How to Check Online Health Claims Before You Act

A friend shares a video about a food that supposedly prevents illness. The speaker sounds confident, the comments are enthusiastic, and a shopping link is one tap away. Before buying anything or changing a health routine, turn the promise into a question you can check.

You do not need to become a researcher. A useful first pass is to identify the claim, find its source, and notice what the evidence cannot tell you.

AI-generated illustration of checking information, not medical evidence.

1. Write down the actual promise

Replace a vague phrase such as better health with the specific result being promised. Is the claim about a nutrient in a food, a change in a laboratory measurement, or fewer illnesses? These are different questions.

Try this fictional example: a snack contains fiber, so an advertisement implies it prevents disease. The first statement can be checked on a label; it does not establish the second. Our Nutrition Facts guide explains what food labels can and cannot show.

Write one sentence: This source claims that doing X produces Y in these people. If you cannot identify X or Y, the promise is not yet clear enough to evaluate.

2. Check who is speaking and who benefits

Open the site’s About page and look for its owner, purpose, contact details, and the author’s relevant qualifications. Check whether advertisements and sponsorships are identified. A seller may provide useful information, but a purchase link is a reason to seek an independent source too.

MedlinePlus recommends checking ownership, funding, editorial processes, and supporting references. A professional-looking page, a familiar person sharing it, or a large audience does not replace those checks.

Look for a clear account of who prepared or reviewed the material. Do not assume an article has received medical review because it displays a health-related logo or links to a government page. Apply these questions to this site as well.

3. Follow the reference and check its date

A reference list is a starting point. Open the source and ask whether it supports the exact statement beside the link. A page about an ingredient is not automatically evidence for a finished product containing it.

Check when the information was written or reviewed. A newly shared post can be based on older material. Age alone does not make information wrong, but current guidance may answer a question differently.

The NIH Office of Dietary Supplements’ evaluation guide advises checking original sources, evidence, and update dates. It also distinguishes research-based information from personal testimonials.

4. Ask what was studied, in whom, and for how long

When a claim relies on research, look for a few basic details:

• Was the work done in people, animals, or laboratory cells?
• How many people participated, and were they similar to the intended audience?
• What was the approach compared with?
• Which outcome was measured, and how long did follow-up last?
• Were side effects, limitations, and funding reported?

MedlinePlus explains why study design and comparison with previous findings matter. A randomized trial randomly assigns people to comparison groups; this helps reduce bias when testing an intervention. The label randomized does not remove the need to inspect the rest of the study.

NCCIH’s health-news checklist also asks whether a result is meaningful in everyday life. A short study cannot settle questions about benefits or harms years later.

For practice, read the evidence and limitations sections of our walking-after-meals guide. Notice how the population, measured outcome, and study duration limit the conclusion.

5. Keep association separate from cause

An association means two things occur together in the data. It does not, by itself, show that one caused the other. Other differences between groups can help explain the pattern.

The CDC’s explanation of study designs describes observational research as recording exposures rather than assigning them, and explains how other factors can affect associations. Such research is valuable; the point is to avoid turning a reported link into a guaranteed effect.

When a headline says linked to, be cautious if the sales page changes that into prevents or cures.

6. Notice when explanation becomes a sales pitch

Pause at promises of guaranteed results, one product for many unrelated diseases, secret cures, or claims that natural means safe. The FDA’s guide to health-fraud warning signs identifies these tactics and warns that testimonials are not a substitute for scientific evidence.

A moving personal story may be sincere without establishing why someone improved. Fraudulent products can also cause harm or delay appropriate diagnosis and treatment.

Product wording deserves scrutiny even in ordinary shopping. Our handwashing guide offers an everyday example: compare antibacterial marketing with the linked public-health guidance rather than assuming the label proves an added benefit.

7. Decide what to do with the uncertainty

Use this simple editorial routine, not a validated scoring system: save the claim, open its source, check a second independent authoritative page, and note what remains unanswered. Two websites repeating the same advertisement are not independent confirmation.

If you still cannot verify it, leave it unshared. For a decision about medicines, symptoms, or treatment, bring the link and your questions to a qualified healthcare professional rather than using the post as instructions.

Protect your privacy too. The NIH evaluation guide advises reading how a site uses personal information before signing up. You do not need to submit a medical history to check a general claim.

Takeaway

Check the promise, the speaker, the source, and the limits of the evidence. Links and checklists cannot establish medical accuracy or professional review. The useful result may be a better question, not an immediate purchase or a change in care.

This article provides general health education and is not medical advice, a diagnosis, or an individualized treatment plan.