Where health claims get challenged

A dietary supplement rarely gets a warning letter because it is secretly bad. It gets one because a single line on a label, an ad, or a product page said something the law does not allow a supplement to say — and nobody caught it before the packaging ran. The FDA and FTC framework for supplement marketing claims (the structure/function rules in 21 CFR 101.93 and the deception standard the FTC applies to health claims) is unusually clear about where this goes wrong. The problems cluster into four places, and each one is a place where an honest brand and a misleading one can produce the very same sentence.

1. The disease claim nobody meant to make

This is the biggest and most common one, and it is a boundary problem, not a wording problem. A supplement may lawfully make a structure/function statement — "supports immune function," "promotes healthy digestion," "helps maintain joint comfort." It may not claim to diagnose, treat, cure, mitigate, or prevent a disease. The moment copy says it "cures anxiety," "prevents diabetes," "fights cancer," or "reverses Alzheimer's," it has stepped out of the supplement lane and into making a drug claim — which makes the product an unapproved new drug, full stop. No disclaimer fixes it; the required 101.93 statement that the product is "not intended to diagnose, treat, cure, or prevent any disease" cannot sit next to a sentence that does exactly that. The trap is that brands drift here casually: a founder describing what the ingredient "helps with" names a condition instead of a function. A mechanical pre-check should catch this first and score it highest, because it is the pattern that turns a label into a letter.

2. The structure/function claim with no disclaimer

Once you are on the right side of the boundary, 21 CFR 101.93(c) imposes a positive duty: every structure/function statement on the label must be accompanied by a specific disclaimer — that the statement has not been evaluated by the Food and Drug Administration, and that the product is not intended to diagnose, treat, cure, or prevent any disease — and the claim must be truthful and not misleading. This is the most fixable failure: the claim itself is lawful, the wording is modest, and the only gap is a sentence that has to be physically present next to it. A checker that recognises a structure/function verb acting on a body system ("supports," "promotes," "helps" + immune/digestion/joint/sleep …) and then looks for the disclaimer language is checking something deterministic and high-value. Note the surface matters: a claim baked into a brand or product name has almost nowhere to carry the disclaimer, so it is scrutinised harder than the same claim in website body copy.

3. "Proven" with nothing behind the word

"Clinically proven," "studies show," "doctor recommended," "scientifically proven to reduce…" — these are read by the FTC as objective substantiation claims, and the standard is competent and reliable scientific evidence that you must already have in hand before the claim runs. The rule is not "don't say proven"; it is "if you say it, have the study, and don't overstate what the study showed." A pre-check cannot see your study binder — that lives outside the text — but it can flag that you asserted proof without naming a basis, which is exactly the moment a marketer reaches for the evidence or, better, narrows the claim to what the evidence supports. When an actual trial is cited, the wording can be fine; when it is bare, it is a gap.

4. The absolutes and the implied FDA stamp

"No side effects," "100% safe," "miracle cure," "guaranteed results" are absolute health claims no responsible supplement can substantiate across every user, and the agencies treat them as red flags. They also often travel with a subtler deception: implying the FDA approved or certified the product. Dietary supplements are not approved by the FDA before marketing; "FDA approved" or "FDA certified" on a bottle misrepresents the regulatory posture and is a classic enforcement target. ("FDA-registered facility," meaning only that a facility registered, is a different statement and must never be phrased as product approval.) Naming nothing that the law does not allow, and removing the word "approved," is the whole fix.

Why the published rules alone do not protect you

The rules genuinely are public — 21 CFR 101.93 is readable today, and the nutrient-content rules in 21 CFR 101.13 and the labelling framework in 101.9 sit alongside it. What publication does not do is fit inside a copywriter's head at the moment a label has to ship. The translation layer — "does this sentence name a disease, carry the disclaimer, or assert proof without a basis?" — is where teams fail, and it is exactly the layer a mechanical check owns.

What a pre-check should report — and not

A good tool reports the wording pattern, the rule it maps to, and a concrete edit, ranked so the disease-boundary and implied-approval cases come first. A dishonest tool claims your label is now "FDA compliant." No text checker can say that, because whether the trial is real, whether a registration exists, and whether the evidence supports the claim all live outside the wording. The honest boundary is not a disclaimer bolted on at the end; it is the whole reason the check is worth trusting.

Heuristic pre-check only — not a health-claims certification or legal advice. No clinical-trial, lab, facility-registration, or FDA-approval verification and no enforcement prediction is performed by any wording-only checker, including the one that links to this article. For a binding view on a specific label, consult 21 CFR 101.93 and the FTC health-claim framework and a qualified professional.