top of page
Search

Diabetes 3.0: Why your brain might be quietly starving in a sea of sugar



Picture a city at night, every window lit, the grid humming - and yet the streetlights flicker out one by one. There is no shortage of electricity; the cables are thick with it. The problem is that the lamps have stopped answering when the current calls. That, in a single image, is the paradox at the heart of one of the most intriguing ideas in modern neuroscience: a brain swimming in fuel and starving all the same.


The brain is a famously expensive organ - roughly two percent of your body weight, burning around a fifth of your daily glucose, never resting, lit up whether you are solving equations or fast asleep. To keep those lamps glowing, brain cells rely on insulin - the same hormone you associate with the pancreas and the diabetic's blood sugar - to shuttle, signal and keep the lights on. And when neurones stop listening to insulin, growing deaf to it the way an overstimulated nerve grows numb, something goes wrong in a way that looks, on the scan and under the microscope, uncomfortably like diabetes.


That resemblance is why, back in 2005, researchers Suzanne de la Monte and Jack Wands at Brown University reached for a deliberately startling phrase. In the brains of people who had died with Alzheimer's they found impaired insulin signalling, fewer insulin receptors and disturbed glucose handling, and proposed that the disease might be understood, at least partly, as a third form of diabetes - a metabolic failure that selectively involves the brain (de la Monte & Wands, 2008). The phrase that has since travelled the internet - Diabetes 3.0, Type 3 diabetes - was born there. Vivid, memorable, and worth stating plainly: it is not an official diagnosis. The American Diabetes Association does not recognise it, and researchers still debate whether it names a cause, a consequence, or simply a striking family resemblance (Shaw, 2017).


So we are standing on contested but genuinely fascinating ground. Let me walk you across it honestly - the firm footing and the soft patches both.



The crime begins twenty years before the first witness


Here is the fact that reframes everything else, and the one most people are never told. By the time someone notices they are losing their keys, or repeating a question, or struggling with a recipe they have made a hundred times, the underlying disease is not new. It is middle-aged.


In families carrying rare genes that guarantee early Alzheimer's, scientists could do something almost no other disease allows: predict roughly when symptoms would arrive, then rewind the tape and watch the biology unfold beforehand. The landmark study of this group, in the New England Journal of Medicine in 2012, found that toxic amyloid begins accumulating around fifteen to twenty years before the first symptom, while the brain's ability to burn glucose starts dimming roughly a decade out (Bateman et al., 2012). The disease process, the authors concluded, begins more than twenty years before dementia is ever diagnosed.


Think of it like rust under a beloved old car: by the time it bubbles up through the paint, the corrosion has been spreading for years. That is why "how long does it take?" has such an unsettling answer - and why it is also quietly hopeful. The long, silent runway is exactly the window in which the metabolic story matters most, because metabolism, unlike your birthday, is something you can act on.



A dose-response you can measure in milligrams


If high blood sugar genuinely menaces the brain, you would expect to see the risk rise as the sugar rises, and not only in people who have crossed the official line into diabetes. That is precisely what one of the most quietly important studies in the field found.


Researchers followed more than two thousand older adults in Seattle, drawing on tens of thousands of glucose measurements over years, and tracked who developed dementia. Among people without diabetes, those with an average glucose of 115 mg/dL had an eighteen percent higher risk than those at 100 mg/dL (Crane et al., 2013). Read that again: both numbers sit in the range a doctor would casually call normal, and the higher one still carried more risk. Among those who did have diabetes, the gradient kept climbing -190 versus 160 mg/dL carried roughly a forty percent higher risk (Crane et al., 2013).


Zoom out to the whole population and the signal holds. A pooled analysis of fourteen studies, more than 2.3 million people, found that type 2 diabetes was associated with around a sixty percent higher risk of dementia (Chatterjee et al., 2016). Sixty percent is not a rounding error. It is the kind of number that, attached to almost any other organ, would have us redesigning public health campaigns.


And the newest twist points away from your fasting number toward what happens after you eat. In late 2025, a UK Biobank study used Mendelian randomisation, - a genetic method designed to probe possible causes rather than mere correlation - and found that genetically predicted higher glucose two hours after a sugar load was associated with a sixty-nine percent higher Alzheimer's risk, while fasting glucose and fasting insulin showed no such signal (Mason et al., 2025). The honest caveats matter: the finding failed to replicate in a second, independent dataset, and the cohort was healthier and less diverse than the general population. So it is a provocative lead, not a verdict. But its quiet implication is the part to keep - the spike after a meal may matter more than the number you fast for in the morning.


Here, though, the honest researcher has to slow you down, because this is the most seductive trap in the subject. Association is not causation. These studies show higher glucose travelling alongside higher dementia risk; they do not prove the sugar is doing the damage, nor that lowering it saves the brain. The lead author of the Seattle study said as much - his data could not tell us whether bringing glucose down would prevent or modify the disease (Crane et al., 2013). High-sugar people differ in a dozen other ways: activity, sleep, inflammation, thirty years of dinners. The number is a flare in the night sky - it tells you something is burning, not what.



Why your neighbour eats the same lunch and stays sharp


Anyone who has watched dementia move through a family asks the same quiet, frightened question: why him and not her? Why the lifelong smoker who stayed lucid to ninety-five, and the careful, vegetable-eating teetotaller who did not?


A large part of the answer is spelled with four letters: APOE. Everyone carries two copies, one from each parent, in a few flavours. The common APOE3 is essentially neutral. The rarer APOE4 is the single strongest common genetic risk factor for late-onset Alzheimer's - it does not sentence you to the disease, and many who develop dementia never carried it, but it tilts the odds hard. Roughly a quarter of people carry one copy; two to three percent carry two (National Institute on Aging, n.d.). For that small double-dose group, the risk runs many times higher than for non-carriers (Max Delbrück Center, 2025).


For years the mechanism was a mystery, which is why a 2025 study offers such a tantalising clue. Working with human stem-cell-derived neurones and mouse models, scientists at the Max Delbrück Center and Aarhus University, publishing in Nature Metabolism, described something that reads almost like a plot twist for our sugar story. As the brain ages, neurones lose some of their ability to run on glucose - so healthy cells learn a clever trick, switching to burning fats for fuel when sugar runs short, the way a hybrid car flips to its second engine on a long hill (Max Delbrück Center, 2025). APOE3 lets neurones make that switch. APOE4 appeared to disrupt that switch in the experimental models, limiting the cell's ability to take up and burn those fats, so that when glucose dips the cell has no backup and edges toward starvation (Max Delbrück Center, 2025). Worth being precise, though: this is a mechanism seen in cells and mice - strikingly plausible, but a clue, not yet a proven account of the living human brain.


Sit with what it would imply. For a person carrying APOE4, the brain may already be walking a metabolic tightrope with the safety net frayed. Add unstable blood sugar, insulin resistance, the metabolic churn of midlife, and you have asked a cell with one engine to climb a mountain. It may be the most elegant explanation yet for why the same plate of food, the same decade of stress, lands so differently on two different brains. Genes load the gun; metabolism, among other things, helps pull the trigger.



The reality check: what happened when we sprayed insulin straight into the brain


If the theory is that the brain has gone deaf to insulin, the obvious experiment is also the most elegant: give the brain insulin directly and see if the lights come back on. You cannot inject it into a vein without crashing the body's blood sugar, but you can deliver it through the nose, where a back-road of nerves carries molecules toward the brain.


This is where the field's intellectual honesty gets tested, and largely passes. The major trial, led by Suzanne Craft and published in JAMA Neurology in 2020, gave people with mild cognitive impairment or early Alzheimer's a year of intranasal insulin or placebo. The headline result was deflating: overall, no meaningful cognitive or functional benefit (Craft et al., 2020). A faulty delivery device muddied the water - the subgroup using a reliable one showed faint hints of benefit - but the cleanest direct test of the "feed the starving brain its insulin" idea did not deliver the triumph the theory predicted. That matters. The relationship between brain insulin and dementia is real and worth chasing - but it is not a solved equation with a tidy answer key. Anyone who tells you otherwise is selling you the ending of a story science is still writing.



What actually moves the needle


So strip away the hype and the despair both. What does the strongest evidence say a person can do - not to cure what may already be decades along, but to load the dice in their favour?


The most encouraging body of work comes from the FINGER trial in Finland, the first large, long-term randomised study of whether changing how people live could protect ageing minds. Over two years, at-risk older adults were guided through a combined programme - a brain-friendly diet, exercise, cognitive training, and tight control of blood pressure and blood sugar - while a comparison group got general advice. The intervention group did measurably better on cognition (Ngandu et al., 2015). A genuine landmark.

It was also, to keep us honest, a modest effect - and critics made a fair point: stacked together, these interventions do not simply add up, and FINGER's combined benefit came out roughly half the size of what some ingredients managed alone in standalone trials (Lampit & Valenzuela, 2015). Lifestyle is powerful, but it is not magic, and not as neatly additive as we would like.


Then, in 2025, came the result the field had been holding its breath for. The US POINTER trial in JAMA tested the same multi-pronged approach across thousands of diverse older Americans, pitting a structured, supported programme against a looser self-guided one. Both improved cognition; the structured version did better - and the benefit held regardless of APOE4 status (Baker et al., 2025). Your genes are not the referee with the final whistle.


Now to the kitchen, and a deflating little cautionary tale. For years the MIND diet - a marriage of the Mediterranean and DASH patterns, heavy on leafy greens, berries, nuts and fish - rode a wave of headlines after observational studies linked it to slower decline. Then someone ran the harder experiment: a proper randomised trial, published in the New England Journal of Medicine in 2023. Over three years, the MIND dieters did no better than the control group on cognition or brain scans, and the lead investigator admitted she had expected a win and was surprised (Barnes et al., 2023). The catch: both groups were eating better and gently cutting calories, both improved, and the trial may have been too short to separate them. The lesson is not that diet does not matter. It is that the gap between associated with and proven to cause is where most health advice quietly falls apart - and the people willing to show you a null result are the ones worth trusting.


Which brings us to supplements, and the most fascinating "it depends" in the field. The VITACOG trial at Oxford gave older people with mild memory problems and high homocysteine - a marker linked to brain shrinkage - a combination of B vitamins (B6, B12, folate), and found it slowed brain atrophy (Smith et al., 2010). A clean win, you might think. But the follow-up revealed the twist that explains a thousand contradictory supplement studies: the B vitamins only worked well in people who already had good omega-3 levels - slowing shrinkage by around forty percent in that subgroup, and doing little in those low in omega-3 (Jernerén et al., 2015). The two nutrients looked less like solo performers than dance partners: the signal was strongest when both parts of the pattern were in place. Which is why "does this supplement work?" is almost always the wrong question, and "for whom, with what, measured how?" is the right one.



The Bredesen question


No honest piece on this subject can dodge the most polarising name in it. Dr Dale Bredesen, formerly of the Buck Institute, built a programme - branded ReCODE™ - on a genuinely appealing premise: that cognitive decline is not one disease with one cause but a final common road reached by many routes (metabolic, vascular, inflammatory, toxic, hormonal), and that you should test each person, find their particular potholes, and fix as many as you can at once. Reducing insulin resistance sits near the centre of it; the protocol targets tight metabolic numbers, aiming for low fasting insulin and a healthy HbA1c, the very levers our whole story has been circling.


Here is the case in its favour. The targets it chases - stable blood sugar, exercise, sleep, nutrient gaps, inflammation - are each backed by real science, and treating the whole person rather than awaiting a single blockbuster drug is sensible for a disease where single-drug trials have a long graveyard. The supporting evidence has grown, too: alongside earlier case series and a single-arm proof-of-concept trial (Rao et al., 2024), December 2025 brought the approach's first randomised controlled trial, which reported a large, statistically significant gain in overall neurocognitive scores for the treated group (Toups et al., 2025).


And here is the case against, which that trial narrows without closing. It remains a preprint - not yet peer-reviewed or independently replicated - it was small (seventy-three patients), and it was run by the protocol's own developers rather than a neutral team. Its gains, moreover, showed up on a computerised cognitive battery but not on the standard Alzheimer's tests, and it did not shift brain volume or the blood biomarkers that define the disease - so "better scores" is not yet the same as "reversed Alzheimer's." Bodies including the Alzheimer Society of Canada have warned that the approach is marketed well ahead of the independent evidence that would settle it, and headline claims now circulating, that it dramatically outperforms licensed drugs, are exactly the kind of leap a careful reader should treat warily (Alzheimer Society of Canada, n.d.).


So where does that leave a thoughtful reader? Roughly here: the protocol's emphasis -assessing the metabolic, vascular, inflammatory and nutritional factors that shape brain health, and acting on the ones that are off - overlaps with the direction mainstream prevention science is already travelling. Its stronger claims, particularly around reversing established disease, are now beginning to be tested in controlled trials, but are not yet confirmed by the independent, peer-reviewed replication that would settle them. Both can be true at once, and a good clinician holds them together without flinching: interested in the framework, candid about the evidence gap.



What the diabetes drugs are teaching us - and what they are not


If high blood sugar harms the brain, surely the drugs that lower it should protect it. Reasonable hypothesis; encouraging early hints. Pooling more than 1.5 million patients, observational studies have linked two diabetes drugs - metformin and the SGLT2 inhibitors - to notably lower rates of dementia and Alzheimer's than other glucose-lowering medicines (Sunwoo et al., 2024). But observational data carries its usual ghost: the people prescribed those particular drugs may simply be healthier in ways the statistics cannot fully scrub out.

Then one of the cleanest tests arrived and humbled the simpler version of the theory. Semaglutide - the GLP-1 drug behind Ozempic and Wegovy - was run through two large phase 3 trials in early Alzheimer's, EVOKE and EVOKE+, nearly 3,800 people between them. It nudged some Alzheimer's biomarkers in the right direction, then did nothing measurable to slow the disease over two years (Cummings et al., 2026). The metabolic theory had its marquee experiment, and the clinical decline was not measurably slowed.


This is the field's most useful corrective. "Blood sugar is involved" does not collapse into "lower the blood sugar and you will save the brain." Metabolism is part of the story, perhaps an important part - but the simple version of Diabetes 3.0 keeps running aground on results like these. Hold the idea firmly enough to act on it, loosely enough to let evidence, not elegance, have the last word.



So what do you actually do on Monday morning?


Strip it all down, and the actions with the best evidence are unglamorous, free or cheap, and almost suspiciously sensible.


Treat your blood sugar like weather you can influence. The villain in this story is less a single sugary treat than chronic instability - the spikes and crashes of a diet built on ultra-processed food and liquid sugar. Walking for ten or fifteen minutes after a meal blunts the glucose surge more than almost any supplement. Eating the vegetables and protein before the bread, finishing meals earlier in the evening, and getting most of your carbohydrate from things that still look like plants are not exotic biohacks; they are the boring fundamentals that move the number.


Build and keep muscle. Skeletal muscle is the body's largest glucose sponge - the more of it you have and use, the more sugar gets pulled out of the blood and away from mischief. Resistance training is not vanity in midlife; it is metabolic insurance for the brain.


Defend your sleep, because poor sleep worsens insulin resistance and impairs the brain's overnight clean-up. Protect your hearing, address mood and isolation, and keep learning hard things - the Lancet Commission's 2024 update estimates that around forty-five percent of dementia cases worldwide are linked to fourteen risk factors we can, in principle, modify across a lifetime - diabetes among them, alongside newer additions like high LDL cholesterol and untreated vision loss (Livingston et al., 2024).


Measure, don't guess. This is where the contradictory supplement studies finally become useful. Rather than swallowing B vitamins or fish oil on faith, the smarter path is to test the things that actually predict who benefits - homocysteine, omega-3 status, fasting insulin and HbA1c, vitamin D - and treat what is genuinely off, in someone who can read the results in context. The VITACOG story is the whole argument: the right nutrient, in the wrong person, does nothing.


And hold on to the timeline. The disease takes its twenty-year run-up, which sounds like a threat but is really the gift - it means midlife is not too early and, as US POINTER showed in older adults, later is not too late (Baker et al., 2025). You are not trying to win in a week. You are trying to change the slope of a very long line.



The myths, briefly buried


"Sugar causes Alzheimer's." Too crude. Unstable glucose is one plausible contributor among several; the evidence is largely associational, and even directly drugging the pathway has so far disappointed (Craft et al., 2020; Cummings et al., 2026).


"It's all in my genes." No. APOE4 raises risk but does not guarantee the disease, most people with dementia are not double carriers, and lifestyle helped across genotypes in US POINTER (Baker et al., 2025).


"A supplement will fix it." It won't - the apparent wins usually turn out to work only in specific people with specific deficiencies (Jernerén et al., 2015).


"My routine bloods were normal, so my brain is fine." Maybe not. The clock can start ticking with glucose still inside the "normal" range - which is exactly why curiosity beats complacency (Crane et al., 2013).



A conversation worth having with Nicolle


If this has made you want to stop guessing and start measuring - your glucose, your insulin, your homocysteine and omega-3 status, and what they mean for the brain you intend to keep sharp for decades - that is exactly the work Nicolle at You Nutrition Clinic does. Nicolle is a registered Nutritionist (BANT), Nutritional Therapy Practitioner (CNHC) and registered Nurse (NMC), with over twenty years as a memory clinic nurse and Alzheimer's clinical trial nurse.


She still works in research, and, following Dr Dale Bredesen's mighty steps, she is ReCODE™ trained. She supports adults living with Alzheimer's and other dementias, those noticing memory changes, and also with anyone determined to lower their risk long before symptoms ever surface through the PreCODE™ program.


Book a consultation with Nicolle at You Nutrition Clinic - and turn this science into your own numbers. To enquire about working with her, email the clinic at admin@younutritionclinic.com. Your future self may be very glad you started the conversation today rather than twenty years from now.


🧩 Connect with us


For research updates, practical tips, and ongoing inspiration, follow us on Instagram:



Stay curious. Stay hopeful. Support your brain. 🧠


Disclaimer: This article is for informational and educational purposes only and does not constitute medical advice. Always consult with a qualified, registered medical doctor (MD) for diagnosis and treatment decisions.



References


Alzheimer Society of Canada. (n.d.). Bredesen Protocol offers false hope of reversing Alzheimer's disease. https://alzheimer.ca/en/whats-happening/news/bredesen-protocol-offers-false-hope-reversing-alzheimers-disease


Baker, L. D., Espeland, M. A., Whitmer, R. A., et al. (2025). Structured vs self-guided multidomain lifestyle interventions for global cognitive function: The US POINTER randomized clinical trial. JAMA. Advance online publication. https://doi.org/10.1001/jama.2025.12923


Barnes, L. L., Dhana, K., Liu, X., et al. (2023). Trial of the MIND diet for prevention of cognitive decline in older persons. New England Journal of Medicine, 389(7), 602–611. https://doi.org/10.1056/NEJMoa2302368


Bateman, R. J., Xiong, C., Benzinger, T. L. S., et al. (2012). Clinical and biomarker changes in dominantly inherited Alzheimer's disease. New England Journal of Medicine, 367(9), 795–804. https://doi.org/10.1056/NEJMoa1202753


Chatterjee, S., Peters, S. A. E., Woodward, M., et al. (2016). Type 2 diabetes as a risk factor for dementia in women compared with men: A pooled analysis of 2.3 million people. Diabetes Care, 39(2), 300–307. https://diabetesjournals.org/care/article/39/2/300/37175


Craft, S., Raman, R., Chow, T. W., et al. (2020). Safety, efficacy, and feasibility of intranasal insulin for the treatment of mild cognitive impairment and Alzheimer disease dementia: A randomized clinical trial. JAMA Neurology, 77(9), 1099–1109. https://jamanetwork.com/journals/jamaneurology/fullarticle/2766822


Crane, P. K., Walker, R., Hubbard, R. A., et al. (2013). Glucose levels and risk of dementia. New England Journal of Medicine, 369(6), 540–548. https://doi.org/10.1056/NEJMoa1215740


Cummings, J. L., et al. (2026). Efficacy and safety of oral semaglutide 14 mg (flexible dose) in early-stage symptomatic Alzheimer's disease (evoke and evoke+): Two phase 3, randomised, placebo-controlled trials. The Lancet. Advance online publication. https://doi.org/10.1016/S0140-6736(26)00459-9


de la Monte, S. M., & Wands, J. R. (2008). Alzheimer's disease is type 3 diabetes—Evidence reviewed. Journal of Diabetes Science and Technology, 2(6), 1101–1113. https://doi.org/10.1177/193229680800200619


Jernerén, F., Elshorbagy, A. K., Oulhaj, A., et al. (2015). Brain atrophy in cognitively impaired elderly: The importance of long-chain ω-3 fatty acids and B vitamin status in a randomized controlled trial. American Journal of Clinical Nutrition, 102(1), 215–221. https://www.sciencedirect.com/science/article/pii/S0002916523277655


Lampit, A., & Valenzuela, M. (2015). Pointing the FINGER at multimodal dementia prevention. The Lancet, 386(10004), 1625–1626. https://doi.org/10.1016/S0140-6736(15)00529-2


Livingston, G., Huntley, J., Liu, K. Y., et al. (2024). Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet, 404(10452), 572–628. https://doi.org/10.1016/S0140-6736(24)01296-0


Mason, A. C., Fatih, N., Sofat, R., Rentsch, C. T., Smeeth, L., Bhaskaran, K., Chaturvedi, N., & Garfield, V. (2025). Disentangling the relationship between glucose, insulin and brain health: A UK Biobank study. Diabetes, Obesity and Metabolism. Advance online publication. https://doi.org/10.1111/dom.70353


Max Delbrück Center. (2025, October 21). Why APOE4 raises Alzheimer's risk [Press release summarising a study published in Nature Metabolism]. https://www.mdc-berlin.de/news/press/why-apoe4-raises-alzheimers-risk


National Institute on Aging. (n.d.). Study reveals how APOE4 gene may increase risk for dementia. Alzheimers.gov. https://www.alzheimers.gov/news/study-reveals-how-apoe4-gene-may-increase-risk-dementia


Ngandu, T., Lehtisalo, J., Solomon, A., et al. (2015). A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): A randomised controlled trial. The Lancet, 385(9984), 2255–2263. https://doi.org/10.1016/S0140-6736(15)60461-5


Rao, R. V., Subramaniam, K. G., Gregory, J., et al. (2024). Sustained cognitive improvement in Alzheimer's disease patients following a precision medicine protocol: Case series.


Shaw, K. M. (2017). 'Type 3' diabetes: A brain insulin-resistant state linked to Alzheimer's disease. Practical Diabetes, 34(6), 187–188. https://doi.org/10.1002/pdi.2115


Smith, A. D., Smith, S. M., de Jager, C. A., et al. (2010). Homocysteine-lowering by B vitamins slows the rate of accelerated brain atrophy in mild cognitive impairment: A randomized controlled trial. PLoS ONE, 5(9), e12244. https://doi.org/10.1371/journal.pone.0012244


Sunwoo, Y., Park, J., Choi, C. Y., Shin, S., & Choi, Y. J. (2024). Risk of dementia and Alzheimer's disease associated with antidiabetic medications: A systematic review and



Toups, K., Tanio, C., Hathaway, A., et al. (2025). Precision medicine treatment of Alzheimer's disease: Successful randomized controlled trial [Preprint, not peer-reviewed]. Preprints.org. https://doi.org/10.20944/preprints202512.2694

 
 
 

Comments

bottom of page