What is continuous glucose monitoring?

CGMs estimate glucose in interstitial fluid and communicate with a supported display. Sampling, transmission and display intervals vary by model; the Stelo app updates every fifteen minutes despite five-minute sensor transmission.

Unlike traditional blood glucose meters that require a finger-stick blood sample to get a single reading, CGM provides a continuous stream of data — showing not just where your glucose is right now, but where it has been over the past hours, and trend arrows that indicate where it is heading.

Key point

CGM does not measure glucose in the blood directly. It measures glucose in interstitial fluid — the fluid surrounding cells in tissue just below the skin. This is closely correlated with blood glucose but can lag behind blood glucose during rapid changes (the lag depends on the device and the situation), which is important to understand when interpreting readings.

1 / 5 / 15 min
Example intervals vary by model and display
Model-specific
Labeled wear duration and replacement instructions
Trial-specific
Benefits depend on device, population and treatment program

How does CGM work?

A CGM system has three main components that work together to deliver continuous glucose data to the user.

1

The sensor

Many CGMs use a filament to sense interstitial glucose; Eversense uses a clinician-inserted optical sensor. Approved wear sites depend on model and age. Do not assume abdominal placement or one sensing chemistry for every CGM.

2

The transmitter

Attached to the sensor, the transmitter reads the electrical signal from the sensor, applies calibration algorithms, converts it to a glucose value in mg/dL or mmol/L, and transmits the data wirelessly via Bluetooth to a display device. Transmitters are either integrated into disposable sensors or reusable components that attach to each new sensor.

3

The receiver / app

Supported displays present sensor information. Trends, sharing and alarms differ by model; Stelo and Lingo do not provide clinical high/low alarms.

Wear and warm-up differ: original G7 supports up to 10 days plus 12-hour grace and a 30-minute warm-up; G7 15 Day up to 15 days plus 12 hours and a 60-minute warm-up; original Libre 3 up to 14 days and 60-minute warm-up. Eversense is clinician-inserted and has separate transmitter and calibration requirements.

CGM vs traditional blood glucose monitoring

Traditional self-monitoring of blood glucose (SMBG) using a finger-stick meter has been the standard for decades. CGM does not fully replace it in all situations, but provides a fundamentally different and richer picture of glucose patterns.

Feature CGM Finger-stick meter
Readings per day Model-specific sensor and display intervals Each manual blood sample
Trend information Yes — direction and rate No — single snapshot
Overnight monitoring Continuous; alarms depend on model Intermittent measurements
Finger-stick required Calibration and confirmation depend on model Every measurement
Real-time alerts Model-specific; no clinical alarms in Stelo/Lingo No continuous alarms
Data sharing Model/app-specific Logs or supported digital export
Cost Sensor, display and any service costs Meter, strips and lancets
Accuracy in rapid change Lag varies by model and situation Capillary blood measurement; meter error also possible
When finger-sticks still matter

Even with CGM, finger-stick verification is recommended when CGM readings seem inconsistent with symptoms, before making critical treatment decisions (such as insulin dosing in some circumstances), and when CGM indicates rapidly changing glucose. Always follow your healthcare provider's guidance on when to verify with a finger-stick.

Benefits of continuous glucose monitoring

The clinical evidence for CGM benefits in people with diabetes is substantial and growing. Beyond diabetes management, CGM is revealing new insights about metabolic health in broader populations.

📉
Improved A1C
In DIAMOND, adults with type 1 diabetes using multiple daily insulin injections had an adjusted between-group HbA1c difference of 0.6 percentage points at 24 weeks with CGM versus usual monitoring. In MOBILE, adults with poorly controlled type 2 diabetes using basal but not mealtime insulin had a 0.4-point between-group difference at eight months. These results apply to those treatment programs and populations.
🛡️
Reduced hypoglycemia
In trials of prescription CGM with low-glucose alerts in adults with type 1 diabetes using multiple daily insulin injections, time spent in hypoglycemia fell: in GOLD-3 (GOLD-3 trial report) daytime time below the studied thresholds was 40–54% lower and nighttime time 48–65% lower than with fingerstick monitoring, depending on the threshold, and in DIAMOND (DIAMOND trial report) the median time below 70 mg/dL was 43 versus 80 minutes per day. These results apply to those populations, devices and thresholds; they do not automatically apply to OTC devices without clinical alarms.
🎯
More time in range
Time in range is the proportion of readings within a defined glucose interval. Consensus recommends over 70% in 70–180 mg/dL for many nonpregnant adults with type 1 or type 2 diabetes, individualized with their care team. Time below range, time above range and variability add information that TIR alone does not supply. CGM provides dense sampling, but TIR has also been studied using seven-point fingerstick profiles; it does not replace laboratory HbA1c.
😴
Overnight safety
Some prescription CGMs have low-glucose alarms and sharing features that can support overnight care. Alarms depend on settings and connectivity and do not guarantee that every low is detected or prevented.
🍽️
Food response insights
Meal logs viewed alongside sensor glucose may reveal associations. A trace cannot prove a food caused a response or identify an optimal diet on its own.
🏃
Exercise optimization
Sensor glucose may be observed alongside exercise, but responses vary. CGM does not guarantee better athletic performance or prevention of hypoglycemia. Follow device instructions and an individualized care plan.
Medical disclaimer

The benefits described above are based on published clinical research in specific patient populations. Individual results vary. CGM is a monitoring tool, not a treatment. Any changes to your diabetes management — including insulin dosing, medication, or diet — should be made in consultation with your healthcare provider.

Who should use CGM?

CGM was originally developed for people with Type 1 diabetes who require intensive insulin therapy. Its indications have expanded substantially and it is now used across a much broader population.

💙
Type 1 diabetes
In DIAMOND, adults with type 1 diabetes using multiple daily insulin injections had an adjusted HbA1c reduction 0.6 percentage points greater with CGM than usual care at 24 weeks. This study does not establish universal benefit or eliminate the need for device-specific instructions and an individualized care plan.
🩺
Type 2 diabetes
MOBILE studied adults with type 2 diabetes using basal but not mealtime insulin and reported an adjusted HbA1c difference of 0.4 percentage points at eight months. Coverage and suitability require separate individual assessment and all applicable plan conditions.
⚠️
Prediabetes
CGM records glucose patterns but is not a diagnostic test for prediabetes and has not been established here as a way to prevent or reverse it. Discuss laboratory testing and an evidence-based prevention plan with a healthcare professional.
⚡
Health optimization
CGM use among people without diabetes has grown rapidly, driven by interest in metabolic health, longevity, and performance optimization. OTC CGM devices now available without a prescription have made this segment accessible. The evidence base for CGM benefit in non-diabetic populations is still developing.

How accurate is CGM?

MARD is the mean absolute relative difference between sensor readings and reference measurements in a particular study. It is an average across paired readings, not an error bound for each reading or a direct clinical ranking across separate studies.

Accuracy depends on the device, glucose range and measurement conditions. A study-wide MARD does not guarantee the accuracy of an individual reading. Follow the current instructions for your specific device and discuss clinical decisions with your healthcare professional. Calibration requirements and any need to confirm readings with a blood glucose meter are device-specific. Follow your current device instructions, including what to do when a reading does not match your symptoms.

Factors that affect CGM accuracy

Rapid glucose changes: Because CGM measures interstitial glucose, it can lag blood glucose during periods of rapid change — such as during or after eating, during intense exercise, or during hypoglycemia; the lag depends on the device and the situation. Trend arrows show the direction of change but are not a universal correction, so follow your device's instructions if a reading does not match how you feel.

Accuracy depends on the device, glucose range and measurement conditions. A study-wide MARD does not guarantee the accuracy of an individual reading. Follow the current instructions for your specific device and discuss clinical decisions with your healthcare professional.

Sensor placement: Accuracy can vary by body site. Follow manufacturer guidelines for approved insertion sites and rotate locations to avoid tissue buildup that may reduce sensor performance.

Certain medications: Some medications, including high-dose acetaminophen (paracetamol), can interfere with CGM accuracy. Check the device's prescribing information for a complete list of potential interferences.

How to get started with CGM

For people with diabetes

Discuss whether CGM is appropriate, the exact model and interpretation with your healthcare professional. Prescription, availability and coverage rules depend on country, product and plan.

For health optimization without diabetes

Several CGM devices are now available over the counter without a prescription in the United States, including Abbott's Lingo and Dexcom's Stelo — both FDA-cleared as over-the-counter medical devices for adults aged 18 and older (Lingo) or people aged 2 years and older (Stelo, since June 2026) who do not use insulin (that includes people without diabetes). These devices are available directly from manufacturers and major pharmacies. They provide continuous glucose data, and medical decisions based on it should involve a qualified healthcare professional.

Important

If you have diabetes or any other medical condition, do not start using a CGM device or make changes to your treatment based on CGM data without first consulting your healthcare provider. CGM data requires clinical context to interpret safely and act upon.

What to expect

Learning to interpret data requires context. Review trends with appropriate guidance, but do not dismiss an isolated low, symptoms or a reading that differs from a blood glucose check; follow device instructions and your care plan.

Frequently asked questions

Does inserting a CGM sensor hurt?
Application, discomfort and skin reactions vary by person and model. Follow the exact device’s placement and application instructions; a clinician inserts implantable Eversense sensors.
Can I shower, swim, or exercise with a CGM?
Water ratings differ: G7, G7 15 Day and Stelo sensors are rated to 2.4m for up to 24 hours; original Libre 3 and Lingo to 1m for up to 30 minutes. Follow each device’s conditions and display connectivity instructions. Exercise use does not guarantee improved performance or prevention of lows.
How much does CGM cost?
CGM costs vary significantly by device, country, and insurance coverage. For prescription CGMs such as Dexcom G7 and original Libre 3, compare the supplier price with your approved coverage and cost-sharing terms. For OTC devices such as Dexcom Stelo and Abbott Lingo, check the selected pack quantity, maximum labeled wear and current purchase terms; a pack does not guarantee a calendar month of continuous readings. Medicare covers CGM when its criteria are met. For private insurance, ask your insurer whether your plan covers the specific device before buying it.
Do I still need to do finger-stick tests if I use CGM?
Calibration and confirmation requirements depend on the model. Some systems are factory calibrated, while Eversense models require scheduled fingerstick calibration. Follow the exact instructions and care plan for warm-up, unavailable readings, symptoms or blood/sensor mismatch; OTC displays do not replace insulin-treatment monitoring.
What is "time in range" and why does it matter?
Time in range is the proportion of readings within a defined glucose interval. Consensus recommends over 70% in 70–180 mg/dL for many nonpregnant adults with type 1 or type 2 diabetes, individualized with their care team. Time below range, time above range and variability add information that TIR alone does not supply. CGM provides dense sampling, but TIR has also been studied using seven-point fingerstick profiles; it does not replace laboratory HbA1c.