Clinically proven and doctor recommended claims: FTC rules

“Clinically proven”, “studies show” and “doctor recommended” are some of the most persuasive words a health product or treatment ad can use. They are also claims in their own right. Each one tells the reader something about the evidence behind the product: that a study was done, that it found a real benefit, or that a qualified professional looked at the product and judged it worth using. If the evidence does not match what the words promise, the ad can be deceptive even when the product works.

This guide explains how the FTC reads claims about the level of scientific support, what kind of evidence it expects behind a health benefit claim, why surveys and patient stories do not count, and how the endorsement guides apply when an ad leans on a doctor, an expert or a professional association. It is written for practices that sell products or cash-pay treatments and for companies that sell health products to consumers. For endorsements by well-known people, see our guide to celebrity and spokesperson endorsements. For claims about your own patients’ results, see our guide to advertising success rates and outcome claims. For natural, organic and green wording, see our guide to natural, organic and eco-friendly claims. When a clinician earns a commission on a recommended product, see affiliate links on health content. Label and reporting duties for products sold under a practice brand are covered in private-label supplements and skincare. For the same claims in third-party marketplace listings, see selling health products on online marketplaces.

The short version

• A claim about the amount or type of evidence, such as “clinically proven” or “scientists agree”, must be true as stated. You need at least the level of support you claim, on top of support for the benefit itself.
• Health benefit claims generally need competent and reliable scientific evidence, which in most cases means randomized, controlled human clinical trials. The study has to match your product, the people you sell to and the benefit you name.
• Customer surveys, testimonials and a practitioner’s own observations of patients are anecdotal. The FTC guidance says they are never enough to substantiate a health product claim.
• Images can make the claim for you. Lab coats, microscopes and medical journals can imply clinical proof, and a fine print line saying no study was done does not fix it.
• “Doctor recommended” is an expert endorsement. The doctor must have the expertise the ad implies and must have actually evaluated the product the way an expert in that field would, 16 CFR 255.3.
• A professional association’s endorsement must reflect the group’s collective judgment reached through a proper process, 16 CFR 255.4. Any payment or business tie with an endorser that consumers would not expect must be clearly disclosed.

Which laws apply

Section 5 of the FTC Act, 15 USC 45(a), declares unfair or deceptive acts or practices in or affecting commerce unlawful. Section 12, 15 USC 52(a), makes it unlawful to disseminate a false advertisement to induce the purchase of food, drugs, devices, services or cosmetics, and 52(b) makes such an ad a deceptive practice under section 5. Services are named in the statute, so these rules reach a practice that advertises a treatment, not only a company that sells a bottle or a device.

The FTC staff Health Products Compliance Guidance, issued in December 2022, replaced the agency’s 1998 dietary supplement advertising guide. It says that most of its examples involve supplements, but the same legal principles apply to the marketing of any health-related product, and it lists foods, over-the-counter drugs, homeopathic products, health equipment, diagnostic tests and health-related apps among the products in past FTC cases. The guidance distills FTC law into two principles: advertising must be truthful and not misleading, and before an ad runs the advertiser must have adequate substantiation for all objective claims it conveys, expressly or by implication.

The guidance also defines advertising broadly. It covers packaging and labels, brochures, websites, social media and influencer posts, press releases and interviews, and talks at trade shows and seminars. A claim a doctor makes on a podcast to promote the practice’s product line is advertising in this sense.

Claims about the level of support

Under the heading “Ads that refer to a specific level of support”, the guidance says marketers of health-related products must have at least the level of support they claim to have. Any assertion about the amount, type or strength of evidence must be accurate, in addition to the basic requirement that the safety or efficacy claim itself is substantiated. Its examples show how ordinary ad phrases are read:

• “Scientists Now Agree!” likely tells consumers that the science has reached consensus. Without evidence of that consensus, the claim is false (Example 16).
• Citing prize-winning research and a large public research budget is deceptive when the prize was for an unrelated use of the ingredient and the government research looked at safety, not efficacy. That holds even if the advertiser has other good evidence of efficacy (Example 17).
• “Clinically tested ingredient” implies not only that the ingredient was tested but that the results prove the benefit, and that the product delivers it. When the product contains other ingredients, the marketer should ask a qualified expert whether the field would require a test of the product itself (Example 18).
• “Clinically proven to improve memory” was not supported by a study that measured only working memory in lab tasks, because the ad’s everyday example involved a different kind of memory (Example 35).

The words do not have to appear for the claim to be made. In Example 1, a brochure with doctors in lab coats looking through microscopes, molecular structures and a stack of medical journals likely implies that a weight-loss product has been clinically proven. In Example 15, medical images, a caduceus and phrases like “research center” conveyed a false claim of scientific proof, and a line in the website terms saying no clinical study had been performed did not correct it, both because it contradicted the claim and because it was not clear and conspicuous.

What counts as competent and reliable scientific evidence

The guidance quotes the FTC’s definition of competent and reliable scientific evidence: tests, analyses, research or studies that have been conducted and evaluated in an objective manner by experts in the relevant disease, condition or function, and that are generally accepted in the profession to yield accurate and reliable results. The research must also be sufficient in quality and quantity, by the standards of the field, when considered in light of the entire body of relevant and reliable scientific evidence.

In practice, the guidance says, substantiation of health benefits will generally need to be randomized, controlled human clinical testing. It names several features the FTC looks at:

1. A control group. Improvement in the treated group alone could come from placebo, natural change in health, practice on the test or other factors. A placebo or sham control helps separate those effects from the product’s.
2. Statistically significant results. The difference between treatment and control must be statistically significant. Studies with several outcome measures should report all of them and adjust for the chance that one positive result is luck.
3. Clinically meaningful results. A statistically significant effect must also be large enough to matter to the consumer’s health.
4. Replication. There is no required number of trials, but a second study by independent researchers adds weight and reduces the risk that hidden bias drove one result.
5. A full write-up. The FTC cannot judge a study from an abstract. The guidance says to review the underlying study with an expert before drafting claims (Example 23), and notes that publication alone is not proof of quality, because peer review varies widely between journals.

Animal and test tube studies were not enough in Example 19, where human research was feasible and was what experts in the field expected. Observational studies can show association but not cause, and the guidance accepts high-quality epidemiologic evidence only in limited cases where experts consider it an acceptable substitute and trials are not feasible.

What does not count: surveys, stories and recommendations

The guidance is direct on this point. Anecdotal evidence about individual consumers’ experiences, including surveys of consumer experiences, is never sufficient to substantiate claims about the effects of a health product. The same applies to a practitioner’s observations of patients: the guidance calls a healthcare practitioner’s observation about a product’s effect on patients anecdotal, because it does not show a causal relationship.

In Example 21, an ad accurately reported that most surveyed customers noticed better memory. Because the survey was only a collection of anecdotes, it did not substantiate any memory benefit. For a practice, the parallel is a line such as “our patients love it” or “we have seen great results in our office” used to imply that a supplement or device works.

Public health recommendations are not substantiation either. The guidance says advisories from a medical organization reflect a judgment on the best evidence available and are not a finding of cause and effect. In Example 22, a medical institution’s accurate advice to keep the mind active did not support a claim that a particular brain training program improves memory. The marketer needed clinical research on its own program.

Match the evidence to the claim

Even good studies can fail to support a claim if they tested something else. The guidance’s examples point to four common gaps:

• Different product. Studies of each ingredient alone may not support a safety or efficacy claim for a combination (Example 37), and research on one extract may not carry over to a product made with a different extraction method and chemical profile (Example 38).
• Different people. Trials in people with a specific condition do not support a claim aimed at healthy adults without first making sure that extrapolation is scientifically sound (Example 34).
• Different conditions. When study subjects also dieted and exercised, the ad must make clear that buyers will need to do the same to see similar results (Example 8).
• The whole body of evidence. A marketer cannot rely on a favorable small study and ignore larger or better studies that found no effect. In Example 31, continuing to promote an early positive study after later larger trials found no difference was deceptive.

Hedging weak science rarely solves the problem. In Example 13, a company with two limited studies said that “promising, preliminary scientific studies show that our product may be effective”. The guidance says the words promising, preliminary and may are unlikely to convey the limits of the science. If stronger evidence points the other way, the guidance says even a qualified claim is likely to be deceptive.

“Doctor recommended” is an expert endorsement

The endorsement guides define an expert as an individual, group or institution with knowledge of a particular subject, gained through experience, study or training, that is superior to what ordinary individuals generally acquire, 16 CFR 255.0(e). For these guides, product includes services, 255.0(d). Two rules in 255.3 then apply whenever an ad presents an endorser as an expert:

1. Real qualifications. The endorser’s qualifications must in fact give them the expertise the ad says or implies they have, 255.3(a).
2. A real evaluation. The endorsement must be supported by an actual exercise of that expertise, including an examination or testing of the product at least as extensive as someone with the same represented expertise would normally need to support the conclusions in the endorsement, 255.3(b). If the ad implies a comparison with other products, the expert must have made that comparison, and if it implies the product is superior, the expert must have found it superior.

The examples in 255.3 are close to health marketing. In Example 2, an endorser of a hearing aid is called “a doctor”. The ad likely implies expertise in hearing, such as a medical doctor with substantial audiology experience or a doctor of audiology. A doctor without that experience might still endorse it only if the ad clearly and conspicuously discloses the nature and limits of their expertise. In Example 6, a medical doctor says a drug will safely lower cholesterol. If what the doctor reviewed was letters from satisfied customers or a rodent study, the endorsement is likely deceptive, and both the advertiser and the doctor are liable.

The FTC health guidance adds Example 42: an infomercial calls an expert a “Doctor” and a leading clinician in joint health, but the person is not licensed to practice medicine, holds a doctorate in another field, has not reviewed the science on the product and was given a paid officer position at the company in return for appearing. The guidance finds the ad likely deceptive on each count, and says the officer position should be clearly disclosed even if the expert were qualified.

For a practice, this has a plain consequence. When your own physician recommends a product your practice sells, that recommendation is still an expert claim about the product. The physician needs the relevant expertise and an evaluation of the evidence, and the practice still needs scientific support for every benefit the recommendation implies, because the guidance says advertisers should not make claims through endorsements that they could not substantiate directly. Our guide to selling supplements in a medical practice covers the professional ethics side.

Associations, institutes and “chosen by hospitals”

Under 16 CFR 255.4(a), an organization’s endorsement must be reached by a process that fairly reflects the collective judgment of the organization. If the organization is presented as expert, it must use experts it recognizes as such, or standards it adopted in advance that suit the products being judged. In Example 1 of 255.4, a mattress endorsed by a chiropractic association needs either an evaluation by experts the association recognizes or compliance with mattress standards the association adopted earlier, not standards built around the advertised mattress.

Two examples in 255.3 cover names and selections. In Example 3, an approval from an “institute” whose name suggests an independent testing body is deceptive if the institute was set up or run by the manufacturer, and may be deceptive even if it is independent, unless it ran valid tests that support the message. In Example 4, a claim that a large hospital selected a non-prescription drug over competitors was deceptive because the hospital chose it for single-dose packaging that consumers could not buy, and the ad did not say so.

Example 3 of 255.4 adds that review and ranking sites that take money in exchange for higher rankings are deceptive, and a product seller who pays for a higher ranking may also be liable. A disclosure of payments does not cure rankings that the payments determine.

Disclose the connection

The guidance says that whenever an expert or consumer endorser is used, the advertiser should clearly and conspicuously disclose any material connection between them, meaning a personal, financial or similar connection that consumers would not reasonably expect and that would affect the weight or credibility of the endorsement. In Example 43, an ad touting a brand as the only one recommended in a best-selling book was deceptive because the manufacturer had paid the author to name it. The disclosure rules in 16 CFR 255.5, and how they apply to paid spokespeople, are covered in our guide to celebrity and spokesperson endorsements.

If a physician who recommends a product has an ownership stake in its maker, receives royalties or is paid to speak about it, that connection belongs next to the recommendation, not in a footer.

Before you use these phrases

• Write down every claim the ad makes, including what the images and the words “proven”, “tested”, “studies show” or “recommended” imply about the evidence.
• Get the full study reports, not abstracts. Have someone with expertise in the relevant field review them against the claim.
• Check that the study tested your product or the same formulation and dose, in people like your buyers, measuring the benefit you name, with a control group.
• Look for studies that found no effect, including ones you or your supplier commissioned. The claim has to hold up against the whole body of evidence.
• Strip out any reliance on customer surveys, reviews, testimonials or staff observations as proof that the product works.
• For “doctor recommended” or “expert approved”, confirm who the doctor or expert is, that their field matches the claim, what they actually reviewed and whether they are paid or connected to the seller.
• For association or institute seals, get the organization’s written process or standards and confirm the organization is independent of the seller.
• Keep the file. Substantiation has to exist before the ad runs.

Use and avoid

• Use: a description that matches the study, such as what was measured, in whom and for how long, next to the claim. Avoid: “clinically proven” for a product when only one ingredient was tested.
• Use: “recommended by Dr. [name], [specialty]” with any financial tie disclosed nearby. Avoid: an unnamed “doctor recommended” seal, or “doctor” for someone whose degree is in an unrelated field.
• Use: patient comments presented as opinions, alongside what a typical buyer can expect. Avoid: survey percentages or reviews offered as proof that a product works.
• Use: public health advice described as general advice. Avoid: implying that a medical body’s general recommendation endorses your product.
• Use: plain wording when the evidence is limited. Avoid: lab coat imagery, journals or “research center” language for a product with no clinical testing.

How we help

We write ad copy, landing pages and Google Ads campaigns for independent practices and keep each claim tied to a source the practice can show. We also build websites, local search and an AI front desk. We are not a healthcare provider and we do not give legal or medical advice; claim substantiation needs your clinical experts and your attorney. Related guides: medical device marketing claims, superlative claims in healthcare ads, Made in USA claims for health products and our guide to healthcare marketing rules.

Sources

• FTC: Health Products Compliance Guidance (December 2022, read November 1, 2026)
• 16 CFR part 255: Guides Concerning the Use of Endorsements and Testimonials in Advertising, sections 255.0, 255.3 and 255.4 (eCFR, up to date as of October 29, 2026)
• 15 USC 45(a): unfair or deceptive acts or practices (Cornell LII, read November 1, 2026)
• 15 USC 52: dissemination of false advertisements (Cornell LII, read November 1, 2026)

More in Health companies, vendors and employers