Key takeaways
- In a 2012 federal review of the research behind weight-loss and immune supplement claims, only 2% of the human studies companies sent had tested the actual product being sold1.
- Lab, animal and survey studies can point to a promising idea, but the FTC says they can’t prove on their own that a product helps people4.
- The most common gaps are a study of one ingredient instead of the product, a different dose or form, an indirect outcome, different people, and a single small trial14.
- Free fact sheets from the NIH Office of Dietary Supplements and NCCIH summarize the evidence for many ingredients and show when each page was last updated812.
- “Not enough evidence” means unproven, not useless, so a low-risk, low-cost, time-limited trial that you tell your clinician about can be a reasonable choice2.
Federal investigators once asked supplement companies a simple question: what’s your proof? In a 2012 report, the HHS Office of Inspector General (OIG) described how it picked 127 weight-loss and immune-support supplements and asked their makers for the evidence behind the claims1. The products were hand-picked, not chosen at random, so they don’t stand for every product on the market1.
Companies sent 1,624 documents covering 72 of the products. About a third were studies in people. Of those human studies, only 2% had tested the product on the shelf. The rest tested an ingredient that may have differed in form, dose or strength1.
None of the 557 human studies met all four elements the FDA recommends for solid scientific support1.
That doesn’t mean those products were frauds. It points to something more common: the claim sounds scientific, but the science underneath is about something slightly different. This guide shows how to spot that gap fast, and what to do when you find it.
“Not proven” is the normal state for most supplements
Supplements don’t need FDA approval before they’re sold, unlike prescription and over-the-counter drugs. Companies are responsible for having evidence that their claims are true, and the FDA doesn’t judge whether a supplement works before it reaches stores2.
Some supplements have solid support for specific uses. The NIH Office of Dietary Supplements (ODS) lists calcium and vitamin D for bone health, folic acid to lower the risk of certain birth defects, and fish-oil omega-3s for some people with heart disease2. For many others, ODS says more research is needed2.
Big reviews show how often promising ideas fall short. In 2019, a team led by Safi Khan reported in the Annals of Internal Medicine on 277 trials involving nearly 1 million people, testing 24 supplements and diets for heart health3.
Omega-3s and folic acid showed some benefits, but with low certainty. Calcium combined with vitamin D was linked to a higher risk of stroke. Multivitamins, antioxidants, vitamin A, vitamin B6 and iron showed no significant effect on deaths or heart outcomes3.
This is different from a scam. A scam lies on purpose. An unproven product may be sold in good faith on early, hopeful evidence. Your job as a shopper is to tell how far the evidence goes.
The evidence ladder, from test tube to systematic review
Not all studies can answer the question you care about, which is whether this product will help a person like you. Each rung on this ladder answers a different question.
| Study type | What it can show | What it can’t show |
|---|---|---|
| Lab (test tube or cell) study | Whether a compound does something to cells | Whether it works, or is safe, in a living body |
| Animal study | Effects in a living system | Whether people respond the same way, at the same dose |
| Observational study (surveys, health records) | Links between a habit and a health outcome | Whether the habit caused the outcome |
| Small pilot trial | Whether a bigger trial is worth running | A reliable answer, since chance plays a big role |
| Randomized controlled trial | Whether the product causes the effect, in the people tested | Results for other doses, forms or groups |
| Systematic review | Whether results hold up across many trials | Anything better than the studies it pools |
| Testimonials | How one person felt | Whether the product, rather than something else, made the difference |
A randomized controlled trial assigns people by chance to the product or a comparison, such as a placebo (a look-alike with no active ingredient). A systematic review gathers every good study on a question and weighs them together.
The Federal Trade Commission (FTC) polices health advertising. Its 2022 guidance says health claims need “competent and reliable scientific evidence,” which generally means randomized, controlled trials in people4.
Animal and test-tube studies aren’t enough by themselves. Observational studies show links, not cause. Personal stories, including customer surveys and practitioner observations, are never enough4.
Industry groups push back. In 2023, the Council for Responsible Nutrition asked the FTC to confirm that randomized trials aren’t a strict requirement for every claim5. Six other trade groups asked the agency to withdraw warning notices it had sent to hundreds of companies5. As of October 2026, the FTC guidance remains posted on its website4.
Five gaps that make a strong claim weak
Most thin evidence isn’t fake. It’s real research that doesn’t quite match the promise on the label. These five gaps cover most cases.
Gap 1: The ingredient was studied, not the product. This is the gap the OIG found 98% of the time1. The FTC warns that mixing ingredients can change how each one works, so studies of single ingredients may not apply to a blend4.
A 2026 case shows how this plays out. The National Advertising Division, a self-regulatory watchdog for advertising, reviewed a mood chocolate that promised to make eaters feel calm and collected7. Studies of two of its ingredients, L-theanine and kanna, weren’t a good fit for that promise.
A saffron study at the same dose as the chocolate’s held up, but only for a narrower, saffron-specific claim. The company dropped most of the challenged claims7.
Gap 2: A different dose, form or extract. The ODS fact sheet on ashwagandha notes that trials used many different extracts, made and standardized in different ways, at different doses8. That makes it hard to say which product or amount works8.
Turmeric has the same problem. Products vary in how much curcumin they contain and often add other ingredients, which makes studies hard to compare12. The FTC gives a similar example: a study of one probiotic strain taken as a capsule may not apply to a different strain sold in a drink4.
Gap 3: The outcome isn’t the one you care about. Some studies measure a lab value instead of how people feel, function or live. The FDA calls these surrogate endpoints9. A surrogate is most trustworthy when many trials have tied it to real health results, as they have for blood pressure and stroke9.
Beta-carotene shows what can go wrong when a plausible theory skips the outcome test. Antioxidant pills seemed like a natural way to protect smokers from lung cancer. In one trial of 29,133 male smokers, beta-carotene pills raised lung cancer risk by 18%10.
A second trial gave beta-carotene plus a form of vitamin A to 18,314 smokers, former smokers and asbestos-exposed workers. It ended early after lung cancer rose 28% and deaths from all causes rose 17%10.
Gap 4: The people studied aren’t like you. The OIG found that about half of the human studies involved groups unlike the people likely to buy the product1.
Vitamin C is a good example. A Cochrane review, summarized by ODS, found that regular vitamin C cut colds in half among marathon runners, skiers and soldiers under extreme physical stress. In the general population, it didn’t significantly lower the chance of catching one11.
Gap 5: One small study, often run by the seller. NIH’s National Center for Complementary and Integrative Health (NCCIH) notes that small studies are more likely to produce results by chance6. The FTC adds that results repeated by independent researchers carry more weight, and that combing through data after a trial for a winning subgroup doesn’t count as proof4. A companion article explains how the sponsor’s money can shape a study.
How the four evidence levels look in real products
Here is how the evidence scale applies to some popular choices, based on federal summaries. Ratings can change as new studies arrive, so check the date on any fact sheet you use.
| Product and use | Rating | Why |
|---|---|---|
| Daily vitamin C to shorten colds | , for a small effect | Colds ran 8% shorter in adults and 14% shorter in children; starting it after symptoms began didn’t help11 |
| Echinacea to prevent colds | May slightly lower the chance of catching a cold; unclear whether it shortens one; several earlier major studies found no benefit1314 | |
| Ashwagandha for stress | Several small, short trials of different extracts suggest lower stress; long-term safety unknown8 | |
| Turmeric for knee arthritis pain | Early research suggests some relief; higher-quality studies needed12 | |
| Turmeric for most other uses | NCCIH calls the evidence insufficient12 | |
| Ginkgo to prevent dementia | A study of more than 3,000 older adults found it didn’t prevent or slow decline14 | |
| Beta-carotene pills for smokers | Two large trials found more lung cancer10 |
Notice how often “promising but early” shows up in this table. That’s the honest middle ground where many natural products sit.
Where to find an honest summary in minutes
You don’t have to read journals to get a fair read on an ingredient. Several free sources, most of them run by the federal government, do the summarizing.
- ODS fact sheets cover vitamins, minerals and many herbs, with a plain-language consumer version and a detailed version for health professionals. Each shows an update date; the ashwagandha sheet, for example, was updated in May 20258.
- NCCIH herb pages summarize what the science says and list safety concerns. The turmeric page was updated in April 202512.
- MedlinePlus, from the National Library of Medicine, keeps an A to Z index of herbs and supplements. As of July 29, 2025, it no longer carries the Natural Medicines database content it once hosted. Its links now point to NCCIH, National Cancer Institute and ODS pages15.
- Cochrane reviews pool trials on a question and include plain-language summaries.
- PubMed, the federal database of medical studies, lets you filter results to randomized controlled trials and recent years.
The 10-minute evidence check
Use this routine before you buy something new for a health reason. A notepad or phone note helps.
- Write down the exact claim (1 minute). What benefit, for whom and how fast? “Supports calm” and “reduces anxiety in adults within two weeks” are very different promises.
- Name the study type (2 minutes). Find the study the company cites. Place it on the ladder above. Lab, animal or survey studies alone mean the claim is still unproven.
- Check the match (2 minutes). Did the study test this product, or one ingredient? Same dose, same form, same extract? Were the people like you in age, health and sex?
- Check the outcome (1 minute). Did the study measure the benefit you want, such as fewer colds or less pain, or a lab marker that may or may not lead there?
- Count the studies and who ran them (2 minutes). One small trial by the seller is a starting point. Several trials, including some by independent teams, are much stronger.
- Read one independent summary (2 minutes). Look up the ingredient on an ODS or NCCIH page and note the date. If the government summary is cautious and the ad is confident, trust the summary.
If you answer “no” or “don’t know” to most of the questions in steps 3 through 5, treat the product as unproven. That isn’t a verdict that it fails. It’s a reminder to pay for it like an experiment, not a sure thing.
When trying something unproven can make sense
Many people who value natural health are comfortable with self-experiment, and many traditional remedies have centuries of use behind them. Tradition is a good reason to study a remedy. It isn’t proof that a particular product works, which is why the FTC doesn’t accept personal stories as substantiation4.
Still, unproven doesn’t have to mean off limits. Trying something with thin evidence can be reasonable when it passes these tests:
- Low risk. Read the safety section of the fact sheet first. Ashwagandha, for example, has been linked to cases of liver injury and to thyroid changes, and its long-term safety isn’t known8.
- Low cost. Buy the smallest size, not a three-month bundle.
- Time-limited. Pick one symptom to track and a set trial period, then stop if nothing changes.
- Never a replacement. ODS advises checking with a provider before using a supplement to treat a condition or in place of a prescribed medicine2.
- Out in the open. Tell your clinician and pharmacist what you take.
Questions readers ask
Does “not enough evidence” mean a product doesn’t work?
No. It means no one has shown it works or fails in good human trials. Some products get tested and fail, as ginkgo did for dementia14. Others hold up. The rating to take most seriously is “evidence says no, or unsafe.”
If something works for me, isn’t that evidence?
Your experience matters to you. But on its own it can’t separate the product from the placebo effect, natural ups and downs, or other changes you made. The FTC notes that improvement in a study can come from placebo effects, time or chance, which is why trials need a comparison group4.
Why don’t supplement companies just run big trials?
Trials are expensive, and the law doesn’t require them before a supplement goes on sale2. Industry groups argue that randomized trials shouldn’t be required for every claim5. The FTC’s position is that health claims generally need them4.
Are drugs held to a higher standard?
Generally, yes. Prescription and over-the-counter drugs need FDA approval before sale, and supplements don’t2. Drug studies can rely on surrogate measures too, though, so the same questions about outcomes apply to both9.
The bottom line
Most health products with thin evidence aren’t scams. They’re bets placed on early research, an ingredient study or tradition, and they’re sold with words that sound more certain than the science.
You can close that gap in about 10 minutes. Ask what was tested, in whom, at what dose and for what outcome, then check one independent summary. If the evidence is thin and the risk is low, you can still choose to try it, as a short, open experiment and never in place of care that’s proven to work.
Related reading
About this article
This article was researched using official and peer-reviewed sources, including federal reports, NIH fact sheets, FTC guidance and a large review in a medical journal, all listed below. It was reviewed in October 2026. It is general education, not medical, legal or financial advice.
Sources
- HHS Office of Inspector General. “Dietary Supplements: Structure/Function Claims Fail To Meet Federal Requirements” (OEI-01-11-00210). U.S. Department of Health and Human Services, 2012. oig.hhs.gov
- NIH Office of Dietary Supplements. “Dietary Supplements: What You Need To Know.” National Institutes of Health, 2023. ods.od.nih.gov
- Khan SU, Khan MU, Riaz H, et al. “Effects of Nutritional Supplements and Dietary Interventions on Cardiovascular Outcomes: An Umbrella Review and Evidence Map.” Annals of Internal Medicine, 2019. pubmed.ncbi.nlm.nih.gov
- Federal Trade Commission. “Health Products Compliance Guidance.” FTC, 2022. ftc.gov
- Nutritional Outlook. “FTC’s high bar for substantiation is inconsistent with existing regulation, say recent citizen petitions from industry.” 2023. nutritionaloutlook.com
- National Center for Complementary and Integrative Health. “How To Make Sense of a Scientific Journal Article: Size of the Study.” National Institutes of Health, accessed 2026. nccih.nih.gov
- NutraIngredients. “Alice Mushrooms case tests limits of ingredient substantiation.” 2026. nutraingredients.com
- NIH Office of Dietary Supplements. “Ashwagandha: Fact Sheet for Health Professionals.” National Institutes of Health, 2025. ods.od.nih.gov
- U.S. Food and Drug Administration. “Surrogate Endpoint Resources for Drug and Biologic Development.” FDA, accessed 2026. fda.gov
- NIH Office of Dietary Supplements. “Vitamin A and Carotenoids: Fact Sheet for Health Professionals.” National Institutes of Health, 2025. ods.od.nih.gov
- NIH Office of Dietary Supplements. “Vitamin C: Fact Sheet for Health Professionals.” National Institutes of Health, 2025. ods.od.nih.gov
- National Center for Complementary and Integrative Health. “Turmeric.” National Institutes of Health, 2025. nccih.nih.gov
- National Center for Complementary and Integrative Health. “Echinacea.” National Institutes of Health, 2024. nccih.nih.gov
- National Center for Complementary and Integrative Health. “Natural Doesn’t Necessarily Mean Safer or Better.” National Institutes of Health, accessed 2026. nccih.nih.gov
- National Library of Medicine. “Herbs and Supplements.” MedlinePlus, accessed 2026. medlineplus.gov