Bacteroides Fragilis

Most people carry this bacterium in their gut without ever knowing it, and for most, that is fine. The reason to pay attention is that a subset of strains carry a toxin gene that has been repeatedly linked to colon polyps and colorectal cancer in human studies, while shifts in the broader population of this bacterium also show up in inflammatory bowel disease, Parkinson's disease, multiple sclerosis, and rheumatoid arthritis.

This is a newer, research-oriented stool test without standardized clinical cutoffs. The most useful question it answers is not whether you carry the bacterium at all, but whether you carry the toxin-producing form (ETBF, short for enterotoxigenic Bacteroides fragilis), and how its abundance compares to your own results over time.

What This Stool Test Actually Measures

The lab uses PCR (polymerase chain reaction, a technique that copies and detects specific stretches of bacterial DNA) to search your stool sample for the genetic fingerprint of B. fragilis (Bacteroides fragilis). Some assays use a general species target like the 16S rRNA gene to estimate how much of the bacterium is present. Other assays separately look for the bft gene, which marks toxin-producing strains. These two readouts mean very different things, and reading the result without knowing which one was measured is the single biggest source of confusion.

Because the test analyzes stool, it reflects bacteria living in or shed from your gut, not anything circulating in your blood. The result is best understood as a snapshot of part of your gut microbial community, not a diagnosis on its own.

Total Abundance Versus Toxin-Producing Strains

Total B. fragilis in stool is a poor standalone marker because it is one of the most common bacteria in healthy human guts. In one foundational PCR study, B. fragilis group bacteria were detected in all 46 healthy volunteers tested. A high total count is not automatically bad, and a low total count is not automatically good.

The toxin gene is where the signal sharpens. Across multiple human stool studies, bft-positive strains are found more often in people with colorectal cancer and pre-cancerous polyps than in healthy controls. One mucosal study found the toxin gene in 47% of people with pre-cancerous or cancerous lesions, versus only about 4% of healthy controls.

Colorectal Cancer and Adenoma

Toxin-producing strains have the strongest documented link to colon cancer biology. A 2025 meta-analysis pooled multiple human studies and found that people with colorectal cancer were about 2.5 times as likely as healthy controls to carry ETBF (odds ratio 2.54, 95% CI 1.63 to 3.98). Other case-control work found the toxin gene in 38% of B. fragilis isolates from colon cancer patients, compared to 12% of isolates from controls.

How ETBF carriage tracks with cancer stage is still unsettled. In one cohort, the toxin was found in 56.5% of stool samples from people with colorectal adenomas, compared to 24% of controls, and detection dropped as the disease progressed (about 35% in stage I/II cancer down to 22% in stage IV), fitting the idea that the toxin may help drive the very early steps of tumor formation. The 2025 meta-analysis actually found the opposite direction when pooling studies together (lower ETBF in stage I/II than in stage III/IV), although its authors flagged that conclusion as unreliable due to large differences between studies. The honest summary is that ETBF is enriched at the adenoma stage, but the stage-by-stage pattern across cancer differs depending on which study you read.

One detail worth knowing: studies suggest the bft-2 subtype of the toxin gene may carry more cancer risk than bft-1. A 2025 strain-level study found bft-2 in 55.8% of colorectal cancer cases versus 16.7% of controls, an odds ratio of about 6.3.

Inflammatory Bowel Disease

Toxin-producing strains also turn up more often in ulcerative colitis. In one study, the bft gene was detected in 51.4% of samples from people with ulcerative colitis, versus 1.6% of non-IBD samples, and was especially common in those with active diarrhea. In Crohn's disease, total B. fragilis levels have been reported as higher than in ulcerative colitis. The takeaway: ETBF positivity in someone with active GI symptoms is a meaningful finding, not background noise.

Microbiome Shifts in Other Conditions

Outside of cancer and IBD, low total B. fragilis often signals broader gut imbalance rather than a specific disease. Lower stool levels have been documented in Parkinson's disease, treatment-naive multiple sclerosis (roughly half the level seen in controls, 0.11 versus 0.22 billion CFU per gram), and rheumatoid arthritis. Gut dysbiosis has also been described in intestinal Behcet's disease, though the specific direction for B. fragitis there is supported by less data than the other conditions. In transplant patients with intestinal graft-versus-host disease, low levels before fecal microbiota transplantation rose in those who responded clinically and stayed low in those who did not, making it a useful tracking marker for microbiome recovery.

Reconciling the Apparent Contradiction

It may sound contradictory that high B. fragilis can mean colon cancer risk while low B. fragilis can mean dysbiosis. The resolution is that this is not a simple good number versus bad number test. The total species is a common, mostly benign gut resident whose abundance changes with diet, antibiotics, and overall microbiome composition. The toxin-producing subset is a separate question with its own risk signal. Interpreting a B. fragilis result without knowing whether the assay measured the total species, the broader B. fragilis group, or the bft toxin gene is the most common interpretive mistake in the published literature.

When Results Can Be Misleading

• Recent antibiotics: courses of clarithromycin, trimethoprim-sulfamethoxazole, and other agents can shift B. fragilis abundance within days, and some changes persist for weeks to months. A stool test pulled during or shortly after antibiotic exposure may not reflect your usual gut state.
• Assay type: TaqMan qPCR and digital PCR are substantially more sensitive than SYBR Green PCR for detecting the bft gene in stool, and bacterial culture sweeps followed by PCR often pick up toxin-positive strains that direct stool extraction misses. The lab method shapes the result.
• Primer coverage: some commonly used primers preferentially detect bft-1 and underestimate bft-2 or bft-3, so a negative toxin result on one panel does not rule out toxin-positive carriage on another.
• Recent illness or GI infection: stool composition can shift transiently with gastrointestinal infection, recent diarrhea, or sample handling problems. A single positive result should be confirmed before driving major decisions.

Why One Reading Is Not Enough

Gut microbial abundance is naturally variable. A single stool PCR is a snapshot that captures one day in a constantly shifting community. The more useful information comes from tracking the same target across multiple time points, especially if you are making a change such as adjusting fiber intake, finishing an antibiotic course, starting metformin, or recovering from a GI illness. A reasonable approach is a baseline test, a follow-up in 3 to 6 months if you are making meaningful diet or medication changes, and at least annual repeat testing if you are using this marker as part of broader gut-health monitoring. Use the same lab and assay each time, because results from different platforms do not directly compare.

What to Do With an Unexpected Result

Because this is a research-tier marker without standardized clinical cutoffs, a single result rarely justifies a major decision by itself. Pair it with the bigger picture. If you have GI symptoms (persistent diarrhea, blood in stool, unexplained abdominal pain, weight loss) and a positive ETBF result, that warrants a gastroenterology workup, not microbiome experimentation. If you are over 45 or have a family history of colorectal cancer and ETBF appears in your stool, it is not a substitute for colonoscopy or established stool-based screening like FIT or Cologuard, but it can be one more reason to make sure those are on schedule. If your result is part of a broader stool panel showing dysbiosis, lab results alone do not point to a specific treatment, and a gastroenterologist, registered dietitian, or functional medicine clinician can help integrate the data with your symptoms.

Companion tests that often add the most clinical context include fecal calprotectin (an inflammation marker in stool), fecal occult blood, and broader multi-organism stool panels that look at other cancer-associated bacteria like Fusobacterium nucleatum and Parvimonas micra. Combining bacterial markers consistently outperforms any single bacterium for distinguishing healthy controls from colorectal disease.