10.4 grams less added sugar per meal (95% CI −13.4 to −7.4, p<0.0001). The icon-plus-text design outperformed the icon-only label for dinner orders, though both designs produced comparable reductions at breakfast, and neither labelling threshold tested showed an advantage over the other.
What Was Studied
This trial investigated whether adding visible warning labels to restaurant menus could reduce how much added sugar consumers selected when placing orders. Two distinct label designs and two policy-relevant labelling thresholds were evaluated, with the primary outcomes being the grams of added sugar ordered across dinner and breakfast menus.
How It Was Studied
This was a five-arm online randomised controlled trial enrolling a national sample of US adults aged 18 years or older who reported eating at a restaurant at least once per month in the preceding year. Participants were matched to US demographic characteristics to improve representativeness. A total of 11,275 individuals were randomised in equal proportions to one of five conditions: a no-label control or one of four label variants combining two design types (a red icon-only label sized at 150% of menu text height, or a black icon-plus-text label at 100% of text height carrying the words “SUGAR WARNING”) with two labelling thresholds (items containing at least 50% or at least 100% of the Daily Value for added sugar). After exclusions for incomplete surveys, duplicate responses, failed attention checks, and implausibly rapid completion times, the analytical sample comprised 10,456 participants. Participants placed hypothetical orders from fast-food, full-service, and café menus to simulate real-world meal contexts.
What Was Observed
- The icon-plus-text label produced the largest reduction in added sugar ordered at dinner — about 11% less, or 10.4 grams fewer, compared to the no-label control (95% CI −13.4 to −7.4, p<0.0001), a highly statistically reliable result across a large sample.
- The icon-only label also significantly reduced dinner-time added sugar selection, though by a smaller margin — approximately 7% less, or 6.6 grams fewer, than the control (95% CI −9.6 to −3.6, p<0.0001). The icon-plus-text label produced a statistically significant additional reduction of 3.7 grams (4%) beyond the icon-only label (95% CI 1.3 to 6.2, p=0.0028).
- At breakfast, both label types similarly reduced added sugar ordered: the icon-only label achieved a 7% reduction (3.6 g; 95% CI −5.8 to −1.3, p=0.0055) and the icon-plus-text label a 5% reduction (2.5 g; 95% CI −4.8 to −0.3, p=0.042), with no statistically significant difference between the two designs in this meal context.
- No significant differences were observed between the two labelling thresholds (≥50% vs. ≥100% of the Daily Value for added sugar) for any outcome, suggesting that threshold selection may be less critical to policy effectiveness than label design.
Why This Matters
Added-sugar warning labels on restaurant menus have been proposed as a population-level policy tool for reducing added-sugar intake, but the comparative effectiveness of different label designs and thresholds had not been established experimentally. This trial directly addresses that gap by demonstrating that label design characteristics — specifically, whether a label includes both an icon and text — influence consumer food selection in a measurable way. The finding that labelling threshold did not significantly affect outcomes offers concrete, actionable guidance relevant to policymakers designing or evaluating such regulations.
How to Read This Result
While this well-powered randomised trial provides strong experimental evidence that sugar warning labels can shift hypothetical food choices, the online simulated ordering task may not fully capture how consumers behave in actual restaurant settings, and the degree to which these effects would translate to real-world purchasing remains uncertain.
Limitations
The central limitation of this study is its reliance on a hypothetical ordering exercise conducted online, rather than direct observation of behaviour in real restaurant environments. Because participants faced no actual cost, social pressure, or physical menu context, the ecological validity of the findings is constrained. Whether the reductions in added sugar selected in this simulated setting would persist — or differ in magnitude — in genuine purchasing situations cannot be determined from these data alone.