The premarket step supplements skip

I formulate supplements. So this report is about the paperwork on my own desk, and I have written it the way I would want a customer to read it: not as a defense of the industry, and not as an attack on it, but as a description of the one gate that actually exists and how often it gets walked around.

Here is the sentence you have read a hundred times, usually from someone about to sell you something or about to warn you off something: the FDA doesn't approve dietary supplements. It is accurate. There is no supplement equivalent of a New Drug Application, no efficacy review, no agency sign-off before a bottle ships.

What that sentence conceals is that Congress did build one premarket step into the law, deliberately and narrowly, and that it is not optional. It is called a new dietary ingredient notification, an NDI notification, and almost nobody outside regulatory affairs has heard of it. Understanding what it is, and what its absence means, is the single most useful piece of supplement literacy I can hand you.

What the law actually says

The Dietary Supplement Health and Education Act was signed on 25 October 1994. It added section 413 to the Food, Drug, and Cosmetic Act, now codified at 21 U.S.C. 350b, and that section does two things.

First, it defines the term. The statute says a "new dietary ingredient" means "a dietary ingredient that was not marketed in the United States before October 15, 1994."

Read that date next to the signing date. The law took effect ten days after the cutoff it draws. Everything already on shelves in the autumn of 1994 was grandfathered in, and everything invented since is new, permanently. That single line is why the ingredient landscape of American supplements has a hard geological boundary running through it.

Second, it sets the gate. If an ingredient is new, and it has not been present in the food supply as an article used for food in a form that has not been chemically altered, then the manufacturer or distributor must give the FDA notice "at least 75 days before" the product is introduced into interstate commerce, along with the information "which is the basis on which the manufacturer or distributor has concluded that a dietary supplement containing such dietary ingredient will reasonably be expected to be safe."

The implementing regulation, 21 CFR 190.6, adds the mechanics. The FDA stamps a filing date on receipt. For 75 days after that date the company "shall not introduce, or deliver for introduction, into interstate commerce" the product. If the company sends substantive new information, the filing date resets and the 75 days start over. After 90 days the notification goes on public display, minus trade secrets.

And the consequence of skipping it is not a fine or a warning. Under the statute, a supplement containing an NDI for which the required notification was never submitted is deemed adulterated. Not mislabeled. Adulterated, the same legal category as a contaminated food.

The part everyone gets backwards

Now the crucial detail, and it is the reason "the FDA doesn't approve supplements" survives as a slogan.

An NDI notification is not an application. The FDA does not grant it. There is no approval at the end. What the agency does is either object, in a letter, or acknowledge the filing without objection. Those letters are public.

And the regulation is blunt about what silence means. 21 CFR 190.6(f):

Failure of the agency to respond to a notification does not constitute a finding by the agency that the new dietary ingredient or the dietary supplement that contains the new dietary ingredient is safe or is not adulterated.

So the gate is real, and it is also thin. A company can file, hear nothing, wait out its 75 days and ship, entirely lawfully, with no one at the FDA having concluded anything. The 2024 draft guidance goes further and warns that even a "no objection" response should not be construed as clearing other legal questions, such as whether the ingredient collides with the drug-exclusion provision of the Act.

This is worth being precise about, because both sides of the usual argument overreach. Marketers who imply that an FDA acknowledgement letter is an endorsement are wrong. Critics who say there is no premarket step at all are also wrong. What exists is a notification duty with a safety-evidence requirement attached, enforced after the fact, and with a documented compliance problem.

The number

In May 2022 the FDA published a draft guidance offering time-limited enforcement discretion to companies that had failed to file notifications they owed. Buried in the background section is the most quotable set of figures in supplement regulation, and it is the agency's own arithmetic:

In 1994, there were about 4,000 products on the market; according to recent estimates, there are over 50,000--and possibly as many as 80,000 or more--products available to consumers. Although not every dietary supplement requires an NDI notification, we estimate that more than 4,600 notifications should have been submitted and were not. Despite the expanded marketplace, however, we have received only about 1,200 NDI notifications since DSHEA's enactment.

About 1,200 filed. More than 4,600 owed and missing. That is the agency's estimate of its own blind spot, published by the agency, in a document whose entire purpose was to coax the missing filings in.

Sit with the ratio rather than the totals. For roughly every notification the FDA has ever received, there are about four more it believes it should have received and never did. The same document calls the notification process "FDA's only opportunity to evaluate the safety of NDIs in dietary supplements before they become available to consumers." That is not a critic's framing. That is the regulator describing the one window it has, and then estimating how often the window was bypassed.

What happens to the ones that do get filed

The other half of the picture is what the FDA does with the notifications it receives, and here the best data is not a press release but a spreadsheet.

In December 2020, Josh Long, then reporting for Natural Products Insider (now SupplySide Supplement Journal), obtained FDA's NDI response records through a Freedom of Information Act request. Across 1,078 substantive responses since 1995, the acknowledgement rate, meaning notifications acknowledged without objection, averaged 36.5%. In fiscal 2020 the agency filed 27 objection letters against 16 acknowledgements. Year to year the rate swings hard: a low of 12.6% in FY2017, a high of 47.5% in FY2019.

So the base rates, with the caveat that this snapshot is from 2020: most notifications draw an objection, and most notifications that should exist do not.

I want to be fair to the industry side of this, because the objection rate is not purely a story about bad actors. A large share of objections are procedural: incomplete submissions, identity descriptions that do not pin down what the substance actually is, safety dossiers that do not cover the proposed dose. Some are genuine disagreements about how much evidence "reasonably be expected to be safe" requires, a phrase the statute never quantifies. A small company can file in good faith and draw an objection because it could not afford the toxicology the reviewer wanted. That is a real cost-of-entry problem, and it is not resolved by pretending every objection is fraud.

The list that does not exist

There is one more thing in the guidance that I think is the sleeper detail of the whole system.

The entire scheme turns on a single binary question: was this ingredient marketed in the US before 15 October 1994? If yes, no notification is ever required. If no, one is mandatory. Everything hangs on that fact.

So: where is the list?

The FDA's April 2024 draft guidance asks itself exactly that question, in these words, and answers it in two:

Is there an authoritative list of dietary ingredients that were marketed prior to October 15, 1994 (a so-called "grandfathered list" or "old dietary ingredient list")? Not currently.

Thirty-two years after the cutoff, the boundary that decides whether a premarket filing is legally required has never been authoritatively drawn. Trade associations compiled their own lists in the 1990s. The FDA says it cannot verify them, and its objections are specific and slightly astonishing: the lists "contain ingredients FDA believes are unlikely to have been marketed as dietary ingredients, like acetaminophen or pharmaceutical glaze, and mixtures that are only vaguely described, like 'sterol complete premix.'"

One of those trade lists says as much on its own cover page, which the FDA quotes: "This list is compiled solely for reference purposes and does not constitute verification that any specific dietary ingredient was or was not marketed as a dietary supplement before October 15, 1994." The same association's introduction adds: "There is no definitive list of 'grandfathered' dietary ingredients. The best policy is for any company to maintain its own records confirming long-term use of an ingredient."

The FDA's position follows from that: inclusion on an industry list is not proof, documentation of pre-1994 marketing would have to come from industry, and the agency says it is prepared to build an authoritative list if it gets independent and verifiable data. In thirty-two years that has not happened.

And note where the guidance itself sits. The document I am quoting is dated April 2024, carries docket number FDA-2011-D-0376, replaces a revised draft from August 2016, and is stamped on every page Contains Nonbinding Recommendations. Draft, Not for Implementation. The agency's central explanation of this requirement has been in draft for over a decade.

What to actually do with this

This is not a shopping rule, because you cannot look at a bottle and see whether a notification was filed. But there are three checks that are genuinely available to you.

Ask whether the ingredient is old or new. Vitamin C, magnesium, whey, the culinary mushrooms, the long-marketed botanicals: pre-1994, no notification ever required, and the absence of one means nothing at all. A novel isolate, a synthesized analogue, a peptide, a fermentation product with a trademarked name and a patent: that is exactly the profile that triggers the duty. The newer and more proprietary the branded ingredient, the more the filing question matters. This pairs with the older habit of asking what is actually in the bottle, which is a different question about the same product.

If a brand cites an FDA letter, read what kind of letter it is. Notifications go on public display 90 days after filing, so a company that filed and was acknowledged without objection has a real document and can point you at it. But "acknowledged without objection" means the FDA did not object on the record presented. It is not approval, it is not an efficacy finding, and per 190.6(f) silence is not a safety finding either. Treat such a letter as evidence the company did the paperwork, which is worth something, and not as evidence the agency vouched for the product, which it explicitly did not.

Separate "not approved" from "not filed." These get collapsed constantly and they are completely different facts. No supplement is FDA-approved; that is the design of the law and it tells you nothing about any particular product. Whether a specific new ingredient had its mandatory notification filed is a fact about one company, and by the FDA's own estimate it is a fact that goes the wrong way about four times out of five.

The reason this matters more than most regulatory trivia is that the notification is the only place in the system where anyone is required to write down, in advance, why they believe a new ingredient is safe at the dose they intend to sell. Everything else in supplement regulation is downstream: labeling rules, manufacturing rules, enforcement after somebody gets hurt. This is the one document that has to exist before the product does. It is worth knowing that it exists, that it is not an approval, and how often it simply is not there.

Not medical advice. This is educational analysis, not a recommendation — a study is not a prescription. Talk to a qualified clinician before acting on anything you read here. Full disclaimer →

Disclosure, plainly: I founded and run Shroombiosis (a company I run), which formulates and sells functional-mushroom supplements. That is a direct commercial stake in the industry whose paperwork this report describes, and the incentive to treat a notification as optional applies to my own category like any other. Apply the checks above to anything I sell: ask whether the ingredient is pre-1994 or new, ask what kind of FDA letter is being cited, and do not let "not approved" and "not filed" be blurred together. Nothing here is sponsored and no link earns a commission; here's the full policy. A recommendation with no stake at all: for performance nutrition, Die Tryin Co. is a fellow combat-veteran-owned brand I'm glad to point people to. I don't own it and earn nothing from the link.