The Impact of Almond and Walnut Consumption on the Human Fecal Metabolome

Abstract

ObjectivesMetabolomic studies can be utilized to generate biomarkers of food intake. Undigested food components affect the fecal microbiota and metabolome. Accordingly, we aimed to identify fecal metabolites unique to almond and walnut consumption. MethodsUntargeted metabolomic analyses were completed on 66 endpoint fecal samples from two separate 3-week randomized, controlled-feeding, crossover studies examining almond (n = 30) and walnut (n = 36) consumption in adults (25–75 yr). Control diets, representative of the typical American diet, were fed at weight maintenance with 0 g/day of nuts. During the treatment arms, the base diet was scaled down to allow isocaloric inclusion of 42 g/day of almonds or walnuts. The Kruskal-Wallis H test was used to determine statistically significant metabolites between treatment and control groups with Benjamini-Hochberg false discovery rate adjustments (reported as q-values). ResultsOf the 318 quantifiable fecal metabolites, 42 were significantly different when comparing the treatment groups to their respective controls after adjustment (q < 0.05). Of these 42 metabolites, 9 were significantly different in both the almond and walnut treatment samples. Two metabolites, palmitoleic acid and p-cresol, were unique to almonds—the relative concentration of palmitoleic acid was higher in the almond group compared to control and p-cresol was lower in almond compared to control. Walnut treatment samples contained 31 unique metabolites, including 15 fatty acyls, the majority of which were higher in the walnut group compared to control. ConclusionsHigher concentrations of fecal fatty acyls in the almond and walnut groups compared to their respective controls support previous findings that the plant cell walls of nuts reduce digestibility, therefore, limiting accessibility of intact lipids. Overall, these results reveal promise in identifying fecal biomarkers of food intake for eventual use in personalized dietary recommendations. Ongoing analyses include utilizing machine learning models to further biomarker panel development through incorporation of baseline data and metagenomic analyses. Funding SourcesThis research was funded by the Foundation for Food and Agriculture Research and the National Center for Supercomputing Applications Faculty Fellowship.