Consent to sharing shopping data for health research: a mixed-methods study

Introduction: Shopping data offer detailed, objective records of diet and lifestyle habits, providing a valuable but under-used resource for health research. While willingness to share personal data has been widely studied, no previous research has directly compared stated willingness with subsequent actual data sharing within the same cohort. Understanding whether willingness predicts actual consent, and how this varies by sociodemographic characteristics, motivations, and barriers, is essential to improve recruitment strategies and the representativeness of collected data. Methods: We employed a mixed-methods approach using the Avon Longitudinal Study of Parents and Children (ALSPAC). Quantitative analyses combined sociodemographic variables with responses from a 2018 survey assessing willingness to consent to shopping data linkage and outcomes from a 2023 loyalty card consent campaign. These were complemented by semi-structured interviews in 2019 exploring attitudes towards sharing shopping data. Results: In 2018, 60.7% of participants indicated willingness to share shopping data (n = 2142, 9230 invited). In 2023, actual consent—requiring provision of loyalty card details for five UK retailers (Sainsbury’s, Tesco, Boots, Morrisons, and Coop)—was obtained from 511 of 6170 invited participants. Demographic characteristics were broadly comparable across samples. Stated willingness was associated with higher odds of actual consent (adjusted OR = 1.48, 95% CI: 1.16–1.90), indicating that willingness is a meaningful predictor of subsequent data sharing. Sociodemographic characteristics and motivations (e.g., societal benefit, scientific interest) showed modest associations with both willingness and consent and moderated the willingness–consent relationship. Interviews reinforced the importance of trust in ALSPAC and clear communication about data use, while highlighting privacy concerns as minor barriers. Conclusion: This study provides the first within-cohort evidence that willingness to share shopping data predicts subsequent data sharing, albeit imperfectly. Willingness surveys can therefore inform recruitment planning, but additional strategies are required to improve realised consent rates. Addressing practical barriers, clarifying data use, and highlighting societal and scientific benefits may help narrow this gap; Targeted engagement, including through informative materials and outreach events, may help to reach groups less likely to consent. These insights can inform data collection in longitudinal cohorts and wider population samples.