Collinsella

Low fiber intake is the clearest diet signal for this stool marker. That matters because the broader Collinsella genus is repeatedly linked with fatty liver disease, higher triglycerides, higher fasting insulin, and insulin resistance.

This is a research-grade measurement, not a diagnosis. There are no agreed-upon cutoffs, and a single number tells you less than the direction it moves after you change what you eat. But as a readout of how your gut is responding to your diet, it is better studied than many single organisms in the microbiome.

What the Test Actually Measures

Collinsella aerofaciens is a bacterial species that lives in the large intestine. Many human studies report the broader Collinsella genus instead, so genus-level findings are related evidence, not a perfect match to a species-specific stool result.

Stool microbiome tests usually look for bacterial DNA in your sample. PCR copies and counts selected DNA targets; sequencing reads many bacterial DNA fragments at once. Labs may report this as an estimated count or as relative abundance, and those numbers should not be compared across methods.

The number is a headcount of bacteria, not a measurement of anything your own body produced. A shift tells you the ecology of your gut changed. It does not directly tell you that your liver, your arteries, or your blood sugar changed.

These bacteria use simple carbohydrates and lactate. They tend to do well when there is little fiber competing for space and fewer fiber-fermenting bacteria crowding them out. The useful thing to know is simple: this marker is partly a reflection of what you feed your gut.

Fatty Liver Disease

The strongest human signal comes from biopsy-proven NASH. NASH is the inflammatory form of fatty liver disease. In a study of 141 adults, the Collinsella genus made up about 0.29% of gut bacteria in healthy controls, 3.45% in people with NASH without cirrhosis, and 4.38% in those with cirrhotic disease. That is roughly a twelvefold difference between healthy people and the non-cirrhotic NASH group.

In that same group, higher genus-level Collinsella tracked with higher total cholesterol and triglycerides, and with lower HDL cholesterol. Gut bacterial diversity was also lower in people with liver disease.

If your result is high and you have never checked liver enzymes, start there. ALT is a liver enzyme released when liver cells are stressed. A raised ALT alongside a high reading is a more meaningful combination than either alone. The study design was cross-sectional. It compared groups at one moment in time, so it cannot tell you whether the bacteria helped drive the liver disease or grew alongside it.

Insulin Resistance and Body Weight

Diet shows the clearest link here. In 130 overweight and obese pregnant women, lower fiber intake went with higher genus-level Collinsella. Higher abundance also tracked with higher fasting insulin and HOMA-IR, the standard calculated estimate of insulin resistance.

A pooled analysis of eight European studies covering 7,415 adults found that a body mass index 5 units higher came with about 11% higher odds of detecting Collinsella in stool. Higher BMI also came with lower overall bacterial diversity.

If your reading is high, the useful companion tests are fasting insulin and HbA1c. HbA1c reflects your average blood sugar over the past two to three months. Those blood tests tell you whether the metabolic pattern linked with the bacteria is actually present in your body.

Why High Is Not Simply Bad

The popular framing of Collinsella as a straightforward villain falls apart quickly. In 147 cancer patients being evaluated for Clostridioides difficile, people who carried the bacterium's DNA but were not producing its toxins had more genus-level Collinsella than those actively producing toxin. Losing Collinsella looked like a marker of active infection risk, not a cause of it.

In that same group, Collinsella abundance did not correlate with calprotectin or lactoferrin. These stool proteins rise when inflammatory white blood cells are active in the gut wall. If Collinsella reliably inflamed the gut lining, you would expect that correlation. It was not there.

Lower levels also show up in other settings. Unaffected first-degree relatives of people with Crohn's disease carried less Collinsella aerofaciens than controls. In pediatric Crohn's disease, lower Collinsella in intestinal tissue, a related but different sample, was part of a profile linked with relapse risk. One Northern China cohort reported lower stool Collinsella in primary Sjögren's syndrome, an autoimmune disease that often causes dry eyes and dry mouth. Children with COVID-19 also had lower stool Collinsella than controls. A country-level analysis of gut microbiome datasets from 953 healthy people found that countries with higher average Collinsella tended to have lower early COVID-19 mortality, but that was ecological evidence, not individual risk prediction.

The way to hold both sets of findings at once: this is not a good-number, bad-number marker. It expands in a low-fiber, simple-carbohydrate environment, which is the same environment often seen with metabolic disease. It may contract when the gut community is disrupted by infection, inflammation, or a diet change. High and low each mean something, but they mean different things, and neither means much without the rest of your clinical picture.

Predicting Who Responds to a Probiotic

One finding here is unusually actionable. In a randomized trial of people with irritable bowel syndrome that was not constipation-predominant, higher baseline Collinsella aerofaciens identified who would improve after 12 weeks on Lacticaseibacillus paracasei DG. The bacterium dropped over treatment, and it dropped only in the people whose symptoms improved.

That is a rare thing in microbiome testing: a measurement taken before treatment that predicts whether the treatment will work. It has been shown for one specific probiotic strain in one specific patient group, so do not generalize it to probiotics in general. But if you have bowel symptoms and are deciding whether that strain is worth a trial, this number carries real information.

Kidney Function and Other Associations

A related species, Collinsella stercoris, distinguished people with chronic kidney disease from controls better than the urine protein-to-creatinine ratio in one discovery-and-validation study of 220 people. That finding concerns a different species than this test, and it has not been replicated widely enough to change how kidney disease is staged.

Women with premenstrual syndrome carried roughly 4.5 times more fecal Collinsella than women without it in a Japanese study of 168 people. Lower levels have been reported in diabetic retinopathy, which runs opposite to the metabolic-disease pattern seen elsewhere. These shifts are condition-specific, not a single risk gradient.

Why a Single Reading Can Fool You

• Stool consistency and transit time: how fast stool moves through you and how much water it holds change the density of every organism counted. This is one of the largest sources of noise in microbiome profiling, and many reports do not correct for it.
• What you ate in the preceding days: because these bacteria respond to simple carbohydrates and fiber, a week of restaurant food before collection can shift the picture away from your usual state. Collect during a representative stretch of your normal eating, not after a holiday or a travel week.
• Recent infection: an active gut infection reshapes the whole community. If you have had a bout of diarrhea recently, wait until you have fully recovered before sampling.
• Lab-to-lab differences: targeted PCR and sequencing do not produce identical numbers. In one pregnancy cohort, genus-level sequencing and targeted PCR for Collinsella correlated only moderately, about 0.63 where 1.0 would be a perfect match. Compare your results only against prior results from the same lab and method.

Tracking Your Trend

One reading is close to uninterpretable on its own. There are no standardized reference ranges, so there is no threshold to compare yourself against. What you do have is yourself, measured twice.

This marker can move. A structured six-week low-calorie diet in people with type 2 diabetes produced an 8.4-fold drop in genus-level Collinsella at six weeks, while many microbiome changes partly drifted back toward baseline by 15 weeks. Twelve weeks on a specific probiotic strain dropped Collinsella aerofaciens in the people whose bowel symptoms improved.

Your own prior value is the only reference range you have, so build one if you are changing diet, weight, or a targeted probiotic. Direction matters more than the isolated number.

What an Unexpected Result Should Make You Do

A high reading is not a finding that stands alone. The useful move is to check whether the host-side pattern it associates with is actually present in you. Order a liver panel with ALT and AST, a fasting lipid panel, fasting insulin, and HbA1c. If all of those are clean, the bacterial finding is mostly a dietary signal. If ALT is raised and triglycerides are up alongside it, you have a coherent picture worth taking to a physician who works on liver and metabolic disease, and liver imaging becomes reasonable.

A low reading with gut symptoms points a different direction. Stool calprotectin helps distinguish inflammation in the gut wall from a shifted bacterial community without inflammation. If calprotectin is up and you have persistent diarrhea, blood in the stool, or unexplained weight loss, that combination warrants a gastroenterologist and likely endoscopy. The Collinsella number becomes a footnote to a more important workup.

If you have bowel symptoms without red flags and your reading is high, the probiotic response data gives you a reasonable experiment to run: a defined trial with Lacticaseibacillus paracasei DG, symptom tracking, and retesting after twelve weeks. That is a better use of this marker than treating the number itself as the problem.