Type 3 Diabetes: What Insulin Has to Do With Alzheimer's

"Type 3 diabetes" is not an official diagnosis. It is shorthand for a research hypothesis: that impaired insulin signalling in the brain may contribute to Alzheimer's disease ¹. A different link is far better established: in an analysis of 2.3 million people, those with type 2 diabetes were about 60% more likely to develop dementia ². And diabetes is one of 14 modifiable risk factors that, according to modelling by the Lancet Commission, together account for up to 45% of dementia cases worldwide ³.

The label is catchy, perhaps too catchy. It suggests Alzheimer's could be treated like diabetes, and so far trials have not borne that out. What remains is an encouraging message nonetheless: metabolism is one of the levers we can pull long before the first memory problems appear.

What does "type 3 diabetes" actually mean?

The term describes the idea that Alzheimer's involves insulin resistance in the brain, meaning nerve cells respond less to the hormone's signals ¹. It was coined in the mid-2000s by Suzanne de la Monte's research group. Their widely cited 2008 paper is a narrative review that mainly summarises lab, animal and tissue studies, so it is background rather than proof ¹.

In the brain, insulin does more than help manage blood sugar. It influences signalling pathways involved in nerve cell survival, in how adaptable synapses are, and therefore in learning and memory ¹. If that signalling breaks down, the hypothesis goes, it could encourage the changes typical of Alzheimer's.

One point matters for perspective: Alzheimer's is not a recognised form of diabetes. "Type 3 diabetes" is a metaphor for a line of research, a useful one as long as nobody takes it literally.

What do studies of the human brain show?

The clearest direct evidence so far comes from a study of brain tissue donated after death. Researchers took samples from the hippocampus, a region central to memory, and stimulated them with insulin in the lab. In people with Alzheimer's who did not have diabetes, the insulin signalling chain responded much more weakly than in comparison tissue ⁴. Markers of this insulin resistance rose from cognitively healthy people to those with mild cognitive impairment to those with Alzheimer's, regardless of diabetes status or the APOE ε4 risk gene, and were linked to poorer memory ⁴.

That is striking because it suggests the brain can be insulin resistant even when blood sugar looks normal. But studies like this have clear limits. Post-mortem tissue is a snapshot from a relatively small number of people. It cannot tell us whether brain insulin resistance helps cause Alzheimer's or results from it; the authors themselves considered it possible that amyloid proteins help trigger the problem ⁴.

How much does type 2 diabetes raise dementia risk?

By around 60%. That was the finding of a pooled analysis of 14 population studies with more than 100,000 dementia cases, and the figure was almost identical for women and men ². The link was strongest for vascular dementia, which results from damage to the brain's blood vessels: women with diabetes had roughly 2.3 times the risk, men 1.7 times ². For non-vascular dementias, mostly Alzheimer's, the increase was about 50% ².

Timing seems to matter too. In the British Whitehall II cohort of about 10,000 people, those whose type 2 diabetes began more than ten years before age 70 had roughly twice the dementia risk of people the same age without diabetes ⁵. Each five years earlier at diagnosis was associated with 24% higher risk ⁵.

Both studies are observational. They show associations, not proof that diabetes itself causes dementia. Shared factors such as high blood pressure, physical inactivity or vascular damage may play a part.

Can insulin or a diabetes drug treat Alzheimer's?

Not so far, at least not in people who already have symptoms. If Alzheimer's really were "diabetes of the brain", treatments that improve insulin action should slow it down. Two large trial programmes tested exactly that.

In the SNIFF trial, 289 people with mild cognitive impairment or early Alzheimer's dementia received insulin as a nasal spray or a placebo for twelve months. The main analysis found no benefit for memory, daily functioning or Alzheimer's biomarkers in spinal fluid, and some of the data were hard to interpret because of technical problems with the spray device ⁶.

Larger still were the evoke and evoke+ trials, with around 3,800 participants in early-stage Alzheimer's. Semaglutide, a GLP-1 receptor agonist used for diabetes and obesity, did not slow disease progression over two years any more than placebo did ⁷.

This weakens the idea that Alzheimer's can be treated like diabetes. It does not rule out that healthy metabolism over decades shapes risk. Treating established disease and preventing it in midlife are two different questions.

Why does metabolism matter before blood sugar rises?

Because type 2 diabetes usually develops over years. Insulin resistance often comes first: muscle, liver and fat cells respond less to insulin, and the pancreas compensates for a while by producing more, until it no longer can ⁸. During this phase, blood sugar may still be in the normal range, and there are usually no symptoms at all.

To check, the American Diabetes Association lists fasting glucose and HbA1c (long-term blood sugar), with an oral glucose tolerance test (OGTT) where needed [8]:

Test	Normal	Prediabetes	Diabetes
Fasting glucose	below 100 mg/dL (5.6 mmol/L)	100–125 mg/dL (5.6–6.9 mmol/L)	126 mg/dL (7.0 mmol/L) or higher
HbA1c	below 5.7% (39 mmol/mol)	5.7–6.4% (39–47 mmol/mol)	6.5% (48 mmol/mol) or higher
2-hour OGTT value	below 140 mg/dL (7.8 mmol/L)	140–199 mg/dL (7.8–11.0 mmol/L)	200 mg/dL (11.1 mmol/L) or higher

Without clear symptoms, a diabetes diagnosis requires two abnormal results ⁸. Cut-offs can vary slightly between guidelines and labs, so interpretation always belongs with a doctor.

One detail from the same guideline stands out to me: evidence that progression from prediabetes to type 2 diabetes can be prevented comes mainly from people with impaired glucose tolerance on an OGTT ⁸. A single normal fasting value tells only part of the story. Blood pressure, LDL cholesterol and body weight complete the picture, and each is a dementia risk factor in its own right ³.

If you want to see your metabolism as a whole, the relevant markers are part of the speak23 Core Health blood test.

What can you actually influence?

More than many people think, even though no single measure is a guarantee. The Lancet Commission lists 14 modifiable factors across the life course, including diabetes, high blood pressure, high LDL cholesterol, obesity, physical inactivity, smoking and untreated hearing loss ³. In its model, the four metabolic and vascular factors together account for roughly 12% of dementia cases, with high midlife LDL cholesterol the largest share at about 7% ³. Such estimates assume the factors are causal and could be eliminated completely. They describe a theoretical potential for populations, not a prediction for any individual.

What is very well established is that metabolism can change. In the Diabetes Prevention Program, an intensive lifestyle programme, aiming for at least 7% weight loss and 150 minutes of activity per week, cut new cases of type 2 diabetes by 58% compared with placebo over an average of 2.8 years; the drug metformin cut them by 31% ⁹.

Whether this also protects the brain is the question behind so-called multidomain trials. In the Finnish FINGER trial, a two-year programme of diet, exercise, cognitive training and vascular risk management improved thinking skills in 1,260 older adults at increased risk slightly more than general health advice did ¹⁰. In 2025, the US POINTER trial of 2,111 people showed that a structured, supported programme benefited cognition more than a self-guided one ¹¹. The effects in both trials were small, and whether they prevent dementia in the long run remains open.

Taken together, the evidence still points to clear, practical steps [3]:

• Know your numbers: have blood pressure, LDL cholesterol, fasting glucose and HbA1c checked regularly, and discuss anything unusual with a doctor.
• Build in movement: physical activity is one of the best-studied levers for both metabolism and the brain ³⁹.
• Tackle weight and smoking: both affect blood sugar and blood vessels, and through them, the brain ³.
• Get support: structured, guided programmes appear to work better than going it alone ¹¹.

For more on insulin, read our article Fasting Insulin: The Lab Value That Can Reveal Diabetes Risk Years Before Blood Sugar Does.

What does this mean for the medicine of tomorrow?

One of the Lancet Commission's most hopeful observations is easy to miss: in high-income countries, the age-specific risk of dementia has fallen, probably partly because vascular risks such as high blood pressure and smoking are better treated today ³. The number of people living with dementia is still rising, mainly because we are living longer ³.

Maybe that is the real point behind the "type 3 diabetes" label. Not that Alzheimer's is a sugar disease, but that brain and metabolism are more closely connected than we long assumed. Part of our brain health is shaped in years when we are not yet worried about it at all. That is not a reason for fear. It is an opportunity.