University of Florida researchers report that older adults with mild cognitive impairment who took glucosamine progressed to dementia faster, with mouse studies pointing to a reversible sugar tagging process in the brain.
A large observational study from University of Florida Health finds that older adults with mild cognitive impairment who reported taking glucosamine, a common over-the-counter joint-pain supplement, progressed to dementia more quickly than those who did not, and points to a specific, reversible metabolic process in the brain that may explain the link.
The work, published 9 June 2026 in the journal Nature Metabolism under the title "Hyperglycosylation is a metabolic driver of Alzheimer's disease," maps a sugar-tagging process inside neurons that may accelerate the disease, and shows that the process can be slowed in animals. Glucosamine, the supplement at the center of the analysis, is best known as a shellfish- or corn-derived pill millions of older Americans take for aching knees and hips, with annual U.S. use measured in the millions.
The clinical arm of the study combed through deidentified electronic health records at UF Health from 2012 to 2024, identifying roughly 1,900 patients with Alzheimer's disease and related dementias (ADRD) and about 2,750 patients with mild cognitive impairment (MCI). About 8% of each group reported using glucosamine. After adjusting for age, sex, and other variables, glucosamine use was associated with roughly 25% higher odds of progressing from MCI to dementia. In patients who already had an Alzheimer's diagnosis, glucosamine use was linked to about 25% higher mortality, a signal that did not appear in the MCI-only group.
The mechanism work is where the picture gets specific. Glucosamine is chemically similar to glucose, the simple sugar the brain runs on, and it crosses the blood-brain barrier. Once inside, it feeds a sugar-tagging process called N-linked glycosylation, in which the cell attaches short sugar chains to newly made proteins. In healthy brains, that process is balanced. In the Alzheimer's brain tissue the team examined, drawn from the UF Neuromedicine Brain and Tissue Bank, it runs in overdrive, a state the authors call hyperglycosylation.
In mouse experiments, the team showed that adding glucosamine to the animals' water raised intracellular protein glycosylation and worsened performance on social-memory tests. Chemically dialing glycosylation back down restored that memory performance, suggesting the pathway is not a one-way slide but a dial that can be turned.
That dial is the part of the finding outside Alzheimer's researchers are watching. The team puts the U.S. patient population at roughly 7 million people with Alzheimer's, alongside the millions more with related dementias, a group for which the last several drug classes to reach the clinic have produced only modest effects. A metabolic lever that responds to chemistry in animals is at least a credible target for human trials.
The authors are emphatic about what their data do not show. "It's an association and not proof of causality," said co-senior author Matthew S. Gentry, who chairs UF's Department of Biochemistry and Molecular Biology, in comments reported by SciTechDaily. Senior author Ramon C. Sun, director of the Center for Advanced Spatial Biomolecule Research and associate director for innovation at the McKnight Brain Institute, has called for clinical trials to test whether the association holds under controlled conditions and, separately, whether reducing glycosylation slows progression in people.
That distinction matters for the millions of older adults already taking glucosamine. The study design, an observational look back at medical records and self-reported supplement use from a single U.S. academic health system, cannot rule out the possibility that the kinds of people who choose to take joint supplements are also the kinds of people who, for unrelated reasons, progress faster. Selection and recall biases are real. The mechanistic work, in mice and in postmortem human tissue, is supportive but not confirmatory of a clinical effect. The next step the authors call for is not a public-health recommendation to stop the supplement but a registered clinical trial.
For now, the practical answer is the one most clinicians are already giving. Do not start or stop glucosamine on the basis of a single observational study, especially without talking to a doctor. The clinical question the field needs answered is whether the metabolic pathway the team has identified is a real therapeutic target in humans. If it is, the supplement aisle is the least interesting place that target might show up.