Impact of supermarket fruit and vegetable placement on store sales, customer purchasing, diet and household waste: A prospective matched-controlled cluster trial

BACKGROUND: Previous product placement trials have been underpowered and limited in outcomes. This study assessed effects of positioning an expanded fruit and vegetable section near entrances on store-level sales, household-level purchasing and waste, and dietary behaviours. METHODS AND FINDINGS: This prospective matched controlled cluster trial (NIHR 17/44/46) involved 36 stores (18 intervention and 18 control) of a discount supermarket chain in England. The study took place between March 2018 and May 2022, and the intervention was implemented for six months. Control stores were matched on store sales, customer profiles and neighbourhood deprivation. Women customers aged 18-60 years, with loyalty cards, who shopped at intervention (n = 280) or control (n = 300) stores agreed to participate. The primary outcome was household purchasing of fresh fruit and vegetables. Secondary outcomes included: i) differences in household purchasing by educational attainment, ii) store sales of fresh fruit and vegetables, iii) dietary quality score for woman, and iv) child aged 2-6 years (if relevant), and v) household fruit and vegetable waste. The proportion of households purchasing fruit and vegetables in intervention compared to control stores was very similar at baseline (-0.1% (95%CI -6.1%, 6.0%)). After 3 months of exposure to the intervention, the proportion was 0.6% (95%CI -5.6%, 6.7%; p = 0.83) and after 6 months the proportion was 3.3% (95%CI -2.5%, 9.2%; p = 0.23). Interrupted time series analyses showed differences in intervention compared to predicted store-level sales of fruit and vegetables were 0.32SDs (95%CI 0.11, 0.53; p = 0.002) at intervention implementation, equivalent to ~2,525 (95% CI 775, 4,115) extra portions per store, per week. The differences were 0.23SDs (95%CI -0.05, 0.52; p = 0.10) at 3 months and 0.18SDs (-0.16, 0.52); p = 0.29) at 6 months post-intervention. Not being able to randomise stores potentially biases the results through unmeasured confounding effects and findings related to intervention dose were not prespecified but determined from process evaluation findings investigating intervention implementation. CONCLUSIONS: This study was conducted during the COVID-19 pandemic and cost-of-living crisis when population level fruit and vegetable sales and intake declined and recruitment to research was challenging. Despite these circumstances, the results of this study show that positioning produce sections near supermarket entrances may improve the nutrition profile of store sales, household purchasing and women's dietary quality. TRIAL REGISTRATION: NCT03573973.