Microbiome Signatures of Nutrients, Foods and Dietary Patterns: Potential for Personalized Nutrition from The PREDICT 1 Study. (29th May 2020)

• Record Type: Journal Article
• Authors: Spector, Tim; Asnicar, Francesco; Berry, Sarah; Valdes, Ana; Franks, Paul; Wolf, Jonathan; Hadjigeorgiou, George; Roy, Caroline Le; Leeming, Emily; Drew, David; Nguyen, Long; Mazidi, Mohsen; Chan, Andrew; Segata, Nicola
• Is Part Of: Current developments in nutrition. Volume 4(2020)Supplement 2
• Page Start: 1587
• Page End: 1587
• Publication Date: 2020-05-29
• Subjects: Nutrition -- Periodicals; Nutritional Physiological Phenomena; Nutrition; Periodicals; Periodicals; Fulltext; Internet Resources; Periodicals; 612.3
• DOI: 10.1093/cdn/nzaa062_044
• Languages: English
• ISSNs: 2475-2991

Abstract

Abstract: Objectives: The human gut microbiome has been linked to risk and severity of a multitude of chronic diseases, but large-scale, high-resolution studies linking it to host diet are lacking. The PREDICT 1 study (NCT03479866) enrolled 1, 102 healthy US and UK adults to examine the genetic, metabolic, microbial, and meal composition/context contributions to metabolic responses to food. Here, we identify microbial features (species, genes, pathways) linked with diet and assess their potential to predict personalized food responses. Methods: Dietary intake was assessed using validated EPIC (UK) and Harvard (US) semi-quantitative food frequency questionnaires (FFQs) to capture habitual intake (1 yr). Nutrient and food consumption (adjusted for energy intake) were calculated, and dietary patterns and diversity of intake were estimated. Shotgun metagenomic profiling was performed on fecal samples collected at baseline from 1, 001 UK and 97 US individuals. Results: We observed strong associations between overall microbial structure, as well as feature-level associations with nutrients, foods, food groups, and established dietary indices. Strongest associations were with daily intake of coffee, meat, and dairy foods, and saturated fatty acids; Spearman's r = 0.45, 0.29, 0.27 and 0.46, respectively. The Healthy Food Diversity Index and the Plant Dietary Index (PDI; comprised of healthy vs. unhealthy PDI, h-PDI, u-PDI) were strongly associated with community structure (r = 0.36, 0.34, and 0.34), highlighting the synergistic impact of dietary diversity, food quality, and microbial outcomes. We identified two clusters of microbial species with consistent, opposed correlations with 'healthy' and 'unhealthy' nutrients, food groups, and dietary indices with clear segregation between the h-PDI vs. u-PDI. These clusters were also coupled to cardiometabolic biomarkers. The associations observed in the UK cohort were reproducible in the independent US cohort. Conclusions: The relationship between a healthy diet, resultant microbial signatures, and cardiometabolic outcomes strongly support the interactions between the foods we eat, the bacteria they enrich, and chronic disease outcomes, highlighting the importance of diet quality and diversity in personalized precision nutrition. Funding Sources: NIHR, Wellcome Trust, Zoe Global Ltd.