Consumption of organic versus conventional fruits and vegetables in relation to cancer risk: findings from the NutriNet-Santé cohort study
Menée à partir de données de la cohorte française "NutriNet-Santé" portant sur 31 179 adultes (durée moyenne de suivi : 7,3 ans ; 75 % de femmes), cette étude analyse l'association entre la substitution des fruits/légumes conventionnels par des fruits/légumes biologiques et le risque de cancer (1 718 cas)
Background: Regular fruits and vegetables (F&V) consumption is linked to a lower risk of certain cancers. However, F&V can contain pesticides, some of which may have carcinogenic properties. Organic foods are thought to reduce pesticide residue exposure, but the relationship between organic consumption and cancer risk is inconsistent. Notably, no study has explored substituting conventional with organic F&V.
Objectives: Therefore, this study examines the association between substituting conventional with organic F&V and cancer risk in the NutriNet-Santé cohort (about 31,000 French adults), using detailed data from a food frequency questionnaire distinguishing between conventional and organic products.
Methods: Substitution of conventional F&V with organic F&V, at fixed total F&V consumption, was examined with multivariable Cox proportional hazards models per 100g/d increment of substitution and by quintiles. Sensitivity analyses as well as marginal structural models, were also implemented to improve causal inference.
Results: A total of 31,179 participants, 75% of whom were females, were included in the analyses. During follow-up (mean=7.3 y (SD=3.1)), 1,718 cancer cases (284 cases of postmenopausal breast cancer) were registered over 227,660 person-years.
Substituting conventional with organic F&V was associated with a lower risk of overall cancer (HR (95%CI) for a 100g/d substitution = 0.98 (0.95,1.00) and a lower risk of postmenopausal breast cancer (HR (95%CI) for a 100g/d substitution = 0.90 (0.85,0.96)) while no association was detected for other cancer locations. When modelled by quintiles, an inverse association was found for post-menopausal breast cancer but not for overall cancer or other locations. When running marginal structural models and testing for additional adjustments findings remained significant for post-menopausal breast cancer.
Conclusions: In the present work, substituting conventional F&V with organic ones was associated with a reduced risk of postmenopausal cancer. The specific role of organic F&V compared to conventional F&V needs further investigation in other contexts.
The American Journal of Clinical Nutrition , article en libre accès, 2026