The human oral microbiome plays an important role in dental and gingival health and has been linked to broader health outcomes. Its composition is known to vary with age, diet, hygiene, smoking, and socioeconomic factors. Yet these environmental explanations leave substantial variation unexplained. Why some individuals consistently harbour specific oral bacteria or show persistent susceptibility to oral disease remains unclear.
A new study published in Nature examines how human and bacterial genetic variation jointly shape the oral microbiome and oral health1. Using large population cohorts with paired host genotypes, oral metagenomic sequencing, and clinical data, Kamitaki et al. (2026) show that oral microbial communities are partly structured by host genetics and by strain-level variation within bacterial species. These findings suggest that the oral microbiome reflects an interaction between co-evolving genomes, rather than a purely environmental imprint.
The authors analyzed oral microbiome samples from large, well-characterized human cohorts with genome-wide genotyping data. Shotgun metagenomic sequencing was used to profile bacterial communities at high resolution, enabling both species-level and strain-level analyses. This approach allows the capture of genetic variation within bacterial species, a feat that 16S-based studies cannot achieve.
Host genetic associations were assessed using genome-wide association analysis linking human variants to microbial abundance and bacterial genetic features. In parallel, bacterial genomes were analyzed to identify strain-specific genetic variants associated with colonisation patterns and host traits. Clinical data enabled links between host-microbe genetic interactions and oral health outcomes by utilizing denture use as a primary proxy for tooth loss and dental caries. The study also identified genetic associations with bleeding gums, though this phenotype showed little genetic overlap with denture risk.
This study identifies 11 human genetic loci associated with variation in the oral microbiome. These loci are enriched in genes related to immune function, epithelial integrity, and saliva composition. Such pathways are biologically plausible mediators of microbial colonisation, influencing bacterial adhesion, nutrient availability, and immune tolerance.
Importantly, the associations are not limited to the presence or absence of bacterial species. Host genetic variants are linked to relative abundances and to bacterial gene dosages within species. This indicates that host genetics can shape which bacteria colonise the mouth and which genetic variants of those bacteria persist.
Heritability estimates suggest that a measurable fraction of oral microbiome variation is attributable to host genetics. While environmental factors remain dominant, genetics provides a stable background that may constrain or bias microbial community structure over time.
A key contribution of the study is its focus on bacterial genetic diversity. The authors show that strains within the same bacterial species differ in their associations with host genotypes and health traits. Specific bacterial genomic regions associated with host genetics were linked to metabolic pathways, such as glycoside hydrolases used for carbohydrate consumption, and surface structures like adhesin proteins (e.g., YadA, CshA, and mucin-binding domains) that facilitate attachment to host mucosal surfaces.
These findings suggest that species-level classification is sufficient for understanding microbiome-health relationships. Two individuals may harbour the same bacterial species, yet carry genetically distinct strains with different functional properties and clinical implications.
The authors connect combined host and microbial genetic variation to oral health phenotypes, including caries and bleeding gums. The study demonstrated that host genetic effects on microbiome composition often have downstream effects on oral health. This was evidenced by the colocalization of genetic signals at loci like AMY1, FUT2, and PITX1, which influenced both oral microbial abundances and the risk of requiring dentures.
The findings at the AMY1 locus highlight a divergence between different oral disease processes. Higher AMY1 copy number, which increases salivary amylase abundance, is associated with a higher risk of denture use, while simultaneously associating with a lower risk of bleeding gums. This pattern suggests that genetic influences on tooth loss and gingival inflammation are at least partly independent and can act in opposite directions. Notably, these associations were specific to AMY1 and were not observed for pancreatic amylase genes, supporting the interpretation that salivary amylase affects oral health primarily through its influence on the oral microbiome rather than through systemic metabolic pathways.
This suggests that disease risk may depend on specific host-microbe genetic combinations, rather than on microbial presence alone. Such interactions could help explain why some individuals develop disease despite similar microbial exposure and behaviours, while others remain unaffected.
This work suggests that the oral microbiome is a genetically structured system shaped by interactions between human and microbial genomes. It supports a model in which host biology influences which bacterial strains can colonise and persist, while bacterial genetic variation modulates how microbes interact with host tissues and immune responses.
The emphasis on strain-level resolution is particularly important. Many microbiome studies report inconsistent or weak associations with disease. This study suggests that this inconsistency may arise from collapsing genetically distinct bacterial strains into single-species categories.
The findings also highlight why microbiome-based interventions have shown variable effects. If host and bacterial genetics jointly shape microbial function, then uniform interventions may produce heterogeneous outcomes across individuals. This is consistent with oral probiotic studies showing heterogeneous and often transient effects, with limited evidence for durable or uniform changes to the resident oral microbiota across populations2.
Understanding oral health through a host-microbe genetic lens raises several challenges and opportunities. Precision approaches to oral disease prevention or treatment may require accounting for both host susceptibility and microbial strain composition. This could influence future diagnostics, risk stratification, or microbiome-targeted therapies.
While this study utilized the diverse All of Us cohort to replicate findings across multiple ancestries, the authors acknowledge that gaps remain. Future research in other large, globally representative cohorts will be needed to determine if these specific host-microbe genetic interactions are found across different human populations and environmental contexts.
Longitudinal data will also be critical. Cross-sectional associations cannot fully resolve causality or how the oral microbiome changes over time, particularly in a system shaped by lifelong exposure and behaviour.
Despite these limitations, the authors establish a clear framework for integrating human and microbial genomics in oral health research. By showing that genetic variation on both sides of the host-microbe interface matters, the field moves beyond descriptive microbiome profiling towards a more mechanistic understanding of disease risk and resilience.