A single-blinded, randomized, parallel intervention to evaluate genetics and omics-based personalized nutrition in general population via an e-commerce tool: The PREVENTOMICS e-commerce study
Lorena Calderón-Pérez,Xavier Escoté,Judit Companys,Juan María Alcaide-Hidalgo,Mireia Bosch,Montserrat Rabassa,Anna Crescenti,Rosa M Valls,Anna Pedret,Rosa Solà,Roger Mariné,Katherine Gil-Cardoso,Miguel A Rodríguez,Héctor Palacios,Antoni Del Pino,María Guirro,Núria Canela,David Suñol,Mar Galofré,Sebastià Galmés,Andreu Palou-March,Francisca Serra,Antoni Caimari,Biotza Gutiérrez,Josep M Del Bas
DOI: https://doi.org/10.1016/j.ajcnut.2024.04.004
Abstract:Background: Personalized nutrition (PN) has been proposed as a strategy to increase the effectiveness of dietary recommendations and ultimately improve health status. Objectives: We aimed to assess whether including omics-based PN in an e-commerce tool improves dietary behavior and metabolic profile in general population. Methods: A 21-wk parallel, single-blinded, randomized intervention involved 193 adults assigned to a control group following Mediterranean diet recommendations (n = 57, completers = 36), PN (n = 70, completers = 45), or personalized plan (PP, n = 68, completers = 53) integrating a behavioral change program with PN recommendations. The intervention used metabolomics, proteomics, and genetic data to assist participants in creating personalized shopping lists in a simulated e-commerce retailer portal. The primary outcome was the Mediterranean diet adherence screener (MEDAS) score; secondary outcomes included biometric and metabolic markers and dietary habits. Results: Volunteers were categorized with a scoring system based on biomarkers of lipid, carbohydrate metabolism, inflammation, oxidative stress, and microbiota, and dietary recommendations delivered accordingly in the PN and PP groups. The intervention significantly increased MEDAS scores in all volunteers (control-3 points; 95% confidence interval [CI]: 2.2, 3.8; PN-2.7 points; 95% CI: 2.0, 3.3; and PP-2.8 points; 95% CI: 2.1, 3.4; q < 0.001). No significant differences were observed in dietary habits or health parameters between PN and control groups after adjustment for multiple comparisons. Nevertheless, personalized recommendations significantly (false discovery rate < 0.05) and selectively enhanced the scores calculated with biomarkers of carbohydrate metabolism (β: -0.37; 95% CI: -0.56, -0.18), oxidative stress (β: -0.37; 95% CI: -0.60, -0.15), microbiota (β: -0.38; 95% CI: -0.63, -0.15), and inflammation (β: -0.78; 95% CI: -1.24, -0.31) compared with control diet. Conclusions: Integration of personalized strategies within an e-commerce-like tool did not enhance adherence to Mediterranean diet or improved health markers compared with general recommendations. The metabotyping approach showed promising results and more research is guaranteed to further promote its application in PN. This trial was registered at clinicaltrials.gov as NCT04641559 (https://clinicaltrials.gov/study/NCT04641559?cond=NCT04641559&rank=1).