Not all evidence is equal. A randomised trial is not the same as a laboratory study. A clinical outcome is not the same as a blood marker. Independent replication is not the same as manufacturer-funded research. Here is how to tell what counts — and what does not.
The strongest evidence is a randomised, double-blind, placebo-controlled trial, in a comparable population, at the dose and in the form used in the product. It is strong because it controls for bias: the randomisation, the blinding and the placebo control mean that the result is unlikely to be explained by anything other than the product. A trial that meets all of these criteria is evidence you can rely on.
The next strongest evidence is a randomised trial that is not double-blind, or not placebo-controlled. It is weaker because it is more vulnerable to bias — but it is still stronger than a laboratory study, because it measures what happens in people, not in a test tube.
The next strongest evidence is an observational study — a study that follows people over time, and compares those who take a supplement with those who do not. It is weaker than a randomised trial, because it cannot control for all of the differences between the groups — but it is still stronger than a laboratory study, because it measures what happens in people.
The next strongest evidence is a laboratory study — a study that measures what an ingredient does to cells, or to animals. It is weaker because it does not measure what happens in people. A laboratory study that shows an ingredient has an effect on cells is not evidence that it has the same effect in people. A product that cites a laboratory study, or an animal study, as proof that it works in humans is overstating its case.
The weakest evidence is a customer review. A customer review is a single person's experience — and a single person's experience is not evidence. A product with a hundred five-star reviews may be a good product — or it may be a product with a hundred fake reviews. A review that cites a randomised trial, in a comparable population, at the dose and in the form used in the product, is stronger than one that cites a customer review. Check the evidence, not the verdict.
The second thing to check is whether the study matches the product. A study of an ingredient, at a different dose, in a different form, in a different population, is not evidence for the product — it is evidence for a different product. An ingredient studied at 500 mg a day, present at 50 mg in the product, is not evidence for the product. Check the dose, the form and the population — not just the ingredient.
The third thing to check is who funded the study. A study funded by the manufacturer, with no independent replication, is less convincing than one funded independently. A study funded by the manufacturer, and replicated independently, is stronger. Check the funding, and the replication — not just the result.
The fourth thing to check is whether the study measured what matters. A study that measures a blood marker is weaker than one that measures a clinical outcome. A study that measures a blood marker, and claims a clinical benefit, is overstating its case. A study that measures a clinical outcome — how people feel, how they function, how they live — is stronger. Check the outcome, not the marker.
The fifth thing to check is whether the study was published. A study that has been published in a peer-reviewed journal is stronger than one that has not. A study that has been published, and replicated, is stronger still. A study that has not been published — or that has been published only on the brand's website — is weaker. Check the publication, not the press release.
A product that cites strong evidence — a randomised, double-blind, placebo-controlled trial, in a comparable population, at the dose and in the form used in the product, funded independently, measuring a clinical outcome, published in a peer-reviewed journal — is a product worth buying. A product that cites weak evidence — a laboratory study, a customer review, a manufacturer-funded trial with no replication — is a product worth questioning.
After the evidence, there are three more steps. First, talk to your doctor — especially if you are taking prescription medicines, or managing a medical condition. Some supplements interact with medicines; some affect blood tests; some are unsafe for people with certain conditions. A five-minute conversation, with a complete list in front of you, is worth far more than a difficult conversation afterwards.
Second, set a baseline and a review date. Before you start, write down how you feel — in a sentence, with a date. A baseline is the only way to judge whether a product has worked. Set a review date — six weeks, eight weeks, twelve weeks — and decide, on that date, whether to continue. A product that does not work, after a fair trial, should be stopped — not swapped for another product that makes the same promise.
Third, keep a record. Write down the product name, the brand, the dose, the batch number, and the date you started. A list is five minutes to set up, and a minute a week to maintain. If a recall is issued, the batch number tells you whether you are affected. If a side effect appears, the batch number helps the brand and the regulator identify the problem.
Independent reviews can help you judge evidence. Thesupplementrating.com publishes evidence-led reviews that start with the label, check doses against research, examine testing disclosure and state their limits plainly — and its safety checklist covers the five checks that matter most before you buy. A review of a botanical memory formula shows the method in practice: label first, dose second, evidence third, verdict last.
Evidence is not a word. It is a hierarchy — and a product that cites strong evidence is a product worth buying. Check the trial, the dose, the form, the population, the funding, the outcome and the publication — and let the evidence, not the marketing, decide.
Top comments (0)