13C and 15N abundance of plasma and saliva protein reflect protein intake from vegetable and animal products

Abstract

Self‐reported dietary intake is often biased. Therefore there is a need for objective biomarkers of diet. We investigated whether isotopic compositions of human protein reflected the proportion of plant and animal protein intake of individuals consuming U.S. foods. We recruited nine each omnivore, lacto‐ova vegetarian, and strict vegan healthy adults. Food frequency questionnaires were administered to assess protein intake. Plasma and saliva protein specimens were analyzed using an isotope ratio mass spectrometry equipped with an elemental analyzer. For plasma protein samples, vegans could be distinguished from omnivores in both 13C (−21.6‰ ± 0.5 vs. −19.3‰ ± 0.5, p < 0. 01) and 15N (8.0‰ ± 0.5 vs. 9.0‰ ± 0.3, p < 0.01) abundances. Lacto‐ova vegetarians were significantly different from vegans for both 13C and 15N abundances. For saliva protein, vegans could be clearly distinguished from omnivores in 13C (−21.5‰ ± 0.6 vs. −19.6‰ ± 0.7, p < 0. 01) and trended in 15N (6.6‰ ± 0.7 vs. 7.3‰ ± 0.7, p = 0.02). The 13C abundances of omnivores were close to values predicted for US foods, but values for vegans contained more 13C than predicted suggesting carbon exchange between macronutrients via intermediary metabolic pathways. The 15N differences displayed the expected trophic level effect. In conclusion, human plasma and saliva 13C and 15N values both reflected the proportion of animal protein in the U.S. diet. Supported by Dairy Management, Inc.