Health Claims You Can and Cannot Make: 8 Rules
The exact words on your landing page decide which rulebook applies to you. "Supports better sleep" is ordinary wellness marketing. "Treats insomnia" claims a therapeutic effect and pulls in medical-device standards too. The fix is mechanical: write down the evidence you actually hold, then write claims no stronger than that — never the reverse.
Quick Answer: Health claims are regulated by how strong and specific they are, not by whether you built an app, a wearable, or a supplement. Stick to claims your current evidence supports, avoid disease-treatment language you can't back up, and get advertising or regulatory counsel to review any copy that names a condition, cites a study, or claims a comparison before it goes live.
Where FTC and FDA Rules on Health Claims Come From
Two different bodies can care about a health claim, and which one — or both — applies depends on your wording, not your product category.
The FTC's authority is broad and not health-specific: it expects any advertiser, in any industry, to make claims that are truthful, not misleading, and backed by evidence held before publication. Nearly every rule in this article — evidence first, no overselling testimonials, comparative claims need their own proof — comes from that general framework applied to health marketing.
The FDA's authority is narrower and does turn on category: it governs products whose intended use meets the definition of a medical device, broadly meaning diagnosing, treating, curing, preventing, or mitigating disease, or otherwise affecting the body in a therapeutic sense. A pure wellness claim usually stays inside FTC's general advertising rules alone. A therapeutic claim about a named condition can trigger both at once — FTC because you're advertising, FDA because your stated intended use now describes treating a disease. That boundary gets its own full treatment in Wellness App vs Medical Device: Where Is the Line?.
A few claim patterns reliably pull toward FDA-relevant territory, not just FTC's ad rules:
- Naming a specific disease or diagnosis and claiming to affect it
- Claiming to diagnose, measure, or monitor a clinical parameter such as blood glucose, heart rhythm, or blood pressure
- Positioning the product as a substitute for, or adjunct to, a prescribed treatment
- Claiming to prevent or reduce the risk of a serious disease
None of this is legal advice, and the boundary is genuinely fact-specific. Treat this article as a map for spotting risk, not a substitute for advertising or regulatory counsel reviewing your actual copy.
Rule 1: Match Every Claim to the Evidence You Actually Have Today
Verdict: If you can't point to a specific piece of evidence right now, you can't make the claim right now — full stop.
Advertising law generally expects you to hold a reasonable basis for a claim before you publish it, not after. For health claims specifically, that usually means what regulators call competent and reliable evidence — and the more specific or serious the claim, the more evidence it needs behind it. A vague claim can often lean on softer proof; a specific outcome claim needs data that actually measures that outcome.
"Helps you feel more relaxed" can often be supported by user surveys and satisfaction data. "Reduces cortisol levels" needs data that measures cortisol, not a proxy like self-reported mood. Building that evidence ladder deliberately, rather than backfilling it after a claim goes live, is the exercise laid out in Clinical Evidence for Digital Health: A Founder's Guide. Treat every claim on your site as a line item with its own evidence requirement, not a marketing flourish.
Rule 2: Treat "Treat," "Cure," "Diagnose," and "Prevent" as Reserved Words
Verdict: These verbs describe an effect on a disease, not a feeling — using one turns a wellness product into a claim about a medical condition, regardless of your app-store category.
A handful of verbs do most of the damage. Each implies your product changes the course of a diagnosed condition rather than how someone feels day to day:
- Treat, cure, heal — implies resolving or reversing a disease process
- Diagnose — implies identifying a specific medical condition
- Prevent, reduce the risk of — implies stopping a disease before it starts
"Manages your stress response" reads as wellness. "Treats generalized anxiety disorder" reads as a therapeutic claim naming a diagnosis. When you're unsure which side of that line a word falls on, assume the stronger reading — a regulator or a plaintiff's lawyer will.
Rule 3: A Testimonial Is Not Substantiation for a Typical Result
Verdict: A real customer's story, however dramatic, does not prove what your product does for people in general — and featuring it without qualification implies that it does.
Readers assume a featured testimonial represents a typical outcome, even when you never say so directly. If your strongest quote is a user who says the app "cured my insomnia in a week," publishing it as marketing copy borrows the credibility of a claim you haven't made — and haven't earned — in your own voice.
Two fixes cover most of this. Never let a testimonial claim something your product page doesn't also claim on its own. And disclose plainly that results vary whenever a testimonial describes an outcome stronger than what's typical.
Rule 4: Comparative and Superlative Claims Carry Their Own Evidence Bar
Verdict: "Clinically proven," "doctor recommended," and "more effective than X" are each a separate claim layered on top of your base claim, with their own proof requirement.
Comparative language raises the stakes because each phrase implies something specific:
- "Clinically proven" implies a real, named clinical study behind the exact claim — not a product that "feels" clinical, or a small internal pilot that showed a trend
- "Doctor recommended" implies an active, current, representative recommendation — not one advisor on your cap table or one physician who happened to download the app
- "More effective than [competitor]" implies a head-to-head comparison — not two separate studies run under different conditions placed side by side
If you can't name the study, the reviewer, or the comparison behind a superlative, don't use the superlative.
Rule 5: Naming a Specific Condition Narrows Your Regulatory Path
Verdict: The moment your claim names a specific diagnosis — anxiety disorder, ADHD, type 2 diabetes — you've moved from general wellness marketing toward medical-device territory.
General wellness claims describe an activity or a lifestyle goal without naming a disease: sleep better, manage everyday stress, build a mindfulness habit, track your mood. The moment a claim names a diagnosed condition and claims to affect it, you're describing an intended use that can meet the definition of a medical device — a different regulator and a different evidence standard.
Where that line falls, feature by feature, is the entire subject of Wellness App vs Medical Device: Where Is the Line?. Read it before you finalize any copy that names a diagnosis.
Rule 6: Evidence Has to Exist Before You Publish the Claim
Verdict: "We're running a study on this" is a roadmap item, not substantiation — a claim is judged on the evidence you held the day you made it, not evidence gathered afterward to defend it.
It's tempting, under competitive pressure, to publish a strong claim now and let the study catch up later, especially when you're confident the results will land your way. That ordering is backward: if the study comes back weaker than expected, later than planned, or not at all, you've been making an unsupported claim the entire time in between.
Publish the claim your current evidence supports. Upgrade the wording only after stronger evidence exists — never the reverse.
Rule 7: Every Channel Carrying the Claim Is Bound by It
Verdict: App-store screenshots, onboarding copy, quoted reviews, and a paid influencer's post all count as advertising — the rules don't stop at your homepage.
Founders often police their landing page carefully and forget everywhere else the claim lives. An app-store description that says "treats anxiety" is judged the same way as a headline on your website. A customer review you choose to feature on your pricing page becomes your claim the moment you feature it.
An influencer or affiliate posting "this cured my insomnia" with your product tagged is your marketing claim too, even though you didn't write the words — and platforms increasingly hold the brand responsible for what a paid partner says. Audit every surface, not just the one your team edits most often.
Rule 8: The Overall Impression Controls, Not Your Technical Wording
Verdict: What matters is what an ordinary reader takes away from the whole page — headline, imagery, and fine print together — not whether one sentence, read narrowly, is technically defensible.
A disclaimer doesn't cure an overpromising headline. A bold "Beat anxiety naturally" next to a small-print "results not guaranteed, not intended to diagnose or treat" still leaves most readers with the impression that the product treats anxiety — and that overall impression is what gets evaluated, not the footnote.
The same logic applies to imagery: a photo implying someone stopped medication because of your app communicates a claim even if no text says it directly. Read your own page the way a skeptical stranger would, top to bottom, and ask what they'd tell a friend it does.
Wellness Claim vs. Medical Claim: The Same Feature, Worded Two Ways
The fastest way to see the line is side by side — the same underlying feature, worded on the wellness side versus the regulated side of the boundary.
| Feature area | Wellness-safe wording | Crosses into regulated territory | Why the line falls there |
|---|---|---|---|
| Sleep | "Supports a calmer wind-down routine" | "Treats insomnia" | Names a diagnosable disorder and claims a therapeutic effect |
| Mood / stress | "Helps you build a daily stress-management habit" | "Reduces symptoms of clinical depression" | Names a specific diagnosis and claims symptom reduction |
| Weight | "Tracks your food and activity choices" | "Reverses obesity" or "cures metabolic syndrome" | Claims to reverse or cure a diagnosed condition |
| Blood sugar | "Logs your meals for your own reference" | "Manages your diabetes" or "replaces glucose monitoring" | Claims a management or monitoring role for a diagnosed disease |
| Pain | "Guides a gentle daily stretching routine" | "Treats chronic back pain" | Names a condition and claims a treatment effect |
| Focus / attention | "Helps you build a focus habit during work blocks" | "Treats ADHD symptoms" | Names a diagnosis and claims symptom relief |
| Evidence framing | "Users report feeling calmer after two weeks of use" | "Clinically proven to reduce anxiety" | Implies a completed clinical study that doesn't exist |
The pattern across every row is the same: wellness wording describes an activity, a habit, or a subjective feeling; regulated wording names a diagnosis and claims a specific effect on it. When you're editing copy, run this test on every sentence — does it name a disease, and does it claim to change that disease's course? If both are true, the wording belongs on the right side of this table, and it needs evidence, and possibly a regulatory path, to match. This is also where claim wording connects to the bigger picture: it's one input into the larger sequence covered in The Complete Guide to Startup Idea Validation in 2025, and it's worth getting right early rather than rewriting an entire funnel after a warning letter.
Key Takeaways
- Evidence comes first, the claim comes second — never the reverse. Write down what you can prove today, then write claims no stronger than that.
- A handful of verbs do most of the regulatory damage. "Treat," "cure," "diagnose," and "prevent" all imply an effect on a diagnosed disease; softer language like "supports" and "helps you build a habit of" usually doesn't.
- Testimonials and before/after stories aren't substantiation for a typical result. Feature them only alongside a "results vary" disclosure, and never let one imply what your own product page doesn't claim.
- Comparative and superlative language carries its own evidence obligation. "Clinically proven" and "doctor recommended" sit on top of your base claim's evidence, not inside it.
- Naming a specific diagnosis moves you from wellness marketing toward medical-device territory. General wellness claims describe habits and feelings; regulated claims name and treat a condition.
- Every channel that carries your claim is bound by it — app-store copy, onboarding screens, quoted reviews, and influencer posts included.
- The overall impression of your page controls, not a single defensible sentence. A footnote disclaimer doesn't cure an overpromising headline.
Frequently Asked Questions
Can a wellness app say it treats anxiety?
Generally no, not without the evidence and regulatory posture that claim requires. "Treats anxiety" names a diagnosis and claims a therapeutic effect, which typically calls for controlled evidence and can raise medical-device questions — a different bar than a wellness claim like "helps you manage everyday stress." Have advertising or regulatory counsel review any copy naming a specific condition before it ships.
What counts as FTC substantiation for a health claim?
Substantiation generally means competent and reliable evidence that existed before you made the claim, with the amount of evidence scaling to how specific and serious the claim is. A soft claim about feeling calmer can often lean on user surveys; a specific claim about a measurable health outcome needs data that actually measures that outcome, not a proxy.
Do I need FDA clearance to make a health claim in my app?
Only if your product's intended use meets the definition of a medical device — broadly, diagnosing, treating, curing, or preventing a disease. Many wellness, tracking, and general-habit apps fall entirely outside that definition; the trigger is almost always the claim you make, not the underlying technology. Confirm your specific product's status with regulatory counsel rather than relying on category assumptions.
Is "clinically proven" a claim I can make without a completed study?
No. "Clinically proven" implies a specific, real clinical study supporting the exact claim you're making, and using the phrase without one is one of the most common overclaims in health-tech marketing. If you can't name the study, don't use the phrase — a softer, honest claim about your actual evidence is safer and still persuasive.