Is dementia actually preventable?
Up to 45% of dementia cases may be preventable. That number comes from the 2024 Lancet Commission report (Livingston and colleagues), which pulled global evidence and landed on 14 modifiable risk factors spread across the life course. Erase all 14 across the population and you cut dementia incidence roughly in half.
That is a population ceiling, not a personal guarantee. But it reframes how you should read the rest of this article. These are levers and control knobs you can turn, not a report card to feel ashamed about.
One clarification on what dementia is. Dementia is a symptom, primarily a loss of episodic memory, sitting underneath a category of diseases. Alzheimer's is the classic example, often arriving with dysregulation of fear or anger alongside the memory loss. Frontotemporal dementia, the type Bruce Willis is dealing with, looks more vascular and hits language areas. Parkinsonian dementias are another cluster. Different diseases, partly convergent mechanisms, one shared headline symptom.
What does early memory trouble actually mean?
If you are experiencing word-finding problems, delayed recall, or tip-of-the-tongue moments, that is usually tiredness before it is aging. And even when it is aging, it is rarely pathological aging. Think of it like losing some bone mass or gaining some body fat with age. Real, worth working against, not a disease process.
Here is what I see in the QEEG when someone has slowed word retrieval. Their alpha waves are losing synchrony, usually in the left hemisphere. Alpha is an idling frequency. My working theory: when the alpha frequency drifts and stays out of sync across the brain, you are not lifting those tissues out of neutral cleanly enough to use them. Retrieving information and monitoring it as it gets handed around both become harder, which shows up as word retrieval lag, short-term memory hiccups, and attention slips.
That pattern is a regulatory feature, not a degenerative disease. You can read more about alpha waves and what they reveal and about restoring mental clarity when brain fog sets in.
What can a brain map show about aging risk?
I do QEEG brain mapping and neurofeedback. The mapping is simple to do. You put a cap on, sit still about 10 minutes eyes closed, then eyes open, and record the resting features your brain always carries. Compared to population averages, you have characteristics that stand out, because people are weird. The point is to see how your brain regulates attention, stress, sleep, and short-term memory, and how that lines up with what you actually experience.
I once mapped an older gentleman with severe anxiety, shakiness, and insomnia after 25-plus years of heavy drinking. His map showed high beta power, high beta coherence (over-connectivity in that fast activation band), and low delta (the rest-and-repair, deep-sleep frequency).
The mechanism behind this is worth understanding. Alcohol drives a GABAergic state, which is why it makes you sleepy and relaxed. Drink chronically and the brain ramps up glutamate, the activation tone, to balance all that exogenous GABA. After withdrawal, the brain cannot downregulate as fast as it learned to upregulate. That mismatch produces the shakiness, the seizures, the cardiovascular and hypertensive risk during rapid withdrawal. Over-arousal reads directly off the map as excess beta.
We trained him for a couple of months. The follow-up map showed the beta dropping away and the coherence normalizing. The cravings fell, the anxiety fell, sleep improved. One day he called the front desk, asked when I would be in, came to the office, and put himself to sleep on the waiting-room couch to prove he had regulatory control over his sleep. That is what shifting a resource looks like.
What are the 14 modifiable risk factors?
The Lancet Commission grouped them across three life stages. The number beside each is the share of risk that factor accounts for in their model.
Early life: less education (5%). This is the one you cannot redo.
Midlife: hearing loss (7%), high LDL cholesterol (7%), depression, traumatic brain injury, physical inactivity, diabetes, high blood pressure, obesity, and excessive alcohol.
Late life: social isolation (5%), air pollution, and untreated vision loss.
Two factors are newer additions in the 2024 report. High LDL cholesterol in midlife, the oxidative-stress pattern, predicted more trouble later. And untreated vision loss joined hearing loss on the sensory side.
Notice the weighting. Midlife carries the most factors at a point when most people assume their dementia risk is decades away. Most dementia flavors take about 30 years to build. The genes load the gun, environment and experience pull the trigger, but the trigger pull is usually slow. Some genetic variants, like the presenilin Alzheimer's genes or certain Parkinsonian predispositions, move fast and show up in your 50s. If that runs in your family, you tend to know it.
How should I think about these factors? Four clusters
Reading the list as four mechanism clusters tells you what to prioritize based on your own history and family pattern.
Cluster 1: the vascular and metabolic group
High LDL, high blood pressure, diabetes, obesity, smoking, physical inactivity, alcohol. This is the biggest chunk of the list and mostly the things your cardiologist already told you. The mechanisms are concrete. Diabetes is an oxidative-stress phenomenon: brain tissue oxidizes faster, driving a diabetic-flavored Alzheimer's process. Blood pressure, smoking, and obesity drive vascular damage. These factors protect the tiny blood vessels that fuel your brain tissue. Work on your stress response too, since chronic activation feeds this cluster. I cover the autonomic side in biohacking fight or flight and the metabolic side in strategic fasting.
Cluster 2: the sensory input group
Hearing and vision. When these tissues fail and you leave them untreated, you lose input the brain needs to stay engaged. The ACHIEVE trial of hearing aids in older adults showed that, in the prescribed higher-risk subgroup, hearing treatment cut progression risk roughly in half, even though the overall trial result did not reach significance. Vision evidence is softer and more observational so far, but a large cohort study showed about a 30% reduction in dementia odds after cataract surgery. Get the glasses, get the cataract surgery, get the hearing aids.
Cluster 3: cognitive and social load
Lifelong education, social connection in later life, and staying out of depression. You cannot redo your schooling, but you can rebuild your social environment and connections. This is where music engagement and active learning matter, since they build the connectivity that acts as protection.
Cluster 4: the hardware
Protect the physical brain. Wear a helmet on the motorcycle, the scooter, the skateboard. Head injury raises dementia risk across the lifespan. Repetitive wear and tear is a real factor, and unlike a concussion you had 20 years ago, the future ones are preventable.
Does lifestyle intervention actually change cognition?
Two trials answer this directly.
The FINGER study (Ngandu and colleagues, Lancet, 2015) was a two-year trial in Finland combining diet, exercise, cognitive training, and vascular risk management. The intervention group's cognition held up substantially better than the control group's. Proof of concept, small-ish, but it worked once.
The American answer came in 2025: the POINTER study, presented at the Alzheimer's Association International Conference and published in JAMA (Baker and colleagues). This was 2,000-plus older adults aged 60 to 79, all at elevated risk, across five US sites, over two years.
The clever design compared two doses of lifestyle rather than lifestyle against nothing. One group got a structured, high-support program: prescribed exercise, prescribed Mediterranean-style diet, scheduled cognitive and social challenges, regular cardiometabolic monitoring. The other group got the same menu and goals as self-guided instructions, no structure, no hand-holding.
Two findings, both important. Both groups improved globally relative to typical aging, including on episodic memory, the primary complaint in dementia. And the dose mattered: more support and more structure produced a bigger effect, strongest in global cognition and executive function. Structure beats willpower.
One caveat. POINTER measured cognition, not dementia risk directly. That is normal aging variance rather than disease trajectory. The Lancet risk-factor work measures the disease side. Your behavior matters, and these two literatures point the same direction.
What does the brain map of a future decliner look like?
An NYU memory-center study (Prichep and colleagues, Neurobiology of Aging) followed 44 people who walked in complaining of memory trouble, mapped them with QEEG, and tracked them seven to nine years. At the end they split the group into progressors (those who converted to Alzheimer's-type dementia) and non-progressors.
The non-progressors had relatively typical brain waves. Slightly elevated theta, slightly low beta, nothing dramatic. The decliners showed large amounts of theta globally, in the temporal lobes and at the back of the head, with globally low beta power.
That high-theta, low-beta signature looks ADHD-like. It does not produce dementia on its own. The variant driven by lack of deep sleep and chronic fatigue can accelerate aging phenomena, however, and I see the same signature across postviral fog, sleep apnea, mild concussion without focal injury, mold exposure, Lyme, chemotherapy, and chronic fatigue with Epstein-Barr. Post-COVID fatigue tends to add bilateral temporal-lobe delta, which I read as a metabolic signature where the brain is not bringing its reserve online. These are big hub tissues that show up dysregulated across many complaints, which is exactly why protecting deep sleep matters. Start with optimizing your sleep, and if the high-theta pattern is yours, SMR neurofeedback trains sleep, focus, and self-control.
What about ADHD and stimulant medication?
Adult ADHD is associated with about a 2.77-fold increase in dementia risk (Levine and colleagues, 2023). Adults with ADHD who take psychostimulant medication show no such elevated risk in the cohort data. The likely mechanism is that stimulants improve perfusion and cortical arousal, increasing blood flow to underactive regions and normalizing the brainwave pattern. Confidence here is medium, but it fits the larger picture: the ADHD-dementia link involves modifiable physiology.
What can I actually do tonight?
A few concrete moves.
For alcohol use disorder, the neurofeedback literature is specific. The Peniston-Kuhlman alpha-theta protocol showed dramatic reductions in relapse. Follow-up work (Sturman-era studies) found that adding SMR training after a course of alpha-theta produced a stronger effect. In clinic I run a course of non-alpha-theta neurofeedback first to stabilize executive function, sleep, and arousal, then move into a rigorous alpha-theta course layered with SMR. The gentleman above did about 30 sessions total.
For the cholinergic side, I like CDP-choline (citicoline). Most forms of choline have poor brain bioavailability; CDP-choline and alpha-GPC get in, and I prefer CDP-choline. It speeds and synchronizes alpha, so word-finding and delayed recall improve within a few days, and it supports remyelination for longer-term myelin repair rather than loss. It is cheap, off-label, well tolerated. Skip the afternoon dose since it can be stimulating, and unless you are deeply symptomatic, take it every other day or five days a week, since cholinergic tone can build into irritability. Talk to your doctor; I am your scientist, not your physician.
For the autonomic and arousal side, HRV biofeedback helps balance sympathetic and parasympathetic tone, and neurofeedback retrains the underlying pattern over a few months. Measuring the brain directly is where it starts. A QEEG shows the electrical fingerprint behind your slow processing or anxiety and gives you specific things to work on.
You do not need to pull all 14 levers. Pick one that seems genuinely easy to change, a small modifiable behavior you could actually keep. Then look honestly at the one that seems impossible, and set it aside for now. Start with the easy one this week. The trajectory bends from there.