Continuous Glucose Monitors: How They Work, Who They're For, and What the Numbers Mean

Short Answer

A continuous glucose monitor (CGM) is a small wearable glucose sensor — usually placed on the arm, belly, or abdomen — that tracks glucose in the interstitial fluid, the fluid between your cells just under the skin. Instead of giving you one blood-sugar snapshot from a fingerstick, it updates every few minutes and sends the pattern to a phone, receiver, or connected device. That trend matters because your body is always moving glucose around: after meals, during exercise, under stress, while you sleep, and when medication is active. A CGM lets you see not only the number, but the direction it's heading. (CDC — Continuous Glucose Monitors)

That's why clinicians often look beyond a single reading and use metrics like time in range — commonly 70–180 mg/dL for many people with diabetes — to understand how much of the day glucose spends in a target zone, while still watching time below and above range separately. The 70–180 mg/dL range is a default CGM reporting range from international consensus guidance, not a personal target for everyone; your clinician may set a different range if you're pregnant, older, at higher risk of hypoglycemia, newly diagnosed, or using glucose-lowering medication. (International Consensus on Time in Range)

CGMs are now a core glucose-management tool for many people with diabetes, especially type 1 diabetes and insulin-treated diabetes, because insulin decisions depend on seeing both lows and fast changes early. NICE guidance, for example, recommends offering adults with type 1 diabetes a choice of real-time CGM or intermittently scanned CGM, based on needs and preferences. In the U.S., over-the-counter CGMs have also expanded access: FDA cleared Dexcom Stelo for adults 18 and older who do not use insulin in March 2024, later clearing Stelo for people age 2 and older who do not use insulin in June 2026; FDA records also list Abbott's Lingo Glucose System as an OTC integrated CGM for adults 18 and older not on insulin. (NICE NG17 — Type 1 diabetes in adults)

A CGM is a glucose device. It is not the same thing as a heart-rate wearable, and a smartwatch app that displays CGM data is not the same as a watch measuring glucose by itself. As of the FDA's safety communication, the agency had not authorized, cleared, or approved any smartwatch or smart ring intended to measure or estimate blood glucose values on its own; devices that claim non-invasive glucose measurement without piercing the skin should not be used for diabetes decisions. (FDA — Do Not Use Smartwatches or Smart Rings to Measure Blood Glucose)

Where Welltory fits is a separate question. Welltory is not a CGM and does not measure glucose — it analyzes heart rate, HRV, sleep, and activity, signals that reflect stress and recovery and can sit alongside glucose data, never in place of it.

CGM vs. fingerstick meter at a glance

A continuous glucose monitor (CGM) is better when you want the "movie" of glucose: how your numbers move after meals, exercise, sleep, stress, or medication. A small sensor sits on the arm or abdomen and estimates glucose in interstitial fluid — the fluid between cells — then sends readings to an app or receiver, often with curves, trend arrows, and alerts. A fingerstick blood glucose meter is the "photo": you prick your finger, place a drop of blood on a strip, and get one capillary blood glucose reading for that moment. It will not show where glucose is heading, but it is still useful for confirmation checks, especially when symptoms do not match a CGM reading or when a clinician asks you to verify a value. (CDC — Continuous Glucose Monitors)

Glucose targets and metrics (set with your clinician)

Your CGM report is most useful when you read it as a pattern, not as a single "good" or "bad" number. Glucose naturally moves after meals, with activity, during illness, and overnight. The point is to see how much of your day your glucose spends in safer zones — and how often it drifts high or low enough that your clinician may want to adjust food timing, medication, insulin, exercise, or alarms.

The 70–180 mg/dL band is the standard time-in-range window used across CGM reporting and consensus guidance. (International Consensus on Time in Range)

For diagnosis, CGM is not the deciding test. Clinicians use lab-based criteria. CDC lists A1C below 5.7% as normal, 5.7–6.4% as prediabetes, and 6.5% or above as diabetes; for fasting blood sugar, 99 mg/dL or below is normal, 100–125 mg/dL is prediabetes, and 126 mg/dL or above is diabetes. A 2-hour oral glucose tolerance test result of 200 mg/dL or above is also in the diabetes range. If a screening result is abnormal, follow up with your clinician so it can be interpreted in context and repeated or confirmed when needed. (CDC — Diabetes Tests)

What a CGM actually measures

A continuous glucose monitor doesn't sample blood the way a fingerstick meter does. A tiny sensor sits just under your skin, in the interstitial fluid — the fluid around your cells — and estimates glucose there while you wear it. That is why the value on the screen is usually called sensor glucose, not a direct blood-glucose measurement. The practical payoff is the curve: instead of one isolated number, you see where glucose is now, whether it's rising or falling, and how quickly that change is happening. That trend can reveal a fast post-meal rise, a slow overnight dip, or a long steady stretch that a single fingerstick could easily miss. (Cleveland Clinic — Continuous Glucose Monitoring)

There is one catch: glucose reaches your bloodstream first, then moves into interstitial fluid. So CGM readings and fingerstick readings won't always match exactly, especially when glucose is changing quickly after food, insulin, exercise, illness, or stress. A commonly cited physiologic lag is about 5–15 minutes; older and review data also show that the displayed sensor value can lag capillary blood by roughly this range because of both tissue movement and device processing. (Physiology Behind Continuous Glucose Monitoring)

Accuracy is usually summarized with MARD — mean absolute relative difference — which describes average distance between CGM readings and a reference glucose method. Lower is generally better, but it's not a promise that every single reading will be off by only that amount. In validation studies, Dexcom G7 in adults showed overall MARDs of 8.2% for arm placement and 9.1% for abdomen placement; FreeStyle Libre 3 showed an overall MARD of 7.8% in a U.S. study of people age 4 and older with diabetes; the 180-day Eversense system reported an overall MARD of 9.1% through sensor wear. These numbers are useful for context, but real-life accuracy can still vary by device, sensor age, placement, glucose range, and rate of change. (Dexcom G7 Accuracy Study)

That's why a fingerstick meter still matters as a backup. FDA device documents for Dexcom G7 say a fingerstick is required when symptoms don't match the displayed glucose value or when no glucose value or trend is shown, and CDC notes that interstitial-fluid glucose can differ from blood glucose, so occasional fingerstick checks may still be needed to confirm accuracy. In plain terms: if the CGM says you're fine but your body says "low," "high," or "something is wrong," trust the body enough to check. (FDA — Dexcom G7 510(k), K213919)

CGM's real-time trend advantage over intermittent testing is well described in the literature: continuous monitoring "may overcome the limitations of intermittent point-of-care (POC) testing by providing real-time glucose trends and reducing treatment delays." (CGM in Non-ICU Hospitalized Adults with T2D)

Time in range: the number that replaced the single reading

The metric that CGMs made mainstream is time in range: the share of the day your glucose spends inside a target window. For many nonpregnant adults with type 1 or type 2 diabetes, that window is commonly set at 70–180 mg/dL. CGM reports usually break the day into three buckets: time in range, time above range — often >180 mg/dL — and time below range — often <70 mg/dL. That shift matters because glucose is not a still photograph. It rises after food, falls with insulin, movement, missed meals, alcohol, illness, and sometimes overnight. A single fingerstick can tell you where you are in one moment; time in range shows how much of your real day your body spent on target. (Clinical Application of Time in Range)

The evidence is strongest for people with diabetes. In non-ICU hospitalized adults with type 2 diabetes, a systematic review and meta-analysis found that CGM improved time in range by a mean difference of about 8.15 percentage points (95% CI +5.76 to +10.55, p < 0.001) compared with standard point-of-care glucose testing, and reduced time spent above range. In type 2 diabetes outside the hospital, when CGM is used to guide lifestyle and nutrition choices, a meta-analysis of 21 randomized trials involving 2,734 adults found better HbA1c (mean difference −0.46%, 95% CI −0.71 to −0.22) and better time in range — not because the sensor "treats" glucose by itself, but because it makes patterns visible enough to act on. (CGM in Non-ICU Hospitalized Adults with T2D; CGM to Guide Lifestyle Choices in T2D)

The device landscape: prescription, OTC, and what a smartwatch can't do

Continuous glucose monitors began as prescription diabetes devices, but they are no longer a niche tool for only a small subset of people with diabetes. In current diabetes care, CGM is moving toward the center of insulin management because insulin decisions depend on timing: glucose can rise after food, fall during or after exercise, dip overnight, or change when you are sick or stressed. A fingerstick gives one point in time. A CGM gives a moving pattern. That pattern is why professional guidance now treats CGM as a standard option for people using insulin — CGM "is now recommended by professional guidelines for all patients with diabetes on insulin therapy" — while CDC patient guidance describes CGMs as especially useful for people with type 1 diabetes, type 2 diabetes using insulin, frequent testing needs, or difficulty keeping glucose in range. (CGM for Type 2 Diabetes Management — scoping review; CDC — Continuous Glucose Monitors)

The newer story is over-the-counter (OTC) CGMs. These are sensors you can buy without a prescription, usually aimed at people who are not using insulin and want to see how meals, exercise, sleep, and daily routines affect glucose. In FDA records, Dexcom Stelo Glucose Biosensor System was cleared OTC for adults 18 and older not using insulin in March 2024; on June 12, 2026, FDA cleared Stelo for people 2 years and older who do not use insulin. The FDA-cleared Stelo sensor is worn for up to 15 days, though the FDA notes pediatric wear time may be shorter. (FDA — First OTC CGM Clearance: Dexcom Stelo)

Abbott Lingo Glucose System is also an OTC integrated CGM for people 18 and older who are not on insulin. Its FDA 510(k) summary says it is meant to continuously measure, record, analyze, and display glucose values, and to help users understand how diet, exercise, and other behavior changes affect glucose excursions — but not to make medical decisions without a qualified healthcare professional. The same FDA document lists a sensor wear time of up to 14 days and notes that Lingo does not provide glucose alarms. (FDA — Abbott Lingo 510(k), K233655)

Prescription CGMs still matter, especially if you use insulin or need alerts for high or low glucose. Common prescription lines include Dexcom G7 and Abbott FreeStyle Libre 3 / FreeStyle Libre 3 Plus. FDA device records list Dexcom G7 as a real-time CGM for diabetes management, and FDA recall communications describe FreeStyle Libre 3 and 3 Plus as real-time CGM devices with alarms for people with diabetes age 4 and older. (FDA — FreeStyle Libre 3 device record, K240902)

Price is the one part of the device landscape that should be treated as a moving target, not a medical spec. FDA clearance records confirm whether a device is cleared, what it is indicated for, who it is for, and key safety limitations; they do not validate the retail price you will see through a manufacturer, pharmacy, insurer, or subscription checkout. CDC's practical cost framing is still useful: coverage is often better for people with type 1 diabetes or insulin-treated type 2 diabetes, while out-of-pocket CGM costs can be high without insurance. (CDC — Continuous Glucose Monitors)

One persistent myth to clear up: no smartwatch — Apple Watch included — measures blood glucose non-invasively. FDA has warned people not to use smartwatches or smart rings that claim to measure blood glucose without piercing the skin, and states that it has not authorized, cleared, or approved any smartwatch or smart ring that measures or estimates glucose on its own. A watch can display data from a paired FDA-authorized glucose device, such as a CGM, but the watch is acting as a screen — not as the glucose sensor in your wrist. (FDA — Do Not Use Smartwatches or Smart Rings to Measure Blood Glucose)

Do you need a CGM if you don't have diabetes?

If you have diabetes — especially if you use insulin — a CGM can be a clinically useful tool, not just a gadget. Current diabetes standards describe CGM as part of diabetes technology and recommend offering diabetes devices with training and individualized support; NIDDK also notes that clinicians may recommend CGM when insulin is needed to manage type 1, type 2, or another form of diabetes. (ADA — Diabetes Technology, 2026)

If you don't have diabetes, the answer is more nuanced. A CGM may help you see patterns you can't feel: how your glucose shifts after a specific breakfast, how a poor night's sleep changes your morning response, or how a stressful day seems to amplify the same meal. That can be genuinely interesting — and sometimes useful for behavior change — because glucose is not a flat line even in healthy metabolism. Your body is constantly absorbing, storing, releasing, and clearing glucose through insulin, liver signaling, muscle uptake, stress hormones, sleep-wake rhythms, and movement.

But "interesting data" is not the same as "proven health benefit for everyone." Research on CGM use in people without diabetes is active, but still early: reviews describe the evidence outside diabetes as limited, with little consensus on what healthy people should do with CGM patterns or whether tracking them improves long-term outcomes. (CGM in People Without Diabetes — review)

So if you try an over-the-counter CGM without diabetes, treat it as a short-term learning tool, not a diagnosis. Use it to notice repeatable patterns — not to label foods as "good" or "bad" after one spike, and not to chase perfectly flat glucose. The FDA-cleared OTC Stelo system, for example, was cleared for people who do not use insulin, including people without diabetes who want to understand how diet and exercise affect glucose; the FDA also says users should not make medical decisions from the device output without talking to a healthcare provider. (FDA — First OTC CGM Clearance: Dexcom Stelo)

If your readings worry you, the next step is not self-treatment. It's a clinician and standard blood testing. Diabetes and prediabetes are diagnosed with blood tests, and NIDDK specifically warns not to diagnose yourself with over-the-counter blood glucose equipment. (NIDDK — Diabetes Tests & Diagnosis)

Glucose dynamics are also being studied in people without diabetes. A 2026 JAMA Network Open cohort of 895 adults without diabetes used masked CGM and smartphone appetite surveys over 9 free-living days and analyzed 7,650 meals; larger post-meal glucose declines were associated with greater hunger and faster initiation of the next eating episode. As the authors note, "glucose level declines induce hunger in laboratory settings, but free-living data are limited." (Postprandial Glucose Decline and Appetite in Adults Without Diabetes)

Where stress, sleep, and recovery fit — the Welltory angle

Glucose regulation doesn't happen in isolation. Your blood sugar is partly a fuel-management system, and fuel management is tied to sleep, threat, effort, and recovery. When you sleep poorly or run under sustained stress, the body leans into "keep energy available" physiology: sympathetic activation rises, the HPA axis becomes more active, and counter-regulatory hormones such as cortisol and catecholamines can push more glucose into circulation while making insulin's job harder. That doesn't mean one bad night "causes diabetes." It means sleep and stress can change the background your glucose is responding to — especially if you already live with diabetes or use insulin. Reviews of sleep, stress, and metabolism describe these shared pathways through the autonomic nervous system, cortisol/HPA-axis activity, and insulin sensitivity. (Metabolic Consequences of Sleep and Circadian Disorders)

This shows up in CGM data. In a cross-sectional study of 155 adults with type 1 diabetes, the group sleeping the recommended 7–9 hours per night had better glycemic-variability markers than people sleeping less or more: mean glucose was 161.2 mg/dL in the 7–9-hour group versus 172.2 mg/dL in the shorter-sleep group and 174.6 mg/dL in the longer-sleep group; coefficient of variation was 35.5% versus 38.6% and 38.8%; and MAGE was 108 mg/dL versus 127.8 and 126.8 mg/dL. The study cannot prove sleep duration caused those differences, but it supports the practical point: sleep sits close enough to glucose regulation that it belongs in the same conversation. (Sleep Duration and Glycemic Variability in Adults with T1D)

This is where a heart-rate and HRV app like Welltory is complementary — not a glucose device. Welltory doesn't measure glucose; it tracks HRV, resting heart rate, sleep, and activity, which can help you see patterns in autonomic load and recovery. HRV is widely studied as a noninvasive marker related to autonomic regulation and stress response, but it is not a substitute for glucose measurement. (Heart Rate Variability: Clinical and Research Applications)

So if your CGM shows a rough glucose day, Welltory may help you ask better "why" questions: Did you sleep less? Was your resting heart rate higher than usual? Did your HRV drop after a stressful day, hard workout, alcohol, illness, or late meal? Those patterns can add context. The glucose number itself still comes from a CGM or meter, never from Welltory. Think of them as two lenses on the same lifestyle: one lens shows glucose directly; the other shows recovery and strain around it.

How to get useful data from a CGM (without over-reading it)

If you use a CGM, treat the first job as getting clean, usable data — not reacting to every dot on the graph. Put the sensor on clean, dry skin, follow the device's instructions for warm-up and calibration, and expect the numbers to move. CGMs read glucose in interstitial fluid, not directly from a fingerstick blood sample, so the display can lag behind what is happening in your blood, especially when glucose is rising or falling quickly. (CDC — Continuous Glucose Monitors)

That's why a surprising reading deserves context. If your CGM says you're low but you feel fine, or you feel shaky, sweaty, confused, or unwell but the CGM looks "normal," use a fingerstick meter as a backup and follow the plan your clinician gave you. A CGM is excellent for watching direction and patterns; a fingerstick is still useful when a specific reading does not fit your body. (Cleveland Clinic — Continuous Glucose Monitoring)

Try to read the curve like a pattern, not a verdict. A single post-meal rise is normal physiology: glucose changes through the day with food, activity, medication, illness, sleep, and routine. The useful question is not "Did I spike once?" but "What keeps happening?" Which breakfasts lead to a long rise? Which late dinners show up overnight? Which stressful days come with higher glucose or more variability? Repeated patterns are what help you and your care team adjust meals, movement, medication timing, or targets. (CDC — Monitoring Your Blood Sugar)

And keep each tool matched to its job. A CGM answers, "What is my glucose doing over time?" A fingerstick helps confirm a specific reading when accuracy matters. An HRV, sleep, and activity app like Welltory answers a different question: "How recovered, strained, or stressed does my body look today?" Those signals do not replace glucose data, but they can help you notice the body context around it — the nights, workouts, stress loads, and recovery dips that may sit next to changes in your curve.

How we made it

Made with AI tools, then edited, fact-checked, and medically reviewed by the Welltory team.

Discounts for blog readers: up to 36% off

See what affects your energy, stress, sleep, and daily state with Welltory

Explore your data