How We Label Evidence in Biohacking
The plain-language labels used across the Biohacking library, from approved medicine to marketing claim, with what each means and how device accuracy differs from proof of benefit.
Content last updated . This page has not been medically reviewed.
Why we use labels
Biohacking covers everything from sleep habits to wearable sensors to experimental claims, and the strength of evidence varies enormously. Different organizations rate evidence in different ways; the GRADE working group, for example, wrote that guidelines are inconsistent in how they rate the quality of evidence and the strength of recommendations. [1] To keep this library consistent, every page uses the same plain labels, explained below. The labels describe the evidence we read, not a recommendation for any person. This is information only, not medical advice.The labels
| Label | What it means here |
|---|---|
| Approved medicine | A regulator has approved a product for a stated use (we name the regulator and the use). |
| Established | Many consistent, well-designed human studies, or long-standing public-health guidance built on them. |
| Promising but limited | Some human evidence with small, short or inconsistent studies, or benefits shown only for some outcomes. |
| Mixed | Good studies disagree, or the answer depends on who was studied and how. |
| Insufficient evidence | Too few or too weak human studies to say. |
| Preclinical (cells or animals only) | Lab or animal work with no human outcome yet. It is never carried over to people. |
| No good evidence of benefit | Studied reasonably well without a benefit. |
| Evidence of harm | Studies or regulators report harm. |
| Marketing claim | Stated by sellers or influencers, with no independent evidence we could read. |
| Proper evidence not found | We searched and found nothing usable. We say where we looked. |
Three rules we add for devices and tracking
1. Direct, estimated or calculated. A sensor may measure something directly (for example, motion). Many numbers are estimates made from that signal (for example, sleep stages), and some are proprietary scores calculated by a company's formula. We say which is which, and we do not give proprietary scores a clinical meaning they have not earned. The American Academy of Sleep Medicine said consumer sleep technologies lack validation and FDA clearance and cannot be used to diagnose or treat sleep disorders at this time. [2]2. Accuracy is not benefit. Showing that a device measures a number well is different from showing that using it improves health. For example, an umbrella review of wearable trials found increased physical activity but limited, inconsistent effects on other outcomes, [3] and one randomized trial found that adding a wearable led to less weight loss than a standard program alone. [4] Label: mixed.
3. A biomarker change is not a health outcome. A lower or higher number on a test does not by itself mean you live longer or feel better. See What Counts as a Meaningful Outcome?.
Study type matters
Placebo effects and spontaneous improvement can both make a treatment look better than it is. A review of three-armed trials found that spontaneous improvement and placebo each contributed importantly to change seen in actively treated people. [6] That is why controlled trials are weighted above personal experience, and why animal or cell findings are labeled preclinical. See Placebo and Expectancy Effects in Wellness Tracking.How these labels relate to the rest of the site
The broader site uses a related set of labels, described in How We Label Evidence in Specialty Topics. For the overall evidence landscape for biohacking, see The Biohacking Research Evidence Landscape.What we do not do
We do not rank brands, claim hands-on testing, give personal protocols or doses, or treat a reading from a consumer device as a diagnosis.Frequently asked questions
Why use labels at all?
Because "studies show" can mean anything from a large trial to one animal experiment, and rating systems differ between organizations. [1]
What does "mixed" mean?
Good studies point in different directions or the effect depends on the group studied. For example, wearable trials increased activity but had inconsistent effects on other outcomes. [3]
Is "FDA general wellness" the same as FDA approval?
No. It describes a policy for low-risk products that promote a healthy lifestyle. [5] It does not say the product works.
Does an accurate device mean a healthier me?
Not necessarily. Accuracy and benefit are different questions. [4]
More details: evidence, limits and cautions
The details behind this page: cautions, limits of the evidence and where sources disagree. Open any note to read it.
Regulatory words are not evidence labels
References
- Guyatt GH et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008. Read 2026-10-22.
- Khosla S et al. Consumer sleep technology: an American Academy of Sleep Medicine position statement. J Clin Sleep Med 2018. Read 2026-10-22.
- Glisic M et al. An umbrella review of systematic reviews of the impact of wrist-worn wearables on health outcomes. Physiol Rev 2026. Read 2026-10-22.
- Jakicic JM et al. Effect of wearable technology combined with a lifestyle intervention on long-term weight loss: the IDEA randomized clinical trial. JAMA 2016. Read 2026-10-22.
- U.S. Food and Drug Administration. General Wellness: Policy for Low Risk Devices (content current as of 01/06/2026) Read 2026-10-22.
- Krogsbøll LT et al. Spontaneous improvement in randomised clinical trials: meta-analysis of three-armed trials comparing no treatment, placebo and active intervention. BMC Med Res Methodol 2009. Read 2026-10-22.