About & how we grade

This site gathers the research on Alzheimer’s in one place, mainstream and alternative, and grades each study the same way: does it work, how strong is the evidence, and who paid for it.

Why this site exists

My mom has Alzheimer’s. Watching it is hard. Then I kept seeing posts and videos on social media promising that a spice blend, a supplement or a pinch of lithium could stop it, each one quoting “a study”. When I went looking, I often couldn’t find the study, or it turned out to be in mice, or tiny, or paid for by the people selling the product. So I built the site I wanted when she was diagnosed: the real studies, the plain results, and who paid for them.

Two ground rules follow from that. A study paid for by the company that sells the product gets flagged, whether the company makes a $26,000-a-year drug or a $44 bottle of pills. And a result in mice is reported as a result in mice.

The five verdicts

The verdict answers one question: based on this study, does the thing help people?

Verdicts
Verdict	What it means
Solid	Good trials in people show a real effect, or a finding is well established.
Promising	Some human evidence points the right way, but it is not settled. It needs bigger or longer trials.
Too early	The evidence comes from mice, cells, case reports, or studies with no comparison group. Interesting, not yet useful.
Didn’t work	It was tested properly in people and no benefit was found.
Harmful	People did worse, or the thing is linked to real harm.

The strength score, 1 to 5

The strength score says how much weight a study can bear. A verdict of “Promising” with a strength of 4 is much more convincing than “Promising” with a strength of 2.

Evidence strength
Score	What it usually means
	Large randomized trials, repeated by different teams, with results that agree.
	At least one large, well-run randomized trial, or a very large natural experiment or pooled review.
	Smaller randomized trials, or large, careful observational studies that follow people over years.
	Small pilot trials, short trials, or observational studies with big weaknesses.
	Animal or lab studies, case reports, or groups of patients with nobody to compare them against.

Some terms, in plain words. A randomized trial assigns people by chance to get the treatment or not, which is the fairest test. An observational study watches what people already do. It can show a link, but not prove that one thing causes another.

Who paid

Money does not make a study wrong. It does tell you which questions to ask. Every study card has one of these four labels, plus a line naming the funder and any conflicts of interest.

Who-paid labels
Label	What it means
Public / nonprofit	Paid for by a government agency, such as the National Institutes of Health, or a university.
Charity	Paid for by a disease foundation or philanthropy.
Industry	Paid for by a drug or device maker, testing its own product.
Seller conflict	Paid for by the maker of a supplement or program, or an author profits from the result, for example by owning the company, holding a patent or selling the program.

We apply the same skepticism to drug companies, supplement makers, device companies and clinic owners. For more on this, see follow the money.

How the sources were checked

The research for this draft was done in September 2026. For each claim, we tried to open the original source: the journal paper, PubMed, or an official page from the FDA, NIH, CMS, WHO or the Alzheimer’s Association.
Many studies on the site were carried over from an earlier version that was checked against journal pages, PubMed and NIH releases.
Some sites blocked automated reading during the research. When we could not open a primary source, we either left the claim out or said plainly on the page that it is “not yet confirmed”.
Some funding lines say “Not yet confirmed” or “Not reported”. Those are pending a check of the paper itself.
Every study card links to the paper or its PubMed record. We do not link to shops.
Percentages come with absolute numbers wherever the source gives them, because “50% more risk” can mean 2 in 100 becoming 3 in 100.

If you spot a mistake, or find the source for something marked as not yet confirmed, please tell us. We would rather be corrected than be wrong.

Medical disclaimer

The site is written for families in the United States, so prices, insurance rules and approvals are US ones unless we say otherwise. Research changes quickly. Each page shows when it was last reviewed.

This is a draft, last reviewed 26 September 2026.