Type 3 Diabetes: Why Alzheimer’s May Be a Metabolic Disease

My patient’s daughter sat across from me, exhausted and scared. Her mother — a sharp, witty woman who had run her own business for 40 years — had been diagnosed with early Alzheimer’s disease six months earlier. The daughter had done everything right: found a good neurologist, enrolled her mother in a memory program, arranged for daily care. But she wanted to know one thing that no one had been able to answer.

“Why did this happen? And could we have prevented it?”

Most doctors would have offered her a gentle but vague answer about genetics, aging, and the mystery of the brain. I gave her a different answer — one that the emerging science increasingly supports, and one that changes everything about how we think about Alzheimer’s disease.

“There’s a growing body of evidence,” I told her, “that your mother’s brain essentially became diabetic. Not her body — her brain. And that this process began decades before her first memory slip.”

The term for what I was describing is Type 3 Diabetes. And if you have a parent, spouse, or sibling with Alzheimer’s — or if you are watching your own memory quietly change and wondering why — this article may be the most important thing you read today.

What Is Type 3 Diabetes?

Let’s start from the beginning, because most people — including most physicians — have never heard this term, despite the fact that it has been appearing in serious scientific literature for over 15 years.

You already know what Type 1 diabetes is: the body doesn’t produce insulin at all, usually due to an autoimmune attack on the pancreas. You also know Type 2 diabetes: the body produces insulin but the cells have become resistant to it — they stop listening to its signal — so blood sugar rises and the whole metabolic system becomes dysregulated.

Now here is the concept that changes everything: your brain has its own insulin system. The brain is not just a passive recipient of blood sugar. It has insulin receptors on neurons. It uses insulin signaling to regulate energy, protect cells, clear toxic proteins, and maintain the architecture of memory. And just like the cells of a Type 2 diabetic’s body, the neurons of an aging brain can become resistant to insulin — losing the ability to use glucose efficiently, even when glucose is abundantly available in the bloodstream.

This brain-specific insulin resistance has been given the name Type 3 Diabetes — or, in the formal literature, Alzheimer’s disease reconceived as a metabolic disorder.

The Research Is Compelling — And It’s Been There for Years

This is not fringe science or a wellness blogger’s theory. The relationship between insulin resistance and Alzheimer’s disease has been documented in peer-reviewed research at some of the most respected institutions in the world — and the evidence is accelerating.

A 2025 systematic review published in a peer-reviewed journal concluded that insulin resistance was consistently identified as a key pathological driver, “impairing brain glucose uptake, amyloid-beta clearance, and tau phosphorylation” in Alzheimer’s disease. Let me translate those three mechanisms into plain language, because each one matters:

A landmark study from Joslin Diabetes Center published in 2025 directly links brain insulin signaling disruption to Alzheimer’s pathology and identifies it as a primary treatment target. Research from the University of Missouri published that same year found that individuals carrying the APOE4 gene — the strongest known genetic risk factor for Alzheimer’s — specifically struggle to convert glucose into brain energy, and that switching to ketone-based fuel may protect brain cells by bypassing this metabolic bottleneck entirely.

Perhaps most compelling: animal studies in which insulin resistance is experimentally induced only in the brain — leaving the rest of the body’s metabolism completely normal — produce animals that develop cognitive deficits, tau pathology, and amyloid accumulation identical to human Alzheimer’s disease. This is not correlation. This is a direct causal demonstration that brain insulin resistance causes Alzheimer’s-like neurodegeneration.

The Timeline That Should Alarm Everyone

Here is the part of this story I find most urgent — and most actionable.

Brain insulin resistance does not appear overnight. It develops slowly, over decades, long before a single memory test would catch it. Research now shows that measurable changes in brain glucose metabolism begin appearing on PET scans 15 to 20 years before the first symptoms of Alzheimer’s disease.

Let that sink in. The patient who receives an Alzheimer’s diagnosis at age 72 may have had a measurably compromised brain energy system since age 52 — or earlier. The “silent decades” are when the damage accumulates. And the silent decades are when intervention is possible.

A 2024 study published in Nature found that insulin resistance — measurable through standard blood testing — was associated with cognitive function decrements in middle-aged adults, independent of blood sugar levels. This is critical: you don’t have to be diabetic, pre-diabetic, or even have elevated blood sugar for insulin resistance to be silently damaging your brain. The insulin resistance effect on cognition operates through its own inflammatory and signaling pathways, entirely separate from glucose levels.

In my clinical practice, I see this constantly. Patients come to me in their 50s and 60s complaining of brain fog, word-finding difficulty, difficulty concentrating, and mental fatigue after minimal cognitive effort. Their blood sugar is normal. Their HbA1c is normal. Their regular doctor has told them everything looks fine. But when I test their fasting insulin levels — a test that most primary care physicians never order — the results tell a very different story.

Why Your Doctor Probably Isn’t Testing for This

The tests your doctor runs are designed to diagnose disease that has already arrived. They are not designed to detect the metabolic breakdown that leads to disease decades down the road.

A standard annual physical measures fasting glucose and HbA1c. These tests will catch Type 2 diabetes once it’s established. They will catch pre-diabetes when the process is fairly advanced. But they will completely miss the early, subtle insulin resistance that is — right now, silently — compromising the brain energy metabolism of millions of middle-aged adults who will be diagnosed with Alzheimer’s in 20 years.

The test you should be asking for is a fasting insulin level. It costs almost nothing. It is available at any standard laboratory. And it measures something entirely different from blood glucose — it measures how hard your pancreas is working to keep your blood sugar normal. When cells become insulin resistant, the pancreas compensates by producing more and more insulin. Blood sugar stays normal — for years — while insulin levels silently rise. That rising insulin, circulating chronically, eventually begins to down-regulate neuronal insulin signaling. The brain, bombarded by elevated insulin, turns down its sensitivity to the signal — exactly as the body’s cells did — and the Type 3 Diabetes process has begun.

How to Calculate Your HOMA-IR — Step by Step

Once you have your fasting insulin result, you can calculate a number called HOMA-IR — Homeostatic Model Assessment of Insulin Resistance. It is one of the most useful and most underused tools in preventive medicine. The math is simple enough to do on your phone.

You need two fasting blood values — drawn after at least 8 hours of no food or caloric beverages:

• •Fasting glucose — in mg/dL (on virtually every standard blood panel)
• •Fasting insulin — in µIU/mL (must be specifically requested — almost never ordered automatically)

A real-world example. Say your labs come back:

• •Fasting glucose: 95 mg/dL — completely normal by standard reference ranges
• •Fasting insulin: 18 µIU/mL — also considered “normal” by most labs

A score of 4.2 indicates significant insulin resistance — yet both individual values looked normal. This is exactly the trap. Each number in isolation told your doctor everything is fine. The calculation tells a completely different story.

How to Interpret Your Score

In my clinical practice, patients who present with brain fog, word-finding difficulty, and mental fatigue — with completely normal glucose and HbA1c — almost universally score between 2.5 and 4.5 on HOMA-IR. Their conventional physicians have no idea, because nobody ordered the fasting insulin.

Print this page and bring it to your next physical. Ask your doctor to add a fasting insulin to your next standard fasting blood draw. Calculate your HOMA-IR yourself. If your score is above 2.0, that is clinically significant — not because it means you have Alzheimer’s, but because it means the metabolic process that precedes cognitive decline by 15 to 20 years may already be underway.

The Alzheimer’s Prevention Diet: What the Evidence Actually Supports

Now for the question every patient and family member wants answered: “Is there anything we can do about this?”

The answer is yes — and it is more powerful than most physicians will tell you, because most physicians are not aware of the metabolic framework I’ve just described.

The intervention with the strongest mechanistic basis for addressing Type 3 Diabetes is therapeutic ketosis. The reason is straightforward once you understand the problem: if the brain can no longer efficiently use glucose, give it a fuel it can use efficiently — even in the presence of insulin resistance.

That fuel is beta-hydroxybutyrate, or BHB. As I described in my article on Ozempic and therapeutic ketosis, BHB is produced by the liver when carbohydrate intake drops low enough to trigger ketogenesis. BHB crosses the blood-brain barrier freely and is metabolized by neurons through a pathway that is completely separate from glucose — one that does not require insulin signaling and is not impaired by insulin resistance.

The research supports this directly. A University of California Davis study published in 2024 found that BHB levels increase almost seven-fold on a ketogenic diet, and that this increase plays a “pivotal role in preventing early memory decline.” An NIH-funded study from the University of Kansas Medical Center found that a ketogenic diet improved daily function and quality of life in patients with Alzheimer’s disease, with benefits appearing within 12 weeks. And a 2024 systematic review in Clinical Nutrition confirmed that ketogenic interventions improved cognitive function in patients with both mild cognitive impairment and Alzheimer’s disease.

The APOE4 Question: Genetics Is Not Destiny

Many patients who come to me are aware of the APOE4 gene — the genetic variant that is the strongest known risk factor for late-onset Alzheimer’s disease. Approximately 25% of the population carries one copy; about 2–3% carry two copies, which elevates lifetime Alzheimer’s risk substantially.

Here is the important nuance that the genetic conversation usually misses: APOE4 is a risk factor, not a sentence. The University of Missouri research cited earlier is especially relevant here — APOE4 carriers, particularly women, have a specific impairment in converting glucose to brain energy. Their primary fuel pathway is compromised by genetics. But their BHB pathway is not. For APOE4 carriers especially, therapeutic ketosis is not just a general brain health strategy. It may be the most targeted, mechanism-specific intervention currently available.

Knowing your APOE4 status is valuable. But knowing your HOMA-IR and fasting insulin is arguably more actionable — because those are the factors you can actually change.

Why Standard Alzheimer’s Treatments Haven’t Worked

I want to address the elephant in the room, because many patients and families come to me after years of trying conventional approaches and finding them inadequate.

The dominant pharmaceutical strategy for Alzheimer’s has focused on reducing amyloid plaques — the protein deposits long considered the hallmark of the disease. After decades of research and billions of dollars in investment, amyloid-targeting drugs have received approval, but their clinical benefit has been modest at best, with concerning side effect profiles.

If Type 3 Diabetes is the correct framework — and the evidence increasingly suggests it is — then the primary intervention needs to address the brain energy crisis, not the amyloid accumulation that results from it. This is the same reason the semaglutide Alzheimer’s trial failed, which I described in detail in my previous article: it addressed the wrong level of the problem.

What I See in My Own Patients

I want to bring this out of the abstract and into the clinical reality I witness regularly.

The patients who come to me with early cognitive symptoms — the foggy thinking, the word-finding failures, the feeling that their mental sharpness has quietly eroded — are almost universally metabolically compromised in ways their other doctors haven’t detected. Elevated fasting insulin. Elevated HOMA-IR. Sometimes measurably impaired glucose metabolism on functional testing. Often a family history that, in hindsight, tells the same metabolic story across generations.

When these patients enter therapeutic ketosis through the SKLeTT Protocol — Specific Ketone Level Titration Therapy — and their blood ketones are carefully titrated to a consistent therapeutic range, the reports I receive are remarkably consistent. Mental clarity returns. The fog lifts. Word retrieval improves. Energy stabilizes. Sleep deepens. Patients describe it as “feeling like myself again” — which is, from a neurological standpoint, precisely what is happening: the brain, finally receiving adequate fuel through the BHB pathway, begins to function as it was designed to.

I am careful not to claim that therapeutic ketosis cures or reverses Alzheimer’s disease. The research does not support that claim, and I will not make it. What the research does support — and what I observe clinically — is that therapeutic ketosis addresses the metabolic root cause of the brain energy crisis that drives Alzheimer’s pathology. And intervention in the early and middle stages of that process can meaningfully change the trajectory.

What You Can Do Right Now

If you have read this far, you are probably thinking about someone specific. A parent. A spouse. Yourself. Here is the most practical guidance I can offer:

The day that daughter asked me why her mother’s Alzheimer’s happened, I couldn’t give her a complete answer. No one can. But I could tell her that the science is showing us, with increasing clarity, that this disease has metabolic roots — roots that reach back decades, roots that are detectable, and roots that are, in the right window of time, addressable.

That is not a guarantee. It is something better: it is a reason to act.

Bibliography & References