General wellness runs on discretion, not exemption

A wearable-device file that logs “general wellness” as though it names one exemption has actually filed two different legal mechanisms under the same label. This blog has already covered the four-condition statutory carve-out Congress wrote for clinical decision support software, at section 520(o)(1)(E) of the FD&C Act. General wellness looks like a cousin of that same idea, and FDA revised its own governing guidance again on January 6, 2026 — but only a narrow slice of what gets called a general wellness product actually rests on a statute at all. The rest of it, hardware and most disease-referenced software claims alike, rests on FDA's own enforcement discretion: a compliance policy the agency chose to adopt, and could in principle narrow, rather than a carve-out Congress wrote into the device definition itself.

Two factors, and both have to hold

FDA's policy for low-risk general wellness devices runs on two conditions together: the product's intended use has to qualify as a general wellness use, and the product has to present a low risk to the safety of users and other persons. A product that clears one and misses the other doesn't get a partial pass — a genuinely low-risk product marketed with a disease-diagnosis claim still needs a device submission, and a product with a plausible general-wellness intended use that happens to be invasive, implanted, or built around a technology like a laser that carries its own dedicated regulatory controls doesn't qualify no matter how modest the marketing claims stay.

What counts as a general wellness use: two categories, not one

The guidance, General Wellness: Policy for Low Risk Devices, sorts a qualifying intended use into two categories, and the distance between them is the whole question. The first covers a claim that never references a disease or condition at all — promoting physical fitness, weight management, self-esteem, or sleep habits. The second covers a claim that does reference a disease or condition, but only where the relationship between a healthy lifestyle and that disease or condition is already well understood and accepted — helping a user live well with type 2 diabetes, or address a risk factor like elevated blood pressure, sits in this category, so long as the claim stops short of a measurable clinical outcome like diagnosis, treatment, cure, mitigation, or prevention. A sensor that estimates a blood-glucose trend to help someone see how a meal affected their afternoon is a different regulatory object than the identical hardware marketed to detect diabetes or replace a clinician's glucose monitor — same device, different intended use, different answer under the policy.

A statute for a sliver of the software, and nothing else

The real split sits underneath the intended-use test, not inside it. Section 520(o)(1)(B) of the FD&C Act, added by the 21st Century Cures Act alongside the CDS exclusion at 520(o)(1)(E), excludes a software function from the device definition outright when it's intended only for maintaining or encouraging a general healthy lifestyle and is unrelated to the diagnosis, cure, mitigation, prevention, or treatment of a disease or condition — the first category above, running in software. Everything else the general wellness policy covers — every non-software product in either category, and any disease-referenced software claim in the second category — qualifies only through FDA's own enforcement discretion: a stated intent by CDRH not to enforce otherwise-applicable device requirements against a product that meets the guidance's own criteria. That's a real difference in what a legal opinion can promise a product team. A statutory exclusion is written into the Act; FDA can't undo it by revising a guidance document on its own schedule. Enforcement discretion is exactly what got revised on January 6, 2026 — the same authority that wrote the policy can narrow it, and has already revised it more than once since it first issued in July 2016.

Where the CDS comparison holds, and where it breaks

The CDS exclusion this blog covered earlier and the general wellness policy get treated as interchangeable in practice, because both let software avoid device status on a short list of conditions and both trace back to the same 2016 statute. The comparison holds for exactly one branch of general wellness — disease-unrelated software — and breaks everywhere else. A CDS function that clears section 520(o)(1)(E)'s four conditions is excluded from the device definition by that statute regardless of whether it touches a disease; a general wellness product that references a disease, software or not, never reaches a statute at all and is leaning on the guidance's own discretion the entire time it's on the market.

Where this meets the file

A labeling-claims tracker built around this policy needs a field most wellness-product files skip: which of the two intended-use categories a given claim falls into, and whether the product is software qualifying for the section 520(o)(1)(B) exclusion or resting on enforcement discretion instead — because only the first of those is durable against a future guidance revision. A claims-review checklist built around the low-risk factor and the two intended-use categories together is previewed in the launch catalog. If your program draws the statute-versus-discretion line differently, the shelf takes that correction directly.