Presentation (CTI)

OBJECTIVES: To compare the effects of six front-of-package (FOP) label designs with the no-label condition on outcomes related to sugar-sweetened beverage (SSB) selection.
METHODS: We conducted an online 7 × 2 factorial randomized controlled trial in Hong Kong. Participants were randomized to one of seven label conditions (i.e., the no-label control condition and six FOP label conditions displaying additional sugar information) and to one of two tax conditions (i.e., tax or no tax). The primary outcome was whether participants selected an SSB from 15 beverages in a hypothetical purchase task. In the primary analysis, each label condition was compared with the control condition, adjusting for tax condition. P values for the six prespecified primary comparisons were adjusted using Dunnett’s method.
RESULTS: Among 2819 participants included in the analysis, the median age was 51.0 years (IQR, 37.0-65.0), and 55.6% (1566/2819) were female. In the no-label control condition, 37.4% (147/393) selected an SSB from the 15 beverages. None of the six label conditions significantly changed SSB selection compared with the control after Dunnett adjustment. The traffic-light label condition showed the largest observed reduction in SSB selection compared with the control (33.2% [128/386] vs 37.4% [147/393]; risk difference, −4.2 percentage points; 95% CI, −10.9 to 2.5; Dunnett-adjusted P = 0.629) and was associated with participants reporting that the packaging made them less likely to choose an SSB (proportional odds ratio, 0.70; 95% CI, 0.54 to 0.90; P = 0.005).
CONCLUSIONS: In this online randomized controlled trial, FOP labels did not significantly reduce SSB selection compared with the no-label condition. The traffic-light label showed the largest observed reduction and was associated with self-reported lower likelihood of selecting an SSB due to the drink packaging.