Continuous glucose monitors for non-diabetics: what the data can and can't tell you

Here is the short version before you spend a few hundred dollars sticking a sensor in your arm. For a metabolically healthy person, a continuous glucose monitor mostly proves a thing your body was always doing: glucose goes up after you eat and comes back down. The wellness industry has turned that ordinary fact into a moral panic about "spikes", and the evidence on healthy people does not support the panic.

CGMs are a fascinating biofeedback gadget and, occasionally, a useful self-experiment. They are not a validated longevity tool for people without diabetes, they can be meaningfully inaccurate, and for some users the monitoring itself becomes the problem. This is informational only, not medical advice, and diagnosed diabetes is a completely different situation we will get to at the end.

How a CGM actually works (and why it matters)

A CGM does not measure the sugar in your blood. It measures glucose in the interstitial fluid, the watery space around the cells just under your skin. That is a proxy for blood glucose, not the same thing, so readings lag and can drift from a finger-prick value, especially when glucose is moving fast, like right after a meal or during a hard session.

For someone with diabetes managing insulin, that proxy is still enormously useful and the alarms are life-saving. For a healthy person obsessing over a line on an app, the gap between "interstitial estimate" and "actual blood sugar" is exactly where the wellness story starts to wobble.

Spikes aren't the enemy

The biggest myth, repeated everywhere, is that any post-meal rise is harmful and your job is to flatten every curve. That is not what the data on healthy people says.

Transient rises after eating are normal physiology. The 2018 "glucotype" research in PLOS Biology found that even people who were normal by standard lab measures showed real spread: the subset classed as "severe glucotype" (about a quarter of the normal-by-bloodwork group) reached prediabetic glucose levels up to roughly 15% of the time and diabetic levels around 2% of the time on a CGM. That is a worst-end subgroup, not the average healthy person, and it is a snapshot of spike intensity, not a diagnosis. A 2025 scoping review put the broader point bluntly: online sources frame spikes as inherently bad, but the medical literature ties health risk to chronically frequent, large swings over the long term, not the occasional bagel.

The harm signal in the research is chronic, frequent, large swings over years, not a single number after lunch.

So when an influencer tells you your oat smoothie "spiked" you, the honest response is: yes, food does that, and most healthy people stay in normal ranges overall anyway.

The accuracy problem

Here is where the wellness use case really strains. A 2025 University of Bath study put consumer CGMs against finger-prick testing in healthy adults, and the gap was not subtle. The CGM overestimated the glycaemic index of a smoothie by about 30% (a GI of 69 versus 53 by finger-prick) and overestimated the time spent above a recommended sugar threshold by nearly 400%. Whole fruit, which finger-prick testing showed was low-GI, got misclassified as medium or high.

Lead author Professor Javier Gonzalez warned the obvious conclusion: for healthy people, leaning on these readings could be misleading and drive unnecessary food choices. If your sensor tells you an apple is "bad", the sensor is more likely wrong than the apple.

Same meal, different day, different number

Even if accuracy were perfect, there is a second problem the marketing skips. The whole premise of a wellness CGM is that it reveals your personal food responses so you can pick winners and losers. But when researchers fed non-diabetic adults the exact same meal on different days, the individual glucose responses were only weakly to moderately reproducible.

That guts the core pitch. If the same meal gives you a different curve on Tuesday than it did on Monday, a single reading is a shaky basis for branding a food good or bad for you specifically. The big personalised-nutrition studies (ZOE's PREDICT 1, for instance) are right that responses differ a lot between people. Whether a CGM can pin down your response precisely enough to act on, from a handful of readings, is a much weaker claim.

The wellness claims vs the actual evidence

Weight loss, more energy, "metabolic optimisation", longevity. The marketing is confident. The evidence is thin.

There is currently no robust evidence that CGM use improves weight, metabolic health, athletic performance, or long-term outcomes in people without diabetes. A 2025 systematic review on CGMs for cardiovascular prevention in non-diabetics found some promise for personalising lifestyle changes, but limited direct evidence on hard endpoints, and the honest framing is biofeedback inside a structured program, not a standalone intervention. For a healthy person, wellness CGM use sits in a near-evidence-free zone.

The major guidelines agree by omission. The American Diabetes Association's Standards of Care address CGM for people with diabetes, not wellness monitoring in healthy people. The 2024 over-the-counter wellness sensors are consumer products, not guideline-endorsed health tools.

Athletes get a special mention because the marketing leans hard on them. Outside of diabetes there is no proven performance benefit, no established causal link between specific glucose levels (short of actual hypoglycaemia) and performance, and accuracy actually gets worse during exercise thanks to that blood-to-interstitial lag.

What it costs

The OTC era arrived fast. The US FDA cleared the first over-the-counter CGM, Dexcom's Stelo, in March 2024, with Abbott's Lingo and Libre Rio following the same year. Lingo is positioned as a general health-and-wellness tool for adults without diabetes. Worth knowing: none of these wellness devices alert you to dangerously high or low glucose, because they are not built for that job.

In the US, wellness subscriptions run roughly USD 84 a month (Lingo) to USD 89 a month (Stelo, about USD 1,068 a year), or roughly USD 49 to 99 for a one-off 14-to-15-day sensor pack. They are HSA/FSA eligible there.

In Singapore, some CGM sensors can be bought from authorised providers and pharmacies without a prescription. Introductory first-sensor pricing has been listed around SGD 27.25 nett for first-time users at screening clinics. Promos and any consult fees change, so confirm the current price directly with the provider before you commit.

Who might actually benefit, and who shouldn't bother

There are real, narrow cases. People at risk of prediabetes or type 2 diabetes can use a CGM as a screening nudge and a behaviour-change cue, ideally inside a structured program with a clinician, not as a solo gadget. Some endurance athletes managing fuelling may get useful free-living data, though even there the causal link between glucose numbers and performance is missing.

For the genuinely healthy person curious about their body, a two-week sensor is a fun experiment. Just hold the results loosely.

And there is a real downside the marketing never mentions. Clinicians warn that constant glucose self-surveillance can fuel food anxiety, over-restriction, and orthorexia-type patterns, with no offsetting benefit for weight or metabolic health in that group. If watching a line makes you scared of fruit, the device is costing you more than dollars.

Bottom line

For a healthy person, a wellness CGM is an interesting biofeedback toy, not a proven health upgrade. It can be inaccurate, it isn't reproducible meal-to-meal, and the "flatten every spike" advice isn't supported by the evidence.

Diabetes is a different conversation

None of the above applies to diagnosed diabetes. There, CGMs are a validated, often life-changing clinical tool, with alarms and real treatment decisions attached. That is a different device used for a different reason under medical care.

If a reading in the "prediabetic" range after lunch worries you, do not diagnose yourself from a sensor blip. Diabetes is confirmed with validated blood tests like HbA1c and fasting glucose. Talk to a clinician. This article is informational, not medical advice.

Sources

The Catalyst Feed

Content TeamIndependent, hands-on coverage of health, fitness & the tech that tracks it. Reviews you can trust — no hype.

#cgm #glucose-monitoring #longevity #biohacking #wellness #metabolic-health #prediabetes