Eur J Nutr. 2022 Aug 4. doi: 10.1007/s00394-022-02970-4. Online ahead of print.
BACKGROUND: Both the nutritional quality of the foods consumed (as nutrient composition) and their ultra-processed nature have been linked to health risks. However, the respective contribution of each of these correlated dimensions or their synergy to the overall diet quality has been rarely explored.
OBJECTIVE: To identify the respective effects of the nutritional quality of the foods consumed, the ultra-processed nature of foods and their cross-effect contributing to the overall quality of the diet.
DESIGN: Cross-sectional observational study.
SETTING: Web-based French NutriNet-Santé cohort study.
PARTICIPANTS: Participants in the NutriNet-Santé cohort study with at least three available 24 h records as baseline dietary data (N = 98 454 participants).
MAIN OUTCOME MEASURES: The overall quality of the diet (qualified using the adherence to the 2017 French national nutrition and health dietary recommendations dietary score PNNS-GS2) was broken down into: (1) an effect of the nutritional quality of the foods consumed (qualified using the modified Foods Standards Agency nutrient profile model (underlying the Nutri-Score) dietary index FSAm-NPS DI); (2) an effect of the ultra-processed nature of the foods consumed (qualified using the proportion of ultra-processed foods consumed UPFp using the NOVA classification), and (3) a cross-effect of both dimensions.
RESULTS: The overall effect from the ‘nutritional quality of the foods consumed’ (FSAm-NPS DI) was 1.10, corresponding to 26% of the total effect; the overall effect from ultra-processed foods consumption was 1.29, corresponding to 30% of the total effect; and cross-effect between nutritional quality of the foods consumed and ultra-processing was at 1.91, corresponding to 44% of total effects.
CONCLUSIONS: Our study provides support to the postulate that nutritional quality and ultra-processing should be considered as two correlated but distinct and complementary dimensions of the diet.