Most people get a CGM, wear it for a week, and end up more anxious than they started. Every meal produces a line that goes up. Some of the lines look alarming. Nothing on the screen tells you which ones matter.
The good news is that the interpretation is simpler than the display suggests. Four ideas cover almost everything you need, and once you have them, the graph stops being a source of worry and becomes the most useful feedback tool available for managing type 2 diabetes.
What the number actually is
A CGM does not measure the glucose in your blood. It measures glucose in the interstitial fluid just under the skin, and glucose reaches that fluid a few minutes after it reaches the bloodstream. In practice the sensor runs roughly five to fifteen minutes behind reality.
This matters more than it sounds. When you watch a sharp drop on the graph, the actual number in your blood has usually already leveled off. When you see a rise begin, it began before you saw it. People who treat the display as live data end up reacting to something that has already happened.
The same lag explains the most common complaint about CGMs, which is that a finger-stick reading disagrees with the sensor. A gap of 15 to 20 mg/dL (about 1 mmol/L) between the two is normal, particularly while glucose is moving quickly. Modern sensors are accurate to within roughly 8 to 10 percent of a lab value on average. That is good enough to see patterns and not good enough to argue over single digits.
What a normal curve looks like
A person without diabetes does not have a flat line. They eat, their glucose rises, and it comes back down. The useful reference points:
- Fasting and pre-meal glucose sits roughly between 70 and 99 mg/dL (3.9–5.5 mmol/L).
- After a meal, the peak usually stays under 140 mg/dL (7.8 mmol/L).
- Glucose returns close to the starting point within two to three hours.
That last point is the one people miss. The shape of a healthy curve is a rounded hill that comes back down, not a line that never moves. If you are aiming for no rise at all after eating, you are aiming at something a person without diabetes does not achieve either.
Height is not the point. Duration is.
This is the single most useful reframe for a new CGM user. What damages tissue over time is not touching a high number, it is the total amount of exposure to high glucose — how high, multiplied by how long.
A peak of 180 mg/dL that is back to baseline within ninety minutes represents far less exposure than a reading of 150 that sits there for four hours. The first looks worse on the graph. The second is the one worth changing.
So when you review a day, do not scan for the highest point. Look for the flat stretches spent above your target, and look at how long each meal takes to come back down. A meal that keeps you elevated for four hours is telling you something. A brief sharp spike that resolves quickly is usually telling you what you already know, which is that you ate carbohydrate.
Time in range: the number worth watching
Your app already calculates this, and it is more informative than any individual reading. Time in range is the percentage of the day your glucose spends between 70 and 180 mg/dL (3.9–10.0 mmol/L).
An international consensus panel published targets in 2019 that are still the standard reference for adults with type 1 or type 2 diabetes:
| Measure | Range | Target |
|---|---|---|
| Time in range | 70–180 mg/dL (3.9–10.0 mmol/L) | More than 70% of the day (about 17 hours) |
| Time above range | Above 180 mg/dL (10.0 mmol/L) | Less than 25% |
| Time well above range | Above 250 mg/dL (13.9 mmol/L) | Less than 5% |
| Time below range | Below 70 mg/dL (3.9 mmol/L) | Less than 4% |
| Time well below range | Below 54 mg/dL (3.0 mmol/L) | Less than 1% |
Time in range tracks closely with A1C, and it arrives weeks earlier. Roughly, each 10 percentage points of time in range corresponds to about a 0.5 percent difference in A1C — a relationship derived largely from type 1 diabetes data, so treat it as a guide to direction rather than a precise conversion for type 2.
Two honest qualifications. Older adults and people at higher risk of hypoglycemia are given looser targets on purpose, because avoiding lows matters more than perfecting highs. And if you are not on insulin or a sulfonylurea, your time below range will usually be close to zero anyway, so the number that will move for you is the one above 180.
Watch the trend, not the day. A single day's time in range tells you about that day's food and sleep. Compare this week to last week instead. Anything shorter than a week is noise dressed up as information.
The other numbers on the report
Most CGM apps produce a summary report. Three items on it are worth understanding:
- GMI, or glucose management indicator. An estimate of what your A1C would be, based on your average sensor glucose. It is useful between blood tests, and it will not match your lab A1C exactly. A gap of a few tenths in either direction is normal.
- Average glucose. Straightforward, but it hides the shape. Two people with the same average can have very different days, one steady and one swinging.
- Coefficient of variation, sometimes shown as variability. This measures how much your glucose swings around its average. Below 36 percent is the usual target. Falling variability is a genuine sign of improvement even when your average has not moved yet.
Readings that are not real
Some of what you see is the sensor, not you. Knowing these prevents a lot of unnecessary alarm:
- Compression lows. If you sleep on the sensor, pressure restricts flow in the tissue underneath and the reading drops, sometimes steeply. The signature is a sudden overnight low that recovers on its own the moment you roll over. If you feel fine and it corrects itself, that is almost always what it was. Wearing the sensor on the arm you do not sleep on solves it.
- The first day. Readings are often noisier in the first 12 to 24 hours after insertion, while the sensor settles into the tissue. Do not draw conclusions from day one.
- Rapid movement. The 5 to 15 minute lag is largest when glucose is changing fast, which is exactly when you are most tempted to check the number.
If a reading is genuinely low and you feel it, treat it and confirm with a finger-stick. If a reading is low and you feel completely normal, check the shape before you act.
A higher morning number is not a mistake
Many people find their highest fasting reading of the week is the one taken before breakfast, and assume something went wrong overnight. Usually nothing did. In the hours before waking, the body releases cortisol and growth hormone to get you going, and both raise glucose. It is called the dawn phenomenon, it happens to people without diabetes too, and it is more pronounced when insulin resistance is present.
Worth noticing, not worth fixing on your own. If your morning numbers are consistently high, that is a conversation with your doctor about timing or medication, not a reason to eat less dinner.
How to actually use it
A CGM only earns its cost if it changes something. The way it does that is by testing one variable at a time against a meal you already eat. Three experiments are worth running in your first month:
- Food order. Eat your usual meal, note the curve. The next time, eat the vegetables and protein first and the starch last. Same food, same amount, different order. The rise is typically noticeably smaller.
- A walk after eating. Same meal on two days. On the second, walk for ten to fifteen minutes starting shortly after you finish. Working muscle takes up glucose without needing insulin, and the difference on the graph is usually obvious.
- Swapping one starch. White rice one day, the same volume of brown rice, beans, or lentils the next. This is where people find out that the food they assumed was the problem was not, and the one they never suspected was.
Three rules make these useful. Change one thing at a time. Repeat each test at least twice, because sleep, stress and illness all move glucose independently of food. And write down what you ate, because you will not remember by the time the pattern emerges.
What to skip. Do not check the app after every meal, do not try to eliminate every rise, and do not judge a food on a single reading. The goal is a set of habits that hold for a year, not a flat line for a weekend.
Where your targets come from
The numbers above are population targets. Yours should be set with your doctor, and they may differ — looser if you are older or at risk of lows, tighter if you are young and newly diagnosed. This matters most if you take insulin or a sulfonylurea, because as your readings improve those medications can drop you too low and the dose may need adjusting. Do not make that change on your own.
Closing thought
A CGM is the only tool that shows you, on the same day, what a specific decision did to your body. That is worth a great deal, but only if you use it to run experiments rather than to grade yourself. The graph is information, not a verdict.
If you take one thing from this page: stop watching the peaks and start watching how long you stay up. Then pick one of the three experiments above and run it this week.
Free resource. The Blood Sugar Wellness Bundle includes a seven-day meal structure and a movement framework built around the same principles — food order, post-meal movement, and one change at a time. No cost, no purchase required.