How to Check a Health or Wellness Claim Before You Act on It
- In a 2023 KFF poll of 2,007 US adults, 96% had heard at least one of ten false health claims, and 24% used social media every week to find health information.
- Ofcom's 2026 media literacy report found 72% of UK adults felt confident judging whether online information is accurate, yet 17% were confident and still mistook a fake profile for a real one.
- UK advertising rules prohibit claims that a food prevents, treats or cures disease, and claims that refer to a rate or amount of weight loss.
- The US Federal Trade Commission expects health claims to rest on competent and reliable scientific evidence, and says testimonials alone do not count.
- A headline that reports a relative risk ("halves the risk") tells you little until you know the absolute numbers behind it.
A video says seed oils are poison. The next one sells a powder for cortisol. A search summary written by software tells you, fluently, something a dietitian would wince at. The supply of health advice is now effectively unlimited, and the supply of checked health advice is not.
This guide is a working method: five questions, a list of warning words, the advertising rules in both countries, and a table showing where to check each kind of claim. It takes about five minutes per claim once you have done it twice.
How much health misinformation is out there?
A great deal, and it varies by topic. A systematic review in the Journal of Medical Internet Research pooled 69 studies of health content on social media. In some studies of posts about smoking products and drugs, up to 87% contained misinformation. For vaccines the figure was 43%.
The studies predate 2019, so they describe the pattern more than the present level.
Exposure is close to universal. In the KFF Health Misinformation Tracking Poll of 2,007 US adults, 96% had heard at least one of ten false claims tested. Fewer than one in five called any of them definitely true. Most people sat in the uncertain middle: probably true, probably false.
In the UK, Ofcom's 2026 Adults' Media Use and Attitudes report found that 72% of adults felt confident judging whether information online is accurate. In a test, 17% were confident and still identified a fake social media profile as genuine. Confidence and skill are different things, and the rest of this page is about skill.
What are the five questions to ask?
1. Who is saying it?
Look for a name, a qualification you can verify and an organisation. In the UK, "dietitian" is a title protected by law and regulated by the Health and Care Professions Council. "Nutritionist" and "wellness coach" are not on that list.
The US National Library of Medicine's MedlinePlus guide suggests favouring government agencies, medical schools and large professional or non-profit bodies.
2. What are they selling?
A discount code, an affiliate link, a course or a supplement line does not make someone wrong. It does give them a reason to sound certain. MedlinePlus is blunt about it: a trustworthy site has one goal, which is to give you good information.
3. What kind of evidence is it?
A personal story shows that something happened to one person. A laboratory or animal study shows that something might be worth testing. A randomised trial in people tests it. A systematic review weighs all the trials together. Wellness marketing leans heavily on the first two.
4. Do independent sources agree?
Check one UK and one US source that sell nothing: the NHS website, an NIH institute, or a Cochrane review. Cochrane produces systematic reviews of health research, and every review includes a plain language summary. If none of them mentions the claim, that silence is informative.
5. Does it apply to you?
A finding in male athletes aged 20 may not transfer to a woman of 55 on blood pressure tablets. Advice that ignores age, pregnancy, medicines and existing conditions was not written with your safety in mind.
Which words are red flags?
MedlinePlus advises caution around dramatic writing, promises of cures and claims that sound too good to be true. In practice, the following words and patterns should slow you down.
- "Detox", "cleanse" or "flush toxins", with no toxin named and no measurement offered.
- "Doctors don't want you to know" or "what the industry is hiding", which asks you to trust the speaker because they are outside the system.
- "Clinically proven" or "science-backed" with no study linked.
- "Always", "never" and "everyone should", applied to food or exercise.
- A named amount of weight loss in a named number of days.
- One cause for every symptom: cortisol, inflammation, gut bacteria or hormones, depending on the month.
- A long ingredient list described as a "proprietary blend", so no dose can be checked.
Balanced information tends to sound duller. It gives ranges, names who was studied, and says what is not yet known.
What are advertisers allowed to claim?
Less than you might think, in both countries. Knowing the rules helps, because a claim that breaks them is a sign that nobody careful has reviewed the advert.
| UK (ASA and CAP Code) | US (Federal Trade Commission) | |
|---|---|---|
| Basic standard | Only health claims listed as authorised on the Great Britain register may be used for foods and supplements | Claims need competent and reliable scientific evidence, judged by experts in the field |
| Disease and benefit claims | Claims that a food prevents, treats or cures human disease are prohibited | Randomised controlled trials in people are generally the evidence expected for health benefit claims |
| Vague claims | "Good for you" style claims are acceptable only alongside a specific authorised claim | Qualifying information must be clear, conspicuous and hard to miss |
| Testimonials and endorsers | Health claims may not refer to the recommendation of an individual health professional | Testimonials are not evidence, and results shown must be typical or clearly qualified |
The CAP Code also bans health claims that refer to a rate or amount of weight loss. The Great Britain nutrition and health claims register, published by the Department of Health and Social Care, lists both authorised and rejected claims, and anyone can search it.
The FTC guidance was issued in December 2022 and sets out what the agency expects of anyone marketing a health product. Rules describe what should not be said. They do not stop it being said, which is why the checking still falls to you.
How do you read a study headline?
With five quick questions, all of which a decent news report will answer in its first few paragraphs. If it does not, the press release was probably the only thing the writer read.
- Was it in people, animals or cells?
- How many people, and who were they?
- How long did it run?
- What was it compared with: a placebo, another treatment, or nothing?
- Is the result given as a relative or an absolute change?
The last question matters most. Take an invented example. A headline says a supplement "halves the risk" of a condition. In the study, 2 in every 1,000 people in the comparison group developed it, against 1 in 1,000 taking the supplement. The relative reduction is 50%. The absolute reduction is one person in a thousand.
Both figures are true, and only one of them sells capsules. Also watch for "linked to" and "associated with". Observational studies can show that two things occur together. They cannot show that one caused the other, however large the sample.
Where should you check each type of claim?
| Claim | What to check | Where |
|---|---|---|
| "This supplement supports immunity, energy or hormones" | Whether the wording is an authorised claim, at what dose, and whether you need the nutrient at all | GB nutrition and health claims register; NHS vitamins and minerals pages; NIH Office of Dietary Supplements fact sheets |
| "This food or diet treats a condition" | Whether any trial in people exists, and what the national guidance says | NHS website; MedlinePlus; Cochrane plain language summaries |
| "A new study proves" | People or animals, numbers, duration, comparison group, absolute change | The study abstract itself, usually on PubMed, not the post about it |
| "Lose a stone (14lb) in a month" | Nothing. A named rate of weight loss breaks UK advertising rules on its face | CAP Code rule 15.6.6; FTC Health Products Compliance Guidance |
| "My results speak for themselves" | Whether anything beyond the testimonial is offered, and whether the person earns from the product | The post's disclosure label; the seller's own site |
| An AI summary or chatbot answer | Whether it links to sources, and whether those sources say what the summary says | Click through to the NHS, NIH or the original paper |
On supplements, the NHS position is that most people should get the nutrients they need from a varied and balanced diet, although some people may need extra. The NIH Office of Dietary Supplements publishes a fact sheet for each common ingredient, in consumer and professional versions.
Can you trust AI summaries and chatbots?
Treat them as a well-read stranger: often right, never accountable. Ofcom found that 54% of UK adults now use AI tools, up from 31% a year earlier, and that 75% of online adults read AI-generated search summaries at least some of the time.
These tools are useful for explaining a term, summarising official guidance you then open, or drafting questions for an appointment. They are poor at knowing your medicines, your history and when they are wrong. A fluent paragraph with no links deserves the same five questions as an influencer's reel.
The same Ofcom report found that only 52% of search engine users correctly identified sponsored results, and 37% were confident but wrong. Before reading any result, check whether it is an advert. It takes two seconds and changes how you read everything beneath it.
How do you cut the noise down?
By choosing sources on purpose. Pick one national health site (the NHS or MedlinePlus), one or two qualified people whose work you have checked against it, and stop there. Unfollow accounts that leave you anxious or reaching for your bank card. Mute words such as "detox" if your platform allows it.
Then set a personal rule for acting. A reasonable one is to wait 48 hours and run the five questions before buying anything or cutting out a food group. Most urgent-sounding wellness advice is not urgent, and the delay costs nothing.
When should you ask a professional instead?
Do not research your way around these
See a GP or pharmacist in the UK, or your doctor or pharmacist in the US, before stopping or changing a prescribed medicine, before taking a supplement alongside medicines, during pregnancy, or if you have a long-term condition.
New symptoms that persist need an examination, not a search. For urgent concerns use NHS 111 or urgent care, and for emergencies call 999 or 911.
In the KFF poll, 93% of people said they trusted their own doctor a great deal or a fair amount, more than any other source tested. It is reasonable to take a screenshot of a claim to an appointment and ask what they make of it. Most clinicians would prefer that to finding out later.
Questions readers ask
Are influencers allowed to give health advice?
Anyone may share an opinion, but once a post promotes a product, advertising rules apply. In the UK, the CAP Code limits food and supplement adverts to authorised health claims and bans claims to prevent, treat or cure disease. In the US, the FTC expects scientific evidence and says a testimonial is not evidence.
How can I tell if a study is reliable?
Check whether it was in people, how many took part, how long it ran and what the comparison was. One small, short trial is a starting point. A systematic review of several trials, such as a Cochrane review, carries more weight. Be wary of any single study described as proof, and look for the absolute numbers behind percentage changes.
Is "natural" safer?
Not as a rule. "Natural" has no bearing on dose, purity or interactions with medicines, and plenty of plants are pharmacologically active. If you take prescribed medicines, ask a pharmacist before adding a herbal product or a high-dose supplement. The NIH Office of Dietary Supplements publishes fact sheets on common ingredients, which are a sensible first read.
What if my friend says it worked for them?
It may well have, and you can be pleased for them. One person's experience cannot separate the product from everything else that changed at the same time, including sleep, season and expectation. That is the job controlled trials exist to do. Ask what else they changed, then run the same five questions.
Sources
- Suarez-Lledo V, Alvarez-Galvez J. Prevalence of health misinformation on social media: systematic review. Journal of Medical Internet Research. 2021;23(1):e17187.
- KFF. KFF Health Misinformation Tracking Poll Pilot. Survey of 2,007 US adults, 23 May to 12 June 2023.
- Ofcom. Adults' Media Use and Attitudes report 2026. 2 April 2026.
- Advertising Standards Authority and Committee of Advertising Practice. CAP Code section 15: food, food supplements and associated health or nutrition claims.
- Department of Health and Social Care. Great Britain nutrition and health claims (NHC) register. Updated 19 May 2026.
- US Federal Trade Commission. Health Products Compliance Guidance. 20 December 2022.
- US National Library of Medicine, MedlinePlus. Evaluating health information.
- Cochrane. Cochrane evidence.
- Health and Care Professions Council. The professions we regulate.
- NHS. Vitamins and minerals; NIH Office of Dietary Supplements. Dietary supplement fact sheets.
© The Reset Edit™ 2026 — Modern Tools + Lifestyle Essentials for Sustainable, Reset Living. All rights reserved.