[Effects of shopping basket-based nudges on vegetable-purchasing behavior: A quasi-experimental study in...

Objectives　The average daily vegetable intake of adults in Japan remains below recommended levels. As foods are primarily purchased at supermarkets (SMs), this study examined whether nudge interventions implemented in SMs are associated with increased vegetable purchases, using a quasi-experimental design.Methods　Among SMs operated by the same company within a single city, three medium-sized stores with similar customer numbers and vegetable sales trends that were located at least 3-km apart were non-randomly assigned to the intervention (stores A and B) or control (store C) condition. In the intervention stores, shopping baskets were equipped with flyers recommending vegetable purchases (Easy-Attractive nudge). Furthermore, Store B displayed a poster that visually presented 70-g portions of popular seasonal vegetables in the produce section (Attractive-Social-Timely nudge). Store C served as the control and received no intervention. The 1-month observation and intervention periods were October and November 2024, respectively. By comparing point-of-sale (POS) data from October to November 2024 with data from the corresponding period in the previous year, we calculated the year-on-year increase rates for daily vegetable sales (primary outcome) as well as non-vegetable and total sales (secondary outcomes), adjusted for relative changes in customer volume. During the final 2 days of the intervention, 50 customers per store were surveyed to obtain their basic demographics, socioeconomic status, awareness of the intervention materials, and impressions. Average treatment effects (ATEs) were estimated using a difference-in-differences approach that modeled outcomes, intervention status, and store and time indicators for the first and second halves of October and November. Store-specific analyses (A versus B) were also performed.Results　No significant differences were observed between stores regarding basic customer characteristics or socioeconomic status. The ATE for vegetable sales was ＋4.0 percentage points (95% confidence interval: 0.6-7.5), indicating a statistically significant increase. In contrast, no significant ATE was observed for non-vegetable sales (－4.4 percentage points: －14.2 to 5.5) or total sales (－3.6 percentage points: －13.0 to 5.9). Store-specific analyses revealed no additional differences between stores B and A. Awareness of shopping basket flyers was 95%. Among those aware, 24% reported purchasing vegetables, 51% expressed positive opinions, and 8% expressed negative opinions. The corresponding figures for the produce-section poster were 36%, 28%, 61%, and 0%, respectively.Conclusions　Easy-Attractive nudge flyers in shopping baskets were associated with increased vegetable purchases. This intervention potentially represents an effective strategy for promoting vegetable selection without increasing the overall customer spending.