CGM without diabetes: what's normal, and is it worth it?

Reviewed 24 sourcesPublished by PepVault

CGMs without diabetes: what the FDA cleared over the counter, normal readings in healthy adults, what a spike means, and what the trials show about benefit.

Woman in a lilac long-sleeve runs a riverside path at sunrise
Last reviewed
Sources
24
Questions
7
Compounds
1

The short answer

Since March 2024 the FDA has cleared over-the-counter CGMs for people not on insulin, so you can wear one without a prescription. In healthy adults, glucose stays between 70 and 140 mg/dL for about 87 to 96% of the day, and short rises above 140 after meals are normal. No trial yet shows a CGM improves health without diabetes, and diagnosis still uses HbA1c and fasting glucose.

  • Three sensors are FDA-cleared without a prescription: Stelo (March 2024), Lingo (May 2024) and Libre Rio (June 2024). None has glucose alerts.
  • Healthy adults spend 87 to 96% of the day between 70 and 140 mg/dL; adults with normal lab tests, average age about 59, spend about 3 hours a day above 140.
  • A CGM does not diagnose diabetes, and its estimated A1C ran 0.59 points above the lab HbA1c in healthy people.
  • The best-tested way to blunt a spike is moving soon after the meal; the evidence that wearing a CGM improves health without diabetes is one 30-person, 30-day trial.

Every number checked at its primary source · · 24 sources · How we research

Evidence: Strong for what normal CGM readings look like (blinded studies of 153 and 560 adults without diabetes); weak for benefit: one 30-person, 30-day trial in prediabetes and no outcome trial in people with normal glucose.

Woman in a lilac long-sleeve runs a riverside path at sunrise

What does a CGM measure, and how accurate is it?

A continuous glucose monitor (CGM) is a coin-sized sensor, usually on the back of the upper arm, with a thin filament just under the skin. It reads glucose in interstitial fluid, the fluid between cells, rather than in blood, and sends the readings to a phone app3. The over-the-counter sensors last 14 to 15 days, and Stelo's app updates every 15 minutes, so the number on screen can be up to 15 minutes old2.

Accuracy has a legal floor. These devices are integrated CGMs (iCGMs), an FDA device class whose rules require more than 87% of readings to fall within 20% of a blood glucose value across the range, and more than 70% within 15% between 70 and 180 mg/dL6. In practice, when the sensor shows 120, most such readings match a blood value of about 104 to 141, and some fall further off. Both makers' accuracy data came from people with diabetes: Dexcom's Stelo study enrolled adults with diabetes, and Abbott reused its FreeStyle Libre 2 data for Lingo, arguing that diabetic glucose swings are the harder test23.

The FDA has authorized no smartwatch or smart ring that measures glucose on its own without piercing the skin, and in February 2024 it told consumers not to buy or use one that claims to7.

A smiling woman in a blue top leaning on her kitchen counter with her phone and a notebook

PepVault for iPhone

Keep your protocol in PepVault

PepVault's bloodwork log keeps your HbA1c, fasting glucose and fasting insulin from each lab panel, flags each against its reference range, which you can set to your lab's, and adds a point to that marker's trend with every new panel.

Download on the App Store

7 days free on the yearly plan for new subscribers.

Which CGMs can you buy without a prescription?

Three sensors have FDA clearance for sale without a prescription, all for people not on insulin and none with glucose alarms. Dexcom's Stelo came first, and the FDA said it is not for people with problematic low blood sugar, because it is not designed to alert to it135.

DeviceClearedCleared forWearRange shownGlucose alerts
Stelo, first clearance12March 5, 202418 and older, not on insulinUp to 15 days70 to 250 mg/dLNone
Stelo, updated clearance4June 12, 20262 and older, not on insulin; under 18 with a caregiverUp to 15 days55 to 350 mg/dLNone
Lingo3May 29, 202418 and older, not on insulinUp to 14 days55 to 200 mg/dLNone
Libre Rio5June 7, 202418 and older, not on insulinUp to 15 days40 to 400 mg/dL (sensor range)None

The June 2026 clearance also names who Stelo is not for: people on dialysis, people with problematic low blood sugar, people changing their medicines without their clinician, and people with a history of eating disorders4. The 2024 Stelo and Lingo clearances said readings are not a basis for medical action without a clinician23; the 2026 Stelo clearance dropped that sentence4.

What do normal CGM readings look like without diabetes?

Two blinded studies set the reference: the wearers could not see their readings, so the numbers show glucose as it runs on its own. One followed 153 healthy, non-obese people aged 7 to 80 on a Dexcom G6 for up to 10 days8; the other, from the Framingham Heart Study, 560 adults with normal glucose tests, average age about 599.

MeasureHealthy, ages 7 to 80 (153 people)8Normal glucose tests, Framingham (560 people)9Target set for people with diabetes11
Average glucose98 to 99 mg/dL (104 over age 60)114.5 mg/dLNo target
Time 70 to 140 mg/dL96% (median)86.8%No target
Time above 140 mg/dL2.1%, about 30 minutes a day12.3%, about 3 hours a dayNo target
Time 70 to 180 mg/dLNot reported97.8%More than 70%
Time above 180 mg/dLNot reported1.3%, over 15 minutes a dayLess than 25%
Time below 70 mg/dL1.1%, about 15 minutes a day0.9%, about 13 minutes a dayLess than 4%

In 57 people, about a quarter of those with normal standard tests had large swings, reaching the prediabetes range up to 15% of the time and the diabetes range 2% of the time10. So never going above 140 is stricter than healthy bodies run. A reading in the 60s is ordinary too: healthy people spent about 15 minutes a day below 708, and for people without diabetes, low blood sugar usually means below 55 mg/dL24.

Can a CGM tell you if you have prediabetes or diabetes?

Not on its own: there are no CGM-based criteria for diagnosing diabetes13. Diagnosis uses blood tests with the American Diabetes Association's fixed cut-offs, as the NIH lists them12. HbA1c is a blood test of your average glucose over about three months.

TestNormalPrediabetesDiabetes
HbA1c12Below 5.7%5.7% to 6.4%6.5% or above
Fasting plasma glucose1299 mg/dL or below100 to 125 mg/dL126 mg/dL or above
2-hour glucose tolerance test12139 mg/dL or below140 to 199 mg/dL200 mg/dL or above

Two CGM numbers mislead people without diabetes. The estimated A1C that CGM apps calculate, the glucose management indicator (GMI), ran 0.59 points higher than the lab HbA1c on average in 153 healthy people, and was off by 0.4 points or more in 71% of them14. Morning glucose also moves from day to day: across 8,315 adults without diabetes, the day-to-day standard deviation was 7.52 mg/dL, and of those whose first CGM morning was normal, 40% had a later morning in the prediabetes range13.

CGM data may still add to a risk picture. In a Spanish population cohort, the 22 of 499 people who developed type 2 diabetes within 5 years had spent a median 7.8% of the time above 140 mg/dL, against 1.9% in the rest, a link that held after adjusting for age, sex, family history, BMI and HbA1c (odds ratio 1.06)15. Followed for a median of 10.8 years, 42 of 513 in the same cohort developed diabetes, and the clearest dividing line was glucose at or above 130 mg/dL for at least 10% of the time16. These are observational links, and your clinician reads your own result.

PepVault's bloodwork log keeps your HbA1c, fasting glucose and fasting insulin from each lab panel, flags each against its reference range, which you can set to your lab's, and adds a point to that marker's trend with every new panel. Log it in PepVault

Does wearing a CGM improve your health if you don't have diabetes?

No trial has shown it yet in people with normal glucose. The closest is a small one in prediabetes: 30 adults with overweight got the same diet counseling and a CGM for 30 days, but only half could see their readings. In the group that saw them, average glucose fell from 129.1 to 121.6 mg/dL, a significant change; in the blinded group it went from 131.1 to 129.5, which was not17. Those are within-group changes, not a head-to-head result, in thirty people over one month, with no weight, lab HbA1c or long-term outcome reported.

The case for choosing food by CGM: people answer the same meal differently. Across 800 people and 46,898 meals, responses to identical meals varied widely18, and in 1,002 UK adults the person-to-person spread in glucose after identical meals was 68%, with the meal's carbs, fat and protein explaining 15.4% of the variance19.

The catch is measurement. When 16 adults without diabetes wore a Dexcom and an Abbott sensor at the same time for 28 days, the two ranked the same 760 meals differently, with a rank agreement of 0.43 on a scale where 1 is full agreement20.

What lowers a glucose spike after a meal?

Moving after you eat is the best-tested lever. A meta-analysis of 8 crossover trials (116 people, 69 of them without type 2 diabetes) found that exercise after a meal lowered the after-meal rise compared with resting, a moderate effect (standardized mean difference 0.55), while exercise before the meal did not, and the sooner after eating, the larger the effect21. In 7 one-day trials, light walking breaks during long sitting lowered after-meal glucose more than sitting through (effect size 0.72) and more than standing breaks22.

A short planned experiment beats glancing at the graph:

  1. Wear one full sensor, 14 to 15 days, eating as usual the first few days to see your baseline.
  2. Log each meal with the time you ate it, so every rise on the graph has a meal beside it.
  3. Repeat a meal you eat often on two days: once sitting afterwards, once with a short walk started soon after eating21.
  4. Judge patterns across several days, not single readings: one reading can be more than 15% off, and two sensors can disagree on which meal was worse620.
  5. Take the summary to your clinician if your morning readings keep sitting above 100 mg/dL or your average runs high; a lab HbA1c or fasting glucose answers the question12.

PepVault's food log keeps each meal with the time you ate it and its carbs, protein and fiber, from a photo or a line of text, so you can line a rise on your sensor's graph up with what you ate. Track your food in PepVault

Commonly reported: surprise at which everyday foods (rice, cereal, juice) send glucose highest, sudden overnight dips that disappear when you stop sleeping on the sensor arm, and sensors that loosen at the edges before the last day.

Is a CGM useful on a GLP-1?

Most people taking a GLP-1 for weight are not on insulin, so the over-the-counter sensors are cleared for them23. The Wegovy label says semaglutide lowers blood glucose and can cause low blood sugar, and that the risk rises when it is taken with insulin or an insulin secretagogue, a drug that makes the pancreas release more insulin, such as a sulfonylurea23. In the label's trial of adults with type 2 diabetes, low blood sugar below 54 mg/dL happened in 6.2% on Wegovy and 2.5% on placebo23.

So the sensors suit watching how meals land while you eat less, and are worth least to someone with normal lab results who will not act on the graph. They are a poor fit for anyone on insulin or prone to lows: none alerts you to a low123.

When to get help Shaking, sweating, nervousness, confusion, dizziness or sudden hunger can be signs of low blood sugar24; if they come with low readings, contact your clinician, especially if you also take insulin or a sulfonylurea.

Track it in the app

Sources

  1. U.S. Food and Drug Administration. FDA clears first over-the-counter continuous glucose monitor (press release, March 5, 2024). fda.gov
  2. U.S. Food and Drug Administration. 510(k) K234070, Stelo Glucose Biosensor System (Dexcom), cleared March 5, 2024: indications for use and 510(k) summary. accessdata.fda.gov
  3. U.S. Food and Drug Administration. 510(k) K233655, Lingo Glucose System (Abbott), cleared May 29, 2024: indications for use and 510(k) summary. accessdata.fda.gov
  4. U.S. Food and Drug Administration. 510(k) K260935, Stelo Glucose Biosensor System (Dexcom), cleared June 12, 2026: indications for use and 510(k) summary. accessdata.fda.gov
  5. U.S. Food and Drug Administration. 510(k) K233861, Libre Rio Continuous Glucose Monitoring System (Abbott), cleared June 7, 2024. accessdata.fda.gov
  6. Code of Federal Regulations. 21 CFR 862.1355, Integrated continuous glucose monitoring system (2025 edition, GovInfo). govinfo.gov
  7. U.S. Food and Drug Administration. Do not use smartwatches or smart rings to measure blood glucose levels: FDA safety communication (February 21, 2024). fda.gov
  8. Shah VN et al. Continuous glucose monitoring profiles in healthy nondiabetic participants: a multicenter prospective study. J Clin Endocrinol Metab. 2019. PubMed 31127824
  9. Spartano NL et al. Defining continuous glucose monitor time in range in a large, community-based cohort without diabetes. J Clin Endocrinol Metab. 2025. PubMed 39257191
  10. Hall H et al. Glucotypes reveal new patterns of glucose dysregulation. PLoS Biol. 2018. PubMed 30040822
  11. Battelino T et al. Clinical targets for continuous glucose monitoring data interpretation: recommendations from the International Consensus on Time in Range. Diabetes Care. 2019. PubMed 31177185
  12. National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). Diabetes tests and diagnosis (criteria from the American Diabetes Association). niddk.nih.gov
  13. Shilo S et al. Continuous glucose monitoring and intrapersonal variability in fasting glucose. Nat Med. 2024. PubMed 38589602
  14. Shah VN et al. Discordance between glucose management indicator and glycated hemoglobin in people without diabetes. Diabetes Technol Ther. 2023. PubMed 36790875
  15. Marco A et al. Time above range for predicting the development of type 2 diabetes. Front Public Health. 2022. PubMed 36568777
  16. Pazos-Couselo M et al. Diabetes risk assessment in adult population without diabetes employing continuous glucose monitoring: a novel approach. Diabetes Res Clin Pract. 2025. PubMed 40449624
  17. Basiri R et al. Enhancing the impact of individualized nutrition therapy with real-time continuous glucose monitoring feedback in overweight and obese individuals with prediabetes. Nutrients. 2024. PubMed 39683399
  18. Zeevi D et al. Personalized nutrition by prediction of glycemic responses. Cell. 2015. PubMed 26590418
  19. Berry SE et al. Human postprandial responses to food and potential for precision nutrition. Nat Med. 2020. PubMed 32528151
  20. Howard R, Guo J, Hall KD. Imprecision nutrition? Different simultaneous continuous glucose monitors provide discordant meal rankings for incremental postprandial glucose in subjects without diabetes. Am J Clin Nutr. 2020. PubMed 32766882
  21. Engeroff T et al. After dinner rest a while, after supper walk a mile? A systematic review with meta-analysis on the acute postprandial glycemic response to exercise before and after meal ingestion. Sports Med. 2023. PubMed 36715875
  22. Buffey AJ et al. The acute effects of interrupting prolonged sitting time in adults with standing and light-intensity walking on biomarkers of cardiometabolic health: a systematic review and meta-analysis. Sports Med. 2022. PubMed 35147898
  23. Wegovy (semaglutide) prescribing information, section 5.4 Hypoglycemia and section 6.1. Novo Nordisk; DailyMed, revised June 2026. dailymed.nlm.nih.gov
  24. MedlinePlus (National Library of Medicine). Low blood sugar (hypoglycemia). medlineplus.gov

Newsletter · Weekly

The weekly peptide briefing.

New guides, the week’s peptide news, one tool. Free, one email a week, unsubscribe anytime.

We’ll email you a link to confirm. Your address is used for the briefing only. Privacy

PepVault for iPhone

Keep your protocol in PepVault

Log every dose, track every vial, and see your protocol come together.

Download on the App Store

7 days free on the yearly plan for new subscribers.

Explore the live demo