Continuous Glucose Monitor

Why glucose is worth looking at after 45

Metabolic drift in midlife is not a character flaw. It is the ordinary arithmetic of losing muscle, sleeping worse, moving less and — for many people — changing hormonal environment.

The scale of it is easy to underestimate. The CDC's most recent national estimates put 115.2 million U.S. adults in the prediabetes range, including 31.3 million people aged 65 and older — 52.1% of that age group. Prediabetes is, by design, the stage that produces no symptoms. Most people carrying it are not aware of it.

Age also changes the shape of the curve, not just the average. A March 2026 analysis in Nature Communications looked at CGM data from 3,634 adults who did not have diabetes or prediabetes, and found that older participants had measurably greater glycemic variability and lower time in a tight range than younger ones — with sex, sleep duration, and carbohydrate and protein intake all showing independent associations. Notably, the same analysis found that the measured factors explained only a modest share of the variation between people. Glucose patterns are real, they shift with age, and they are not fully explained by anything we can put on a questionnaire.

For women, the perimenopausal transition adds its own layer — changes in body composition, sleep and insulin sensitivity that often arrive before anything shows up on an annual fasting glucose. For men, declining testosterone tracks with the same visceral fat and insulin-sensitivity story. In both cases the standard once-a-year blood draw is a poor instrument for catching a moving target.

What the evidence actually says

This is the section most CGM pages leave out, so we will be direct.

In people who already have a glucose problem, CGM has real support. A systematic review and meta-analysis published in the European Journal of Medical Research in January 2026 pooled 23 studies covering 1,074 participants without diabetes and found a significant improvement in mean glucose (SMD −0.54). But the finding underneath the headline is the one that matters: the benefit appeared in people with prediabetes, and no appreciable glycemic benefit was observed in healthy, normoglycemic participants. The same analysis found no significant change in BMI. The authors' own framing is that a CGM works as "a precision biofeedback tool integrated within structured lifestyle programs" — not as a weight-loss intervention on its own.

In metabolically healthy people, the honest answer is that we do not know. A 2026 narrative review in JAMA Internal Medicine examined CGM use beyond diabetes and concluded there is no scientific evidence that continuous glucose monitoring improves health or prevents disease in people without diabetes. The authors went further and named a specific risk: normal glucose fluctuations being over-interpreted, leading to unnecessary dietary restriction or anxiety. That is a real harm, and it is more likely when a device arrives in the mail with an app that labels ordinary rises as problems.

The professional guidelines have not moved ahead of the data. The American Diabetes Association's Standards of Care in Diabetes—2026 recommends CGM for people with diabetes on insulin or on therapies that can cause hypoglycemia. It acknowledges that over-the-counter monitors are now available to anyone, including people without diabetes or with prediabetes who want to assess their glycemic responses — but it does not recommend the practice as standard care, because the outcome evidence is not there yet.

So here is our position. A CGM is a strong tool for a specific person: an adult in or near the prediabetes range, or someone on a metabolic or weight program where the data will change a decision. It is a weak tool for a metabolically healthy person looking for something to optimize. We would rather tell you that before you wear one than after.

Who this is likely to help — and who it probably isn't

Situation by situation, this is where a wear usually earns its place — and where it usually does not.

What we look at, and what we deliberately don't

When your wear is finished, the pattern is most useful read alongside your labs. In practice, four things carry most of the signal.

Where you spend your time. How much of the day sits in a steady band versus swinging. Time-in-range is a well-established metric in diabetes care; the standard 70–180 mg/dL band was validated for people being treated for diabetes and does not transfer cleanly to someone without it, which is why your own pattern matters more than a target designed for a different population.

How far you travel after meals, and how fast you come back. The height of the peak matters less than how long the elevation lasts and whether it settles. Two people can hit the same number and have very different curves.

Overnight and early morning. The fasting reading your annual draw captures is one point on an eight-hour line. Seeing the whole line sometimes explains a "normal" fasting glucose that never felt normal.

Day-to-day variability. How much the same breakfast produces a different result on a bad night's sleep, a travel week, or a stressful stretch. This is often the most actionable finding and the one a single blood draw can never show.

What we do not do is hand you a target to chase. We do not set glucose goals for people without diabetes, we do not tell you to eliminate foods based on a single spike, and we do not treat the app's color coding as clinical guidance. A CGM that makes you afraid of an apple has made your health worse, not better.

Alongside a weight or metabolic program

If you are on or considering a GLP-1, a CGM answers questions the scale cannot. Weight is a slow, noisy signal that moves for a dozen reasons. Glucose response moves earlier. Seeing both lets your clinician tell whether the metabolic side of the program is doing what it should, and whether symptoms you are attributing to the medication are actually glucose-related — which, more often than people expect, they are not.

It also does something quieter and more useful: it tells you which of your ordinary meals your body handles well. Most people finish a wear with two or three durable changes rather than a new set of rules, and durable is the entire point.

Alongside hormone therapy

Glucose handling and hormonal status are not separate systems. Where someone is being worked up for perimenopausal symptoms, or is early in testosterone therapy, glucose patterns are part of the picture — not because a CGM diagnoses anything hormonal, but because reading the two together sometimes explains a symptom that neither explains alone. This only works if both are on the table at the same time, which is the argument for measuring before treating rather than after.

What a CGM cannot do

It cannot diagnose diabetes or prediabetes. Those diagnoses are made from A1c, fasting plasma glucose or an oral glucose tolerance test, and a CGM does not replace any of them. If your pattern suggests a problem, the next step is a confirmatory blood test, not a longer wear.

It cannot tell you whether a food is good or bad. It shows how your body handled that food, that day, in that context.

It cannot substitute for a relationship with a clinician who knows your history. That comes with a treatment program, where your prescribing clinician uses the data alongside the labs your treatment depends on.