The research team enrolled participants in six distinct groups to capture the progression spectrum. These included a low-risk group of individuals under 45 years of age with no intestinal lesions detected by colonoscopy, a high-risk group of individuals aged 45 or older who likewise showed no lesions, patients with advanced adenomas, and patients with colorectal cancer at stages I and II, stage III, and stage IV. This staged design allowed the investigators to ask a question that has long fascinated microbiome researchers: do the microbial signatures of colorectal cancer emerge abruptly, or do they shift gradually and measurably as the disease moves from precancerous lesions to invasive and metastatic stages?

The metagenomic results were striking in their directional consistency. Several dominant bacterial genera showed progressively lower relative abundance as cancer stage advanced. These declining taxa included UBA7182, Lachnoclostridium B., Faecalibacillus, Fusicatenibacter, and Anaerobutyricum—genera that are broadly associated with a healthy, fermentative gut environment and the production of beneficial short-chain fatty acids. Their steady erosion across stage groups suggests that the metabolic functions these organisms perform, such as butyrate production that nourishes colonocytes and supports anti-inflammatory signaling, may gradually diminish as the tumor microenvironment evolves. Conversely, two genera moved in the opposite direction: Intestinimonas and Bacteroides showed higher relative abundance in more advanced disease groups, hinting at a compositional takeover in which opportunistic or stress-tolerant organisms replace the beneficial core community.

Perhaps the most consequential transition occurred between the high-risk group and the advanced adenoma group. At the boundary between healthy tissue and early neoplastic transformation, the researchers observed a marked decline in the detection rate of low-abundance taxa at the sequencing depth used in the study, together with a measurable reduction in overall microbial diversity. In ecological terms, the arrival of advanced adenomas appears to coincide with a simplification of the gut bacterial ecosystem—a loss of rare species that may function as sensitive early-warning indicators. If confirmed in prospective cohorts, this diversity collapse could become a focal point for the development of early-detection strategies, since a dwindling rare biosphere may register in stool-based assays before symptoms ever appear.

The metabolomic arm of the study added a layer of biochemical context, though with an important interpretive caveat. When the researchers compared the low-risk and high-risk groups—defined strictly by age at 45 years—they found metabolomic differences that appeared to be driven largely by age and age-associated factors rather than by early-disease biology. This distinction matters because it guards against overinterpreting metabolic shifts between the two healthy comparison groups as evidence of preclinical cancer. Within the disease-related analyses, putatively annotated metabolite features, classified at confidence Levels 2 and 3 of the Metabolomics Standards Initiative, mapped onto eight candidate KEGG pathway modules, pointing to perturbed biochemical routes that accompany malignant progression, including pathways connected to energy metabolism and the pentose phosphate pathway, a route with established roles in rapid cell proliferation.

The ionomic analysis, a less common but technically demanding component of multi-omics studies, profiled elemental concentrations in the samples and revealed stage-specific signatures of essential and trace elements. Strontium, iron, and phosphorus reached their highest levels in stage III colorectal cancer, a stage characterized by lymph node involvement and often intensive systemic change. Beryllium predominated in the low-risk individuals, while sulfur was enriched in both the low-risk and high-risk groups. Although the biological significance of these elemental patterns remains to be fully worked out, ions such as iron are known to influence both host physiology and bacterial competition in the gut, and the study’s findings suggest that ionic profiles shift in tandem with microbial and metabolic changes as the disease progresses.

To synthesize these three data layers—microbial taxa, metabolites, and elemental profiles—the researchers applied integrated analytical frameworks, including multi-omics factor analysis and machine learning classifiers such as support vector machines, alongside conventional ordination techniques like principal component analysis and Bray-Curtis dissimilarity measures. The integrated evidence converged on a central conclusion: microbiota, metabolites, and ionic profiles differ across colorectal cancer stage groups, including the earliest disease-stage groups. This convergence across independent molecular domains strengthens the case that the gut microbiome ecosystem is not a passive bystander in colorectal cancer but a dynamic system whose architecture is reshaped in step with tumor progression.

The authors are careful to frame their findings as hypothesis-generating rather than diagnostic. Because the study was cross-sectional, with no within-individual longitudinal sampling, it cannot directly observe progression within a single patient; the stage-associated differences describe correlations between groups, not causal trajectories in individuals. The researchers also flag two specific limitations that warrant particular caution against over-interpretation. First, the low-risk and high-risk groups differ by age—younger than 45 versus 45 and older—so differences between them are confounded by age and cannot be fully statistically adjusted. Second, no microbiome positive or negative controls, such as mock communities, extraction blanks, or no-template controls, were included, meaning reagent and background contamination cannot be fully excluded and the low-abundance findings must be treated as exploratory. Before any diagnostic application, the patterns described here would need to be validated in independent prospective cohorts.

Even with these caveats, the scale and breadth of the study mark a significant advance in cancer microbiome research. Colorectal cancer remains one of the most common and deadly malignancies worldwide, and growing evidence links its etiology to complex interactions among gut microbiota alterations, metabolic dysregulation, and disturbances in essential ions. By simultaneously mapping all three dimensions across a staged cohort of nearly a thousand samples, the work provides a resource for researchers seeking microbial or metabolic biomarkers of disease stage, and it sharpens the scientific conversation about how aging, microbial ecology, and tumorigenesis intertwine. The observed interplay between the microbiome and senescence—the biological aging process—emerges as a particularly promising frontier, and the authors position their correlation-level findings as a foundation for future longitudinal studies that could ultimately translate these ecosystem-level signatures into tools for earlier detection and better risk stratification of colorectal cancer.

Subject of Research: Stage-associated changes in the gut microbiome ecosystem across colorectal cancer progression

Article Title: Multi-omics analysis reveals stage-associated differences in the gut microbiome ecosystem across stages of colorectal cancer in a cross-sectional cohort

Article References: Shuwen, H., Jian, C., Yinhang, W., Yating, X., Caiyun, C., Zefeng, W., Shuwen, L., Peng, Q., Xi, Y., & Wei, W. (2026). Multi-omics analysis reveals stage-associated differences in the gut microbiome ecosystem across stages of colorectal cancer in a cross-sectional cohort. BMC Medicine. https://doi.org/10.1186/s12916-026-05218-8

Image Credits: AI Generated

DOI: 10.1186/s12916-026-05218-8

Keywords: colorectal cancer, gut microbiome, metagenomics, metabolomics, ionomics, microbial diversity, advanced adenoma, cancer staging, BMC Medicine, multi-omics, gut microbiota, biomarkers