Sulfur Microbial Diet Score and Colorectal Cancer Risk

A named dietary pattern linked to sulfur-metabolizing gut bacteria has been associated with higher colorectal cancer risk in large prospective U.S. cohorts. The sulfur microbial diet (SMD) score does not prove that a burger changes your microbiome and causes cancer. It does offer a more specific warning than generic advice to "eat better": diets higher in processed meat and lower in plant foods may favor microbial pathways that produce hydrogen sulfide (H₂S), a compound that can injure colonic tissue at high concentrations.

The awkward part is that the biology does not point to one villainous "sulfate food." A large gene-mapping study found that gut microbes more often carry H₂S-producing routes that use organic sulfur compounds, including cysteine, methionine, and taurine, than the classic sulfate-reduction pathway. That makes the sulfur story broader, less tidy, and harder to reduce to a grocery-list rule.

A diet score built around gut microbes

In a 2022 prospective cohort study by Wang et al., researchers developed the SMD score across the Health Professionals Follow-Up Study, Nurses' Health Study, and Nurses' Health Study II, totaling about 215,000 participants. They used sparse canonical correlation analysis to connect 33 food groups with 43 stool-resident sulfur-metabolizing bacterial species.

The resulting score loaded positively on processed meats and low-calorie beverages and inversely on plant-based food groups. Participants with higher scores had a higher prospective risk of colorectal cancer after multivariable adjustment. Earlier work in a male cohort indicated that the association concentrated in distal colorectal cancers, where luminal H₂S exposure is thought to peak.

That is an intriguing result, but it needs careful wording. The published material available for this study does not report the hazard ratios by quintile, confidence intervals, or subsite-specific estimates. We therefore know the reported direction of the association, not its precise size or statistical precision. The cohorts also consisted largely of White health professionals, and an SMD score can travel with other colorectal-cancer risk factors: less fiber, more heme iron, more N-nitroso compounds, and differences in lifestyle that statistical adjustment cannot fully erase.

Low-calorie beverages add another complication. They contribute to the score, yet the proposed sulfur mechanism does not clearly explain that component. The index may capture a broader dietary pattern as well as a sulfur-related microbial signal.

Why sulfur biology is more complicated than sulfate

The old shorthand was simple: sulfate-reducing bacteria make H₂S, so avoid sulfate. The evidence does not support making diet decisions on that premise alone. Wolf et al.'s 2022 metagenomic survey examined 514 Human Microbiome Project reference gut genomes and 16,936 metagenome-assembled genomes from five colorectal-cancer cohorts in Austria, the United States, China, and France. Those cohorts included 265 healthy controls, 112 people with adenomas, and 290 with carcinoma.

Among the sulfidogenic genes they identified, organic routes dominated. Cystathionine-beta-synthase accounted for 27.5% of identified sequences, d-cysteine desulfhydrase for 17.9%, methionine-gamma-lyase for 14.7%, and cystathionine-beta-lyase for 11.8%. Together, cysteine- and methionine-related routes accounted for about 74.7% of the identified sulfidogenic gene sequences. The canonical dissimilatory sulfite reductase genes, dsrAB, accounted for 18.2%.

More strikingly, dsrAB appeared in only 121 of 16,936 genomes, or fewer than 1%, and in 17.4% of subjects. Anaerobic sulfite reductase genes were more prevalent. Sulfidogenic capacity appeared across six bacterial phyla, including Proteobacteria, Firmicutes, Bacteroidetes, and Fusobacteria. The study also identified Bilophila wadsworthia, a taurine-metabolizing organism, as a colorectal-cancer indicator.

This does not mean every food containing cysteine, methionine, taurine, or sulfate is dangerous. Gene carriage shows metabolic potential, not enzyme activity or the amount of H₂S produced in a living colon. Wolf et al. did not pair gene abundance with measured luminal H₂S. Some bacteria also resist easy moral sorting: the researchers found dsrAB in Roseburia species, which are known butyrate producers. Trying to eradicate "sulfur bacteria" would be a remarkably blunt instrument.

H₂S itself is dose-dependent. At micromolar concentrations it can have cytoprotective effects; at millimolar colonic concentrations, it can inhibit cytochrome c oxidase and contribute to reductive stress and genotoxicity. A 2024 review of sulfidogenic gut microbes places this biology in the context of inflammatory bowel disease, colorectal cancer, and cardiovascular disease, but mechanistic plausibility is still not a human prevention trial.

Early-onset colorectal cancer adds urgency, not proof

Colorectal cancer is rising among younger adults in several countries. Moon et al.'s 2023 review reported an annual increase in colorectal-cancer incidence among people younger than 50 of 4.3% from 2012 to 2015 and 2.2% from 2015 to 2019 across 19 countries on five continents. It cited projections to 2030 of a 90% increase in colon cancer and a 124.2% increase in rectal cancer among people aged 20 to 34.

The anatomy is suggestive: 42% of early-onset cases occurred in the rectum and 31% in the distal colon, the same regions implicated by the H₂S exposure hypothesis and the earlier SMD work. But these are ecological trend data and a narrative review, not a trial assigning young adults to sulfur-promoting diets. Obesity, antibiotics, early-life microbiome disruption, screening patterns, and other changes could contribute. The location match is a clue, not a causal verdict.

Signals beyond cancer—and competing explanations

A 2025 prospective study by Deng et al. linked a gut sulfur metabolism-promoting dietary pattern in low-income U.S. adults with higher all-cause mortality and altered circulating metabolites. This expands the association beyond affluent health-professional cohorts, but it remains observational. The reported material does not identify which metabolite classes explain the association, so claims about a systemic sulfur mechanism would outrun the evidence.

Inflammation may also change the meaning of a sulfur-rich fecal profile. Walker et al.'s 2021 study of inflammatory bowel disease found that fecal sulfur metabolites tracked disease state, while leaving the direction of causation uncertain. Mucosal inflammation may encourage sulfidogenic activity, rather than simply result from it. The same problem applies to cross-sectional cancer metagenomes: tumor-associated microbial changes can be consequence, cause, or both.

Randomized trials support the broader idea that diet and micronutrients can alter colorectal-cancer-relevant microbiome features, but they do not validate the sulfur pathway. In a trial of healthy volunteers, berry supplementation changed gut microbiota, increased fecal polyphenol metabolites, and reduced colon-cancer-cell viability in fecal-water assays. Those assays are surrogate outcomes, not cancer incidence.

Two other randomized trials point to non-sulfur routes. A double-blind magnesium trial enriched taxa associated with vitamin D synthesis and colorectal-cancer inhibition, including Carnobacterium maltaromaticum and Faecalibacterium prausnitzii. A vitamin D3 trial in healthy adults altered microbiota stability and metabolite networks. These studies do not refute the SMD finding. They do show why the association cannot yet be assigned specifically to microbial sulfur metabolism: magnesium and vitamin D status may track with meat-heavy, plant-poor diets and affect the microbiome through other pathways.

What to do with this now

The highest-confidence move is still ordinary, if not glamorous: reduce processed meat and increase the variety of plant foods you eat. That recommendation fits the SMD cohort association and a wider body of colorectal-health research without requiring you to diagnose your own hydrogen-sulfide pathway from a lunch receipt.

• Do not focus narrowly on inorganic sulfate. Gut H₂S production often relies on organic sulfur pathways involving cysteine, methionine, and taurine.
• Do not treat all plant-derived sulfur as automatically protective. A 2025 study of sulfoquinovose, a compound in green plant foods, found that certain microbial partnerships can convert it to H₂S. Context and dose matter.
• Assess magnesium and vitamin D independently with a clinician when appropriate. Their microbiome effects do not establish cancer prevention, but they are plausible parallel factors rather than footnotes to the sulfur story.

The SMD score is a useful risk marker and a biologically plausible hypothesis—not a proven switch that turns colorectal cancer on or off. Human trials that directly lower colonic H₂S or sulfidogenic activity and measure validated cancer-risk outcomes are still missing.