Adjusting for protein quality by food source may affect nutrient density metrics
Adam Drewnowski
Abstract
Methods to assess nutrient density of foods, commonly known as nutrient profiling (NP), typically include protein as a component. In this study, the goal was to apply a correction for protein quality by food source to selected NP algorithms. Analyses of 378 component foods of the Fred Hutch food frequency questionnaire showed that animal-source foods (ie, meat, eggs, and dairy) along with some soy products and nuts were the only foods that provided > 20% of the daily value (DV) of protein per 100 g or per 100 kcal. Most beans, pulses, legumes, grains, and vegetables provided <10% DV of protein per 100 g or per 100 kcal. Adjusting for protein quality using a simplified Protein Digestibility Corrected Amino Acid Score (PDCAAS) had consequences for point-based NP models (namely, Nutri-Score) and for continuous nutrient density scores (namely, Nutrient Rich Foods). Quantitative methods that use protein content to capture nutrient density may require a protein-quality adjustment, especially when adapted for use in low- and middle-income countries where protein quality is an issue of public health concern.