CGMs for People Without Diabetes: What the Evidence Says

Last verified November 2, 2026. Every FDA record, price and study result on this page links to the source we checked. Where we couldn't confirm something, we say so.

We have not worn or measured these sensors ourselves. This page is general information, not medical advice.

The short answer

Since 2024 you can buy an FDA-cleared continuous glucose monitor (CGM) without a prescription. That makes it legal and easy, not useful. In people without diabetes, no trial has shown that wearing one improves health or prevents disease, and a 2026 JAMA Internal Medicine review said so directly. Single meal readings are noisy and often don't repeat. If you're curious, a two- to four-week experiment costs $54 to $99 and can be educational. If you're worried about diabetes, ask for an HbA1c or fasting glucose test instead. If you have a history of disordered eating, talk to a clinician first.

Status snapshot: what FDA cleared

FDA created an over-the-counter route for CGMs in 2024. All three products below were cleared through 510(k) under 21 CFR 862.1355. We read each record on FDA's database.

Device	FDA record	Who it's cleared for	What it tells you
Dexcom Stelo	K234070, cleared March 5, 2024, product code SAF3. Pediatric extension announced June 12, 20262	People who don't use insulin: adults at first, now age 2 and up (children under adult supervision). Includes people without diabetes who want to see how diet and exercise affect glucose12	The first OTC CGM. Not for people with problematic low blood sugar, because it doesn't alert for lows1.
Abbott Lingo	K233655, cleared May 29, 2024, product code SAF4	Adults 18 and over not on insulin, to understand how diet and exercise affect glucose5	Marketed for wellness. Displays glucose from 55 to 200 mg/dL5.
Abbott Libre Rio	K233861, cleared June 7, 2024, product code SBH6	Classed as an OTC CGM "for non-intensive glucose management"6	Abbott's OTC sensor aimed at people with type 2 diabetes not on insulin, not at wellness users.

What the clearance does and doesn't mean. FDA reviewed whether these sensors measure glucose accurately enough to meet its standards for integrated CGMs. Lingo's file says it met those standards using data from the FreeStyle Libre 2 study⁵. FDA did not review whether wearing one makes healthy people healthier. All three labels say not to take medical action based on the readings without a clinician³⁵. Services such as Levels and Nutrisense are apps and coaching wrapped around these sensors, not separately cleared devices.

Evidence map: is it useful without diabetes?

Grades: A consistent randomized trials or meta-analyses in people without diabetes. B supported by randomized trials that are small or varied. C few, mixed or indirect studies. D no outcome evidence, only plausibility or observational data. X a safety signal or caution.

Claim or use	Grade	Best evidence	Notes
Improves health or prevents disease in healthy people	D	A 2026 JAMA Internal Medicine review found no scientific evidence that CGMs improve health or prevent disease in people without diabetes13.	No long-term outcome trials exist in this group.
Weight loss or behavior change	C	Meta-analysis of 25 trials (2,996 people, mostly with diabetes): CGM feedback lowered HbA1c by 0.28%, but effects on weight and BMI were not significant12. Only 3 trials were in adults with obesity and without diabetes.	Most trials paired the CGM with coaching, so the sensor's own effect is unclear.
Finding "your" best and worst foods	Cto D	People do vary a lot in their response to the same meal (800 people in Israel15; 1,002 in the UK PREDICT study16). But when NIH researchers fed 30 adults without diabetes the same meal twice, a week apart, CGM responses were unreliable: reliability (ICC) was 0.28 for an Abbott sensor and 0.17 for a Dexcom10.	One reading per meal is not enough to label a food good or bad for you. Repeat tests and average them.
Measuring glucose accurately in the normal range	B−	The sensors meet FDA's accuracy standards5. But in a 15-person crossover trial, CGM read about 0.9 mmol/L (about 16 mg/dL) higher than finger-stick blood after meals and stayed high about four times longer, changing one smoothie from moderate to high glycemic index11. Accuracy is generally lowest at low readings14.	Fine for trends and big differences; poor for small ones or for single "spike" sizes.
Spotting early glucose problems	C	In 57 adults, some people who were normal on standard tests spent time in prediabetic ranges on CGM17. There is still no agreed CGM definition of "abnormal" for people without diabetes14.	A finding needs confirming with HbA1c, fasting glucose or a glucose tolerance test.
Athletic performance and fueling	D	A 2024 expert review found studies too small and flawed to show how to use CGM data to improve performance, and sensors are less precise during exercise18.	Interesting for experiments; not proven to help.
Anxiety or restrictive eating	Xcaution	FDA tells people with a history of disordered eating to talk to a provider before using Stelo2. The 2026 review warns that normal swings get overinterpreted, leading to needless food restriction and anxiety13.	If tracking makes you fear normal foods, stop.

The pattern: CGMs are good at showing that glucose rises after meals and falls with activity. They are not good at ranking individual meals, and nobody has shown that healthy people who wear one live longer or stay healthier.

What normal looks like

Most wellness apps color any rise above a set line as a "spike". It helps to know what healthy people actually show. In a 2019 study of 153 healthy people aged 7 to 80 wearing a blinded Dexcom G6, the median time between 70 and 140 mg/dL was 96%. Glucose went above 140 mg/dL for a median of about 30 minutes a day. Mean glucose was about 98 to 99 mg/dL, and 104 mg/dL in people over 60⁹. In a Stanford study, a bowl of cornflakes and milk pushed glucose above 140 mg/dL in about 80% of the 30 adults given standardized breakfasts, most of whom had no diabetes diagnosis¹⁷.

So brief rises above 140 after a meal are normal. Repeated very high readings, or high fasting readings, are a reason to get a blood test, not to change your diet on your own.

If you want to try one: options compared

The evidence doesn't support a "best CGM for non-diabetics" ranking, so we don't rank. If you've decided to run a short experiment, we only list options that meet these criteria:

1. The sensor is FDA-cleared for over-the-counter use by people without diabetes, and we read its FDA record.
2. The price, wear time and refund terms are published, and you can buy a short trial without a long contract.
3. We checked the price on the seller's site on the date shown.
4. Commission plays no part.

Cheapest short trial, with a refund: Abbott Lingo

On November 2, 2026 hellolingo.com sold one 14-day sensor for $54 and two for $89, with no auto-renewal on either. It offers a 30-day money-back guarantee, works with iPhone and Android, and is HSA/FSA eligible⁷. Abbott says about 77% of sensors last the full 14 days and about 15% end before day 11⁷. The app shows readings from 55 to 200 mg/dL⁵, so it won't show exact values outside that range.

Longer wear per sensor: Dexcom Stelo

On November 2, 2026 stelo.com sold a one-month pack (two 15-day sensors) for $99 one time, or $89 a month on a subscription you can pause or cancel. All sales are final, so there's no refund if you don't like it⁷⁸. It needs iOS 18.6 or Android 13 or later, has a 30-minute warm-up, and is HSA/FSA eligible⁷. Dexcom says about 78% of sensors last the full 15 days¹⁹. It's the only OTC CGM cleared for children, under adult supervision²; we see no reason for a healthy child to wear one.

Coaching services: Nutrisense and Levels

Both were still operating on November 2, 2026. Nutrisense listed CGM programs from about $152 to $249 a month, using Stelo sensors, with dietitian coaching included and an early-cancellation fee on 3- and 6-month plans²⁰. Levels changed its pricing in April 2026; its older bundles ran from $288 to $1,999 a year, mixing a month or two of CGM with lab panels, and it sells Stelo as an add-on²¹. We couldn't open Levels' current pricing page. Neither service has a published trial showing better outcomes than wearing the sensor alone, so we don't list them as options: you'd pay several times the sensor price for coaching whose added value is unmeasured.

Side by side

Price checked November 2, 2026 on each seller's US site.

	Abbott Lingo	Dexcom Stelo
FDA record	K233655 (May 2024)4	K234070 (March 2024)3
Ages	18 and over, not on insulin	2 and over, not on insulin2
Wear per sensor (claimed)	14 days	15 days, plus a 12-hour grace period
Sensors lasting full term (maker's figure)	About 77%7	About 78%19
Shortest purchase	$54 for 14 days7	$99 for about 30 days7
Cost per 4 weeks, no contract	$89	$99 (or $89 on a cancelable subscription)8
Refunds	30-day money-back guarantee	All sales final
Independent accuracy data in people without diabetes	NIH duplicate-meal analysis used an older Abbott sensor (Libre Pro)10	NIH analysis used an older Dexcom (G4)10; 2019 healthy reference ranges used Dexcom G69

We found no independent head-to-head accuracy study of Lingo against Stelo, and neither FDA file we read publishes an accuracy figure.

How to run a useful two-week experiment

1. Pick one or two questions. For example: does a walk after dinner blunt the rise? Does my usual breakfast push me well above 140 mg/dL?
2. Repeat each test. Eat the same meal on at least two or three days and compare the average, because single responses don't repeat well¹⁰.
3. Compare against healthy norms, not the app's colors. Use the ranges above: most of the day between 70 and 140, with short rises after meals⁹.
4. Allow for lag. Sensors read fluid under the skin, which trails blood by several minutes after meals¹¹¹⁸.
5. Stop and get a blood test if your fasting readings are often high. A CGM can't diagnose diabetes or prediabetes.

Safety and who should avoid it

• Insulin users and people with frequent or severe lows: OTC sensors are not for you. They don't alert for dangerous lows¹. Use a prescription CGM through your care team.
• People on dialysis: not appropriate for Stelo, per FDA².
• History of disordered eating: talk with your clinician first².
• Medicines: don't change any medicine based on CGM readings without your prescriber¹.
• Skin: the trial side effects were mild: local infection, skin irritation, and pain or discomfort at the site¹.

Myths

• "Any spike above 140 is damaging." Healthy people spend about 30 minutes a day above 140 mg/dL⁹. No study we found shows that short post-meal rises in healthy people cause harm.
• "One test tells you which foods are bad for you." The same meal can give quite different responses a week apart¹⁰.
• "FDA-cleared means it's proven to help healthy people." Clearance covers measurement accuracy and labeling, not health benefit⁵¹³.
• "A CGM reading equals a blood test." In one trial, CGM read higher than finger-stick blood after meals and stayed high longer¹¹.
• "A flat line is the goal." Glucose is supposed to rise after you eat carbohydrate. Normal people don't have flat lines⁹.

FAQ

Can I buy a CGM without a prescription?

Yes. Dexcom Stelo and Abbott Lingo are FDA-cleared for over-the-counter use by people not on insulin¹⁴.

Is a CGM worth it if I don't have diabetes?

For health outcomes, there's no evidence it is¹³. As a short, cheap experiment to learn how your meals and walks affect glucose, it can be educational, as long as you don't overread single meals.

Will a CGM help me lose weight?

The trials so far don't show it. In the largest meta-analysis, CGM feedback didn't significantly change weight or BMI¹².

Can a CGM tell me if I have prediabetes?

No. It can hint at a problem, but there's no agreed CGM cutoff for people without diabetes¹⁴. Prediabetes is diagnosed with HbA1c, fasting glucose or a glucose tolerance test.

Stelo or Lingo?

For a two-week trial, Lingo is cheaper and refundable⁷. Stelo lasts a day longer per sensor and is the only one cleared for children, but sales are final²⁷. We found no independent study comparing their accuracy.

Are Levels and Nutrisense still around?

Yes, as of November 2, 2026. Both now sell Dexcom's Stelo with their apps and coaching, at a higher price than the sensor alone²⁰²¹.

Should athletes wear one?

It's unproven. Sensors are less precise during exercise, and there's no validated way to turn CGM data into better fueling or performance¹⁸.

Looking at other health trackers? See our smart ring picks and best recovery wearables.

References

FDA records were checked on FDA's 510(k) database on November 2, 2026. PubMed could not be opened directly from our tools, so some study details come from the journal page, a repository or a published summary, as noted.

1. US Food and Drug Administration. FDA clears first over-the-counter continuous glucose monitor. 5 March 2024. Source
2. US Food and Drug Administration. FDA clears first over-the-counter continuous glucose monitor for children. 12 June 2026. Source
3. 510(k) K234070: Stelo Glucose Biosensor System, Dexcom, product code SAF, 21 CFR 862.1355, cleared 5 March 2024. FDA decision letter and indications. Source
4. 510(k) K233655: Lingo Glucose System, Abbott, product code SAF, decision 29 May 2024. FDA 510(k) database. Source
5. 510(k) K233655 summary: Lingo indications for use, accuracy shown against iCGM special controls using FreeStyle Libre 2 study data, reportable range 55 to 200 mg/dL. Source
6. 510(k) K233861: Libre Rio Continuous Glucose Monitoring System, Abbott, product code SBH, decision 7 June 2024. FDA 510(k) database. Source
7. Seller pages checked 2 November 2026: Lingo plans (hellolingo.com) and Stelo one-time purchase (stelo.com). Lingo · Stelo
8. Stelo monthly subscription page. Checked 2 November 2026. Source
9. Shah VN, DuBose SN, Li Z, et al. Continuous glucose monitoring profiles in healthy nondiabetic participants: a multicenter prospective study. Journal of Clinical Endocrinology and Metabolism, 2019 (PMID 31127824). Source
10. Hengist A, Ong JA, McNeel K, Guo J, Hall KD. Imprecision nutrition? Duplicate meals result in unreliable individual glycemic responses measured by continuous glucose monitors across four dietary patterns in adults without diabetes. American Journal of Clinical Nutrition, 2025;121(1):74-82. 30 adults, 1,189 duplicate-meal responses; ICC 0.28 (Abbott Libre Pro) and 0.17 (Dexcom G4). doi:10.1016/j.ajcnut.2024.10.007 Source · Preprint
11. Hutchins K, Betts J, Thompson D, Hengist A, Gonzalez J. Continuous glucose monitor overestimates glycemia, with the magnitude of bias varying by postprandial test and individual: a randomized crossover trial. American Journal of Clinical Nutrition, 2025;121(5):1025-1034. doi:10.1016/j.ajcnut.2025.02.024 Source · Summary
12. Richardson KM, Jospe MR, Bohlen LC, et al. A systematic review and meta-analysis of randomised controlled trials on using continuous glucose monitoring to drive behaviour change in people with and without diabetes. International Journal of Behavioral Nutrition and Physical Activity, 2024;21. doi:10.1186/s12966-024-01692-6 Source
13. Continuous glucose monitoring in type 2 diabetes and beyond. Narrative review, JAMA Internal Medicine, published 4 August 2026; co-authored by Minna Johansson, University of Gothenburg (full author list not verified). doi:10.1001/jamainternmed.2026.2772. Details via the University of Gothenburg news release. Source · Summary
14. Klonoff DC, Nguyen KT, Xu NY, et al. Use of continuous glucose monitors by people without diabetes: an idea whose time has come? Journal of Diabetes Science and Technology, 2023;17(6):1686-1697. doi:10.1177/19322968221110830 Source
15. Zeevi D, Korem T, Zmora N, et al. Personalized nutrition by prediction of glycemic responses. Cell, 2015;163(5):1079-1094. doi:10.1016/j.cell.2015.11.001 Source
16. Berry SE, Valdes AM, Drew DA, et al. Human postprandial responses to food and potential for precision nutrition. Nature Medicine, 2020;26:964-973. doi:10.1038/s41591-020-0934-0 Source
17. Hall H, Perelman D, Breschi A, et al. Glucotypes reveal new patterns of glucose dysregulation. PLoS Biology, 2018;16(7):e2005143. 57 adults; the cornflakes figure (80% of a 30-person standardized-meal substudy) is from the Stanford release and senior author Michael Snyder. doi:10.1371/journal.pbio.2005143 Source · Summary
18. Flockhart M, Larsen FJ. Continuous glucose monitoring in endurance athletes: interpretation and relevance of measurements for improving performance and health. Sports Medicine, 2024;54:247-255. Source
19. Dexcom provider page for Stelo: wear time and share of sensors lasting the full 15 days. Checked 2 November 2026. Source
20. Nutrisense CGM program plans. Checked 2 November 2026. Source
21. Levels support. Levels Classic, Core and Complete plans (legacy plans for members who joined before April 2026). Checked 2 November 2026. Source