Contrary to ultra-processed foods, the consumption of unprocessed or minimally processed foods is associated with favorable patterns of protein intake, diet quality and lower cardiometabolic risk in French adults (INCA3).

Purpose: While the consumption of ultra-processed foods is steadily increasing, there is a growing interest in more sustainable diets that would include more plant protein. We aimed to study associations between the degree of food processing, patterns of protein intake, diet quality and cardiometabolic risk.

Methods: Using the NOVA classification, we assessed the proportion of energy from unprocessed/minimally processed foods (MPFp), processed foods (PFp) and ultra-processed foods (UPFp) in the diets of 1774 adults (18-79 years) from the latest cross-sectional French national survey (INCA3, 2014-2015). We studied the associations between MPFp, PFp and UPFp with protein intakes, diet quality (using the PANDiet scoring system, the global (PDI), healthful (hPDI) and unhealthful (uPDI) plant-based diet indices) and risk of cardiometabolic death (using the EpiDiet model).

Results: MPFp was positively associated with animal protein intake and plant protein diversity, whereas PFp was positively associated with plant protein intake and negatively with plant protein diversity. The PANDiet was positively associated with MPFp (β = 0.14, P < 0.0001) but negatively with UPFp (β = - 0.05, P < 0.0001). These associations were modified by adjustment for protein intakes and plant protein diversity. As estimated with comparative risk assessment modeling between extreme tertiles of intake, mortality from cardiometabolic diseases would be decreased with higher MPFp (e.g. by 31% for ischemic heart diseases) and increased with higher UPFp (by 42%) and PFp (by 11%).

Conclusions: In the French population, in contrast with UPFp, higher MPFp was associated with higher animal protein intake, better plant protein diversity, higher diet quality and markedly lower cardiometabolic risk.