Blood levels of imidazole propionate, a molecule made by gut bacteria and abbreviated ImP, tracked with how quickly memory and thinking skills slipped in 1,196 cognitively unimpaired adults whose average age at the start was 61.2. Those in the top quartile for ImP showed steeper decline than those in the bottom quartile. The work comes from University of Wisconsin-Madison teams led by Alzheimer’s imaging scientist Barbara Bendlin and microbiologist Federico Rey.
The Wisconsin group tested the question in two ways: by following people over time, and by giving the compound to mice bred to develop Alzheimer’s-type pathology. The two lines of evidence do different jobs, and only one of them can speak to cause.
The quartile comparison behind “fastest decline”
The phrase has a precise definition in the paper, published in Nature Communications. Participants from the Wisconsin Registry for Alzheimer’s Prevention and the Wisconsin Alzheimer’s Disease Research Center were sorted into quartiles by plasma ImP, and the analysis concentrated on the highest and lowest groups. Cognition was scored with three components of a modified Preclinical Alzheimer’s Cognitive Composite: animal naming, letter fluency and the Trail Making Test part B. The high-ImP group started with lower scores and declined faster with age on all three measures. “Fastest” therefore means the top quarter of the distribution, not a diagnosed subgroup of rapid decliners.
The university’s summary adds that ImP was measured in blood at baseline, before the decline was assessed, which is the basis for saying the compound preceded the slide. Rey put the finding plainly: people with the highest ImP levels “also experienced much faster cognitive decline.”
Tau markers and a gene on chromosome 12
In the same cohort, higher ImP went along with higher plasma p-tau217 and with neurofilament light chain, a marker of damaged neurons, according to the journal paper. A genetic variant, rs7969761 in the SLC6A13 gene, was tied to higher ImP; the School of Medicine and Public Health’s release says about 43 percent of participants carried a variant linked to substantially higher levels, possibly because it affects how the kidneys filter the compound.
A Mendelian randomization analysis, which uses inherited variants as a stand-in for lifelong exposure, estimated that each standard-deviation increase in genetically predicted ImP carried an odds ratio of 1.16 for Alzheimer’s disease. The effect is modest, with a reported P value of 0.003, and the paper’s own lead variant carries an odds ratio of only 1.02 for Alzheimer’s risk when looked at on its own.
Two other details frame who was studied. Plasma ImP was higher in men than in women across the cohort, and a smaller microbiome subset of 294 people with stool metagenomic data, drawn from the MARS cohort, was used to look at the bacteria that make the compound. Everyone in the main cohort was cognitively unimpaired when ImP was first measured, so the study speaks to decline in people who were still functioning normally, not to patients already diagnosed with dementia. Cognitive scores were also tracked as a decline over time, so a participant needed several visits to count toward the longitudinal comparison.
Plaques and tangles in ImP-treated mice
The animal work is where the causal claim lives. In 5XFAD mice, a strain that develops amyloid plaques, ImP supplementation increased plaque number and area. In PS19 mice, which model tau pathology, it drove tau hyperphosphorylation at several disease-relevant sites, and blocking the enzyme GSK3-beta prevented the effect in cultured neurons. According to the university release, when ImP reaches the brain in mice it increases buildup of abnormal beta-amyloid and tau, leading to neuron death.
Human data cannot show that. The people in the registry were observed, not assigned to high or low ImP, so the quartile finding is an association that the genetic and mouse work make more plausible. The Waisman Center listing of the earlier preprint, whose author list spans Wisconsin and collaborating groups including Gothenburg’s Henrik Zetterberg and Kaj Blennow, shows the work circulated as a bioRxiv manuscript in 2025 before the journal version.
Histidine metabolism and the drug target
ImP comes from bacterial metabolism of histidine, an essential amino acid abundant in protein-rich foods, so SciTechDaily’s rundown of the work says dietary modification alone looks insufficient. The Wisconsin team points instead to a drug. The university release notes the work builds on a decade-old finding by Bendlin and Rey that the gut microbiomes of people with Alzheimer’s differ from those of healthy people. Bendlin said that an inhibitor able to lower ImP levels in the blood could hopefully reduce the risk of Alzheimer’s.
No such inhibitor exists in a trial yet. What the paper does supply is a number to aim at: an odds ratio of 1.16 per standard deviation, and a top-quartile group that began about 61 years old and already scored lower.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- The FBI warns a new phishing trick can hijack your account without your password
- The NSA says three phone features should be off whenever you aren’t using them
- Hawaii’s magnitude 6.0 Kona earthquake is now a federal major disaster
- Comcast is shutting down its email service, and Xfinity customers must act to keep it
