A large-scale investigation across five industrialized nations has revealed that associations between human diet and the gut microbiome are largely population-specific rather than universally conserved. Coordinated under The Human Diets and Microbiome Initiative, the cross-sectional study examined 1,976 adult participants recruited through the Microsetta platform across the United States, the United Kingdom, Spain, Mexico, and Japan. By standardizing recruitment, biospecimen handling, dietary documentation, and shotgun metagenomic sequencing across international sites, researchers aimed to isolate genuine biological divergence from technical artifacts that frequently confound multi-cohort evaluations. The resulting datasets demonstrated that while basic ecological principles govern host-microbial communities globally, the specific microbial lineages, predictive dietary models, and functional genomic capacities of intestinal bacteria diverge substantially across distinct cultural boundaries and geographic environments.
Previous multi-population investigations of the human gastrointestinal ecosystem frequently identified stark taxonomic contrasts when comparing traditional hunter-gatherer or agrarian communities against Westernized urban societies. However, comparable evaluations across distinct industrialized populations have remained scarce, in part because methodological inconsistencies in DNA extraction protocols, sequencing instruments, and dietary collection instruments obscure true ecological differences. In this coordinated effort, fecal specimens were gathered using standardized collection tubes containing 95 percent ethanol, stored at ambient room temperature, and processed through centralized bead-beating and automated purification pipelines prior to shotgun metagenomics on Illumina NovaSeq instruments. Quality-filtering procedures pruned initial feature sets to 1,976 rigorously screened individuals and 3,375 microbial operational genomic units, ensuring that subsequent cross-border evaluations rested on uniform technical benchmarks.
Initial evaluations revealed marked geographic differences in participant demographic traits, nutrient profiles, food consumption, and microbiome community structures across the five cohorts. Significant cross-country variations appeared in age distributions, self-reported cosmetics usage, and body mass index, which ranged from an average of 21.7 in Japan to 25.7 in Mexico. Principal component analyses demonstrated robust clustering by country across all measured domains, including host metadata, nutrient intake, and dietary food groups. In the microbial domain, separation between nations was driven substantially by obligate anaerobes within the class Clostridia along with multiple operational genomic units corresponding to Escherichia coli. Permutational multivariate analysis of variance demonstrated that geographic origin accounted for approximately 28 percent of microbial variance, outperforming covariates such as age, sex, body mass index, cosmetic usage, and sequencing depth in explained variation.
To evaluate complex multidimensional relationships connecting intestinal taxa with host traits and nutritional intakes, researchers applied sparse canonical correlation analysis across cohorts. Metagenomic features demonstrated strong multivariate correlations with clinical and demographic metadata, dietary nutrient values, and specific food categories, with canonical correlation coefficients reaching 0.70, 0.83, and 0.72, respectively. These global associations were heavily influenced by between-country divergence, with participant profiles from Japan and Spain exhibiting the greatest divergence from other nations. Specific operational genomic units displaying widespread contributions across domains included Clostridium leptum, Butyribacter intestini, Fontimonas thermaphila, and Mycobacterium cosmeticum. On the nutritional side, variables exhibiting influential loadings included dietary protein, magnesium, niacin equivalents, and water-insoluble fiber, alongside culturally distinct foods such as garlic, pasta, rice, and noodles.
Within-country evaluations similarly revealed that relationships between specific foods and intestinal microorganisms were heavily influenced by localized dietary practices. Although overall overlap of top microbial drivers across nations was limited, select taxa such as Enterocloster clostrioformis, Campylobacter concisus, Cloacibacillus evryensis, Acidaminococcus intestini, and Bacteroides pectinophilus were identified across two or three cohorts. Following up on observed correlations with citrus intake, investigators confirmed that Bacteroides pectinophilus abundance correlated positively with measured dietary pectin intake within the three national cohorts possessing available pectin data. More broadly, however, dominant correlations reflected regionally distinctive or culturally prominent food items, demonstrating that microbial abundance profiles do not merely track absolute consumption frequency but respond to complex, population-specific nutritional combinations.
Diet quality was formally quantified using the Healthy Eating Index 2015, which translates nutrient and food intake into a standardized composite score based on established dietary guidelines. Average total diet quality scores differed significantly across cohorts, with Spain reporting the highest average compliance scores, followed in descending order by the United Kingdom, the United States, Mexico, and Japan. Participants from Spain scored highest on adequacy metrics for greens, beans, fatty acid ratios, and seafood and plant proteins, while recording favorable moderation scores for lower intake of refined grains, added sugars, and sodium. Conversely, participants from Japan recorded lower scores for fruit, whole grains, and greens, alongside higher sodium and refined grain consumption, while reporting comparatively lower saturated fat intake than western counterparts.
When testing whether dietary quality metrics correlated with intestinal microbiome composition within each nation, researchers observed pronounced differences in explanatory power. Multivariate tests adjusted for age, sex, and body mass index revealed significant links between overall diet quality scores and metagenomic profiles in the United States, Mexico, and Spain, whereas participants from Japan and the United Kingdom showed limited associations. Machine-learning regressors trained via Elastic Net algorithms using microbial presence-absence profiles accurately predicted diet quality scores within individual cohorts but exhibited sharp declines in predictive performance and explained variance when cross-applied to foreign cohorts. Even though participant microbiomes from Spain and Mexico exhibited greater taxonomic similarity to one another than to other countries, cross-prediction between them was lower than models evaluated against the United States.
Further analyses using Bayesian regression models identified 48 operational genomic units that showed consistent associations with overall diet quality across all five studied countries. Phylogenetic mapping showed that these universally shared positive associations concentrated within a distinct clade of the class Clostridia, though this group appeared generally depleted in Mexican participants compared to other cohorts. Beyond this conserved core, the vast majority of microbes linked to dietary health were strictly population-specific. Investigation of country-restricted organisms identified three specific Bacillus subtilis genomic units that achieved greater than 60 percent prevalence in Japan while remaining at or below 10 percent prevalence elsewhere. Within the Japanese cohort, these units displayed continuous genomic sequence coverage and correlated strongly with self-reported consumption of natto, a traditional fermented soybean food.
To investigate whether gut microbes exhibit country-specific genomic microevolution, the researchers reconstructed metagenome-assembled genomes, yielding 73 dereplicated species bins containing at least 30 high-quality assemblies. Pairwise genomic distance matrices evaluated through permutation testing revealed significant geographic structure in 35 bacterial species, including lineages within Prevotella and the Lachnospiraceae family. Genome-wide association analyses identified geographically differentiated variants overrepresented in pathways for carbohydrate metabolism, amino acid biosynthesis, cofactor production, and energy utilization. For example, Bifidobacterium longum genomes from Mexico demonstrated widespread loss of galactose-metabolism genes, matching high regional lactose intolerance prevalence, while Faecalibacterium longum strains from Japan lacked specific histidine biosynthesis genes. Microbial folate biosynthesis pathways inversely matched population folate intake, showing enrichment in Spain and Mexico where dietary folate was lowest.
The study authors emphasized several structural limitations inherent to their crowdsourced, cross-sectional cohort design. The participant population was heavily skewed toward adult females and individuals with higher average dietary awareness, potentially constraining generalizability across broader socioeconomic segments. While country-specific food frequency questionnaires appropriately captured regional culinary items, harmonizing disparate national nutritional databases into unified nutrient classifications and applying standardized dietary index scoring introduces potential classification bias. Furthermore, the cross-sectional nature of sample collection precludes drawing definitive causal inferences regarding whether specific dietary exposures directly induce microbial lineage shifts. Nevertheless, the findings challenge assumptions that universal microbial biomarkers can reliably index dietary health across diverse global populations, underscoring the necessity of culturally stratified approaches in precision nutrition.
Subject of Research: Geographic and dietary drivers of human gut microbiome composition and genomic variation across five countries
Article Title: Five-country analysis of geographic and dietary drivers of gut microbiome composition and genomic variation
Article References: Khatib, L., Oles, R., Hernández, A. R., Myers, T., Zhao, J., Kumar, M., Laroya, A., Tolosa-Enguis, V., Matsunami, N., Suzuki, D., Chiba, N., Gilbert, K. E., Das, P., Sanders, J. G., Patel, L., Litwin, N., Nowinski, B., Birmingham, A., Farmer, S., … Knight, R. (2026). Five-country analysis of geographic and dietary drivers of gut microbiome composition and genomic variation. Nature Communications, 17(1), Article 11004. https://doi.org/10.1038/s41467-026-77928-9
Image Credits: AI Generated
DOI: 10.1038/s41467-026-77928-9
Keywords: gut microbiome, shotgun metagenomics, dietary patterns, The Human Diets and Microbiome Initiative, metagenome-assembled genomes, Healthy Eating Index, Bifidobacterium longum, Faecalibacterium longum, Bacillus subtilis, geographic variation, sparse canonical correlation analysis, folate biosynthesis, microbial genomics, nutritional ecology, operational genomic units, food frequency questionnaires
News Source: Daisy Hatcher. (November 16, 2026). Five-Nation Study Finds Geography and Diet Shape Gut Microbiomes. Scienmag.



