Dietary supplement statistics
Reference, maintained weekly from official sources
Recalls and adverse event reports, counted from the US Food and Drug Administration’s own data. Nothing here is an estimate, a projection, or a number taken from another publisher. Quote any of it with credit.
How many dietary supplement recalls has the FDA recorded?
585 enforcement reports in total. 2017 was the heaviest full year in this window with 114, and 2026 is still running.
| Year | All recalls | Class I |
|---|---|---|
| 2016 | 9 | |
| 2017 | 15 | |
| 2018 | 7 | |
| 2019 | 1 | |
| 2020 | 4 | |
| 2021 | 16 | |
| 2022 | 1 | |
| 2023 | 1 | |
| 2024 | 2 | |
| 2025 | 2 | |
| 2026 (part year) | 9 |
Are the most serious supplement recalls increasing?
2026 has recorded 9 Class I supplement recalls, more than 2022 to 2025 combined (6). And 2026 is not a complete year.
Class I is the FDA’s most serious category, used where exposure to the product is judged likely to cause serious harm or death. For all severities together, 38 recalls in 2025 was the highest annual count since 2017, when there were 114.
How serious are supplement recalls, by class?
| Classification | Reports | Share |
|---|---|---|
| Class II | 61.2% | |
| Class I | 20% | |
| Class III | 18.8% |
Which states do recalled supplements come from?
Florida accounts for 137 of 585 reports, 23.4%, well ahead of California. This is the recalling firm’s location, not where the product was sold.
| State | Reports | Share |
|---|---|---|
| FL | 23.4% | |
| CA | 13.2% | |
| NY | 7.7% | |
| PA | 6.8% | |
| GA | 5% | |
| UT | 4.8% | |
| TX | 4.1% | |
| CO | 3.9% | |
| ME | 3.6% | |
| OR | 3.4% |
Do firms recall supplements voluntarily, or does the FDA make them?
504 of 585 reports, 86.2%, were initiated by the firm rather than mandated by the FDA.
How many adverse events are reported for supplements?
Read this before quoting any number in this section. These are voluntary reports to the FDA. A report means somebody told the FDA that a problem happened after using a product. It does not mean the product caused it, and the FDA does not verify causation. Anyone writing “supplements caused X deaths” from this data has got it wrong, and so have we if we ever phrase it that way.
With that said: 54,906 reports sit in FDA’s CAERS database under its supplement industry category. 13,025 of them record a hospitalisation as an outcome and 1,327 record a death. One report can carry several outcomes, so these do not sum to the total.
| Reported outcome | Reports |
|---|---|
| Other Serious or Important Medical Event | |
| Hospitalization | |
| Visited a Health Care Provider | |
| Visited Emergency Room | |
| Other Serious Outcome | |
| Life Threatening | |
| Other Outcome | |
| Disability | |
| Death | |
| Required Intervention | |
| Allergic Reaction | |
| Injury | |
| Congenital Anomaly |
What do people actually report?
Choking is the single most reported reaction, ahead of every systemic symptom. That points at the physical form of the product, the gummies, the horse-sized tablets, rather than at anything pharmacological.
| Reaction | Reports |
|---|---|
| Choking | |
| Nausea | |
| Vomiting | |
| Diarrhoea | |
| Dyspnoea | |
| Dizziness | |
| Headache | |
| Hypersensitivity | |
| Abdominal Pain Upper | |
| Dysphagia |
How fast are reports about the weight-loss drugs rising?
Reports, not confirmed harms. Volume tracks how widely a drug is used and how much attention it gets, so a rise here is mostly a rise in use. It does not mean the drug caused what was reported. Semaglutide reports rose as semaglutide prescriptions rose. Read these as a measure of how much a drug is being taken and talked about, not of how dangerous it is.
Reports to the FDA naming semaglutide, the drug in Ozempic and Wegovy, went from 1,595 in 2019 to 20,013 in 2025.
| Drug | Known as | Total reports |
|---|---|---|
| Semaglutide | Ozempic, Wegovy | |
| Tirzepatide | Mounjaro, Zepbound | |
| Metformin | type 2 diabetes | |
| Finasteride | hair loss | |
| Minoxidil | hair loss |
Semaglutide reports by year
| Year | Reports |
|---|---|
| 2019 | |
| 2020 | |
| 2021 | |
| 2022 | |
| 2023 | |
| 2024 | |
| 2025 |
Are teenagers using steroids more, and are they sleeping less?
Steroid use among US high school students has fallen, not risen. It peaked at 6.1% in 2003 and was 2.9% in 2017. Sleep went the other way: 68.9% of students were not getting eight hours in 2007, rising to 74.6% by 2017, which is close to three in four.
These are CDC’s most recent figures in this dataset, and that is 2017. The survey has run since, but those results are not in this release. Say the year when you quote them, and do not call them current.
| Year | Percent | 95% CI | Sample |
|---|---|---|---|
| 2003 | 4.7 to 7.8 | 15,044 | |
| 2005 | 3.5 to 4.6 | 13,822 | |
| 2007 | 3.4 to 4.6 | 13,844 | |
| 2009 | 2.9 to 3.8 | 16,320 | |
| 2011 | 3.2 to 4.1 | 14,824 | |
| 2013 | 2.7 to 3.6 | 13,454 | |
| 2015 | 2.8 to 4.3 | 15,165 | |
| 2017 | 2.5 to 3.3 | 12,068 |
| Year | Percent | 95% CI | Sample |
|---|---|---|---|
| 2007 | 66.9 to 70.9 | 12,154 | |
| 2009 | 67.5 to 70.7 | 14,782 | |
| 2011 | 67.3 to 69.9 | 12,198 | |
| 2013 | 66.8 to 69.8 | 12,335 | |
| 2015 | 70.4 to 74.9 | 14,534 | |
| 2017 | 73.1 to 76.0 | 11,811 |
Method, and what these numbers are not
Recalls. openFDA food enforcement reports (US Food and Drug Administration), openFDA, dataset updated 2026-10-07. Query: product_description:"dietary supplement". Counts enforcement reports whose product description contains the phrase "dietary supplement". FDA publishes no supplement-only category in this dataset, so this is a phrase match rather than an official category total.
Adverse events. CAERS, the FDA Adverse Event Reporting System for foods and dietary supplements, openFDA, dataset updated 2026-07-07. Query: products.industry_name:"Vit/Min/Prot/Unconv Diet(Human/Animal)". These are voluntary reports. A report means somebody told the FDA that a problem happened after using a product. It does not mean the product caused it, and FDA does not verify causation. Never write that supplements caused these outcomes.
Drug reports. FAERS, the FDA Adverse Event Reporting System for drugs, openFDA, dataset updated 2026-07-30. Reports, not confirmed harms. Volume tracks how widely a drug is used and how much attention it gets, so a rise here is mostly a rise in use. It does not mean the drug caused what was reported.
Teenagers. CDC Youth Risk Behavior Survey, US high school students, national totals, CDC. National totals only, with sex, race and grade all set to Total. The most recent year in this dataset is the one shown; CDC has run the survey since, but those results are not in this release.
A recall is one enforcement report, not one unit, one brand or one illness. The latest year is partial, so do not set it against a complete year without saying so.
What we deliberately leave out. We do not publish per-ingredient product counts from the NIH Dietary Supplement Label Database. Its search returns counts that move by roughly a tenth depending on the query term used alongside the ingredient, and on that basis about a third of all on-market labels appear to contain collagen, which cannot be right. We would rather carry fewer numbers than one we cannot defend.
Every figure is regenerated from the source by a script in our repository and refreshed weekly, because openFDA republishes roughly weekly.
Questions people ask about these numbers
How many dietary supplement recalls are there each year?
The FDA has published 585 enforcement reports for products described as dietary supplements since 2012. The count swings a lot year to year, from 10 to 114, because one firm recalling many products can lift a whole year. Read the trend across several years, not one.
What share of supplement recalls are the most serious kind?
117 of 585 enforcement reports are Class I, about 20 percent. Class I means the FDA judged there was a reasonable probability of serious harm or death. Class II, the largest group, means temporary or medically reversible harm. The class is the FDA's own judgement, recorded at the time.
How many adverse event reports does the FDA receive about supplements?
54,906 reports naming a dietary supplement product sit in the FDA's CAERS database. The most reported reaction is choking, with 6,203 reports. A report is somebody telling the FDA that something happened after taking a product. It is not a finding that the product caused it.
Does a recall mean a supplement is dangerous?
Not on its own. A recall means a product was removed from sale, and the reasons range from an undeclared allergen or a labelling error to genuine contamination. The FDA's class tells you how serious it judged the risk. Most entries in this data are Class II, meaning temporary or reversible harm, not the most serious category.
Which drugs generate the most adverse event reports?
Of the medicines tracked on this page, Metformin has the most at 439,165 reports. Report volume follows how many people take a drug as much as how risky it is, so a widely prescribed medicine will generate more reports than a rare one that is more dangerous. These counts cannot be compared without that in mind.
Where does this data come from, and can I quote it?
Every figure comes from the US Food and Drug Administration's own openFDA databases and from the CDC, with the dataset date shown beside it. Nothing here is modelled, estimated or bought from a third party. Quote any of it with credit to Health Trend Wire and a link to this page.
Why do these numbers differ from figures I have seen elsewhere?
Usually because of what is being counted. These figures count enforcement reports and adverse event reports whose text matches a specific product description, over a stated period. A different phrase, a different date range or a different database gives a different number. The exact query behind each table is stated in the method section.
Citing this page
Free to quote with credit to Health Trend Wire and a link to this page. If you need a cut we do not show, or you think a figure is wrong, write to us and we will check it against the source and correct it in the open. Our method and corrections policy explains how.