How continuous ketone monitors work

Last checked September 28, 2026

Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you. It never changes what we recommend. How we make money.

This site gives general information, not medical advice. Talk to your doctor or diabetes team before changing how you monitor or treat any condition.

A continuous ketone monitor (CKM) works like a continuous glucose monitor, with a different enzyme. A thin, flexible filament sits just under the skin, usually on the back of the upper arm, and measures ketones in the fluid around the cells every few minutes.

What it measures

The body makes three ketones: beta-hydroxybutyrate (BHB), acetoacetate and acetone (Dhatariya, 2017). Continuous sensors, like finger-prick blood meters, measure BHB. The sensor reads it in interstitial fluid, the fluid between cells, rather than in blood (Fante et al., 2025).

How the sensor turns ketones into a number

In the first published human study of a CKM, Abbott’s sensor used an enzyme, β-hydroxybutyrate dehydrogenase, wired to an electrode: the enzyme reacts with BHB and the electrode measures the resulting current, which rises with the ketone level. The sensor gave a straight-line response from 0 to 8 mmol/L (Alva et al., 2021). SiBio describes its sensor the same way: an electrochemical sensor whose signals are processed by algorithms into blood-equivalent ketone values (SiBio).

Calibration

The 2021 study used a single calibration for the whole 14 days, and the signal changed by only 2.1% over that time (Alva et al., 2021). Abbott has since described how it made the ketone part of its dual sensor factory-calibrated, so users do not calibrate it themselves (Hoss et al., 2025).

Readings, warm-up and alarms

Why a sensor reading can differ from a blood test

A sensor reads the fluid between cells, not blood, so its number will not always match a finger-prick taken at the same moment, especially while levels are changing. Published figures on how far ketone sensors trail blood are still scarce. The 2021 Abbott study reported an average sensor response time of 228 seconds (about 4 minutes) to a change in ketone concentration (Alva et al., 2021). A 2025 review notes that sensor accuracy, calibration and clinical validation are still open questions for CKMs (Fante et al., 2025).

This is why the FDA says ketone results should be read together with glucose readings and symptoms (FDA), and why the international expert recommendations suggest keeping a backup blood ketone meter (summary). What accuracy studies found.

More guides

Sources