How we research, grade and review

Every guide in the library follows the same process, so you can judge what you are reading. Here is how we choose sources, how we grade evidence, what our review labels mean and how we correct mistakes.

Where our information comes from

Each guide starts from a question readers ask, such as whether ashwagandha helps with stress or whether magnesium helps with sleep. We then look for the strongest sources we can find, in roughly this order:

Systematic reviews and large, well-run trials in people, found through PubMed and the Cochrane Library.
Government health agencies, such as the National Institutes of Health and its National Center for Complementary and Integrative Health and Office of Dietary Supplements, the Centers for Disease Control and Prevention, and the Food and Drug Administration.
European and international regulators and expert bodies, such as the European Medicines Agency and the World Health Organization.
Smaller trials, laboratory and animal studies, and traditional-use records, which we label for what they are.

Sources are numbered where they are used and listed at the end of each guide, so any statement can be traced. When a trial was paid for by the maker of the product it tested, we try to say so.

How we grade the evidence

We rate each use of a remedy separately, because a plant can be well supported for one purpose and untested for another. Every rating answers one question: what do good studies in people show for this specific use?

Strong evidence. Consistent results from good human trials or reviews.
Promising but early. Positive results from small, early or mixed trials. Worth knowing about and worth raising with your clinician.
Not enough evidence. Too few good human studies to say yet. Not shown is different from shown not to work.
Evidence says no, or unsafe. Tested properly and failed, or linked to documented harm. These are two different findings, and the guide’s scorecard and text say which one applies.

Safety information sits in each guide’s text: who should avoid a remedy, which medicines it can interact with, and when to see a clinician. A gentle reputation is not the same as proof of safety, so we treat food amounts, teas, oils, extracts and concentrated supplements as different things.

Respect for tradition

We describe how healing traditions such as Ayurveda, Traditional Chinese Medicine and Western herbalism have used a plant or practice, because that history matters to many readers and often guided the research that followed. We keep that history separate from the evidence about a particular health outcome, so you can value both without confusing one for the other.

What the labels on each guide mean

Last updated is the month the guide was written or last substantially revised.
Links checked means the source links were tested on that date (see the October 2026 check below).
Clinical review. Guides that discuss medicines, doses, pregnancy, children or health conditions say whether a licensed clinician has reviewed them. Many have not yet been. Read those guides as preparation for a conversation with your own clinician, not as a substitute for one.
Older article. A few articles were published before the library was rebuilt in 2026. They carry a notice, sit in a separate group on topic pages and are kept out of search engines until they are re-checked.

The October 2026 source check

In October 2026 we checked every source link in the library, nearly 14,000 links in all. Links to PubMed records and DOIs were matched against the title of the study they claimed to support. Where a link pointed to a missing page, we replaced it with a verified link to the same document or study, removed the link but kept the reference (for pages from organizations, news reports and summaries), or removed the reference when it named a specific study we could not locate. Where a journal’s website blocked automated checks, we replaced the link with a PubMed or DOI link once we had confirmed it pointed to the same study. A small number of links on sites that block checkers could not be confirmed and were left in place.

The same check found that several references in our older articles, published before the 2026 rebuild, led to unrelated or nonexistent studies, including three in the hawthorn and chamomile articles. We removed those references, along with the two hawthorn claims that relied on them. The older chamomile article now points readers to our current, fully sourced chamomile guides, and the remaining older articles are labeled until they are re-checked.

Corrections

We want to know when we get something wrong. Use the “Report a correction” link at the end of any guide, the contact form, or email cs@healthresearchusa.net. We check every report against the sources. Significant corrections are noted on the guide with the date of the change.

Independence

The site does not currently carry advertising or affiliate links, and ratings come only from the published evidence listed in each guide. See our funding and disclosures page for details.