Anyone can write “studies show.” This page explains what actually happens between picking a topic and publishing an article — so you can judge for yourself how much weight to give what you read here.
Where we look first
Not all sources are equal, and we work down a rough hierarchy.
At the top are the primary and government-backed sources: the NIH and its Office of Dietary Supplements, the FDA, the CDC, and PubMed for the underlying research itself. These are where we go before anything else.
Next come established medical institutions and professional bodies — the kind of organizations that publish clinical guidance and have a reputation to protect.
Peer-reviewed studies are read directly whenever they’re central to a claim. Reading someone else’s summary of a study is how errors get copied around the internet, and it’s how a cautious “may be associated with” turns into a confident “proven to.”
What we don’t treat as a source: supplement manufacturer websites, sales pages, affiliate marketing material, and other blogs.
Reading a study properly
Finding a study that supports a claim is easy. Working out whether it means anything is the actual job.
A few things we look at every time:
Who was studied. Research done in mice, or in cells in a dish, tells you very little about what a supplement does in a human body. It’s a starting point for further research, not a basis for a recommendation.
How many people. Twelve participants and twelve hundred participants support very different levels of confidence.
What dose. This one catches people out constantly. A study showing a benefit at 1,200 mg says nothing about a product containing 50 mg — but plenty of marketing cites exactly that kind of study.
How it was designed. Randomized controlled trials sit well above observational studies, which in turn sit above surveys and anecdote. Studies funded by the company selling the product get read with extra care.
Whether anyone replicated it. A single promising result is a lead. Several independent studies pointing the same direction is evidence.
How we describe the strength of evidence
Vague language hides weak evidence, so we try to be specific about how solid something actually is:
- Well supported — multiple good-quality human trials point the same way, and mainstream medical bodies generally accept it.
- Some evidence — human research exists and looks promising, but it’s limited, mixed, or the studies are small.
- Early or preliminary — mostly animal studies, lab work, or a single small human trial. Interesting, not actionable.
- Not supported — the claim is widely marketed but the research doesn’t back it, or the evidence points the other way.
When we can’t find credible evidence for a claim, we say so. “There isn’t good research on this” is a legitimate finding and a useful one.
Supplement reviews specifically
Supplement products are where marketing and evidence diverge most sharply, so these get a particular process.
We start with the ingredient list, not the sales page, and research each active ingredient independently — what it’s been studied for, at what dose, and with what results.
Then we compare that research to the dose in the actual product. A formula can contain a genuinely well-researched ingredient at a fraction of the studied amount, which makes the citation technically true and practically meaningless.
We look at what the company claims versus what the ingredients could plausibly do. Claims about curing, treating, or reversing conditions are red flags, and legally supplements aren’t permitted to make them.
We also note safety information: known side effects, drug interactions, and groups who should avoid the ingredient — pregnant women, people on blood thinners, anyone with liver conditions.
Where a product is heavily marketed but thinly evidenced, the review says that. That happens more often than not, and it’s the reason these reviews exist.
What we can’t do
We don’t run lab tests on products, and we don’t independently verify that what’s on a label is what’s in the bottle. Supplements aren’t approved by the FDA before sale the way medicines are, and third-party testing varies enormously between brands.
We also can’t tell you whether something is right for you. Everything here is general information drawn from published research. Your medications, conditions, and history change the picture in ways no article can account for — which is why the answer to “should I take this” always involves your doctor or pharmacist.
Tell us when we’re wrong
If you find a claim on this site that the research doesn’t support, or a study we’ve read incorrectly, get in touch. Corrections get made and dated.