“New study proves this supplement changes everything.” You have five minutes, an overflowing browser and no wish to become a research scientist before breakfast. You still deserve a useful way to decide whether the claim is worth your attention.
Start by finding the exact promise and the original evidence. You do not have to settle the whole supplement debate. You need to know whether this particular claim is strong enough to change this particular decision.
First, underline the promised result
“Supports performance” can mean almost anything. Faster over a fixed distance? One extra repetition? A change in a blood marker? More muscle after months of training? Write the advertised result as a plain sentence with a measurable outcome.
Here is an invented advertisement: “Plant formula builds 20% more muscle in just 12 weeks.” Before asking whether the product is good, ask what “20% more” refers to. Is it a change within one group, a difference from placebo, or a result in a small subset? Is “muscle” actually a measurement at one spot on an arm? The headline has not answered those questions.
Follow the link once
A useful reference should lead to the study, not another sales page repeating the same sentence. Check the authors, journal and date. If the page only says “research shows,” look for a reference list. NIH guidance recommends considering who provides online information, why it exists, how claims are supported and when the material was updated. [1]
This does not mean commercial websites are automatically wrong. It means that a seller’s confidence is not an additional experiment. Ten shops repeating one study still leave you with one study.
If you cannot find the source after a short, reasonable search, stop. “Unverified” is enough to justify postponing an optional purchase. You are not obliged to spend an evening repairing someone else’s incomplete advertisement.
Use four boxes to translate the study
| Box | What to write | Why it matters |
|---|---|---|
| People | Who took part, and how many? | A small study in trained young men does not answer every question for older beginners. |
| Product | Ingredient, amount and duration | A mixture, food and isolated ingredient are different interventions. |
| Comparison | What did the other group receive? | Improvement over time may reflect the shared training programme. |
| Result | The difference between groups on the relevant outcome | This is closer to the effect being attributed to the product. |
These boxes simplify questions that transparent trial reporting is supposed to make answerable. The CONSORT 2025 guidance asks researchers to report methods and results clearly, including outcomes, participant flow and open-science information. A beautifully formatted paper is still not automatically a trustworthy or relevant trial. [4]
Watch for the quiet switch
Claims often become overstated through a small change in wording. “Increased a marker linked to recovery” becomes “gets you ready for tomorrow.” “Tested alongside supervised training” becomes “builds muscle.” “Some participants improved” becomes “works especially well for people like you.”
Return to the invented 20% claim. Suppose the supplement group’s measurement rose by 20%, while the comparison group’s rose by 18%. The extra improvement is two percentage points, not 20 percentage points. You would still need to know the uncertainty around that difference and whether it matters in practice. These numbers are an arithmetic example, not a real trial.
Also ask whether the product on sale is the product studied. NCCIH notes that retail supplements can differ materially from research products. A study on one ingredient does not establish the effect of every blend that contains it. [3]
Give funding the right amount of attention
Read the funding and conflict-of-interest statements. A manufacturer’s involvement is relevant context, especially if the investigators are employed by the sponsor or the sponsor helped choose the comparison. It is a reason to examine methods and independent replication carefully.
It is not, by itself, a calculation showing the result is false. Conversely, public funding cannot rescue a poorly designed study. Keep the question concrete: does the design support the claim, and has another group found a similar result?
When only an abstract is available, recognise the boundary. You may be able to check the sample and headline findings, but not allocation procedures, missing data or detailed harms. An honest summary should say what was actually read.
A quality seal answers a different question
It is sensible to care about what is in a product. But independent quality testing and clinical effectiveness are different matters. NIH explains that quality seals do not guarantee that a supplement is safe or effective. Check what the specific programme certifies instead of treating the symbol as proof of better results. [2]
This matters at the checkout. Paying more for verified ingredient content may address a quality concern; paying more because a seal supposedly proves muscle growth is a different, unsupported inference.
Make the final decision smaller
You now have three reasonable outcomes. The claim is relevant and adequately supported, so you can consider cost and suitability. The evidence is interesting but too uncertain for your needs, so you wait. Or the claim does not match your goal, so you move on.
None requires declaring the product universally brilliant or useless. If your actual difficulty is getting to the gym around childcare, a small laboratory performance benefit may be less useful than a programme that fits the hours available. If your protein intake is already organised, another powder may solve no current problem.
For deeper examples, read the subgroup trap in muscle-building claims and who was actually studied?
WeightStart’s supplement plans are another offer to evaluate with the same questions: what is included, what problem does it address and what is the evidence for the recommendation? Today’s practical step is to save your four-box note beside one product you are considering. If the essential boxes remain empty, your money can stay where it is.
Sources
Updated 2 October 2026. The four-box audit and numerical example are editorial tools. The references explain evidence reporting and supplement evaluation.
- NIH Office of Dietary Supplements. How To Evaluate Health Information on the Internet: Questions and Answers. Public guidance, accessed 2 October 2026.
- NIH Office of Dietary Supplements. Dietary Supplements: What You Need to Know. Public guidance, accessed 2 October 2026.
- NCCIH. Using Dietary Supplements Wisely. Public guidance, accessed 2 October 2026.
- Hopewell S, Chan AW, Collins GS, et al. CONSORT 2025 statement: updated guideline for reporting randomised trials. BMJ. 2025;389:e081123. doi:10.1136/bmj-2024-081123.