How Sugar on Brain Cells May Be Driving Alzheimer’s and How a Common Joint Supplement May Be Making It Worse

By Kyle Proffitt

July 7, 2026 | While most of the research on Alzheimer’s Disease (AD) in the last few decades has focused on amyloid-beta plaques and the tau tangles—the misfolded debris cluttering and hindering normal brain function—a group at the University of Florida has identified a new angle of pursuit by performing an unbiased survey of all the different fats, sugars, and small molecule metabolites in the brains of Alzheimer’s patients. The standout result is a general increase in sugars stuck to the cell surfaces in AD samples, and the work led to the surprising, and potentially quite impactful, result that the common joint supplement glucosamine appears to accelerate AD progression. The study was published in June in Nature Metabolism (DOI: 10.1038/s42255-026-01538-4). 

Hyperglycosylation “is really just a bunch of sugars that have been added to the outside of your cells,” explained corresponding author Ramon Sun, Department of Biochemistry and Molecular Biology, College of Medicine, University of Florida. “When you have too much of it, the hypothesis is that your brain is starting to degenerate, and your neurons are no longer firing correctly, and you’re beginning to show signs of cognitive decline and eventually dementia.” But what is cause, and what is effect?  

New Methods Open New Avenues of Inquiry 

Sun’s lab is highly interested in method development, and in this case, methods they created have enabled the new discovery. “We try to leverage new technologies and just have a fresh, new take on essentially old chronic diseases that no one has really fully solved,” he said. Last May, they published a report demonstrating this spatial biomolecule detection technique, applied to the brains of normal and AD model mice. A high-speed laser scans across a section of tissue, vaporizing ionized molecules into a mass spectrometer, where different species are resolved by mass, shape, and size and quantified. “We can collect metabolomics, lipidomics, and glycomics at the same time from the same piece of tissue with really high precision,” Sun said. “Then you can reconstruct all of the signals bioinformatically.”  

In the newer work, they optimized sample preparation to enhance N-glycan (sugar chains attached to proteins) detection, enabling measurement of “a couple hundred different types of sugars,” and looked specifically at post-mortem human AD and normal brain samples. “Our study is really, I think, one of the earlier ones that basically didn’t even pay any attention to the cellular or protein aggregates side of Alzheimer’s disease,” Sun explained. “We really just went for the metabolism aspect.” 

They found both upregulated and downregulated lipid and metabolite species between AD and normal brain samples, but what really caught their attention was a striking increase of N-glycan content across gray and white matter in AD samples. After expanding their study sample size, they observed a progressive increase of glycosylation with worsening AD stage, particularly in gray matter.  

Sun explained that these sugars on brain cells “play a normal functional and physiological role … they’re actually required for your brain to function.” The sugar appendages help neurons recognize each other and conduct routine synaptic signaling. Their importance is evidenced by many congenital glycosylation disorders that are strongly associated with neurological defects. But what is their role in AD? 

Loss-of-Function and Gain-of-Function Experiments 

The researchers turned back to mice to study how hyperglycosylation might drive AD. Working with two different mouse dementia models—one amyloid-beta and one tau-driven —they identified the same patterns of increased hyperglycosylation. They also noted that the hyperglycosylation was pronounced in the cortex, hippocampus, and thalamus, “regions associated with memory, cognitive processing, and neuroinflammation, aligning with known patterns of neurodegeneration in AD,” according to the paper. Isotope labeling experiments in the AD mouse models revealed that glycan biosynthesis was increased, as opposed to a faulty glycan degradation apparatus accounting for the elevated cell-surface levels. The group also identified upregulated genes for glycan biosynthetic enzymes in both mouse AD models and human AD samples, and additional detailed experiments revealed neurons as the predominant cell population experiencing these increased glycan levels. “The basic building block machinery is hyperactivated,” Sun said. 

Next, the researchers tested whether interventions could influence cell glycosylation and disease pathology. An shRNA [short hairpin RNA]-based knockdown of PGM3, a pivotal enzyme for preparing N-glycan building blocks, reduced brain hyperglycosylation and improved performance in social interaction memory tests. A drug targeting a different enzyme directly responsible for sugar attachment to proteins (which had to be injected directly into the brain because it does not otherwise cross the blood-brain barrier) produced similar improvements. Interestingly, these interventions did not affect amyloid-beta formation or general inflammation. Improved memory without appreciable changes to these mainstream markers of AD pathology raises important questions about what’s actually driving the disease. 

As a further demonstration of the functional relevance, the researchers were able to worsen disease progression by feeding mice daily glucosamine. Glucosamine is able to cross the blood-brain barrier and feed directly into the hexosamine pathway, adding fuel for increased protein glycosylation. This intervention increased global N-glycan levels, and mice exhibited worsened memory. “Once you give oral gavage of glucosamine to these mouse models, it really does accelerate dementia and memory loss in these mice,” Sun said. Importantly, normal mice don’t experience these same effects; their glycosylation levels and memory scores are unaffected by glucosamine supplementation.  

Glucosamine: Check with Your Doctor 

Glucosamine was a convenient agent to directly supply the hexosamine pathway for N-glycan attachment to proteins, but it is also, of course, a well-known supplement for joint pain. It made perfect sense for the researchers to ask whether glucosamine supplementation might influence or accelerate AD pathology. They accessed records from the University of Florida health system covering over 50,000 patients with AD-related dementia (ADRD), a group that includes AD and also vascular dementia, Lewy body dementia, and frontotemporal dementia. The researchers compared these patients’ data against those with mild cognitive impairment (MCI) and found that in both groups, about 8% regularly used glucosamine. With a 10-year survival analysis, they identified a nearly 25% increased mortality rate for glucosamine users in the ADRD group, whereas mortality was not significantly impacted in the MCI group. However, as a general rule, about 5% of MCI patients per year transition to ADRD. Interestingly, of glucosamine users in the MCI group, there was a 25% increased likelihood of transitioning to ADRD in the 10-year follow-up; the total number of individuals transitioning from MCI to ADRD in this time period increased from about 48% to about 60% when taking glucosamine. 

Sun’s theory here is that it’s difficult to diagnose early-stage AD, and in the case of patients diagnosed with MCI but who actually have AD, “for those people taking glucosamine, it’s getting much, much worse” and more quickly revealing the disease. The primary conclusion from these findings is that there’s an AD-specific vulnerability to glycan overload.  “The good news is that glucosamine doesn’t really impact anything if you don’t have the disease already; it’s a very disease-specific factor,” Sun said. 

Sun’s recommendation, for now, is a little guarded: a prospective trial is warranted, but until one is performed, he would “exercise caution for people who have Alzheimer’s disease who are also taking glucosamine; I would say talk to your neurologist or primary care doctor to assess whether you really need it.” If glucosamine were a really effective treatment, this advice would be harder to give, but analyses have consistently failed to identify efficacy for glucosamine/chondroitin sulfate supplements greater than placebo. The impact of this advice could be huge. Based on the number of individuals with AD in the US and the 8% likely to be using glucosamine, this could mean that “over 1 million patients may be unknowingly exacerbating their disease progression through glucosamine supplementation,” according to the paper. 

Outstanding Questions 

Several questions remain. The team has not pinned down why the glycosylation machinery becomes hyperactive in Alzheimer’s brains in the first place, nor exactly how excess surface sugar translates into neuronal dysfunction. There is at least a theory about the effects: “if you have too much sugar on the outside of your cell, you can totally imagine it’s going to start to aggregate and disrupt all the cellular processes,” Sun said.  

They have also not directly tested the link between dietary or circulating glucose and brain glycosylation, although the same biosynthetic pathway can be fed from glucose as well as glucosamine—a plausible biochemical bridge to the long-observed epidemiological link between diabetes and Alzheimer’s risk. There’s an interesting wrinkle to this idea though. Other studies have shown not increased but reduced glucose uptake in AD brains, even years before clinical symptoms present. Because these processes are occurring in brain cells with low glucose levels, there is a suggestion that the excess glycosylation and the upregulated glycan biosynthetic enzymes are compensatory mechanisms.  

The full metabolomics dataset underlying the paper has been released; “all of these are publicly available data,” Sun noted. “People can download the data and just start playing it with themselves … there’s a lot of different pathways that are unregulated during Alzheimer’s disease, and we just chose one to validate our findings.” 

What Happens Next 

A prospective glucosamine trial in Alzheimer’s patients is the obvious next move. Until then, the human data remain correlational, and the mouse data carry the causal weight, a limitation Sun acknowledged while arguing that the convergence of human screening, mouse mechanism, and clinical-record outcomes is enough to warrant both clinical caution and a real trial. 

On treatment, the most immediately actionable step is to stop adding glucosamine to a brain that is already hyperglycosylated. The harder, longer path is a brain-penetrant drug to throttle the glycosylation pathway, but the blood-brain barrier is “usually the biggest challenger of finding drugs that hits the brain,” Sun said. Gene therapy and siRNA approaches are also on the table. The lab itself is not pursuing drug discovery. “We’re working on something different currently that could also just modulate brain metabolism,” he said. This includes lifestyle changes—diet, exercise, sleep apnea, oxygenation—and how those factors shape disease trajectory.  There is a through line in which Sun is interested in why some individuals with a disease will worsen quickly while others fare much better. As the glucosamine story shows, “whatever you do on a daily basis does impact—either reverse or accelerate—how fast or slow your disease is progressing,” he said.