Prevotella/Bacteroides Ratio

Two people can eat the same high-fiber diet for months and get different results. In one Danish trial analysis, people with one gut pattern lost about 3.8 kg more body weight during a 24-week diet study than people with the other pattern. Same diet study. Different microbes.

The P/B ratio is the number used to separate those two groups. It is not a disease test and it has no validated normal range. The best-supported use is narrower: it may help predict fat-loss response to high-fiber, whole-grain diets in a few Danish trial analyses.

What the Number Actually Is

Prevotella and Bacteroides are two common bacterial genera that help define one of the main ways adult gut communities differ. They often show an inverse pattern, so when one is abundant the other tends not to be. The test divides the amount of one by the amount of the other, using DNA read out of a stool sample.

Neither genus comes from your body. Both are part of the microbial population living in your colon, and both belong to a broader bacterial branch now called Bacteroidota. Older papers and lab reports often call it Bacteroidetes. That naming detail is why this result gets confused with the Firmicutes-to-Bacteroidetes ratio, which is a different and much broader measure.

A higher ratio points toward a Prevotella-leaning community, which tends to be better suited to breaking down plant fiber and complex carbohydrates. A lower ratio points toward a Bacteroides-leaning community, which tends to handle animal-protein-rich diets and bile acids differently. Researchers often call these enterotype-like patterns. They are broad clusters, not diagnoses, and neither one is simply good or bad.

Fiber Response and Fat Loss

This is the strongest evidence behind the test, and it comes from randomized diet trials analyzed by baseline gut type.

In a 24-week post-hoc analysis in overweight adults, eating above rather than below 30 g of fiber per 10 MJ was linked to 8.3 kg of weight loss in high-ratio participants and 3.2 kg in low-ratio participants, a difference of about 5 kg. An earlier 26-week trial found high-ratio participants lost 3.15 kg more body fat on an ad libitum whole-grain diet than on an average Danish diet, while low-ratio participants showed no meaningful difference between the two diets.

Read that second finding carefully. For the low-ratio group, switching to whole grains did not move body fat much. That is not a reason to skip fiber, which has benefits unrelated to weight. But it does suggest that if you have been eating whole grains expecting the scale to move and it hasn't, your gut ecology may be part of the answer.

The effect is not universal, and it does not carry over cleanly to blood sugar. In a crossover trial using barley fiber, baseline Prevotella abundance did not separate metabolic responders from non-responders, and the barley itself improved glucose tolerance across groups. An earlier, smaller study pointed the other way: people whose blood sugar improved on barley bread had more Prevotella, and transplanting their stool bacteria into mice reproduced part of the effect. So the blood sugar picture is genuinely mixed, while the fat-loss signal has been the more consistent one.

Weight Regain After a Diet

The same trait that may help you lose fat on fiber can work against you when fiber drops.

In a weight-loss maintenance analysis, people with a high ratio regained about 1.5 kg more weight than low-ratio participants, and the effect was strongest when their glucose control was already impaired and fiber intake was low. A high-Prevotella gut appears to be more dependent on fiber, not just more capable with it.

This resolves something that otherwise looks contradictory. A high ratio is not a protective trait and a low ratio is not a defect. The ratio tells you how sensitive your weight may be to dietary fiber in both directions. If you have a high ratio and you are coming off a diet, keeping fiber high may matter more for you than it does for someone with a low ratio.

Metabolic and Obesity Associations

Outside feeding trials, the picture gets murkier. This is where popular gut-health narratives often get the test wrong.

Prevotella is often described as the healthy, plant-based genus and Bacteroides as the Western, unhealthy one. The observational data does not support that framing. In a cross-sectional study of 287 adults, a higher ratio was associated with higher BMI and with differences in brain reward circuitry, independent of sex, race, and diet. A Japanese cohort study reported a Prevotella-enriched cluster with roughly double the odds of obesity and diabetes, a single-cohort finding that has not been widely replicated.

So the same direction of the ratio looks helpful in Danish fiber trials and unfavorable in some Asian metabolic cohorts. Reviews of this literature describe the associations as conflicting across populations. The reconciliation is that this is not a good-number, bad-number marker. It is a phenotype indicator: it tells you which kind of microbial community you have, and each community carries different strengths and vulnerabilities depending on what you eat and where you live. A high ratio in someone eating a lot of fiber is a different biological situation from a high ratio in someone eating very little fiber, even though the number can look similar.

Autoimmune and Inflammatory Conditions

Several inflammatory conditions show shifted proportions of these two genera, though the evidence is mostly associative and does not show that the microbes cause the disease.

In ankylosing spondylitis, stool metagenomic sequencing found enrichment of Prevotella copri and Prevotella melaninogenica. The Bacteroides side is less consistent: some studies report depletion of Bacteroides species, while others report increases in related Bacteroidaceae, so the genus-level picture varies between cohorts. In Graves' disease the direction conflicts outright. Some studies report more Prevotella and less Bacteroides, others report the opposite, and results diverge between Asian and European cohorts. One Graves' orbitopathy study found Prevotella copri and Prevotellaceae tracking with thyrotropin receptor antibody levels. These are immune signals in people who already have disease, not screening findings for healthy adults.

Some of the biology is plausible. Bacteroides fragilis can produce propionic acid, and lab plus mouse work in Graves' disease suggests this may support immune cells that keep inflammation restrained, including one mouse study where the bacterium plus propionate added to the effect of low-dose methimazole. That is not the same thing as proving that changing your stool ratio prevents or treats autoimmune disease in people.

None of this makes the ratio a screening test for autoimmune disease. Multi-species models that include Bacteroides and Prevotella alongside other bacteria have separated Hashimoto's thyroiditis cases from controls in small cohorts. The ratio on its own has not been assigned a validated sensitivity or specificity for any autoimmune condition.

Colorectal Cancer Signals

In a cohort of 333 people undergoing colorectal cancer surgery, higher preoperative stool Prevotella abundance independently predicted better progression-free survival. A five-bacteria stool score outperformed conventional clinical predictors including carcinoembryonic antigen and lymphatic invasion after adjustment for cancer stage. In a separate diagnostic study, adding Bacteroides to a panel with Fusobacterium and Parvimonas improved separation between cancer patients and controls, and the panel improved again when Faecalibacterium was added.

These are prognostic and diagnostic panels, not the standalone ratio. A fecal occult blood test detects bleeding; these microbial shifts reflect a changed gut environment. They are complementary, and neither replaces colonoscopy.

Why One Reading Can Be Useful

Most biomarkers need serial tracking because they move constantly. This one appears more stable than many stool markers, though it is not immovable.

A 6-month randomized trial of the New Nordic Diet found that enterotypes defined by the P/B ratio usually did not shift, despite a substantial change in what people ate; only two participants moved between groups. Other work in older adults suggests prebiotic supplementation can change specific taxa without reshaping the whole community. Longitudinal data also suggest switches between the two dominant community states are uncommon in free-living adults.

That stability is the reason to test, if you test at all. If the number barely moves, it functions less like a lab value to chase and more like a baseline trait you can plan around. The more important question is whether the dietary strategy it points toward actually works for you.

Track the downstream results. Weight, waist, fasting glucose, and insulin respond to food changes on a timescale of weeks. Those are the numbers that tell you whether the strategy suggested by your gut type is working.

When Results Can Be Misleading

• Stool consistency: transit time and stool form correlate strongly with enterotype classification. Looser stools are more often classified as Prevotella-dominant. A sample collected during a bout of loose stools may not represent your usual state.
• Recent antibiotics: systemic antibiotics can lower gut microbiome diversity and shift genus-level abundance for weeks. A sample collected during or soon after antibiotics may reflect the drug exposure rather than your baseline ecology.
• Sodium intake: a small human stool study found high sodium intake tracked with lower Bacteroides and higher Prevotella while broader phylum-level measures changed little. That single finding has not been replicated, but a very salty stretch can make the ratio harder to interpret.
• Laboratory method: DNA extraction and sequencing amplification vary between runs and between labs. Comparing a result from one company to a result from another is not reliable. Use the same lab if you plan to retest.

Stool also reflects the far end of the colon, not the mucosal surfaces higher up, so the sample captures only part of the picture.

What to Do With an Unexpected Result

Since there is no threshold to cross, the useful question is not whether your number is high or low but what combination of findings you are looking at.

If you have a high ratio and you are trying to lose fat, the trial evidence points toward pushing fiber and whole grains harder than a generic recommendation would. Pair the result with fasting glucose, insulin, and HOMA-IR: high ratio plus impaired glucose handling is the specific combination that predicted the worst weight regain when fiber dropped, and it argues for treating fiber as maintenance rather than as a diet phase.

If you have a low ratio, the same trials suggest whole grains alone may be less likely to drive fat loss for you. That is information about which lever to pull, not permission to skip fiber. Look harder at calorie strategy, protein, and resistance training, and consider working with a registered dietitian who can build around the finding rather than fighting it.

If the result comes alongside real symptoms, persistent diarrhea, blood in stool, unexplained weight loss, or joint pain, the ratio is not the thing to chase. Those need a gastroenterologist and appropriate workup, including colonoscopy where indicated. Stool short-chain fatty acids and fecal pH can add useful context by showing what your microbes are producing, and fecal calprotectin distinguishes inflammatory from functional bowel symptoms in a way no enterotype ratio can.

Where This Fits in the Evidence

No randomized trial has shown that measuring this ratio in asymptomatic adults detects early disease, changes management, or improves outcomes beyond standard panels. No professional body recognizes it as a clinical parameter, and there are no validated reference ranges.

What it has is a specific randomized-trial signal about who loses more fat on fiber-rich, whole-grain diets. That is narrower than the marketing around gut testing suggests. It is also more useful than most exploratory stool markers ever get. Treat it as a planning input for diet, not as a verdict on your health.