Warm scene: an arm with a small round CGM sensor and a smartphone showing a gentle glucose curve. The sensor measures glucose in the interstitial fluid around the clock.
Sensor on the arm
Live curve
Measure around the clock – without constant finger pricks.
CGM stands for “continuous glucose monitoring”. A small sensor is worn on the skin (usually the upper arm); a fine filament reaches into the tissue beneath. Instead of one-off finger pricks, the system delivers a value every few minutes and draws a continuous curve from them.
What exactly is measured
Important to understand: a CGM doesn’t measure sugar directly in the blood, but in the fluid between the cells (interstitial fluid). From this signal an algorithm calculates the estimated glucose value.1
Why the curve “lags” slightly
Evidence level: established
Because sugar has to pass from the blood into the tissue first, the sensor value follows blood sugar with a small delay. Studies put this physiological lag at about 5 to 10 minutes.12 It’s most noticeable during fast changes – for example right after eating or during exercise.
Two curves: blood (blue) rises first after eating; the interstitial fluid (green) that the sensor measures follows with about five to ten minutes’ delay. That’s why the CGM reading can lag slightly during fast changes.
Why the sensor lags slightly
The sensor measures not in blood, but in interstitial fluid
Time after eating →
approx. 5–10 min lag
Blood (finger prick)
Interstitial fluid (sensor)
How accurate is a CGM?
Accuracy is often given as MARD (mean absolute relative difference from a reference value). Modern systems reach a MARD of about 10%.13 That’s good for everyday use and spotting trends, but in critical situations it doesn’t replace a blood measurement prescribed by your doctor.
Time in Range: the key everyday metric
Evidence level: established (international consensus)
From CGM data you can read how much time the glucose value spends in the target range – the “Time in Range” (TIR). The 2019 international consensus recommends a target range of 70–180 mg/dl for most people with diabetes, in which you should spend at least 70% of the day (about 17 hours).4 Below 70 mg/dl should be under 4%, below 54 mg/dl under 1% of the time.4
The 2019 international consensus recommends that most people with diabetes spend at least 70 percent of the day (about 17 hours) in the 70 to 180 mg/dl target range. Less than 4 percent below 70 and less than 1 percent below 54 mg/dl. 70 percent Time in Range corresponds to roughly an HbA1c of 7 percent.
Time in Range: as much time in target as possible
Recommendation of the 2019 international consensus
above 180 mg/dl – too high
70–180 mg/dl – target rangeat least 70% of the day (~17 h)
below 70 mg/dl – under 4% of the time
below 54 mg/dl – under 1%
70% Time in Range corresponds to roughly an HbA1c of 7%. Individual targets are discussed with your doctor.
Key takeaways
- A CGM measures glucose in interstitial fluid, not directly in blood.
- This creates a physiological lag of about 5–10 minutes – mainly during fast changes.
- Accuracy of modern systems is around a MARD of 10%.
- Time in Range: target 70–180 mg/dl, at least 70% of the day.
- TIR and HbA1c complement each other – daily distribution plus long-term average.
Sources
Sources as of: 21 Jul 2026.
- CGM measures interstitial fluid; physiological lag ~5–10 min; MARD. Clinical Diabetes / review. diabetesjournals.org
- Interstitial vs. blood glucose lag; kinetics. Study. pmc.ncbi.nlm.nih.gov
- CGM accuracy and MARD of modern systems. Review. ncbi.nlm.nih.gov
- Time in Range international consensus 2019: 70–180 mg/dl, ≥70% of day; targets. Diabetes Care. diabetesjournals.org