Is this real, or are they selling hope?
- Ask who is selling it. That single question sorts most of what you will be shown. The 10-second bias check.
- "Peer-reviewed" and "published" are not the same thing. And a study with a comparison group tells you more than one without, and both tell you more than an ad.
- What lowers risk is mostly free. Blood pressure, hearing, sleep, movement, staying connected.
- ReCODE and other root-cause programmes: the plain guide, with pros and cons.
- Near Phoenix? The ReCODE and functional-medicine doctors we checked, with their licences, Medicare and costs.
Somewhere around 2am, everyone caring for someone with dementia types the same thing into a search bar: is there anything that helps? What comes back is a wall of pills, drinks, helmets, and "protocols."
No supplement, drink, or device has yet been proven in large trials to prevent, treat, or cure dementia. (A few prescription drugs can modestly slow early Alzheimer's for some people, worth asking the doctor about, but they aren't cures, aren't for every type, and aren't anything sold online.) The things that help most are further down this page, and most are free.
You don't need to become an expert on every fad. You need one sorting rule that still works on the next product you've never heard of. Here it is, in three buckets.
Steer away
A real risk here, or a regulator has already stepped in.
- "Memory" supplements like Prevagen. A federal court ordered its maker (December 2024) to stop claiming it improves memory or is "clinically proven"; the company is appealing, so treat the memory claim as unproven, and note the company's own study found no benefit. If the best-selling memory pill can't back the promise, be gentle with yourself for having hoped it would.
- Methylene blue (the blue-dye trend). Not proven for dementia, and dangerous: taken with common antidepressants (SSRIs and others) it can trigger serotonin syndrome, a life-threatening reaction. Not worth the risk.
- Any restrictive diet forced on someone who's losing weight or refusing food. In dementia, keeping weight on and meals pleasant beats any "brain diet." Don't take food they'll eat away from them for a theory.
Being studied, not proven
Interesting, maybe harmless, but don't pay big or pin your hopes on it.
These show up in hopeful headlines. The honest status is the same for all of them: early, short, or mixed studies; no proof it prevents or treats dementia. If it's cheap and safe and you want to try it, that's your call. Just don't let it replace the doctor or the green list below, and don't spend money you need elsewhere.
- Lion's mane · red-light / "photobiomodulation" helmets · hyperbaric oxygen (HBOT) · hydrogen water
- Coconut / MCT oil · curcumin (turmeric) · ginkgo biloba · omega-3 / fish oil · high-dose B vitamins
A few of these interact with medicines (see the caution box). "Being studied" is not "shown to work." It's fine to wait for the answer instead of paying for it now.
The Bredesen protocol (ReCODE) does not belong in this bucket. It is a doctor-led programme with its own randomized trial, so it has its own section below.
Worth your energy
Cheap or free, low-harm, and backed by evidence.
None of these are a cure, and none are a guarantee. But they're the steps experts agree lower risk (for vascular dementia especially, they can buy longer steady stretches), and several also make daily life better right now. Across a whole population, researchers estimate that addressing about 14 everyday factors could delay or prevent up to ~45% of dementia. That is a population estimate, not a personal promise. But it means these are the real levers:
- Treat hearing loss. The single best-evidenced one. In a large trial, hearing aids didn't change much for the average participant, but among older adults already at higher risk, those given hearing aids showed about half as much cognitive decline. That's a population finding about lowering risk, not a promise for any one person, and hearing aids earn their keep anyway by bringing them back into the conversation.
- Treat vision loss. An updated pair of glasses or a cataract fixed keeps them oriented and engaged.
- A daily walk / physical movement. The most consistently protective habit there is, and it burns off restlessness too.
- Blood pressure and cholesterol under control. Especially for vascular dementia, this is brain care as much as heart care. It's the closest thing to "prevention" that exists.
- Stay socially connected. Visits, a day program, a phone call rhythm. Isolation is a risk factor; company is a protection.
- Don't smoke, keep alcohol light, and treat diabetes and depression.
- A Mediterranean-style plate. Vegetables, fish, olive oil, whole grains. Reasonable and heart-healthy; think "worth doing," not "miracle."
The unglamorous truth: the real "recovery story" is blood pressure controlled, sleep fixed, hearing working, the next stroke prevented. Families who do that quietly get some of the longest, steadiest stretches of anyone.
The 10-second bias check
When you can't tell, ask one question: who profits if I believe this? That single test sorts most of the internet. Then scan for the tells. A pitch usually wears several:
- "Clinically proven" with no named study you could look up.
- A countdown timer, "only 3 left," or an auto-renewing subscription.
- "Doctors don't want you to know": real science doesn't need a conspiracy.
- Testimonials instead of trials: one moving story where a study should be.
- The words cure or guaranteed, with no named study behind them.
- It's sold by the same person telling you it works.
Wanting to fix this is love looking for somewhere to go. Give it somewhere good. Two true things, held together: "I know how much you want to help her, and I love you for it. And what helps most costs little: her hearing checked, a walk together, her blood pressure handled." Honor the feeling, share one honest fact, and point the hope at the green list and the doctor, not the checkout page.
Bring the full bottle list to a pharmacist and ask by name. The ones that bite. methylene blue + antidepressants (serotonin syndrome); ginkgo, high-dose fish oil, or vitamin E + blood thinners (bleeding); high-dose vitamin B6 (nerve damage over time); and any supplement started without telling the doctor who manages their prescriptions. And never let a supplement or diet replace a medicine that's working.
ReCODE (the Bredesen protocol)
A doctor tests for the many things that can drive dementia and treats each one, alongside food, exercise, sleep and supplements. Many people in its studies improved. In its 2025 randomized trial, the overall score on a computer thinking test rose on ReCODE and fell on usual care. It costs several thousand dollars a year, and Medicare rarely pays.
The full ReCODE guide, with videos.
Near Phoenix? The doctors we checked.
- We lean on big randomized trials. Personal, many-part care rarely fits that test, so the habit can make us undervalue clinical reasoning and what is helping real people now.
- We lean on mainstream US sources (the Alzheimer's Association, NIA, FDA). They have their own pulls: drug-trial funding, legal caution, and a system that pays for prescriptions more than for lifestyle care.
- This page was researched with an AI assistant. Its training leans the same way, toward establishment sources and cautious wording. An earlier version of this page said reversal was not shown, which went further than the evidence.
The bottom line
There is real hope here, and real cost. Some people do better when the causes behind their symptoms are found and treated. The everyday moves help almost everyone: a daily walk, sleep, blood pressure handled, hearing aids, people around the table. Be wary of anyone selling a guaranteed cure, and just as wary of anyone who says nothing can be done. Then protect the money too: paperwork and scams.
Two trustworthy, ad-free places: the FDA's list of unproven Alzheimer's products, and the Alzheimer's Association's plain-language take on alternative treatments. Or just call the 24/7 helpline (1-800-272-3900) and ask. They field this question all day and sell nothing.