Wellness App or Medical Device? Where FDA Draws the Line

Every health app founder eventually asks whether they need FDA clearance, and most of them are actually asking the wrong question. The real question is narrower and more answerable: does your product make a claim that crosses FDA's general wellness line. The EU asks a parallel question, and MDR Rule 11 and the FDA device rules set out both tests together.

In short: FDA's general wellness policy — a policy of enforcement discretion, not a statutory exemption — covers low-risk products that promote a healthy lifestyle. The line turns on your claims, not your technology: a "wellness" label does not override clinical marketing language, and a claim to diagnose, treat, cure, mitigate, or prevent a disease makes the same product a regulated medical device. A disease reference is not automatically fatal, though: the policy expressly permits "a healthy lifestyle may help reduce the risk of, or help you live well with," claims for certain chronic diseases.

The two-question wellness test

First, the mechanism, because it matters more than founders expect. The general wellness line comes from FDA's guidance General Wellness: Policy for Low Risk Devices (first finalised in 2016, most recently revised January 2026), and it is a policy of enforcement discretion: FDA states it does not intend to examine low-risk general wellness products to determine whether they are devices or whether they comply. It is not a statutory exemption, a clearance, or an affirmative legal status you can hand to a diligence team — FDA can narrow the policy at its own discretion. What decides whether your product is a device in the first place is the definition in section 201(h) of the Federal Food, Drug, and Cosmetic Act, applied to your intended use.

The policy itself sets out a test with two parts, and a product needs to satisfy both to qualify. The first question is about claims, and it is more nuanced than "never mention a disease." The guidance recognises two categories of general-wellness intended use. Category 1 covers claims about general health and healthy lifestyle with no disease reference at all — sleep quality, fitness, relaxation, stress management. Category 2 expressly permits claims that relate the role of a healthy lifestyle to helping reduce the risk of, or helping users live well with, certain chronic diseases or conditions — those where it is well understood and accepted that a healthy lifestyle plays a part, the guidance's own examples including weight, high blood pressure, and type 2 diabetes. What crosses the line is a claim to diagnose, treat, cure, mitigate, or prevent a disease, or a disease reference untethered from that healthy-lifestyle framing. The second question is about risk: even where claims stay within those categories, does the product pose risks to user safety if device regulation, such as design controls or adverse event reporting, were not applied.

Both conditions matter, but the claims question is where nearly all the practical action sits. Most consumer health technology is low-risk by design, and the deciding factor in FDA's actual enforcement pattern is overwhelmingly what the product claims to do, not what its underlying technology is capable of. A company can build genuinely sophisticated signal processing into a wearable and still sit comfortably inside general wellness, as long as its claims stay in general-health language or the permitted healthy-lifestyle framing described above. The inverse is also true: a technically simple product can talk its way into device territory with a handful of poorly chosen sentences on a landing page.

Claims audit: how founders talk themselves into device status

The gap between a wellness product and a medical device is rarely a technology gap. It's a claims gap, and founders cross it gradually, usually without a single deliberate decision to do so. It typically starts with genuinely general wellness language: "track your sleep," "understand your activity patterns," "monitor your heart rate during exercise." As the product matures and the growth team looks for language that converts better, claims drift toward clinical specificity: "identify signs of poor sleep quality" becomes "detect signs of sleep apnoea"; "monitor your heart rate" becomes "detect irregular heart rhythms"; "track your mood" becomes "screen for signs of depression."

Each individual step in this drift can feel like a small, defensible copywriting improvement. Collectively, they cross from general wellness into detection and diagnostic claims about specific conditions — precisely the diagnose/treat/prevent territory the policy does not cover. FDA's test looks at the claims as they actually exist across your product, marketing site, app store listing, and any clinical or scientific claims you make in investor or press material, not at your original intended positioning. A useful internal discipline is a standing claims audit: anyone with the authority to publish user-facing copy should check new claims against a documented, current list of what your regulatory status actually supports, rather than relying on memory of what the product was originally supposed to say.

This drift rarely happens in one place at once. A common pattern: the app itself stays conservative, but a blog post written by a growth marketer references "early detection," a paid social ad promises the product can "catch problems before they become serious," and an app store description written eighteen months ago by a founder who has since left the company still uses language nobody has revisited. FDA's test doesn't care which of these was the "real" positioning. It looks at the overall impression your claims create, across every surface a user or regulator might encounter, and the most aggressive claim anywhere in that set effectively sets your regulatory status.

Worked examples across the line

A sleep-tracking app that reports total sleep time, time in each sleep stage, and general sleep quality trends, with copy encouraging "better sleep habits," sits clearly on the wellness side. The same app adding "possible signs of sleep apnoea detected, consult a doctor" has crossed into device territory, because it now references a specific medical condition and implies a diagnostic function.

A heart rate variability app that reports HRV trends and correlates them with self-reported stress or recovery, framed around general wellness and training readiness, sits on the wellness side. The same underlying HRV data, reframed as "detect signs of atrial fibrillation" or "screen for cardiac arrhythmia," is a device, and depending on the clinical significance of a missed or false detection, likely warrants Class II-equivalent scrutiny.

A mood and mental wellbeing journaling app, encouraging reflection and general emotional awareness, sits on the wellness side. The same app claiming to "screen for depression" or "identify signs of anxiety disorder" has crossed the line, because it now references specific diagnosable conditions rather than general emotional wellbeing.

A glucose-adjacent nutrition app that helps users understand how different foods affect their reported energy levels, without referencing diabetes or blood glucose management as a medical condition, sits on the wellness side. The same app marketed toward "managing your blood sugar" as a health condition, particularly if it ingests continuous glucose monitor data and offers condition-specific guidance, moves toward device territory.

A posture and movement app that tracks general activity, standing time, and movement variety through the day, encouraging users to "move more and sit less," sits on the wellness side. That same app claiming to "identify signs of musculoskeletal disorder risk" or "detect early indicators of chronic back pain" has moved toward device territory, since it now implies a diagnostic assessment of a specific condition rather than general activity encouragement.

A fertility and cycle-tracking app that helps users understand their own reported cycle patterns and general reproductive wellness, without claiming clinical accuracy for a specific medical purpose, sits on the wellness side. The same app marketed as a method of contraception, or as capable of detecting a specific reproductive health condition such as polycystic ovary syndrome, has crossed into device territory, since contraceptive claims and disease-detection claims both trigger direct FDA jurisdiction regardless of how the rest of the app is positioned.

The claims audit worksheet

The exercise described above, writing down every distinct health-related claim your product makes and checking it against the two-question test, works best as an actual document your team maintains, not a one-time mental exercise. The table below is a template you can copy directly into a shared document. Add a row for every distinct claim you can find across your app, website, app store listings, paid ads, investor deck, and any press material you've solicited or reviewed. Two worked rows are filled in as examples.

A few notes on using this worksheet well. Pull claims from every surface, not just the product itself: app store listings, onboarding screens, push notification copy, blog posts, paid social ads, investor decks, and press releases you've pitched all count. Write the claim exactly as it appears, not a paraphrased or softened version, since FDA's test looks at actual wording and overall impression. Where a claim is ambiguous, treat the ambiguity itself as a flag rather than resolving it in your own favor by default. And revisit the worksheet whenever marketing, product, or growth teams ship new copy, not just once a year, since claims drift is a continuous process rather than a single event.

Some teams find it useful to assign an owner to this document, typically whoever has final sign-off on public-facing copy, so the worksheet doesn't quietly go stale the way a one-time compliance memo often does.

The CDS carve-out next door

A separate, narrower carve-out exists for clinical decision support software — and its legal mechanism is different in a way worth being precise about. Unlike general wellness, which rests on FDA enforcement discretion, qualifying CDS is statutorily excluded from the device definition under section 520(o)(1)(E) of the FD&C Act, added by the 21st Century Cures Act and interpreted narrowly in FDA's Clinical Decision Support Software final guidance (September 2022, most recently revised January 2026). All four statutory criteria must be met: the software must not directly acquire, process, or analyse medical images or signals (with "signals" reaching patterns and signals from signal-acquisition systems, not just raw waveforms); it must support rather than replace clinical judgment; the clinician using it must be able to independently review the basis for its output rather than accepting a black-box recommendation; and it must be directed to a healthcare professional, not to a patient or caregiver acting without one. Does My Health AI Need FDA Clearance? covers this carve-out and the SaMD categories that sit outside it in more depth. It's a materially narrower and more technical carve-out than general wellness, and most AI-driven tools that interpret complex data do not qualify for it even where they might otherwise seem clinical-decision-support-adjacent.

The distinction matters because founders sometimes reach for the CDS carve-out as a fallback when they realize their product doesn't fit general wellness, without checking whether it actually meets all four criteria. A tool that analyzes an ECG waveform directly, for instance, fails the first criterion regardless of how much clinical judgment sits downstream of it, because it is performing direct signal analysis rather than supporting a clinician's independent review of raw data. Confirming CDS eligibility requires checking your product against all four parts of the test, not just the parts that happen to fit.

What happens if you get it wrong

Marketing a product as general wellness while its actual claims meet the device definition in FD&C Act section 201(h) does not exempt the product from regulation. It means the product is an unauthorised medical device being marketed without the clearance or authorization the law requires. And crossing the line carries downstream weight beyond the premarket submission itself: a device also owes quality-system compliance under FDA's QMSR (in force since 2 February 2026), design controls, and adverse-event reporting — 510(k) vs De Novo vs PMA: Which FDA Pathway covers what the device route actually entails. In practice, this risk surfaces in a few ways. An FDA enforcement action is possible, though these are more commonly triggered by complaints, adverse events, or a company's own public claims than by systematic wellness-app auditing. Difficulty in a future fundraising or acquisition due diligence process is a second, more common exposure, since sophisticated investors or acquirers increasingly check regulatory status against actual marketing claims. Difficulty adding genuinely clinical features later is a third: a product with a wellness-only regulatory history and infrastructure is not positioned to quickly add validated diagnostic claims without the underlying evidence and design controls a device pathway requires.

The more common and more expensive failure mode we see is not a dramatic enforcement action. It's a startup that spends two years building on a wellness foundation, only to discover during a Series B due diligence process that its actual marketing claims have drifted into device territory, forcing a scramble to either walk back claims or retroactively build the regulatory infrastructure that should have existed from the point the claims changed.

Consider a hypothetical but common shape of this problem: a company launches a wellness-positioned wearable, raises a seed round on general-health claims, and eighteen months later a growth hire who never saw the original regulatory reasoning starts running performance marketing that describes the product as detecting "early warning signs" of a named cardiac condition, because that language tested well in ad copy. Nobody made a deliberate decision to become a medical device company. But by the time a Series B investor's diligence team pulls the ad library, the company's actual public claims no longer match its regulatory posture, and the fix, walking back live ad copy, revisiting app store language, and possibly briefing existing users, is materially more disruptive and more visible than it would have been if the drift had been caught by an internal audit six months earlier.

When to stop guessing

If you are working through a specific claim and genuinely unsure which side of the line it sits on, that uncertainty is itself a signal worth acting on, because ambiguous claims are exactly the kind that both regulators and sophisticated counterparties scrutinise most closely. The practical move is to write down every distinct health-related claim your product currently makes, across every surface, app store listing, website, investor materials, press coverage you've solicited, and check each one against the two-question test above individually, since a product's overall regulatory status is really the sum of its most aggressive individual claim. The worksheet above is meant to make that process concrete rather than something you do once in your head and then forget.

Frequently asked questions

Does adding a disclaimer, "not intended to diagnose," protect a device-level claim? No. FDA's test looks at your actual claims and their overall impression, not at a disclaimer appended to them. A prominent claim that a product "detects early signs of" a disease, paired with a small disclaimer saying it isn't intended to diagnose, does not neutralise the claim for regulatory purposes.

Can we make disease-related claims in investor materials but not in consumer marketing? This is a common but risky pattern. FDA and other parties, including plaintiffs' counsel and due diligence teams, can and do consider claims made anywhere a company communicates about its product's capabilities, not only consumer-facing copy, when assessing actual intended use.

Is a product safer to market as general wellness even if it has genuine clinical capability? Not necessarily safer from a business standpoint. Understating genuine clinical capability to stay inside the wellness policy can leave value on the table — as can stripping every disease word from your copy unnecessarily, since Category 2 of FDA's policy permits well-framed healthy-lifestyle claims about certain chronic diseases — and create its own risk if internal or external communications describe the product's real capabilities inconsistently with its public wellness positioning.

Does FDA pre-clear wellness positioning before launch? No, there is no formal pre-clearance process for general wellness products, since by definition they fall outside FDA's device jurisdiction. This is also why the discipline has to be internal and ongoing rather than something a one-time regulatory sign-off can settle permanently.

If our claims change, do we need to redo a regulatory assessment? Yes. Because FDA's test is claims-based, any material addition of a disease-referencing or diagnostic claim is a regulatory event, not just a marketing update, and should trigger a fresh look at your product's status, ideally before the new claim goes live rather than after.

Who inside a startup should actually own the claims worksheet? In practice, whoever has final sign-off on public-facing copy, which is sometimes the founder, sometimes a head of marketing, and sometimes a regulatory or quality lead once one exists. The specific title matters less than making sure one person is accountable for reviewing new claims against the worksheet before they go live, rather than the review happening informally or after the fact.

Does an investor or acquirer actually check this during due diligence? Increasingly, yes. Diligence teams for health-tech deals routinely pull a company's app store listings, ad libraries, and website copy and compare it against the company's stated regulatory status. A mismatch doesn't automatically kill a deal, but it raises questions about internal controls and can slow a process down at exactly the point a founder wants it to move fast.

The free MedTech Compass can give you an AI-generated first read on where your current claims likely sit. It is not a validated regulatory determination, and claims this consequential to your business are worth a second, expert look before you commit to a positioning either way.

---

Where next: Does My Health AI Need FDA Clearance? · MDR Rule 11: Why Software Lands in Class IIa · 510(k) vs De Novo vs PMA: Which FDA Pathway

Find out in minutes where your current claims likely sit. Start the MedTech Compass →