Effect of an Internet-based, personalized nutrition randomized trial on dietary changes associated with the Mediterranean diet: the Food4Me Study

Background: Little is known about the efficacy of personalized nutrition (PN) interventions for improving consumption of a Mediterranean diet (MedDiet). Objective: The objective was to evaluate the effect of a PN intervention on dietary changes associated with the MedDiet. Design: Participants (n = 1607) were recruited into a 6-mo, Internet-based, PN randomized controlled trial (Food4Me) designed to evaluate the effect of PN on dietary change. Participants were randomly assigned to receive conventional dietary advice [control; level 0 (L0)] or PN advice on the basis of current diet [level 1 (L1)], diet and phenotype [level 2 (L2)], or diet, phenotype, and genotype [level 3 (L3)]. Dietary intakes from food-frequency questionnaires at baseline and at 6 mo were converted to a MedDiet score. Linear regression compared participant characteristics between high (>5) and low (<= 5) MedDiet scores. Differences in MedDiet scores between treatment arms at month 6 were evaluated by using contrast analyses. Results: At baseline, high MedDiet scorers had a 0.5 lower body mass index (in kg/m(2); P = 0.007) and a 0.03 higher physical activity level (P = 0.003) than did low scorers. MedDiet scores at month 6 were greater in individuals randomly assigned to receive PN (L1, L2, and L3) than in controls (PN compared with controls: 5.20 +/- 0.05 and 5.48 +/- 0.07, respectively; P = 0.002). There was no significant difference in MedDiet scores at month 6 between PN advice on the basis of L1 compared with L2 and L3. However, differences in MedDiet scores at month 6 were greater in L3 than in L2 (L3 compared with L2: 5.63 +/- 0.10 and 5.38 +/- 0.10, respectively; P = 0.029). Conclusions: Higher MedDiet scores at baseline were associated with healthier lifestyles and lower adiposity. After the intervention, MedDiet scores were greater in individuals randomly assigned to receive PN than in controls, with the addition of DNA-based dietary advice resulting in the largest differences in MedDiet scores. Although differences were significant, their clinical relevance is modest.