Two Gut Bacteria Linked to an Alzheimer’s Metabolic Signature, Study Finds

If you or someone you love is living with Alzheimer's disease, or with the memory lapses of mild cognitive impairment (MCI), you've probably already read that gut bacteria might be involved somehow. A new modeling study adds a concrete mechanism to that hunch: it links two specific gut bacteria species to a cluster of blood chemicals already known to shift in Alzheimer's disease, and shows that a person's APOE gene, the single strongest genetic risk factor for the disease, changes how much influence those bacteria appear to have. The findings don't prove gut microbes cause Alzheimer's, but they sketch a testable pathway connecting what's happening in the gut to what shows up in the blood of people with memory problems.

Key takeaways:

• Researchers built computer models of gut-microbiome metabolism for 383 people and found that two bacteria species, Bacteroides uniformis and Bacteroides thetaiotaomicron, were the strongest predicted drivers of five blood-metabolite changes linked to Alzheimer's disease.
• The predicted effect differed depending on a person's APOE gene variant, the gene most strongly tied to Alzheimer's risk.
• This is a computational modeling study, not a direct measurement: when the researchers checked three of the predicted metabolite changes against real blood samples, the predictions didn't match the measured levels.
• The approach is a research tool, not a diagnostic test or a treatment, and no intervention was tested on anyone.

Two Gut Bacteria Linked to Alzheimer's Metabolic Shifts

The core finding comes from a team led by Tim Hensen, published in the journal Gut Microbes and first posted as a preprint on bioRxiv. Using data from 383 people, the researchers built personalized computer models of how each person's gut bacteria interact with the rest of the body's metabolism (whole-body metabolic modeling, a method that uses flux balance analysis, a way of simulating how biochemical reactions could flow through a system based on which genes and enzymes are present). The models predicted that five blood metabolites already known from past research to be altered in Alzheimer's disease were produced at a higher capacity in people with Alzheimer's dementia or MCI: S-adenosyl-L-methionine (a compound used in dozens of cellular reactions), L-arginine, creatine, taurine, and formate. When the team traced which microbes contributed most to these predicted shifts, two species stood out: Bacteroides uniformis and Bacteroides thetaiotaomicron, both common residents of a healthy human gut that were more abundant in the Alzheimer's and MCI groups than in the healthy controls.

Why This Matters for Alzheimer's and MCI

This should matter to anyone managing their own memory problems, or a family member's, because it points to a specific, testable channel between the gut and the blood, rather than a vague appeal to "gut health." The paper analyzed people already diagnosed with Alzheimer's dementia or MCI, the stage of cognitive decline that sometimes, but not always, progresses to Alzheimer's. For someone whose testing has turned up MCI, or whose parent has an Alzheimer's diagnosis, the result reinforces that whatever is happening in the gut tracks with disease stage, and that a person's APOE variant (a gene governing how the body transports cholesterol and fat in the brain, and the largest inherited risk factor for late-onset Alzheimer's) appears to shape that gut-to-blood relationship even before diagnosis.

Inside the 383-Person Whole-Body Modeling Study

The researchers built gut microbiome-personalized whole-body metabolic models for 383 people: 34 with Alzheimer's dementia, 51 with MCI and a probable Alzheimer's diagnosis, and 298 healthy controls. Each model was built from that person's own shotgun metagenomic sequencing (a technique that reads nearly all the DNA in a stool sample, rather than a handful of marker genes), mapped onto a reference database of microbial metabolism called AGORA2/APOLLO, then run through flux balance analysis to simulate how much of 38 curated blood metabolites the gut microbiome could plausibly help produce. To identify which microbes mattered most, the team paired that modeling output with a statistical method called LASSO-based stability selection, which repeatedly tests which bacteria behave as reliable predictors rather than one-off noise. A subset of 168 participants, including 28 with MCI and 10 with Alzheimer's dementia, had real blood chemistry data on file, which the researchers used to check the model's predictions against actual measurements.

What This Means If You're Worried About Memory Loss

For a reader managing their own brain health, this study is not a green light to buy a Bacteroides supplement or start megadosing arginine and taurine. No probiotic, diet, or supplement was tested, and the models describe a predicted capacity for microbes to influence blood chemistry, not a proven treatment effect. What is worth asking a doctor about, if it comes up, is whether metabolic or genetic risk markers like APOE status already factor into how your care team thinks about Alzheimer's risk. The honest takeaway is that this adds one data point to a growing picture of a gut-brain connection in Alzheimer's, not a new rule for what to eat or take.

The Study's Big Caveat: Predictions Didn't Match Blood Tests

The researchers' own validation check is the most important limitation here. When they compared the model's predicted metabolite changes against actual measured plasma levels in the 168-person subset, the three metabolites they could test showed no significant association between predicted flux and measured blood levels, a flux-plasma mismatch the authors report directly. In plain terms, the model's predictions about chemical production didn't line up with what blood tests actually found for those compounds. The authors also note they had no dietary intake data, a major influence on blood metabolites the models couldn't account for, and that mapping each person's gut species onto the modeling database reduced the average species identified per sample from 149 to 61, a real loss of detail. The models also assume a chemical steady state that doesn't capture how microbes and the body interact minute to minute, and can't represent the increased intestinal permeability seen in Alzheimer's-linked gut changes.

What Comes Next for Microbiome Modeling in Alzheimer's

The authors frame this as a hypothesis-generating step, and the obvious next move is the one their own validation check points to: testing more of the model's predicted metabolites against real, measured blood chemistry in larger groups, ideally with dietary records to rule out food as a confound. The approach follows a wider trend in Alzheimer's research of tying specific microbial metabolites to brain pathology and of untangling the molecular mechanisms behind microbiome changes seen across many Alzheimer's studies. If the two Bacteroides species hold up as consistent markers in future cohorts, they could become candidates for microbiome-based monitoring tools. For now there's no timeline attached, and no clinical trial underway based on these results.

Common Questions

Does this study show that gut bacteria cause Alzheimer's disease? No. The study used computer modeling to find associations between gut bacteria and blood chemistry already linked to Alzheimer's; it did not test whether changing those bacteria changes disease risk or progression.

What is a whole-body metabolic model? It's a computer simulation, built from a person's own gut microbiome DNA data, that estimates how microbial and human metabolism together could produce or use particular chemicals in the blood.

What is APOE and why does it keep coming up in Alzheimer's research? APOE is a gene involved in handling cholesterol and fat in the brain. Certain inherited versions, especially APOE epsilon4, are the strongest known genetic risk factor for late-onset Alzheimer's, which is why researchers keep checking whether findings like this one differ by APOE status.

Should I get my gut bacteria tested for Bacteroides levels? No clinical test or treatment recommendation comes out of this study. The bacteria identified here were part of a research model, not a validated diagnostic marker.

References

1. Tim Hensen, Lora Khatib, Lucas Patel, et al. (2026). Personalized whole-body modeling links gut microbiota to metabolic perturbations in Alzheimer’s disease. Gut Microbes. https://doi.org/10.1080/19490976.2026.2732659
2. Tim Hensen, Lora Khatib, Lucas Patel, et al. (2025). Personalised whole-body modelling links gut microbiota to metabolic perturbations in Alzheimer’s disease. https://doi.org/10.1101/2025.10.28.685084
3. Gut microbial metabolites and the brain-gut axis in Alzheimer's disease: A review. NIH. https://pmc.ncbi.nlm.nih.gov/articles/PMC12505531/
4. Federico Rey, Vaibhav Vemuganti, Jea Kang, et al. (2025). Gut bacterial metabolite imidazole propionate potentiates Alzheimer's disease pathology. https://doi.org/10.21203/rs.3.rs-7077445/v1
5. Yashar Vaziri, Jaleh Bagheri Hamzyan Olia, Cigir Biray Avci, Alireza Nourazarian (2025). Molecular mechanisms of gut microbiota dysbiosis and metabolites in Alzheimer’s disease pathogenesis: implications for precision therapeutics. Molecular Brain. https://doi.org/10.1186/s13041-025-01263-1