The obesity gut-bacteria ratio did not hold up in 5,846 Finns

If you suspect your weight, your metabolism and your gut are tangled together, you have probably run into the claim that people with obesity carry the "wrong" balance of gut bacteria. A large Finnish analysis in iScience put that claim to the test in nearly 6,000 people. The famous ratio did not track body weight at all, and what did show up was smaller and more nuanced than the usual headlines suggest.

Key takeaways

• In 5,846 Finnish adults, the Bacteroidota-to-Firmicutes ratio (the balance between two dominant gut bacterial groups) was not associated with body mass index (BMI) or waist-hip ratio.
• Higher BMI did go with lower Shannon diversity (a score of how varied and evenly balanced gut species are), but BMI explained only 0.3% of the variation in overall gut microbiome composition.
• The data are a single snapshot, so they cannot show whether gut bacteria drive weight gain, follow it, or neither.

Firmicutes to Bacteroidetes ratio and obesity: no link

In a random sample of 5,846 Finnish adults, the ratio of Bacteroidota to Firmicutes (the two biggest bacterial groups in the human gut) was not associated with BMI or waist-hip ratio. The team, led by Li-Fang Yeo at Turku University Hospital with Rob Knight of the University of California San Diego among the authors, found a slope of 0.002 per standard deviation of BMI (p = 0.90) and 0.005 for waist-hip ratio (p = 0.58). Those are effectively zero.

The ratio has a long history. A 2006 Nature study in mice tied obesity to shifts in these two bacterial groups and to a gut community better at harvesting energy from food. Human data have been messier. The authors cite a 2023 Nature Microbiology piece on microbiome myths, along with earlier meta-analyses, for the conclusion that obesity is not linearly associated with this ratio, and say their sample is large enough to confirm it.

Gut diversity and BMI: a real but tiny signal

Higher BMI tracked with lower Shannon diversity, yet BMI accounted for only 0.3% of the variation in gut community composition. Shannon diversity dropped by 0.10 per standard deviation increase in BMI (p = 3.7e-10) after adjusting for age, sex, diabetes, cardiovascular disease, diet quality, physical activity, smoking and alcohol. Plain species richness (a simple count of species) was not significantly associated with BMI (p = 0.35), and lean and obese microbiomes showed no obvious clustering when plotted.

For anyone with obesity, type 2 diabetes or other metabolic problems who hopes a stool test will explain a stubborn weight issue, that matters: body size leaves only a faint fingerprint on the microbiome. Individual species did shift with BMI, with 132 staying significant after internal validation, mostly from the Oscillospiraceae, Lachnospiraceae and Bacteroidaceae families. Leaner participants had more bacteria linked to breaking down cellulose (plant fiber), such as Ruminococcus champanellensis. Akkermansia, a genus tested as a supplement in a small trial the authors cite, was not significantly associated with BMI or waist-hip ratio.

FINRISK 2002 cohort: internal and external validation

The study used FINRISK 2002, a population health survey in which 13,500 people aged 25 to 74 were invited and 7,355 provided a stool sample. After excluding pregnancy, recent antibiotics, very low sequencing counts and missing covariate data, 5,846 participants remained. Their stool DNA was read with shallow shotgun metagenomics (sequencing all the DNA in a sample, not one marker gene).

To cut false positives, the researchers used eastern Finland (3,906 people) for discovery and western Finland (1,940) for validation, then tested the survivors in an external cohort, FINRISK 2007, of 275 people. Only 43 of the 132 internally validated species were significant there, all in the same direction. The authors note earlier studies ranged from 20 to 1,674 participants and cite an mBio analysis finding that only one of ten obesity-microbiome studies was adequately powered.

What this gut microbiome finding could mean for you

A Firmicutes-to-Bacteroidetes number on a microbiome report is not a useful stand-in for body weight based on this dataset. If a test or article claims your ratio explains your weight, this study gives no support for reading it that way. Lower diversity went with higher BMI on average, but the effect is far too small to say anything about any one person.

Nothing here tells you to take, avoid or change anything. The diversity link persisted after adjusting for a diet-quality score and physical activity, but that is an association, not a treatment target. Earlier work is mixed too: a 2021 systematic review of 32 cross-sectional studies found lower Shannon diversity in obese adults in nine of 22 studies, while a meta-analysis of seven found no significant difference. If you are managing obesity, diabetes or a gut condition, ask your clinician what a stool test can and cannot tell you about your situation.

Cross-sectional design and shallow sequencing: the limits

This is a cross-sectional study, so it cannot say whether microbes influence weight or weight influences microbes. Stool was sampled once, in 2002, and frozen at -20 degrees C for about 15 years before sequencing. Shallow sequencing also makes functional annotation hard: the pathway results are predicted from the DNA, not measured, and the counts were highly sparse. Of 61 predicted pathways that passed internal validation, only 11 replicated externally. Treat those as hypotheses.

All participants were Finnish, although the authors say Finnish gut microbiomes resemble other Western cohorts. The external cohort was small at 275 people. The models did not adjust for metformin use, and Christensenella, a genus tied to obesity in lab work, was too rare to assess. For waist-hip ratio, a test of sample spread was significant (p = 0.001), so that result may reflect dispersion. Two authors disclose industry consulting ties, while the work was funded by public and foundation grants.

Repeated stool sampling: what comes next

The authors say repeated stool sampling and BMI measurements from the same participants are currently underway, which could show whether microbiome differences come before or after weight change. They also say targeted gut metabolomics (measuring the chemicals microbes actually produce) would be the best way to test the predicted pathway findings. No timeline for either is given.

Common Questions

Is the Firmicutes-to-Bacteroidetes ratio a good marker of obesity? In this study of 5,846 people, no. The ratio was not associated with BMI or waist-hip ratio.

Do lean people have more diverse gut bacteria? Slightly, on average. Shannon diversity was lower with higher BMI, but simple species richness was not, and BMI explained only 0.3% of overall composition.

Does this mean gut bacteria do not affect weight? No. A cross-sectional study cannot test cause and effect. The authors note that mouse studies showed the obesity phenotype could be transferred by gut microbiota transplant, but this human dataset cannot confirm that pathway.

References

1. Li-Fang Yeo, Joonatan Palmu, Aki S. Havulinna, et al. (2026). Population scale analysis reveals gut microbiome diversity and functional shifts linked to obesity. iScience. https://doi.org/10.1016/j.isci.2026.117043
2. D.H. Bessesen (2007). An obesity-associated gut microbiome with increased capacity for energy harvest. Yearbook of Endocrinology. https://doi.org/10.1016/s0084-3741(08)70093-3
3. Alan W. Walker, Lesley Hoyles (2023). Human microbiome myths and misconceptions. Nature Microbiology. https://doi.org/10.1038/s41564-023-01426-7
4. Mariona Pinart, Andreas Dötsch, Kristina Schlicht, et al. (2021). Gut Microbiome Composition in Obese and Non-Obese Persons: A Systematic Review and Meta-Analysis. Nutrients. https://doi.org/10.3390/nu14010012