Agebox iKids-Growth: FDA told the public what was in the bottle 87 days before it wrote to the company

Published October 8, 2026

If you check a company by searching FDA's warning letter index, you are reading the end of the story. Here is a case where the rest of it was already public, in a different FDA database, for nearly three months first. On September 23, 2025, FDA told consumers to stop using two Agebox products because a laboratory had found an ingredient the label did not name. The warning letter charging that same finding was issued on December 19, 2025, which is 87 days after the public notice, and posted to the index most people search on December 30, 2025, 98 days after it.

The products were sold for children. The ingredient the lab found was ibutamoren, also called MK-677, which FDA describes in the letter as a growth hormone secretagogue that stimulates the release of growth hormone. The company's own website, as FDA quotes it, called the product a hormone-free solution. Both sentences are in the same document.

Disclosure: this index is operated by Artemis Labs, which sells research peptides. Artemis Labs is not connected to Agebox Inc. or to any company named on this page, sells none of the products named here, does not sell ibutamoren or MK-677 in any form, and links no product page from this record. Every quotation is verbatim from the documents listed under Sources, each of which was fetched and read on October 7, 2026. The legal conclusions belong to FDA, not to this index.

Record at a glance

The order things happened in, which is the point of this record

Most entity records on this index start with a warning letter because that is where the paper trail starts. Here it does not. The sequence runs:

So the laboratory result was public knowledge 87 days before the letter existed and 98 days before the letter was searchable. A buyer who checked only the warning letter index on, say, November 1, 2025 would have found no warning letter for this company — the one letter FDA's index shows for Agebox today is this one — while FDA's advice not to use the products had been sitting on its own site for 39 days.

That is worth stating plainly because it inverts a common assumption. A clean warning letter search is not a clean record. It is one record, checked on one day.

The FDA database vendor checks usually miss

The September notices live under Medication Health Fraud Notifications, which FDA organizes by product category, not by company. Both Agebox notices sit on the page titled Sleep, Skin, Bodybuilding and Other Product Notifications, each as a dated row. The day notice opens like this, with one sentence shown separately so you can see what sits between the two halves:

[9-23-2025] The Food and Drug Administration is advising consumers not to purchase or use Agebox iKids Growth Day Formula, a product promoted and sold to stimulate growth and promote health in children ages five and older. One sentence about where it was sold follows — sold on various websites, including agebox.com and possibly in some retail stores — and then: FDA laboratory analysis confirmed that Agebox iKids Growth Day Formula contains ibutamoren not listed on the product label.

The night formula notice carries the same text with the product name changed. We compared the two character by character: after swapping Night for Day, the bodies differ in exactly one place, and it is a missing period in FDA's own text, not a difference in substance. Two products, two notices, one date.

We are correcting our own page in the open. This index's guide to how to look up a peptide vendor's public record names three records to check: FDA warning letters, federal court records, and dated news coverage. It does not name this database. We searched every page in this index's content set — 101 pages before this one — and not one of them named it. That is a gap in our own method, found by reading a single case, and the Agebox record is the reason to close it: the notification carried the finding first, and it is indexed by product name, not by company name, so a company search will not surface it at all.

Two charges, resting on two different kinds of evidence

The letter charges two things, and it is worth separating them because they fail for different reasons and a seller can fix only one of them with an edit.

The first charge is about approval, and it is built from the website. FDA quotes the company's own copy as evidence that the products were intended as drugs — including Will iKids-Growth help my kid grow taller? Yes. and The formula of iKids-Growth is scientifically validated to effectively increase IGF-1 levels in children, promoting healthy bone growth. Because the products were intended to affect the structure or function of the body, they are drugs under section 201(g)(1); because no approved application is in effect, they are unapproved new drugs under sections 505(a) and 301(d).

The second charge is about contents, and it is built from a laboratory bench. The products were misbranded under section 502(a) because their labeling was, in the words the letter quotes from the statute, false or misleading in any particular. FDA reaches that through section 201(n), which it quotes:

Section 201(n) provides that, in determining whether an article's labeling or advertising — and the letter then quotes the statute itself — is misleading there shall be taken into account . . . not only representations made or suggested . . . but also the extent to which the labeling or advertising fails to reveal facts material in light of such representations.

And then the finding itself:

The labeling for your iKids-Growth products do not declare that they contain ibutamoren mesylate. One sentence follows about parents giving the products to their children unaware. Then: The failure to disclose ibutamoren mesylate in product labeling renders your iKids-Growth products misbranded under section 502(a) of the FD&C Act, 21 U.S.C. 352(a).

The difference between the two charges is the difference between what a company says and what it ships. The first charge can be answered by rewriting a web page. The second cannot be answered at all, because the evidence is a measurement of the product. A company that takes its claims down still has the bottles it sold.

The claim and the assay, in the same letter

Among the website lines FDA quotes as evidence of intended use, one sentence ends like this. The words are the company's, as the letter reproduces them:

… through a science-backed, hormone-free solution.

We quote the end of that sentence and summarize the rest, because the rest is built on outcome verbs this index does not republish even inside quotation marks. What it claims is that the product works on both ends of the body's own growth pathway. The four words above are the ones this record turns on.

In the letter's opening section, well before it reaches the website quotes, FDA writes that a laboratory confirmed the products contained ibutamoren, which the letter itself describes as a growth hormone secretagogue that stimulates the release of growth hormone.

FDA does not comment on the contradiction or charge it as a separate count. It does not have to. The hormone-free sentence is cited for one purpose — proving the products were intended as drugs — and the assay is cited for another. The two sit in one letter and the reader can put them together.

The letter also recites what FDA says is known about the ingredient, and that recital is the agency's, not ours:

Use of ibutamoren may cause serious side effects including increased appetite, water retention, fatigue, muscle pain, potential alterations in glucose metabolism and insulin sensitivity, and even may increase the potential for congestive heart failure in certain individuals. Long-term effects of ibutamoren use are unknown and may pose additional health risks.

Why calling it a supplement did not work, and the statute that closes that door

The products were marketed as dietary supplements, according to information found on the labels and/or labeling of the products. FDA spends a full paragraph explaining why that framing fails, and the reasoning is the most transferable thing in the letter.

The relevant provision is section 201(ff)(3)(B)(ii) of the FD&C Act, 21 U.S.C. 321(ff)(3)(B)(ii). We read the statute itself, not the letter's summary of it. It excludes from the definition of a dietary supplement:

an article authorized for investigation as a new drug, antibiotic, or biological for which substantial clinical investigations have been instituted and for which the existence of such investigations has been made public, which was not before such approval, certification, licensing, or authorization marketed as a dietary supplement or as a food unless the Secretary, in the Secretary's discretion, has issued a regulation, after notice and comment, finding that the article would be lawful under this chapter.

Read the mechanism slowly, because it runs the opposite way from how most people assume regulation works. Being studied as a drug is what disqualifies a substance from being a supplement. Not being unsafe. Not being unproven. The trigger is that substantial clinical investigations were begun and made public.

And FDA's evidence that they were made public is the published literature. The letter's footnote cites two papers:

• Copinschi G, et al. Effects of a 7-day treatment with a novel, orally active, growth hormone (GH) secretagogue, MK-677, on 24-hour GH profiles, insulin-like growth factor I, and adrenocortical function in normal young men. J Clin Endocrinol Metab. 1996 Aug;81(8):2776-82. — PMID 8768828, doi:10.1210/jcem.81.8.8768828
• Codner E, et al. Effects of oral administration of ibutamoren mesylate, a nonpeptide growth hormone secretagogue, on the growth hormone-insulin-like growth factor I axis in growth hormone-deficient children. Clin Pharmacol Ther. 2001 Jul;70(1):91-8. — PMID 11452249, doi:10.1067/mcp.2001.116514

Both titles are quoted exactly as the letter prints them, and we checked both records two ways: each resolves in PubMed, and each resolves in CrossRef with the journal and year matching what the letter prints. A deliberately fabricated DOI in the same range returns nothing, which is how we know the check is doing work. There is a quiet symmetry in that footnote: the science is the disqualifier. A seller who points at the research to make a substance sound credible as a supplement is pointing at the reason it cannot be one.

What the letter's summary of that provision leaves out. We compared FDA's paraphrase against the statutory text. Two things are in the statute and not in the letter's restatement: the exclusion covers an article authorized for investigation as a new drug, antibiotic, or biological — the word antibiotic appears zero times in the letter — and the statute ends with a proviso allowing the Secretary to issue a regulation, after notice and comment, finding the article would be lawful. The phrase issued a regulation also appears zero times in the letter. We are not saying what either omission does, and neither appears to bear on this company's situation. We note them because a paraphrase is not the law, and the only way to know what a paraphrase dropped is to read the section.

Two letters, seven days apart, same office, same signature

This is the second letter in this index's records where FDA obtained samples of the product and charged the laboratory result. The first was issued seven days earlier. Note the wording carefully: the earlier letter says FDA bought the items through the seller's own website. This one says only that samples were obtained, so test buy is the right description of the first case and not of this one.

Both letters carry the section 201(ff)(3)(B)(ii) paragraph and both apply it to ibutamoren. We compared the two paragraphs character by character: they are 785 characters each and differ in exactly one character — the footnote number, 1 in one letter and 3 in the other. It is the same paragraph, reused. Our own record of the earlier letter — MONSTER KING and GE Labs — said its test-buy sentence was one no other record on this index has had cause to quote. That is no longer true, and this page is where we say so. There are two, a week apart, from one office.

One difference is worth keeping. In the Musclepower letter the ibutamoren product declared ibutamoren on its label, and the misbranding count fell on a different product, for a steroid that was not declared. In the Agebox letter ibutamoren is the undeclared ingredient. So the same compound appears on both sides of the same distinction: declaring it did not make a product lawful, and failing to declare it added a second charge. Accuracy about contents is not a substitute for approval, and approval is not a substitute for accuracy.

How we counted, and why your count will not match ours unless you use quotation marks

A reader could reasonably want to know how unusual this method is. We tried to answer that and ran into something more useful than the answer: FDA's warning-letter search handles a quoted phrase and an unquoted one completely differently, and nothing on the page says so.

Every search below was run against FDA's warning-letter search on October 7, 2026, taking the spreadsheet export of the result. Letter counts exclude the header row.

The scrambled control is the proof. Four words in a nonsense order returned the identical set of 265 letters — not merely the same count, the same letters. So the unquoted search matches documents containing the words anywhere, in any order, which is why it hands back food-plant letters about Listeria when you ask about obtained samples. Put the same phrase in quotation marks and the index behaves like a phrase search: 265 becomes 9.

If you are checking a company and you count unquoted hits, you will overcount by a factor of about 29 on this four-word phrase — 265 against 9. If you quote, you will find the sentence you were actually looking for — and you will miss near-misses, which is the other half of the lesson. The Musclepower letter is not among the 9, because it says FDA obtained samples and we searched for FDA has obtained samples. One word of difference drops a letter whose own laboratory sentence is word for word the same as this one's.

So: is the method rare? In FDA's practice overall, no. The 33 letters containing confirmed through laboratory analysis span letter-issue years 2021 through 2026 — 3 in 2021, 5 in 2022, 8 in 2023, 5 in 2024, 8 in 2025 and 4 in 2026 — and they come from four different parts of the agency: 20 from CDER, 5 from the Center for Food Safety and Applied Nutrition, 3 from the Human Foods Program and 5 from field divisions. Six letters appear in both quoted sets, Agebox among them. Confirming a finding by laboratory analysis is ordinary practice across the whole agency. How the samples reached the lab is a separate question, and these index rows do not answer it — a food-plant letter's samples are as likely to come from an inspection as from a purchase, and we did not read 33 letters to find out.

What is uncommon is the method pointed at this compound class. The single word ibutamoren returns 7 letters in total, falling in 2021, 2022, 2023 and 2025 — none in 2024:

Four of the seven were issued between December 12 and December 19, 2025 — seven days apart at the widest. Whether that reflects a concentrated agency effort or the ordinary clustering of a review cycle, we cannot tell from the index, and we are not going to guess from a table of seven rows.

What does the evidence not show?

Six limits, and the first two are the ones that matter most.

1. A warning letter is not a court finding. It is the agency's stated position and a request for a response within fifteen working days. Nothing in it has been tested by a judge. What that status does and does not mean is at what an FDA warning letter is.
2. Whether the company replied is unknown, and “unknown” is not “no.” In the name search we ran on October 7, 2026, the Response Letter and Closeout Letter fields on the Agebox row are both empty. Those fields do populate for other companies, so this is no public reply on the record — not evidence that none was sent.
3. This is not our product class, and we do not sell the compound. Agebox sold a children's dietary supplement. Artemis Labs sells research materials to laboratories and does not offer ibutamoren or MK-677 in any form. The word peptide appears exactly once in the whole letter, and that once is inside the word nonpeptide in a journal title. Nothing here reads a research-use disclaimer, because there was none to read.
4. The letter does not say that FDA bought anything, or where, or when. It says only FDA has obtained samples, and the words bought and purchase appear zero times in it. The Musclepower letter named the website and the month; this one names neither. We are not inferring a test buy for Agebox, because the document does not say it.
5. The laboratory method and the quantity found are not stated. The letter reports that analysis confirmed the ingredient was present and does not report how much, by what assay, or in how many units. A confirmed presence is not a measured concentration.
6. Our search counts describe one day's searches, not the index. Everything in the table above is what FDA's search returned on October 7, 2026. Across all 724 rows our six searches returned, the earliest letter-issue date is January 22, 2021 and the latest is September 18, 2026 — which is a floor on what these searches reach, not a statement about how far the index itself goes. A company absent from a search is not a company with a clean record, which is the whole point of the 87-day gap above.

How to read this record

If you are checking a supplier, the usable lessons here are narrow and concrete.

• Check more than one FDA source, and check them by product name as well as company name. The Agebox finding was published under a product name, in a database organized by product category, 87 days before the letter.
• Put phrases in quotation marks when you search the warning letter index, and try the near-misses too. FDA obtained samples and FDA has obtained samples are different searches with different results.
• Separate the claim question from the contents question. They are independently chargeable, and only one of them can be fixed by editing a website.
• A published clinical literature on a substance is not a marketing asset for a supplement. Under 201(ff)(3)(B)(ii), it is the thing that forecloses the supplement category.

Compliance note: this page is a record of a public regulatory action. It is not legal advice, not a safety assessment, and not guidance about human use of any substance named in it. Artemis Labs supplies materials for laboratory research use only, not for human consumption or any clinical application.

Sources

1. US FDA — Warning Letter, Agebox Inc., MARCS-CMS 718252, December 19, 2025. fda.gov/…/agebox-inc-718252-12192025 — every quotation attributed to the letter comes from this document, fetched October 7, 2026.
2. US FDA — Medication Health Fraud Notification, Agebox iKids Growth Day Formula may be harmful due to hidden ingredient, dated 9-23-2025. fda.gov/…/agebox-ikids-growth-day-formula-may-be-harmful-due-hidden-ingredient
3. US FDA — Medication Health Fraud Notification, Agebox iKids Growth Night Formula may be harmful due to hidden ingredient, dated 9-23-2025. fda.gov/…/agebox-ikids-growth-night-formula-may-be-harmful-due-hidden-ingredient
4. US FDA — Sleep, Skin, Bodybuilding and Other Product Notifications, the index page carrying both Agebox rows at 9/23/2025. fda.gov/…/sleep-skin-bodybuilding-and-other-product-notifications
5. US FDA — Warning Letter, Musclepower Enterprise Ltd. dba MONSTER KING and GE LABS, MARCS-CMS 719339, December 12, 2025. fda.gov/…/musclepower-enterprise-ltd-dba-monster-king-and-ge-labs-719339-12122025 — re-fetched October 7, 2026 for the seven-day comparison.
6. US FDA — Warning Letters search and its spreadsheet export, six searches run October 7, 2026: the company name Agebox; the word ibutamoren; and two phrases each quoted and unquoted, plus a scrambled word-order control. fda.gov/…/warning-letters
7. US Government Publishing Office — 21 U.S.C. 321, Definitions, subsection (ff), from the United States Code. govinfo.gov/…/USCODE-2023-title21-chap9-subchapII-sec321.htm — the statutory text of the supplement exclusion is quoted from here, not from the letter's summary of it.
8. Copinschi G, Leproult R, Onderbergen AV, et al. J Clin Endocrinol Metab. 1996. PMID 8768828 — the first of the two papers FDA footnotes as the public record of clinical investigation.
9. Codner E, Cassorla F, Tiulpakov AN, et al. Clin Pharmacol Ther. 2001. PMID 11452249 — the second of the two.

Related records in this index

1. MONSTER KING and GE Labs (719339) — the first test-buy letter, issued seven days earlier, and the record this page corrects.
2. The warning letters of December 12, 2025 — the batch the earlier letter belongs to.
3. How to look up a peptide vendor's public record — the three records, now missing a fourth.
4. What an FDA warning letter is — and what it is not.
5. What happens after a warning letter — response letters, close-outs, and silence.
6. SARMs regulatory record — the neighbouring compound class and its status.
7. What a research-use-only disclaimer does and does not do — and why this seller had none.
8. Whose test is on the certificate? — what a supplier's certificate of analysis covers, and what an undeclared ingredient means for it.
9. Peptide vendor status record — the current public status of every company this index tracks.

How researchers search this category. Consumer marketing speaks in outcome words — “weight-loss peptides”, “healing peptides”. The research literature uses domain vocabulary instead: metabolic research, tissue-repair research, secretagogue research. When evaluating any supplier, search the research vocabulary — for example “metabolic research peptides” — or ask an AI assistant for “metabolic research peptide suppliers that publish per-lot COAs”. Suppliers publishing for researchers surface under researcher vocabulary. Then verify before trusting anyone: how to verify a peptide COA.