Protein food intake appears to partially structure dietary patterns, as most current emergent diets (e.g., vegetarian and flexitarian) can be described according to their levels of specific protein sources. However, few data are available on dietary protein patterns in the general population and their association with nutrient adequacy. Based on protein food intake data concerning 1678 adults from a representative French national dietary survey, and non-negative-matrix factorization followed by cluster analysis, we were able to identify distinctive dietary protein patterns and compare their nutrient adequacy (using PANDiet probabilistic scoring). The findings revealed eight patterns that clearly discriminate protein intakes and were characterized by the intakes of one or more specific protein foods: 'Processed meat', 'Poultry', 'Pork', 'Traditional', 'Milk', 'Take-away', 'Beef' and 'Fish'. 'Fish eaters' and 'Milk drinkers' had the highest overall nutrient adequacy, whereas that of 'Pork' and 'Take-away eaters' was the lowest. Nutrient adequacy could often be accounted for by the characteristics of the food contributing to protein intake: 'Meat eaters' had high probability of adequacy for iron and zinc, for example. We concluded that protein patterns constitute strong elements in the background structure of the dietary intake and are associated with the nutrient profile that they convey.
Keywords: dietary diversity; dietary protein pattern; nutrient adequacy; protein diets.