Supplement clinical studies: how to find them, read them and cite them honestly on a VSL page

Supplement clinical studies are the human trials, ideally randomized and controlled, that test whether an ingredient does what a claim says. Start with the NIH Office of Dietary Supplements fact sheets and Cochrane reviews for the overall picture, then pull the trials themselves from PubMed. A study supports your claim only when its dose, form, population and outcome match your product and your words.

Where to find supplement clinical studies

Work from summaries down to single trials. A summary tells you what the whole body of evidence says before one exciting paper talks you into something the rest contradicts. Three sources are worth checking for any ingredient.

NIH Office of Dietary Supplements (ODS). Congress set up ODS under the Dietary Supplement Health and Education Act of 1994 (About ODS). It publishes evidence-based fact sheets for health professionals and consumers on specific vitamins, minerals, herbs and other supplements, covering health effects, safety, recommended amounts and interactions with medicines (ODS fact sheets). Start with the health-professional version for your ingredient: it is written for people who read studies.

Cochrane Reviews. A Cochrane Review is a systematic review of research in health care and health policy, published in the Cochrane Database of Systematic Reviews and updated when new evidence becomes available (Cochrane Library). If one covers your ingredient and outcome, it shows what the trials found together, and its summary of findings tables give key information on the quality of the evidence.

PubMed. PubMed is a free resource for searching biomedical and life sciences literature, developed and maintained by the National Center for Biotechnology Information at the US National Library of Medicine, part of NIH. It holds more than 40 million citations and abstracts and often links to the full text when a source such as PubMed Central or the publisher has it (About PubMed).

A PubMed search that works for most ingredients: type the ingredient plus the outcome you want to claim, such as "magnesium sleep quality". Then use the filter sidebar: the article type filter (the user guide gives Clinical Trial and Review as examples), the systematic review filter, free full text, and, under Additional filters, the Humans species filter (PubMed user guide). The default sort is Best Match, so scan past the first page before you conclude there is nothing.

The ODS website is the gentlest of the three starting points, and a short NIH video with ODS experts shows what you will find there:

What counts as good evidence for a supplement claim?

For advertising, the bar is set by the FTC's Health Products Compliance Guidance, published in December 2022: health claims need competent and reliable scientific evidence, meaning tests, analyses, research or studies conducted and evaluated objectively by experts (FTC). The guidance is explicit about how it weighs study types:

• Randomized, controlled human clinical trials (RCTs) are the most reliable form of evidence, and as a general matter, health benefits need to be substantiated by them.
• Animal studies have only limited value in predicting effects in humans, and animal and in vitro (test-tube) studies aren't sufficient to substantiate a health claim without confirmation in human RCTs.
• Epidemiological or observational studies can show an association, but they don't prove a causal link.
• Quality beats quantity. Claims are judged against the entire body of relevant and reliable evidence, not only the studies you picked.
• Testimonials aren't science. Individual experiences aren't a substitute for research.

In practice, the three kinds of study a PubMed search returns sort like this:

How the FTC weighs three kinds of study

StrongestRandomized controlled trial

• People are assigned at random to the ingredient or a control, such as a placebo
• Built to test cause and effect
• The FTC calls it the most reliable form of evidence

Shows a linkObservational study

• Follows what people already do, such as their diet
• Can show an association
• Doesn't prove the ingredient caused the result

Not enough aloneAnimal or in vitro study

• Tests in animals, or in cells in a lab
• Useful for ideas about how something might work
• Not sufficient for a health claim without human RCTs

The FTC side of supplement advertising has its own post:

How to read a supplement study in ten minutes

Open the full text. The FTC says it can't evaluate the quality of a study from an abstract or an informal summary, and neither can you. Then work through these questions in order, with the list next to the paper:

Reading a supplement study

• Design: randomized, placebo-controlled and blinded?
• Population: who took part, and how many?
• Health status: healthy people, or patients with a disease?
• Dose: how much a day, compared with your serving?
• Form: the same extract or compound, taken the same way?
• Duration: how long, and does your claim promise results sooner?
• Outcome: what was measured, and is it the benefit you claim?
• Effect: statistically significant and large enough to matter to a person?
• Funding: who paid for the study, and who ran it?
• Context: does the rest of the evidence agree?

0 of 10 done

Three of the questions need extra care.

Population. The FTC asks whether the study population reflects the people your ad targets. In one of its examples, studies in people with significantly impaired blood circulation to the brain don't support a claim that a tea improves mental alertness in healthy adults. If the participants had a diagnosed disease, a claim built on the trial also drifts toward a disease claim, which is covered below.

Outcome. The guidance asks how well the outcomes tested relate to the specific benefit advertised. Its example: gains on working-memory tasks in a lab don't support a general claim that a product improves memory. Write down the exact outcome the trial measured, and compare it word for word with your claim.

Effect size. Any statistically significant result must translate into a benefit that is clinically meaningful for consumers. A difference the participants couldn't notice is a weak basis for a promise in a VSL.

Publication doesn't settle quality either. The FTC notes that a published study isn't a guarantee of quality, or proof that a product is effective for the benefit advertised.

How to cite studies on a VSL page

Citations do two jobs on a sales page: they let a skeptical buyer check you, and they force you to match each claim to a source. Five rules:

1. One claim, one citation, side by side. Put the reference number next to the claim it supports, on the page and in the video, where it should stay on screen while the claim is spoken and long enough to read on a phone.
2. Describe the study accurately. Say who took part, how much they took, for how long and what changed, using the paper's own numbers.
3. Link the record. Point each reference to its PubMed record or the journal's page, so anyone can read it.
4. Don't pad. A long list of loosely related papers implies support you don't have. Every reference should back a claim you actually make.
5. Keep a claims file. For each claim: the exact wording, the studies, why they match your dose and form, and who reviewed it and when.

Describing a study in one sentence

Watch the sentence that introduces a study: that is where a careful paper turns into an overclaim. Compare these fictional lines:

Overclaims vs descriptions

Fictional lines, in matching order

Don'tOverclaims

• "Scientists proved it melts stress away"
• "Clinically proven formula"
• "Backed by 50 studies"
• "Works in days"

DoDescribes

• "In an 8-week trial in 60 healthy adults, 600 mg a day of this extract improved stress questionnaire scores more than placebo [3]"
• "Contains the same extract, at the same daily amount, used in two human trials [1, 2]"
• "Supported by two randomized trials in healthy adults [1, 2]"
• "Measured after 8 weeks of daily use [3]"

The versions that describe are longer and less exciting, and they survive a reviewer who opens the paper.

On a RevvyScales page, the scientific references block holds your list of sources, and you fill it in from the theme editor (comments and references). A references list helps buyers check you, but it isn't substantiation by itself: each study still has to support the claim it sits next to, which is why the VSL compliance guide treats references and claims together.

Does a study on an ingredient prove your product works?

Only when your product matches what was studied. The FTC's questions are specific: Is the ingredient or combination of ingredients in your product the same as in the study? How do the dosage and formulation compare? Is the product taken the same way? When conditions differ, you have to judge whether it is appropriate to extrapolate, and the guidance gives the example of two stimulant ingredients that may affect the body differently together than they do individually.

A fictional example shows how fast the gaps add up. Your capsule contains 300 mg of a botanical extract, call it Extract X, alongside two other ingredients. The best trials you found gave 600 mg a day of a different, standardized extract of the same plant, on its own, for eight weeks, to adults diagnosed with a sleep disorder:

The honest options are to change the product so it matches the evidence, to narrow the claim to what the evidence supports for your product, or to drop the claim.

Wording matters as much as the match. In the FTC's reading, "clinically tested ingredient" implies more than that a test took place. "Clinically proven" describes your product only if your product, as sold, was tested and the results support the claim. When only the ingredients were studied, say exactly that, for example: "Contains 600 mg of [the same standardized extract], the daily amount used in the human trials cited below."

When a citation becomes a disease claim

Citations can create claims of their own. FDA's rule on supplement claims counts the citation of a publication or reference as a way to make an implied disease claim when the citation refers to a disease use and, in the context of the labeling as a whole, implies treatment or prevention of a disease (21 CFR 101.93). A trial in patients with high blood pressure, cited under a line about heart health, can read as a claim that your product treats hypertension.

Three habits keep you on the right side:

• Prefer trials in healthy people that measured the structure or function you describe.
• When the best evidence comes from patients, get regulatory advice before you cite it, and never place it next to a structure/function line it seems to support.
• Read your references the way a regulator would: as part of the page's message, not a footnote.

The FTC side applies at the same time. Under its guidance, you need adequate substantiation for every objective claim, express or implied, before the ad runs, so the research comes first and the script second. For the line between the two kinds of claim, see structure/function claims. If your VSL explains how the product works, that mechanism is a claim too, and our guide to the unique mechanism shows how to present one without drifting into pseudo-science.

A claim-to-citation worksheet

Do this before the script is final, not after the VSL is edited. The order matters:

From claim to citation

1. 1Write the exact claimThe words you want in the VSL
2. 2Read the summariesODS fact sheet, Cochrane review
3. 3Pull the human trialsPubMed, then the full text
4. 4Check the matchDose, form, population, outcome
5. 5Cite it beside the claimA numbered reference with a link

Keep one row per claim in a sheet like this one (the row is fictional):

When the evidence is thin

Sometimes the worksheet's honest answer is that the evidence doesn't support the claim you wanted. You have four moves, in this order:

1. Change the claim to what the evidence does support, even if it is narrower.
2. Change the product so its dose and form match the trials.
3. Test your product in a properly designed human trial, run by an independent research team, before you claim the results.
4. Drop the claim and sell on what you can document: the formula, lot testing, the guarantee.

What you can't do is fill the gap with testimonials. The FTC is plain that individual experiences aren't a substitute for scientific research, however honest the customer.

Then make sure what is in the bottle matches what the trials used. The certificate of analysis for each lot is where you confirm the dose you are citing.