Diabetes drug and nasal insulin show brain benefits in Alzheimer's trial — as GLP-1 drugs fail

Two widely available medications — a diabetes pill and insulin delivered through the nose — have shown promising effects on memory, brain blood flow and toxic protein levels in people with early Alzheimer's disease, according to a clinical trial published in Alzheimer's & Dementia. The findings arrive the same week that Novo Nordisk's high-profile attempt to repurpose its blockbuster weight-loss drug semaglutide for Alzheimer's ended in failure, and as a separate large-scale review identified an unexpected group of drugs — antibiotics, antivirals and vaccines — as associated with lower dementia risk.

The Wake Forest University School of Medicine trial, led by professor Suzanne Craft, enrolled 47 older adults with an average age of 70 who had mild cognitive impairment or early Alzheimer's disease. Over four weeks, participants were randomly assigned to receive intranasal insulin alone, the diabetes drug empagliflozin (sold as Jardiance) alone, both medications together, or a placebo. Critically, none of the participants had diabetes — making this the first time empagliflozin has been tested in non-diabetic Alzheimer's patients, according to EurekAlert.

Our study suggests that targeting metabolism can change the course of Alzheimer's disease. For the first time, we found that empagliflozin, an established diabetes and heart medication, reduced markers of brain injury while restoring blood flow in critical brain regions. We also confirmed that delivering insulin directly to the brain with a newly validated device enhances cognition, neurovascular health and immune function.— Suzanne Craft, Ph.D., lead investigator and professor of medicine, Wake Forest University School of Medicine

The two drugs worked through different but overlapping mechanisms. Empagliflozin significantly lowered cerebrospinal fluid tau — the protein that forms toxic tangles in Alzheimer's brains — and reduced neurogranin and vascular markers linked to disease progression. It also increased HDL cholesterol, demonstrating metabolic benefits even without diabetes. Intranasal insulin, delivered at 160 IU daily through a cartridge pump device developed by Aptar Pharma, improved performance on sensitive cognitive tests and increased the structural integrity of white matter connections in the brain. It also reduced plasma GFAP, a marker of dysfunction in the support cells that maintain connections between blood vessels and brain cells.

Immune and inflammatory proteins in cerebrospinal fluid and blood were altered by both medications, a finding the researchers interpret as evidence that the drugs activate protective immune responses while curbing harmful inflammation. Proteins linked to the nasal-olfactory plexus were especially affected by intranasal insulin; the researchers describe that plexus as a newly identified pathway joining the brain's waste-clearance system to immune systems elsewhere in the body. Both drugs were safe and well-tolerated; compliance rates exceeded 97% and participants rated the nasal device 4.6 out of 5.0 for ease of use.

Brain insulin resistance is a common feature of Alzheimer's disease, and it typically coincides with vascular problems that limit blood flow and nutrient delivery to the brain. Together, those disruptions speed the buildup of amyloid plaques and tau tangles while impairing the brain's ability to clear them. The metabolic approach tested at Wake Forest targets these upstream problems — something the recently approved anti-amyloid drugs, which the researchers note have modest benefits and are unavailable to many patients due to side effects and contraindications, do not address.

We plan to build on these promising results with larger, longer studies in people with early and preclinical Alzheimer's disease. Because empagliflozin or intranasal insulin improved tau tangles, cognition, neurovascular health and immune function, we believe these treatments could offer real therapeutic potential, either on their own or in combination with other Alzheimer's therapies.— Suzanne Craft, Ph.D., lead investigator, Wake Forest University School of Medicine

That same week, Novo Nordisk reported results from two large phase 3 trials — named Evoke and Evoke+ — which together enrolled more than 3,800 people with early-stage Alzheimer's to determine whether semaglutide, the active ingredient in Ozempic and Wegovy, could preserve brain function. After two years of follow-up, semaglutide was not significantly better than a placebo on that measure, according to BioPharma Dive. Novo said it has decided to discontinue the one-year extension periods of both trials. Full results will be presented at a conference in March.

Based on the significant unmet need in Alzheimer's disease as well as a number of indicative data points, we felt we had a responsibility to explore semaglutide's potential, despite a low likelihood of success.— Martin Holst Lange, chief scientific officer and head of research and development, Novo Nordisk

Novo shares fell 8% on the news. The failure follows an earlier mid-stage trial of a different Novo GLP-1 drug, liraglutide, which also missed its main goal in Alzheimer's, though some secondary endpoints hinted at slight effects on cognition and grey matter preservation. Howard Fillit, co-founder and chief science officer at the Alzheimer's Drug Discovery Foundation, pointed to semaglutide's improvement of some Alzheimer's-related biological markers in the Evoke trials as a reason it might yet prove useful in combination therapy. 'Existing anti-amyloid drugs slow cognitive decline by around 30%,' Fillit said, 'so therapies aimed at other pathways will be crucial as we chip away at the remaining 70%.'

Research published in Alzheimer's and Dementia: Translational Research & Clinical Interventions introduces yet another angle on the disease. Drawing on health data from more than 130 million individuals and over one million dementia cases across 14 studies, a systematic review conducted by researchers from the universities of Cambridge and Exeter — among them CPFT Research and Development Director Dr Ben Underwood and Dr Ilianna Lourida from the University of Exeter — found that antibiotics, antivirals, vaccines and anti-inflammatory drugs such as ibuprofen were associated with a reduced risk of dementia.

The infection-related finding was described as unexpected. The study, as reported by the NHS trust, lends support to the idea that common dementias may be set off by viral or bacterial infections, and it fits with growing interest in whether vaccines such as the BCG vaccine for tuberculosis might lower dementia risk. The review also found conflicting evidence for several drug classes, including some blood pressure medications and antidepressants, with some associated with decreased risk and others with increased risk — a pattern the researchers attribute partly to differences in how studies were conducted and the fact that different medications within the same class often target different biological mechanisms.

We urgently need new treatments to slow the progress of dementia, if not to prevent it. If we can find drugs that are already licensed for other conditions, then we can get them into trials and – crucially – may be able to make them available to patients much, much faster than we could do for an entirely new drug.— Dr Ben Underwood, Research and Development Director, CPFT

Because a particular drug is associated with an altered risk of dementia, it doesn't necessarily mean that it causes or indeed helps in dementia. We know that diabetes increases your risk of dementia, for example, so anyone on medication to manage their glucose levels would naturally also be at a higher risk of dementia – but that doesn't mean the drug increases your risk.— Dr Ilianna Lourida, National Institute for Health and Care Research Applied Research Collaboration South West Peninsula (PenARC), University of Exeter

Taken together, the three bodies of research point toward a field in flux. The GLP-1 failure narrows one much-discussed avenue, while the metabolic trial and the observational review open others — particularly around infection, inflammation and the brain's own insulin system. Craft's team at Wake Forest plans larger and longer trials. The UK government, meanwhile, is supporting development of an Alzheimer's trial platform specifically designed to evaluate repurposed drugs rapidly, according to the NHS trust's account of the Cambridge-Exeter review.

Why it matters — With no cure for Alzheimer's and the most advanced approved drugs offering only modest benefits, these findings — positive for metabolic and anti-infective approaches, negative for GLP-1s — are actively reshaping where researchers and drug developers focus next.

Sources differ on Whether diabetes medication is associated with increased or decreased dementia risk: Some diabetes medications associated with decreased dementia risk (cpft.nhs.uk) vs Some diabetes medications associated with increased dementia risk (cpft.nhs.uk)

Reported by nhs.uk, cpft.nhs.uk, eurekalert.org, biopharmadive.com, ukdri.ac.uk