Although plant proteins are often considered to have less nutritional quality because of their suboptimal amino acid (AA) content, the wide variety of their sources, both conventional and emerging, suggests potential opportunities from complementarity between food sources. This study therefore aimed to explore whether, and to what extent, combinations of protein ingredients could reproduce an AA profile set as a nutritional objective, and to identify theoretical solutions and limitations. We collected compositional data on protein ingredients and raw plant foods (n = 151), and then ran several series of linear optimization to identify protein ingredient mixes that maximized the content in indispensable AA and reproduced various objective profiles: a "balanced profile," based on AA requirements for adults; "animal profiles" corresponding to conventional animal protein compositions, and a "cardioprotective profile," which has been associated with a lower cardiovascular risk. We assumed a very good digestibility of plant protein isolates. As expected, obtaining a balanced profile was obvious, but we also identified numerous plant protein mixtures that met demanding AA profiles. Only for particularly demanding profiles, such as mimicking a particular animal protein, did solutions require the use of protein fractions from more specific sources such as pea or canola. Optimal plant blends could mimic animal proteins such as egg white, cow milk, chicken, whey or casein with a similarity reaching 94.2, 98.8, 86.4, 92.4, and 98.0%, respectively. The limiting constraints were mainly isoleucine, lysine, and histidine target contents. These different solutions offer potential for the formulation of mixtures adapted to specific populations or the design of plant-based substitutes. Some ingredients are not commercially available but they could be developed.
Keywords: amino acid profile; cardiometabolic health; indispensable amino acids; linear optimization; plant-based protein isolate; protein blend.
Copyright © 2022 Dimina, Rémond, Huneau and Mariotti.