The continuous glucose monitor went from niche endocrinology tool to consumer product in about three years. A non-diabetic man can now buy one over the counter, stick it on his arm, and watch his glucose move in real time on his phone.
The question we get more than any other about CGMs: is the data useful, or is it just a new way to be anxious about your breakfast?
The honest answer is "both, and which one depends on how you use it."
What is a CGM?
A continuous glucose monitor is a small sensor worn on the back of the upper arm that measures interstitial glucose every one to five minutes for roughly 14 days. It is the same technology used in type 1 diabetes management, repackaged for consumers. Common brands include Dexcom Stelo, Abbott Lingo, Levels, and Nutrisense.
The sensor measures interstitial fluid glucose, not blood glucose directly. The two correlate closely but lag by about 10 to 15 minutes. That lag matters less for trend reading and more if you are trying to time a specific meal response down to the minute.
What does the data actually show
For a metabolically healthy non-diabetic man, the data shows a few consistent patterns:
- Fasting glucose between roughly 75 and 95 mg/dL.
- Post-meal peaks between 110 and 140 mg/dL for most meals.
- Higher peaks (140 to 170 mg/dL) for high-glycemic meals (white rice, juice, dessert) that resolve in 90 to 120 minutes.
- Modest overnight drift, occasional dips into the low 70s.
- Reliable spikes after stress, poor sleep, and certain individual foods that surprise you.
None of this is pathological. A non-diabetic body is supposed to mount a glucose response to food. The peaks are not the problem. The shape of the curve over time, the height of the average, and the frequency of large excursions are where the actual signal lives.
What is hype
A few claims show up in CGM marketing that the data does not support:
- "Spikes cause weight gain." Glucose excursions in a non-diabetic person are not what drives fat gain. Sustained calorie surplus is.
- "Flat lines are optimal." A flat glucose curve in a non-diabetic is not a goal. The body is supposed to respond to food.
- "Your personalized food list will be revealed." Most "spike" responses are reproducible within a person but not magic. A bagel will spike most men. You did not need a sensor to know that.
- "Optimize cognition through glycemic control." The evidence here is thin in non-diabetics.
The marketing collapses the reasonable claim (variability and average can be useful trend data) into the unreasonable claim (flatlining your glucose unlocks a new tier of human performance).
When a 2-week wear gives useful information
There are specific scenarios where we will recommend a CGM:
- Fasting insulin or A1c trending up, but not yet diabetic. The CGM shows post-meal excursions that the A1c lags behind.
- A man planning to start a GLP-1 protocol who wants a baseline glucose pattern before and after.
- An aging endurance athlete sorting out under-fueling versus over-fueling during long sessions.
- Suspected reactive hypoglycemia, where symptoms (shakiness, fatigue) follow meals.
- A man with a strong family history of type 2 diabetes who wants the earliest possible signal.
Outside these cases, the marginal value is lower. A normal man with normal labs wearing a CGM mostly learns that white rice spikes him and steel-cut oats do not. Useful, but not a fortnight of skin-puncture useful.
Common misinterpretations
When we read CGM data with patients, the same misreads show up:
- Treating a 140 mg/dL peak as alarming. It is not, in a non-diabetic, if it resolves in 90 minutes.
- Chasing flat lines with food restriction. A small reduction in average glucose at the cost of a major calorie deficit and lost muscle is a net loss.
- Overreacting to overnight drift. Sleeping near 75 mg/dL is normal physiology, not impending hypoglycemia.
- Ignoring stress and sleep effects. A poor night's sleep raises next-day glucose more than most "bad" foods.
- Mistaking sensor noise for signal. The first 24 hours of any sensor is often inaccurate. The last 24 hours can drift too. Read the middle of the wear.
What numbers are worth tracking
If you wear a CGM, the metrics worth pulling:
- Average glucose for the wear. Under 100 mg/dL is a reasonable target for a non-diabetic.
- Time in range (70 to 140 mg/dL). Above 90% is healthy. Above 95% is excellent.
- Glucose variability (standard deviation or coefficient of variation). Lower is better for most men.
- Largest excursions and what preceded them. This is where the food learning lives.
These four numbers compress two weeks of data into something useful at a clinical visit.
CGM versus other metabolic markers
A CGM is not a substitute for the static labs. Fasting insulin, ApoB, A1c, liver enzymes, and a DEXA tell you more about long-term metabolic trajectory than 14 days of glucose data ever will. The CGM is a complementary read, not a primary one.
We will sometimes use a CGM as the trigger for a deeper metabolic panel if a patient sees excursions that surprise him. Mostly, the lab work comes first and the CGM comes second, if at all.
Where this lands
For most healthy men, a single two-week CGM wear is a useful experiment. You learn what your body actually does with the meals you eat. You stop guessing.
What we do not recommend is wearing one indefinitely. The data anxiety is real. Past the first two weeks, the marginal information curve flattens, and the cost (financial and psychological) starts to outrun the benefit. Wear it. Read it. Take what is useful. Put it down.
If you wear one and the data is clean (average under 100, time in range above 95%, low variability), that is a reasonable answer. You did not need the device to be healthy, but you now have a baseline to compare against in five years.
