A few years ago, a continuous glucose monitor was something you wore because a doctor told you to. Today it is something people buy the way they buy a sleep tracker: a small disc on the back of the arm, a phone app streaming a live glucose line, and a growing conviction that watching that line is the key to metabolic health. The devices are now sold direct to consumers who have never had an abnormal blood sugar test in their life.1 The obvious question, and the one almost nobody answers honestly, is whether a healthy person actually gets anything out of it.
The short version: a CGM is a genuinely useful learning tool for a few weeks, and a device of rapidly diminishing returns after that. The longer version is worth understanding, because the way these devices are marketed encourages exactly the wrong mental model of what your glucose is doing.
What a CGM Actually Measures
A continuous glucose monitor does not measure blood glucose. It measures glucose in the interstitial fluid, the thin layer of fluid surrounding the cells just under your skin. Glucose diffuses from your blood into that fluid, which means the sensor reading trails your actual blood glucose by roughly 5 to 15 minutes.2 In accuracy studies, current sensors differ from lab-measured blood glucose by about 8 to 9 percent on average, and about 1 reading in 20 is off by more than 20 percent.3 This lag and error are harmless when you are looking at trends, but they are the reason you should never treat a single CGM number as a precise measurement. The value of the device is the shape of the line over hours, not any one point on it.
That distinction matters most in exactly the population now buying these devices. In someone with diabetes, the glucose swings are large enough that even an imperfect sensor tells a clear story. In a metabolically healthy person, the swings are small, so sensor noise makes up a much larger share of what you see. A jagged 15-point wiggle that looks alarming on the app may be within the margin of error of the device itself.3
What "Normal" Looks Like When You're Not Diabetic
Here is the part the marketing tends to skip. In people without diabetes, glucose is a tightly regulated variable. In healthy people it sits between about 70 and 140 mg/dL for roughly 96% of the day, averaging around 100 mg/dL.4 After a meal it rises, and how high depends on what you ate, but a typical healthy peak stays under roughly 140 mg/dL, arrives within about an hour of eating, and then falls back. Overnight, it settles into a flat, low baseline.5
Crucially, a spike is not a disease. When a healthy person eats a bagel with jam, their glucose will climb, sometimes past 140 mg/dL for a short window, and then insulin brings it back.64 That is the system working, not failing. A great deal of CGM anxiety comes from people discovering that food raises blood sugar, which is what food is supposed to do. The meaningful signals are not individual peaks but patterns: peaks that are consistently very high, peaks that take a long time to come back down, or a fasting and overnight baseline that keeps creeping upward over weeks. Those patterns are worth paying attention to. A single dramatic line after dessert is not.
The Case For Wearing One (For a While)
None of this means a CGM is useless for a healthy person. Its real value is as a short, personal experiment. Population averages tell you that white rice raises glucose more than lentils; they do not tell you how your body handles a specific breakfast, or what a poor night of sleep does to your glucose the next morning.7 A CGM turns those abstractions into your own data.
Worn for two to four weeks, a monitor can teach a few things that are genuinely actionable. You learn which of your habitual meals produce an outsized response, and often the fix is small, such as eating protein and vegetables before the starch, or walking for ten minutes after eating, both of which visibly blunt the post-meal rise.89 You learn that a short walk after dinner can flatten a peak that would otherwise last an hour. You may see that a night of bad sleep makes your glucose rise more after breakfast the next morning, a useful reminder that metabolic health is not only about food.10 For someone who suspects their diet is worse than they admit, seeing the line respond in real time can be a useful behavior-change tool, though trials of CGM feedback, mostly in people with diabetes, show modest effects.11
The key word is learn. Once you have mapped your own responses, the device is largely repeating lessons you already have. This is why the honest recommendation for most healthy people is a short stint, not a permanent subscription. The metabolically healthy body does not change its glucose handling much week to week, so after the first month there is little new to see.
Where a CGM Falls Short as a Standalone Signal
The biggest limitation is that glucose is a downstream marker. By the time your fasting glucose or your post-meal peaks are clearly abnormal, insulin resistance has usually been developing for years. The body defends normal glucose by pushing out more and more insulin, so glucose can look fine on a CGM long after the underlying system has started to struggle.12 This is why a CGM cannot, on its own, tell you whether your metabolism is healthy, and why guidelines do not use it to screen for or diagnose prediabetes.13 It is measuring the variable your body works hardest to keep normal.
The markers that move earlier live in a standard blood panel, not on your arm. Fasting insulin rises before fasting glucose does. The HOMA-IR score, which combines fasting glucose and fasting insulin, flags insulin resistance well before glucose crosses any diabetic threshold; you can estimate yours with the HOMA-IR calculator.12 And HbA1c, which reflects your average glucose over roughly three months, is one of the markers clinicians actually use to define prediabetes.13 A CGM is a real-time window; these blood markers are the longer memory. Read together they are far more informative than either alone.
How to Get Real Value Without Overreacting
If you decide to try one, treat it as an experiment with a start and an end date. Wear it for two to four weeks, not indefinitely. Spend the first few days just observing your normal patterns before you start changing anything. Then run small tests: the same breakfast with and without a post-meal walk, or protein-first versus carb-first ordering, and watch the difference. Pay attention to the overnight and fasting baseline, which is more informative than any single daytime spike. And resist the urge to chase a perfectly flat line, because aggressively avoiding every glucose rise can push people toward needlessly restrictive eating, which is its own problem.
Then pair what you learned with a proper blood panel. If your CGM showed consistently high peaks or a rising baseline, that is a reason to check fasting insulin, HbA1c, and a fasting glucose, and to read them against functional ranges rather than just the lab's disease-detection cutoffs. If your numbers were unremarkable and you now understand your own responses, you have gotten what the device had to offer, and you can take it off.
This is exactly the gap Merios is built to close: it keeps your lab panels, your HbA1c trend, and your wearable data in one place, so a two-week glucose experiment becomes a data point in the long arc of your metabolic health rather than a stressful line you watch in isolation. The CGM shows you today; your bloodwork trend shows you the direction. The direction is what matters.
The Bottom Line
For a person without diabetes, a continuous glucose monitor is a good teacher and a poor permanent habit. A few weeks of wearing one can genuinely change how you eat, move, and sleep, because it makes the consequences visible. But glucose is the variable your body guards most carefully, so a normal CGM line is not a clean bill of metabolic health, and an occasional spike is not a warning of disease. Use it to learn your patterns, act on the big ones, confirm the picture with the blood markers that move earlier, and then stop staring at the line. Your metabolism is a trend, not a livestream.