📚 Part of our Blood Sugar Guide — an evidence-based guide.
Continuous glucose monitors have moved from a diabetes device to a wellness product. In the US you can now buy one over the counter, wear it for two weeks, and watch your blood sugar respond to everything you eat. The question is whether that information makes you healthier — and the honest answer is that it depends heavily on what you do with it.
What a CGM actually measures
A small filament sits under the skin and samples glucose in interstitial fluid — the fluid between cells — not blood. That distinction matters more than the marketing suggests.
Interstitial glucose lags blood glucose by roughly 5 to 15 minutes, and the gap widens when levels are changing fast. So the spike you see after a meal is real but time-shifted, and a reading during a rapid rise or fall can be meaningfully off.
Accuracy is also imperfect. Consumer sensors typically run within about 8–10% of a laboratory value on average, with individual readings sometimes further out. Sensors can differ from each other on the same arm, on the same day. Pressure on the sensor during sleep produces false lows — the well-known “compression low” that alarms people at 3 a.m.
What the data actually shows in people without diabetes
Here is the part most CGM marketing omits: glucose spikes after meals are normal.
Studies placing CGMs on healthy adults have found post-meal peaks frequently reaching 140 mg/dL and occasionally higher, in people with entirely normal A1C and no metabolic disease. Seeing 150 after a bowl of rice is not evidence of a problem.
There is also no established threshold in people without diabetes at which a given spike causes harm. The claim that “flattening your curve” prevents disease is plausible — but it is extrapolated from diabetes research, not demonstrated in healthy people. No trial has shown that non-diabetic adults who minimise glucose variability live longer or get less heart disease.
That gap between what is measured and what is proven is where a lot of anxiety and a lot of money currently sit.
Where a CGM genuinely helps
If you have prediabetes. This is the strongest case. It makes an invisible condition visible, and the feedback loop — eat, walk, watch the curve change — changes behaviour in a way that a once-a-year A1C never does. Roughly one in three US adults has prediabetes and most do not know it.
As a two-week experiment, not a permanent habit. Most of the useful information arrives in the first fortnight: which of your regular meals spike you hardest, whether a walk after dinner blunts them, what a bad night’s sleep does to your morning numbers.
To test specific interventions. Eating vegetables and protein before starch (meal order), walking for ten minutes after eating (the trial), or identifying whether your morning highs come from the dawn phenomenon or something else (how to tell).
If you have unexplained symptoms — shakiness, fatigue or palpitations a few hours after eating — where the timing is genuinely informative.
Where it does harm
Food fear. The most common bad outcome is not a wrong reading but a wrong reaction: eliminating fruit, oats, beans or whole grains because they produce a visible bump. These are foods with strong evidence for long-term metabolic and cardiovascular benefit. Trading them for something that spikes less is a poor exchange.
In people with disordered eating, a device generating continuous numerical judgement about food is a genuine risk, not a theoretical one.
Chasing noise. Sensor variability, compression lows and the interstitial lag all produce readings that look like signal and are not.
Cost without a decision attached. Ongoing subscriptions run to hundreds of dollars a year. If the reading is not going to change what you do, it is entertainment.
How to use one well, if you use one
Decide in advance what questions you are answering, and wear it for two to four weeks rather than indefinitely. Judge meals by the whole pattern — how high, how long, and how fast it returns to baseline — not by the peak alone. Test one variable at a time. Expect spikes after carbohydrate; that is physiology, not failure. And weigh it against a $10 A1C test, which for most people without symptoms answers the underlying question adequately (what A1C means).
When to see a doctor
See a doctor rather than buying a sensor if you have symptoms of high blood sugar — excessive thirst, frequent urination, unexplained weight loss, blurred vision. A CGM is not a diagnostic tool for diabetes; A1C, fasting glucose and OGTT are. And if a sensor consistently shows fasting readings above 100 or frequent readings above 140, take that to a doctor for proper testing rather than acting on it alone.
The honest bottom line
A CGM is a good teaching tool and a poor permanent accessory. It genuinely helps people with prediabetes and anyone willing to run a structured two-week experiment. For a metabolically healthy adult, it mostly demonstrates that food raises blood sugar — which is normal — and the claim that flattening those normal spikes prevents disease remains unproven. Borrow the lesson, then take the sensor off.
Frequently asked questions
Should healthy people use a continuous glucose monitor?
There's little strong evidence that CGMs improve health in people without diabetes. They can be educational and motivating, but they can also cause anxiety and over-restriction, and they cost money. It's optional, not essential.
What can a CGM show a non-diabetic person?
It reveals how your blood sugar responds to specific foods, meals, exercise, sleep, and stress in real time, which some people find useful for spotting personal patterns.
Are glucose spikes bad if I don't have diabetes?
Some rise after meals is completely normal in healthy people. Large, frequent spikes may matter over time, but a single reading out of context isn't a diagnosis. Focus on overall diet quality rather than chasing a flat line.