A coin-sized sensor on the back of your arm, streaming your glucose to your phone every few minutes, around the clock. A few years ago that was specialist medical equipment. Today, in the US, you can buy one over the counter. Continuous glucose monitors — CGMs — are having a moment, and with the moment comes a lot of noise. Here's what one actually tells you, calmly.
Most people aren't told this, but a CGM is not a tiny blood-glucose lab strapped to your arm. It's a different kind of measurement, with real strengths, honest limitations, and one growing area where the marketing has moved faster than the evidence. Let's take them in order.
What a CGM actually measures (it's not blood)
A CGM's thin filament doesn't sit in your bloodstream. It sits just under the skin, in the fluid that surrounds your cells — the interstitial fluid. Glucose has to drift from your blood into that fluid before the sensor sees it, which makes interstitial glucose a related-but-distinct signal from the blood value a fingerstick or lab test gives you (Cengiz & Tamborlane 2009).
Two practical consequences follow. First, there's a time lag: one careful analysis found sensor readings trail blood glucose by around 9–10 minutes on average, partly physiology and partly the device itself (Schmelzeisen-Redeker 2015). When your glucose is changing quickly — during a meal, exercise, or a rapid drop — the CGM is describing the recent past, not this exact second.
Second, there's a margin of error. In the validation study of one widely used factory-calibrated sensor, readings differed from reference values by about 11% on average (Bailey 2015); newer sensors have improved on that, but no CGM is lab-grade. A reading of 104 versus 112 is not a meaningful difference. The magic of a CGM was never any single number — it's the shape of the line: which way you're heading, how fast, and how often you visit the high and low neighborhoods.
A CGM measures glucose in the fluid under your skin, lagging blood by several minutes, with a real margin of error. It's a trend instrument, not a lab test. Read the curve, not the digits.
What the numbers mean
Context keeps CGM data calm instead of alarming. When researchers put CGMs on 153 healthy people without diabetes, average glucose sat around 98–99 mg/dL and they spent a median of 96% of the time between 70 and 140 mg/dL (Shah 2019). If you think in mmol/L rather than mg/dL, our glucose unit converter will translate those figures for you. Notice what that also says: even metabolically healthy people spend some time outside that band. A rise after a meal is not a malfunction. It's what eating looks like.
For people managing diabetes, the key CGM metric is time in range. An international consensus recommends that most adults with type 1 or type 2 diabetes aim to spend more than 70% of the day between 70 and 180 mg/dL, with only small slivers of time below 70 (Battelino 2019). Worth being precise here, because wellness apps blur it: 70–180 is the clinical target range for diabetes management; 70–140 is simply where healthy people happen to spend most of their time. They answer different questions, and holding yourself to the wrong one is a recipe for unnecessary worry.
Where CGMs genuinely earn their keep
The strongest evidence is in people who use insulin. In the DIAMOND randomized trial, adults with type 1 diabetes on injections who wore a CGM lowered their HbA1c by about 1 percentage point over 24 weeks — roughly 0.6 points more than the fingerstick group (Beck 2017). In the MOBILE trial, adults with type 2 diabetes on basal insulin saw HbA1c fall from 9.1% to 8.0% with CGM versus 9.0% to 8.4% with a standard meter, and time in range improved by about 15 percentage points (Martens 2021).
That makes sense: when you dose insulin, seeing the direction and speed of your glucose — and getting an alarm before a low — is directly actionable, day after day. For insulin users, a CGM is a genuinely well-supported tool, and how to use one is a conversation to have with your care team.
The honest part: CGMs without diabetes
Here's where we'd rather be straight with you than exciting. For people without diabetes, the evidence that wearing a CGM improves health is thin. A 2023 review put it plainly: outside of diabetes there is very little consensus on what the data means, and whether CGM feedback actually changes behavior in healthy users is simply not known yet (Klonoff 2023).
There's also a quieter risk worth naming: over-interpretation. If you don't know that healthy people spike after meals, that sensors read the past, and that an 11% error band exists, a CGM can turn every sandwich into a verdict. We've seen this pattern before with other health wearables — data without context tends to produce anxiety, not insight. Your body isn't broken because a line went up after lunch.
None of this means curiosity is wrong. Short-term CGM use can genuinely show you your patterns — how your glucose responds to a walk after dinner, a short night's sleep, or the order you eat a meal in. Just hold it as an experiment, not a report card.
If you're curious anyway: a calm starter's guide
- Know your options. In the US, over-the-counter sensors (such as Dexcom's Stelo and Abbott's Lingo) are aimed at people not using insulin, alongside the prescription systems used in diabetes care. If you have diabetes — especially on insulin or medications that can cause lows — go the prescription route with your clinician, since alarms and accuracy standards differ.
- Expect noisy edges. The first day of a new sensor tends to read less reliably, and sleeping on the sensor arm can produce false overnight lows ("compression lows"). One odd reading, especially at night, is not an emergency by itself — but real symptoms of low blood sugar always outrank the gadget.
- Watch weeks, not minutes. Averages, overnight patterns, and how quickly you return to baseline after meals carry more signal than any single spike.
- Don't self-diagnose. A CGM is not a diagnostic tool. If your data consistently looks unusual — fasting values drifting high, long elevated stretches — that's a fasting glucose and HbA1c conversation with your doctor, who can see the whole picture.
For insulin users, CGMs have solid trial evidence behind them. For everyone else, they're an interesting mirror with unproven benefits — useful for learning your patterns, risky only if you treat every wiggle as a verdict.
The calm summary
A CGM reads glucose in the fluid under your skin, a few minutes behind your blood and with a real margin of error — which makes it a superb trend instrument and a mediocre oracle. Healthy people spend most of their time between 70 and 140 mg/dL and still spike after meals; people managing diabetes aim for more than 70% of the day between 70 and 180. The proven wins belong to insulin users. For the curious, a CGM can teach you how your own habits move your glucose — as long as you read it like a weather chart, not a judgment.
This is education, not medical advice. Decisions about monitoring — and anything your data seems to be telling you — belong in a conversation with your healthcare provider.
Calmly, Mr Blood Sugar.