Category: The Evidence
Most supplement content is advertising in an informational costume. This is how to tell the difference: the tiers of evidence, the tells of a conflicted source, and the sources worth trusting. It also happens to be our own editorial standard.
Category: Research Reviews | Reading time: ~15 min | Level: Beginner to Intermediate
Try to research a supplement honestly and you run into a strange problem: almost everything you find is trying to sell you something. Type a herb into a search bar and the top results are shops, shops with blogs, and blogs that are really shops. The article confidently explaining why the herb works is, more often than not, hosted by the company hoping you will add it to your basket by the final paragraph. The information and the advert have merged so thoroughly that most people cannot tell where one ends and the other begins.
This is a guide to telling them apart. Not a list of good and bad supplements, because that list changes and you should be able to build it yourself, but the underlying skill: how to read a claim, weigh the evidence behind it, and notice who profits from you believing it. Learn this once and you never again have to take a wellness site's word for anything. You can go and check.
We will be upfront about something. This article doubles as our own editorial standard. Everything it tells you to demand of a source, grade evidence honestly, disclose conflicts, admit uncertainty, is what we try to hold ourselves to across this library. So read it two ways: as a tool for judging anyone's supplement claims, and as the standard you are entitled to hold us to as well.
Advertising has learned to look like education. The most effective supplement marketing does not shout; it explains. It walks you through a plausible mechanism, cites a study or two, uses careful-sounding language, and only reveals its purpose with a discreet buy button. This works because it borrows the form of honest information while keeping the function of a sales pitch.
The single most useful move you can make is to separate two questions that marketing deliberately blurs. First: what does the evidence actually say? Second: who benefits if I believe this? A trustworthy source lets those two questions have different answers, and will happily tell you something inconvenient to its own interests. A sales page, by construction, cannot. Its answer to the first question is always whatever serves the second. Once you start reading with both questions in mind, a lot of confident content quietly reveals what it really is.
Not all evidence is equal, and the gap between the strongest and the weakest is enormous. Marketing thrives by quoting the weakest as if it were the strongest, so knowing the order is half the battle.
From strongest to weakest, roughly:
Here is the tell to watch for. A great deal of supplement hype takes a finding from the bottom of this list, an in-vitro result, an animal study, a single tiny trial, and presents it with the confidence appropriate to the top. When a page says a herb kills cancer cells, check whether that happened in a person or in a dish. Usually it is the dish, and the distinction is the whole story.
Evidence quality tells you how much to trust a finding. Incentive tells you how much to trust the person reporting it.
Ask three questions of any source. Who funded the study being cited? A trial paid for by the company selling the product is not automatically wrong, but it belongs under heavier scrutiny, especially if the favourable result only ever appears in manufacturer-funded work and independent studies do not replicate it. Who funded the article? A summary hosted by a shop that stocks the ingredient has an interest baked in. And does the writer profit from the conclusion, through sales, commission or affiliate links? A conclusion that reliably points at a purchase is a conclusion to distrust.
None of this means industry involvement makes something false. That overcorrection is its own mistake, and it would discard a lot of genuinely good research, since industry funds much of what gets studied. The skill is weighting, not blanket rejection: scrutinise conflicted sources harder, check whether independent evidence agrees, and be most sceptical of claims that only ever appear where someone stands to gain. A source that discloses its conflicts openly has earned more trust than one that hides them, even before you read a word of the science.
A short list of places whose incentive is not to sell you a bottle, and which are worth learning to use.
The US National Center for Complementary and Integrative Health, part of the National Institutes of Health, publishes plain-language, evidence-based summaries of herbs and supplements, and is refreshingly willing to say when the evidence is weak or absent [1]. It is a sensible first stop for a balanced overview.
Cochrane is an independent global network that produces systematic reviews to a high methodological standard, deliberately structured to reduce bias [2]. When a Cochrane review exists on your question, it is among the most reliable summaries you can find, and its habit of concluding that evidence is insufficient is a feature.
PubMed, run by the US National Library of Medicine, is the searchable index of the primary biomedical literature [3]. You can read study abstracts directly and see for yourself the sample size, duration and design that a marketing page glossed over. You do not need to be a scientist to check whether a cited study was in humans and how many took part.
Regulators publish too. The European Medicines Agency's Committee on Herbal Medicinal Products issues monographs that state, for each herb, whether its use is well established or merely traditional, which is often a bracingly honest counterweight to marketing [4].
None of these is perfect or complete, and all of them are drier than a sales page. That dryness is the point. They are describing reality, which hedges, rather than selling a product, which does not.
There is a counterintuitive lesson underneath all of this, and it is the one we try hardest to live by. A source that grades evidence down, that says this is weak, this is only tested in animals, nobody really knows yet, is more trustworthy than one that is always positive, not less.
The reason is simple. Reality is uneven. Some supplements have strong evidence, many have thin evidence, and plenty have essentially none. A source that reflects that unevenness, praising the well-supported and dismissing the weak, is tracking the truth. A source that is enthusiastic about everything is tracking its inventory. So when you see a site cheerfully recommend a dozen herbs with equal confidence, that uniformity is itself the warning sign. Genuine evidence does not distribute itself so conveniently.
This is why we tell you when the evidence for something is limited, name the limitations of studies we cite, and, as in our creatine article, plainly recommend things we do not even sell. It costs us the occasional sale. It is meant to. That cost is the proof that the information is not an advert.
You do not need a science degree to protect yourself. A short routine handles most of it.
When you read a supplement claim, find the actual evidence behind it, not the confident sentence but the study. Check whether it was in humans, how many people, for how long, and who paid. Ask who benefits from you believing the claim, and raise your scepticism if the answer is the same party making it. Cross-check the claim against one of the incentive-free sources above. And treat certainty as a warning: the more confident and downside-free a claim sounds, the more it deserves suspicion, because honest evidence almost always arrives with caveats.
Above all, prize sources that are willing to disappoint you. A page that admits a popular herb does not really work, or that the science is not in yet, is doing something a sales pitch structurally cannot. That willingness is the single best signal that you are reading information rather than advertising.
We built this library to be the kind of source this article tells you to look for, and we would rather be judged by that standard than exempt from it. Our range is deliberately small, a set of remedies we are willing to defend, and the writing exists to inform rather than to move every ingredient we could stock. When the evidence for something is weak, we say so. When a study has limitations, we name them. When something we do not sell works, we tell you it works.
If you want to put this into practice, our label forensics guide shows you how to check whether a product is what it claims, and our piece on talking to your GP about supplements covers the conversation that no amount of online reading replaces. You can also ask Remy to reason through a claim with you, or browse the Remedy Library to see the evidence grade we attach to each ingredient, including the ones we grade down.
There is no product to buy at the end of this article, which is rather the point.
The FAQ entries above cover the practical questions: how to judge a study, what the evidence tiers are, how to spot a conflict of interest, and which sources are worth trusting. The short version: separate what the evidence says from who profits, weight claims by evidence tier, lean on incentive-free sources, and treat confident, downside-free positivity as the tell that you are reading an advert.
1. National Center for Complementary and Integrative Health (2024). Evidence-based plain-language summaries of complementary and integrative approaches. nccih.nih.gov. Government health information body. 2. Cochrane (2024). Systematic reviews of healthcare interventions. Cochrane Library, cochranelibrary.com. Independent research network. 3. US National Library of Medicine (2024). PubMed and MEDLINE, the searchable index of biomedical literature. pubmed.ncbi.nlm.nih.gov. Primary literature database. 4. European Medicines Agency, Committee on Herbal Medicinal Products (2024). European Union herbal monographs. ema.europa.eu. Regulator.
Ask a few quick questions. Was it a randomised controlled trial in humans, or a test-tube or animal study being described as if it applies to people? How many participants, and for how long? Who funded it, and did they have a stake in the result? Was it published in a peer-reviewed journal or just posted by the seller? A reliable claim survives all of these. Most marketing claims fall at the first.
Roughly, from strongest to weakest: systematic reviews and meta-analyses of randomised trials, then individual randomised controlled trials, then observational studies in people, then animal studies, then test-tube (in vitro) studies, then expert opinion and mechanism, then anecdote and testimonial. Higher tiers are more likely to reflect a real effect in real people. A lot of supplement hype quotes a test-tube finding as if it were a proven human benefit.
Look at who benefits from the conclusion. If the article is on a site that sells the thing it praises, if the study was funded by the manufacturer, or if the writer earns a commission on the product, the conclusion is not neutral. Conflicts do not automatically make something false, but they should raise your scrutiny, and a source that never discloses them is a source to distrust.
Sources whose incentive is not to sell you a supplement. The US National Center for Complementary and Integrative Health writes plain-language, evidence-based summaries. Cochrane produces rigorous systematic reviews. PubMed lets you read the primary studies and their abstracts directly. Regulators such as the European Medicines Agency publish herbal monographs. None of these is trying to sell you a bottle.
Because reality is less exciting than marketing. Honest sources hedge, grade down, and say when the evidence is weak or absent, which reads as boring next to a confident sales page. That restraint is the feature, not the flaw. A source that is always certain and always positive is telling you about its incentives, not about the science.
A rough one: distrust confidence and certainty, especially when the confident party is selling something. If a page tells you a herb definitely does something dramatic, has no downsides, and conveniently links to buy it, treat all three as warnings. Genuine evidence almost always comes with caveats. The presence of honest hedging is itself a decent shortcut for trustworthiness.
No, and dismissing them all is its own error. Plenty of good research is industry-funded because that is who pays for it. The point is not to discard those studies but to weight them, look at the methods, check whether independent work agrees, and be more sceptical of a result that only ever appears in studies paid for by the seller. Funding is a reason to scrutinise, not automatically to reject.
By being a source people come back to. Our range is a small set of remedies we can stand behind, and the library exists to build trust rather than to push every ingredient. If we inflated weak evidence to sell more, we would win once and lose the reader. Grading honestly, including about things we do not sell, is a slower business model and a more durable one.
Then that is the answer, and a good source will say so. Large parts of the supplement field are genuinely under-studied, and the correct response is not to invent a confident verdict but to say the evidence is thin and let you decide with your eyes open. A source willing to say nobody knows yet is usually more trustworthy than one that always has an answer.