Type 3
Diabetes?

A nickname, not a diagnosis.

The term "type 3 diabetes" was put forward in 2008 by a group of researchers who had noticed how much the Alzheimer's brain looked, metabolically, like a diabetic organ — and who argued the resemblance was worth taking seriously.[1] It was a deliberately provocative label, meant to shake loose a new way of thinking about a disease that has defeated almost every treatment aimed at it.

So before anything else, the honest caveat, stated plainly: "type 3 diabetes" is a research nickname, not a diagnosis. No doctor will write it on your notes. It is not in any diagnostic manual. And Alzheimer's is genuinely multifactorial — age, genetics (especially the APOE4 gene), the amyloid and tau proteins, blood-vessel health and inflammation all play real parts. Insulin and metabolism are one thread in that weave, not the whole cloth.

Metabolic health and brain health travel together.

Start with the part that is well-established and replicated across very large studies: people with type 2 diabetes have a substantially higher risk of developing dementia — on the order of around 60% higher than people without it, in a pooled analysis of more than 2.3 million people.[3] The link holds across countries and decades.

That's an association — a risk factor, not a verdict. It doesn't prove that diabetes causes the dementia; shared roots (blood-vessel damage, high blood sugar, the whole metabolic fire) could drive both. But it's a strong, consistent signal that whatever is burning in the body in type 2 diabetes is not leaving the brain untouched.

The insulin machinery falters inside the brain.

The brain was long thought to be "insulin-independent" — it takes up glucose without needing insulin the way muscle does. But the brain is full of insulin receptors, and insulin does real jobs there: in memory, in how neurons connect, in their upkeep. And when researchers examined the actual brain tissue of people who had died with Alzheimer's, they found the insulin-signalling pathway measurably broken — the same kind of resistance seen in a diabetic muscle or liver, but in neurons.[2]

This is the observation at the heart of the whole "type 3" idea, and the observation itself is solid: in the Alzheimer's brain, the cells that should respond to insulin have gone partly deaf to it, and the degree of that deafness tracks with how much the person's thinking had declined.[2] A brain struggling to hear its own fuel-and-maintenance signal is a brain in trouble.

Cause, or consequence, or companion?

Here's the hard question the nickname leaps over. We can see insulin resistance in the Alzheimer's brain — but is it a cause of the damage, a consequence of it, or a companion riding alongside the real culprits? A sick, degenerating neuron might lose its insulin response because it's dying, not die because it lost it. Both stories fit the snapshot.

That uncertainty isn't a weakness to hide — it's the actual state of the science, and you deserve it straight.

So they tried giving the brain insulin.

If brain insulin resistance drives Alzheimer's, there's an obvious thing to try: get more insulin to the brain and see if thinking improves. You can do it with a nasal spray, which carries insulin up toward the brain while mostly bypassing the bloodstream. It's the most direct test of the whole idea — and it's been run as a proper randomised trial.

The largest such trial, published in 2020, did not find a benefit on its main measure of thinking and memory.[4] The result was muddied by a malfunctioning spray device partway through, so it isn't a clean "no" either — but it is certainly not the clean "yes" the hypothesis would have predicted. The honest summary is: the most direct test we have, so far, has not worked.

Hopeful, not magical.

So where does an honest reading leave you? Not with a cure, and not with a reason to despair — somewhere more useful than either.

• You can't "cut sugar to prevent Alzheimer's." No one has shown that any diet prevents or reverses the disease. Claiming otherwise is an overclaim, however well-meant.
• But metabolic health and brain health clearly travel together. Looking after blood sugar, blood pressure, weight, sleep and activity is among the few things genuinely linked to lower dementia risk across the research — and it's good for the rest of you regardless. That's a sound bet even while the "type 3" question stays open.
• The brain is not sealed off from the fire. Whatever the final verdict on cause, a body that's metabolically on fire is not giving its most fuel-hungry organ the easiest conditions. Cooling that fire is reasonable, low-risk, and yours to do.

The honest shape of a hard subject.

A firm link between metabolic disease and dementia. Real, measurable insulin trouble inside the Alzheimer's brain. A bold hypothesis that ties them together — still unproven, and not yet borne out by its most direct test. That's the honest shape of it: enough to take the metabolic fire seriously for the brain, not enough to promise anything. On this subject above all, that restraint is the respect it's owed.

Serious about the fire. Honest about the limits. Gentle with the people living this.