Written and medically reviewed by Dr Zeeshan Afzal (MBBS) — Medical Doctor & Medical Content Lead, Welzo
Evidence-based · Honest about what's proven and what isn't · UK-focused
Last updated: July 2026 · Reading time: ~13 minutes
Few health claims have travelled further on thinner evidence than "your gut bacteria are why you can't lose weight." The idea has a real scientific origin, a genuinely fascinating set of animal experiments behind it, and a human evidence base considerably weaker than the marketing suggests. It also has one recent, well-conducted trial result that deserves more attention than it has had. This guide covers all three. For the wider science see our digestive health hub, and the practical options in the Welzo gut health supplement range.
Table of contents
The foundational experiments are genuinely striking. Researchers took gut bacteria from obese mice and transplanted them into germ-free mice — animals raised with no microbiome at all. The recipients gained more fat than those receiving bacteria from lean donors, even when food intake was controlled. Later work repeated this using microbiota from human twin pairs discordant for obesity, with similar results.
That establishes something important: gut bacteria can causally influence adiposity, at least in a controlled animal model. It is why the field exists and why it is worth taking seriously.
Germ-free mice are an extreme model. They have no competing microbial community, are genetically near-identical, live in sterile conditions and eat a fixed diet. Humans have decades of accumulated microbial history, enormous genetic variation, and eat what they like. The gap between "bacteria can affect fat gain in a sterile mouse" and "changing your bacteria will change your weight" is where most of the commercial claims live.
You will have seen this one: obese people supposedly have a higher ratio of Firmicutes to Bacteroidetes, extracting more calories from the same food. It often comes with a specific number — an extra 150 calories a day or similar. It is the single most repeated claim in gut-and-weight content, and it has not held up.
Sze and Schloss conducted a meta-analysis pooling ten studies following PRISMA guidelines. They found modest pooled associations between obesity and alpha diversity measures — Shannon diversity, observed species counts and evenness. But no significant association was observed for the Bacteroidetes-to-Firmicutes ratio, nor for the individual relative abundances of either phylum. When they applied machine learning models to classify people as obese or non-obese from their microbiome, predictive power was poor. Their conclusion was that any association is weak and its detection confounded by large interpersonal variation and insufficient sample sizes.
Later reviews have reinforced this, noting the ratio is heavily influenced by diet, ethnicity and lifestyle, and that the field has moved towards functional characteristics rather than taxonomic ratios. If a product or test leans on your Firmicutes/Bacteroidetes ratio, that is a reason for scepticism, not confidence.
| Popular claim | Status | What the evidence says |
|---|---|---|
| Obese people have a high Firmicutes/Bacteroidetes ratio | Not supported | Failed to replicate in pooled analysis of 10 studies |
| Gut bacteria can affect fat gain | Supported in animals | Germ-free transplant studies are consistent |
| Lower diversity is associated with obesity | Weakly supported | Real but small; poor predictive power for individuals |
| A faecal transplant can make you lose weight | Not supported | Randomised trials show no weight or BMI benefit |
| Your microbiome explains why diets fail | Unsupported | No test predicts individual weight outcomes |
Reduced microbiome diversity does appear more often in people with obesity, and that finding survived meta-analysis where the phylum ratio did not. But "more often" is doing a lot of work — the distributions overlap heavily, and you cannot tell an individual's weight status from their microbiome. Diversity is a population-level marker, not a personal diagnosis.
The productive shift in this field has been from who is there to what they do. Bacterial fermentation of fibre produces short-chain fatty acids, which stimulate release of the appetite-regulating hormones GLP-1 and PYY from the gut wall. That is a genuine, mechanistically coherent route by which gut bacteria influence how full you feel — and it is diet-dependent, which is the practical point.
Bacteria do extract energy from otherwise indigestible fibre, so microbial composition can shift the calories you absorb. The honest framing is that this is a modest contributor operating alongside far larger determinants: total intake, food composition, activity, sleep, medication, genetics and socioeconomic circumstance. It is a factor, not the factor.
If the mouse studies translated directly, faecal microbiota transplantation should produce weight loss. It does not. A meta-analysis of randomised trials found no significant differences in weight, BMI, waist circumference, total body fat or lean mass — with no heterogeneity between studies, meaning the null result was consistent rather than muddled.
A double-blind randomised trial in adolescents with obesity found no impact of encapsulated FMT on its primary outcome of BMI or on total body fat over six months; a four-year follow-up still found no BMI difference, though some secondary body-composition and metabolic markers favoured FMT. FMT for weight remains experimental, is not licensed for this in the UK, and carries real risks.
Meta-analyses land in an awkward middle. One analysis in adults with overweight or obesity and related metabolic disease found no significant reduction in body weight (−0.26 kg) but a significant BMI reduction of 0.73 kg/m² and small decreases in waist and hip circumference — while explicitly cautioning that more research is needed before recommending probiotics as a therapeutic strategy.
In a cleaner test, a 2025 meta-analysis of 13 trials in 693 patients undergoing bariatric surgery found no significant difference in percentage excess weight loss, post-operative BMI or BMI reduction, with subgroup analyses showing no clinically meaningful effects. Probiotics have legitimate uses — see the best probiotics in the UK and do probiotics work — but weight loss is not one of the well-supported ones.
This is the exception worth knowing about, and the nuance matters.
The 2019 proof-of-concept trial published in Nature Medicine enrolled 40 overweight or obese insulin-resistant adults (32 completed) on three months of live or pasteurised A. muciniphila. Pasteurised bacteria improved insulin sensitivity by roughly 29%, reduced fasting insulin by about 34% and lowered total cholesterol by 8.7%. Body weight fell by 2.27 kg versus placebo — but at P = 0.091, which is not statistically significant. Read carefully, that trial was a metabolic result, not a weight-loss result.
The more interesting finding arrived in 2026. A randomised controlled trial put 90 adults with overweight or obesity through an 8-week low-energy diet to achieve at least 8% weight loss, then 24 weeks of ad libitum eating with pasteurised A. muciniphila MucT or placebo. Weight regain was significantly lower on MucT (1.2 kg versus 3.2 kg on placebo, P = 0.012), with greater net weight loss from baseline (3.1 kg, P = 0.009). Response related to participants' starting Akkermansia levels, and no serious treatment-related adverse events occurred.
That is a properly designed trial with a clinically meaningful endpoint — weight maintenance is where most diets fail. It is still one trial of 90 people and needs replication. Our guides to pasteurised Akkermansia and Akkermansia vs conventional probiotics explain why the heat-treated form behaves differently.
Berberine has a reasonable evidence base for glycaemic and insulin measures. An umbrella meta-analysis found it improved fasting glucose, HbA1c, HOMA-IR, fasting insulin and inflammatory markers in type 2 diabetes and PCOS, while cautioning about high heterogeneity and the small number of meta-analyses per subgroup. That is a metabolic story with relevance to blood sugar regulation, not a licensed weight treatment. It interacts with many medicines and is unsuitable in pregnancy or breastfeeding — see berberine interactions and berberine vs metformin, and talk to your GP first.
Strip away the overclaiming and a useful picture remains — it is just less exotic than a bacterial ratio.
Fibre is where gut health and weight legitimately intersect. It slows gastric emptying, adds bulk, and feeds the bacteria that produce short-chain fatty acids and trigger GLP-1 and PYY release. Most UK adults fall well short of the recommended 30g a day. Practical routes: 30 plants a week, high-fibre foods in the UK, how to increase fibre comfortably, resistant starch, prebiotic foods and soluble fibres like modified citrus pectin.
Highly processed foods are typically energy-dense, fibre-poor, quick to eat and easy to overconsume — and are associated with reduced microbial diversity. This is one of the more robust dietary findings in the field, and it operates through appetite and intake as much as through bacteria. See ultra-processed food and the gut, sweeteners and gut health and the best foods for gut health.
Human trials link increased fermented food intake to greater microbial diversity and lower inflammatory markers, and polyphenol-rich foods feed beneficial bacteria. Neither is a weight-loss intervention, but both improve the dietary pattern that is.
Short sleep alters appetite hormones and microbial composition, and increases next-day intake — see gut health and sleep. Alcohol adds energy-dense calories and disrupts the microbiome. These are less interesting than bacterial taxonomy and considerably more actionable.
Compromised gut barrier function allows bacterial products into the circulation, contributing to low-grade inflammation associated with insulin resistance. This is a plausible mechanism linking dysbiosis to metabolic dysfunction — and notably, barrier support is exactly what the Akkermansia trials were built around.
GLP-1 receptor agonists work partly by mimicking a hormone your gut already produces in response to food — including in response to short-chain fatty acids from fibre fermentation. It is a neat illustration that gut signalling genuinely influences appetite. It is not evidence that a supplement replicates a licensed medicine: these drugs achieve their effects at pharmacological doses no food or probiotic reaches.
Practically, these medicines commonly cause nausea, constipation and other gut symptoms, and adequate fibre and fluid intake matters more than usual while taking them. Metformin has its own well-known gastrointestinal profile — see metformin and stomach side effects. If you are prescribed any of these, discuss supplements with your prescriber rather than adding them independently, and never stop prescribed medication without advice.
For weight specifically, no. This is where the evidence is most decisive: the pooled analyses show microbiome profiles cannot reliably classify people by weight status, let alone predict who will respond to which diet. Personalised diet claims built on microbiome sequencing are ahead of the science. Our gut microbiome test guide covers what these tests can and cannot do. If you have persistent digestive symptoms, clinically validated testing arranged through your GP is a different and more useful proposition.
See your GP if you have:
If your relationship with food, weight or your body feels distressing or preoccupying, that is worth talking to your GP about too — the UK eating disorder charity Beat also offers support. Weight is genuinely difficult, it is influenced by far more than willpower, and asking for help is reasonable. See also when to see a GP about stomach symptoms. Source: NHS — Unintentional weight loss.
In germ-free mice, yes — transplanting microbiota from obese donors produces more fat gain than lean-donor microbiota, even with controlled intake. In humans the effect appears far smaller and less consistent. Gut bacteria are one modest contributor among many much larger determinants of body weight.
The phyla are real; the claim built on them has not held up. A meta-analysis of ten studies found no significant association between obesity and the Bacteroidetes-to-Firmicutes ratio or either phylum's abundance, and machine learning models predicted weight status poorly. The ratio also varies with diet, ethnicity and lifestyle, making it unreliable as a marker.
Not reliably. Pooled analyses have found no significant body weight reduction, with small effects on BMI and waist circumference of uncertain clinical relevance, and authors cautioning against recommending them for this purpose. A 2025 meta-analysis of 13 trials in bariatric surgery patients found no benefit for weight loss at all.
It has the best human evidence of any single strain, but the finding is about maintenance rather than loss. In a 2026 randomised trial, adults who had lost at least 8% of their weight on a low-energy diet regained 1.2 kg on pasteurised A. muciniphila versus 3.2 kg on placebo over 24 weeks. The earlier 2019 pilot showed clear insulin-sensitivity benefits, but its weight change did not reach statistical significance.
Counterintuitively, the heat-treated form has produced the stronger results in human trials — the 2019 study found pasteurised bacteria improved insulin sensitivity and lowered fasting insulin significantly, while live bacteria trended in the same direction without matching those effects. A heat-stable bacterial membrane protein is thought to drive much of the benefit.
No, on current evidence. Meta-analysis of randomised trials found no significant effect on weight, BMI, waist circumference or body fat, with consistent null results. A trial in adolescents with obesity missed its primary BMI endpoint at both six months and four years. FMT is not licensed for obesity in the UK and carries genuine risks.
No. Microbiome profiles cannot reliably distinguish people by weight status in pooled data, let alone predict individual dietary responses. Personalised nutrition marketing based on stool sequencing is currently ahead of the evidence.
Plausibly yes, and this is the most mechanistically credible link. Bacterial fermentation of fibre produces short-chain fatty acids, which stimulate release of the appetite-regulating hormones GLP-1 and PYY from the gut lining. This is also why fibre intake, rather than any particular bacterium, is the practical lever.
UK guidance is 30g of fibre daily for adults, and most people fall well short. Variety matters alongside quantity, since different fibres feed different microbes — aiming for 30 or more different plant foods weekly is a useful framing. Increase gradually with plenty of fluid to avoid bloating.
Almost certainly not in the way the marketing implies, and it is worth saying plainly: weight regain after dieting is the normal physiological response, driven by well-documented changes in appetite hormones and energy expenditure. That is not a personal failing and not a bacterial defect. If weight management is difficult, your GP can discuss NHS weight management services and, where appropriate, licensed treatments.
Medical disclaimer: This article is for general information and education only and does not constitute medical advice, diagnosis or treatment. It should not replace consultation with a qualified healthcare professional. No supplement described here is a treatment for obesity or a substitute for licensed medical care. Always seek advice from your GP or a suitably qualified clinician before starting new supplements — particularly if you are pregnant, breastfeeding, taking any medication including weight-loss or diabetes medicines, or have a medical condition. Berberine is unsuitable in pregnancy and breastfeeding and interacts with many medicines. Never stop prescribed medication without medical advice. Unexplained weight loss needs prompt medical assessment.
About the author
Dr Zeeshan Afzal (MBBS) is a practising medical doctor and Welzo's Medical Content Lead. He writes and reviews Welzo's health content to ensure it is accurate, evidence-based and aligned with current UK clinical guidance, translating complex research into practical advice patients can trust.
Ocenione doskonałe przez 26,523+ recenzji
Gwarancja 100% zwrotu pieniędzy
Nasza gwarancja zwrotu pieniędzy zapewnia, że otrzymasz swoje przedmioty lub otrzymasz pełny zwrot pieniędzy.
