Glucose Monitors Are Booming Among Non-Diabetics. The Evidence Isn't There Yet
A 2026 JAMA Internal Medicine review found no evidence that glucose monitors improve outcomes in people without diabetes. Here is what that means before you buy one.
There is a small sensor on a lot of upper arms right now, and behind it is a quiet bet: that if you can see your glucose moving in real time, you will make better choices and live better. It is an appealing bet. It is also, so far, one nobody has shown pays off for people who do not have diabetes.
That is the plain conclusion of a 2026 narrative review in JAMA Internal Medicine, which found no evidence that continuous glucose monitors improve health outcomes in people without diabetes. The timing is awkward, because the devices have never been easier to buy. The FDA cleared over-the-counter CGMs from Dexcom and Abbott in March 2024, and market commentary this year points to adoption among non-diabetic consumers rising by something like 15 percent. That figure comes from industry trend coverage, which is a weaker kind of source than a trial, so hold it loosely.
What the JAMA review found, and what it did not
A narrative review is not a single experiment. It is a researcher gathering the existing literature and asking what it adds up to. The answer here was not that CGMs are harmful, or that they are useless. The answer was narrower and, I think, more honest: the evidence that wearing one improves outcomes in people without diabetes is not there yet.
Those two statements sound alike and are not. "No evidence of benefit" is different from "evidence of no benefit." The first says the question has not been properly answered. The second says it has. We are in the first situation. Nobody has run the long, well-controlled studies that would tell a healthy person whether watching their glucose changes their weight, their blood pressure, their heart disease risk, or their lifespan.
Marketing tends to quietly swap one for the other, and the leap runs the opposite way too. Skeptics who say CGMs are proven worthless are overreaching in the mirror image. What we can say is smaller: the case has not been made.
Why a glucose spike in a healthy person is not what it looks like
The thing that makes a CGM graph feel urgent is the spike: the line climbing after a bowl of rice, a banana, a slice of cake. Wellness content has taught a lot of people to read that climb as damage in progress.
It is closer to ordinary digestion. In someone whose metabolism works normally, blood glucose rises after eating carbohydrate, insulin responds, and the level comes back down, usually within a couple of hours. That is the system doing its job. A rise followed by a return is what a healthy response looks like.
The problem in diabetes is different. Either insulin is not being made, or the body does not respond to it well, so glucose climbs higher and stays high. That sustained elevation is what damages blood vessels and nerves over years. A brief, normal excursion in a metabolically healthy person is not that. Treating the two as the same thing is the core confusion, and it is the one sold most often.
There is also plain variability. Sleep the night before, stress that morning, a hard workout, what else was on the plate, even the time of day can move the same meal's curve. Two readings of the same food can look like two different foods.
What CGMs are actually validated for
Continuous glucose monitors were built for diabetes management, especially for people who use insulin and need to avoid dangerous highs and lows. In that setting they have a real track record, and for many patients they are close to life-changing. The research base for that use is substantial.
The step from there to general wellness optimization is the one that has not been tested. It assumes that tracking a healthy person's glucose curve and tweaking food accordingly produces a better outcome than eating sensibly without the data. That is a hypothesis. It may even turn out to be true for some people. But a reasonable hypothesis is not a result.
The sensors also have limits worth knowing. They measure glucose in the fluid beneath the skin, not in blood directly, so readings lag and carry some error. That is fine for spotting trends in someone managing diabetes. It is a lot of precision to read into a small wobble in a healthy person.
A reality check: what is proven versus what the marketing implies
| What CGMs are proven to do | What the marketing implies |
|---|---|
| Help people with diabetes, especially those on insulin, track glucose and avoid dangerous highs and lows. Source: FDA indications and long-standing diabetes guidelines. | Optimize the glucose of anyone, diabetic or not, toward better health. Source: no outcome trial in non-diabetics; the 2026 JAMA Internal Medicine review found no evidence of improved outcomes. |
| Show how glucose moves across a day, including after meals. Source: sensor design and validation studies. | Reveal which foods are "bad for you" based on your personal spikes. Source: spikes in healthy people are normal digestion; no evidence that avoiding them improves outcomes. |
| Be cleared for over-the-counter purchase by adults not on insulin. Source: FDA, March 2024. | Be proven useful for general wellness because they are cleared. Source: clearance concerns safety and accuracy as a device, not that using one improves health. |
| Give motivated users feedback that can feel instructive. Source: self-reports and early small studies. | Produce lasting dietary change and better long-term health. Source: not established. |
The third row is the one I would underline. A clearance tells you the device is reasonably safe and measures what it says it measures. It does not tell you that wearing it will make you healthier.
The real cost: money, and the quieter kind
The money is the easy part to count. These sensors are replaced every couple of weeks, and the subscription-style pricing adds up to something on the order of a hundred dollars a month, though prices change, so check the current number before you commit. That is real money spent on data of unproven value.
The other cost is attention. There is a pattern people who have worn sleep trackers will recognize, sometimes called orthosomnia: a person becomes anxious about their sleep score, sleeps worse because of the anxiety, and the number gets worse, which feeds the worry. Glucose has a version of that. Data without clinical context is easy to misread, and a healthy person watching a graph can turn a normal bump into a small crisis.
I have watched this loop in my own life with other numbers. Some metrics inform you. Others install a low hum of vigilance you did not have before. It is worth asking, before wearing a sensor, which kind you are inviting in.
There is a quieter food cost as well. If a graph makes oats, fruit, or a shared dessert feel dangerous, you can end up with a more anxious relationship to eating without any measurable gain.
Where a CGM might still make sense for a healthy person
None of this makes the devices a scam. There are reasonable uses, as long as they are held modestly.
- Short-term curiosity. Wearing one for a single two-week sensor to see how your own body responds to a few meals can be educational, as long as you treat it as learning and not as a verdict.
- Prediabetes risk, or a strong family history. Here a conversation with a clinician comes first. Standard blood tests, such as fasting glucose and A1C, are cheap and well validated. A CGM might be a supplement under guidance, not a substitute for them.
- A clear question. "Does my glucose drop oddly after exercise?" is a question a sensor can help answer. "Am I optimizing?" is not.
And where it does not make sense: wearing one continuously for months, with no question and no plan, as a source of daily judgment about what you eat.
What to do with the impulse
The impulse behind the purchase is a good one. People want to understand their bodies. If that is you, the unglamorous inputs have far stronger evidence behind them: regular movement, enough sleep, fiber, protein, and not eating a large refined meal and then sitting still for hours. A walk after dinner does more for most people's glucose than any graph about it.
As a software engineer I am drawn to instrumentation. More data feels like more control. But in systems and in bodies, measuring something is only useful when you know what decision the measurement is supposed to change. A dashboard without a decision attached is just a dashboard.
FAQ
Does the JAMA review say CGMs are bad for healthy people?
No. It found no evidence that they improve outcomes in people without diabetes. That is not the same as showing they cause harm or that they will never help.
Is a glucose spike after eating dangerous?
In a metabolically healthy person, a rise after a meal followed by a return to baseline is normal. The concern in diabetes is a sustained high level, which is a different pattern.
Should I ask my doctor before buying one?
If you have prediabetes, a family history, or symptoms that worry you, yes. Standard blood tests are cheap, validated, and a better first step than a sensor.
Are over-the-counter CGMs accurate?
They are cleared by the FDA and measure glucose in the fluid under the skin, so readings lag and carry some error. They are reasonable for spotting trends, less so for reading meaning into small fluctuations.
Is a short trial worth it?
For curiosity, possibly. Go in with a specific question, wear it for one sensor cycle, and treat what you see as information about one stretch of your life, not as a rulebook.