Ketone Testing & Measurement
How accurate are the different ways to measure ketosis, and what do the numbers mean?
Copper Keto Companion gathers the research on how ketosis is actually measured here — blood, urine, and breath — including where each method is unreliable. The studies use blood beta-hydroxybutyrate as the comparison point, and the cheaper methods trade accuracy for convenience.
Contents — 5 entries
- 📄 Urine Dipsticks Are Not Accurate for Detecting Mild Ketosis (Substudy)
- 📄 A High-Resolution Breath Acetone Meter for Ketosis Monitoring
- 📄 Capillary Blood Tests May Overestimate Ketosis (Triangulation Study)
- 📄 Update on Measuring Ketones (Review)
- 📄 The Glucose Ketone Index Calculator (GKI) for Metabolic Therapy
📄 Urine Dipsticks Are Not Accurate for Detecting Mild Ketosis (Substudy)
Gibson AA, et al. — Obesity Science & Practice, 2020 · Obes Sci Pract, 2020
Copper Keto Companion research surfaced this report because it covers how reliably urine strips detect ketosis. In 50 postmenopausal women with obesity on a severely energy-restricted diet (263 paired samples), urine dipsticks detected only 35%, 52%, and 76% of cases at blood beta-hydroxybutyrate thresholds of 0.3, 0.5, and 1.0 mmol/L — missing roughly half of mild ketosis at 0.5 mmol/L — while staying highly specific (97 to 100%). The sample was small and limited to postmenopausal women.
What it examines: a comparison of urine dipstick readings against blood beta-hydroxybutyrate during dieting. Why it's in the Copper Keto Companion research index: a substudy of how often urine strips miss mild ketosis versus a blood meter.
📄 A High-Resolution Breath Acetone Meter for Ketosis Monitoring
Suntrup DJ, et al. — PeerJ, 2020 · PeerJ, 2020
Copper Keto Companion research surfaced this report because it covers how well breath testing tracks blood ketones. Across 21 adults and 1,214 paired readings, single breath-acetone and blood-BHB measurements correlated only moderately (R-squared 0.57), while daily ketone exposure tracked closely (R-squared 0.80), and a breath reading predicted whether blood BHB was above 1.5 mmol/L with high accuracy (ROC AUC 0.935). A 1-to-5-hour lag between breath and blood weakens the point-by-point match.
What it examines: a validation of a breath-acetone device against point-of-care blood ketones. Why it's in the Copper Keto Companion research index: a validation study of breath-acetone testing against blood beta-hydroxybutyrate.
📄 Capillary Blood Tests May Overestimate Ketosis (Triangulation Study)
Norgren J, et al. — Am J Physiol Endocrinol Metab, 2020 · pubmed / 31821040
Copper Keto Companion research surfaced this report because it covers how finger-prick blood ketone readings compare with venous blood and plasma. In 15 healthy volunteers put into ketosis with medium-chain fats (180 paired measurements), capillary readings ran higher than venous blood from the same device — a slope of 0.64 to 0.78, meaning capillary values were roughly 20 to 35% higher. All three methods correlated strongly with plasma ketones (r of 0.91 to 0.99), so they track change well, though the authors note cutoffs may need adjusting by method. It is a small study, amended by a 2022 corrigendum.
What it examines: a comparison of capillary, venous, and plasma beta-hydroxybutyrate measurements. Why it's in the Copper Keto Companion research index: a validation study of how capillary blood ketone readings compare with venous and plasma.
📄 Update on Measuring Ketones (Review)
Huang J, et al. — Journal of Diabetes Science and Technology, 2023 · J Diabetes Sci Technol, 2023
Copper Keto Companion research surfaced this report because it covers how the three ketone measurements differ in what they capture. This 2023 review explains that blood meters read current beta-hydroxybutyrate (the basis for diagnosing diabetic ketoacidosis), urine strips report an average since the last void and lag behind, and breath acetone ranges from about 1 ppm in healthy people to over 1250 ppm in ketoacidosis. It notes nutritional ketosis is commonly set at 0.5 mmol/L or higher and that no breath device is FDA-cleared. As a narrative review, it synthesizes prior work rather than adding new data.
What it examines: a review of blood, urine, and breath ketone measurement methods. Why it's in the Copper Keto Companion research index: a review comparing what blood, urine, and breath ketone tests each measure.
📄 The Glucose Ketone Index Calculator (GKI) for Metabolic Therapy
Meidenbauer JJ, et al. — Nutrition & Metabolism, 2015 · pubmed / 25798181
Copper Keto Companion research surfaced this report because it covers a combined glucose-and-ketone metric used in clinical ketogenic therapy. The 2015 paper introduced the glucose ketone index (GKI) — blood glucose divided by blood ketones, both in mmol/L — as a single number to track ketogenic metabolic therapy in brain cancer, proposing GKI values approaching 1.0 as the therapeutic target. It was developed for preclinical models and clinical cancer management, not for general dieting, and reports a tool rather than a diet trial.
What it examines: the introduction of the glucose ketone index for monitoring therapeutic ketosis. Why it's in the Copper Keto Companion research index: the source paper defining the glucose ketone index (GKI) for clinical ketogenic therapy.
All 5 sources last verified June 2026.
Frequently Asked Questions
What's the most accurate way to measure ketosis?
The studies here use blood beta-hydroxybutyrate as the comparison point. A 2023 review (Huang) notes blood meters read the current ketone level directly, urine strips lag and miss mild ketosis (Gibson, 2020, caught about half at 0.5 mmol/L), and breath acetone tracks blood only moderately moment to moment (Suntrup, 2020). This summarizes research and is not medical advice.
Do urine ketone strips stop working over time?
They become less sensitive. A 2020 substudy (Gibson) found urine dipsticks detected only about half of mild ketosis (52% at a blood level of 0.5 mmol/L), and reviews note the body excretes less acetoacetate as it adapts. This is a research summary, not advice.
How accurate is breath ketone testing?
Moderately, and better across a day than at a single moment. A 2020 validation (Suntrup) found single breath and blood readings correlated at R-squared 0.57, while daily ketone exposure correlated at 0.80, with a 1-to-5-hour lag between the two.
What is the glucose ketone index (GKI) for?
Clinical ketogenic therapy, not general dieting. The 2015 paper that introduced it (Meidenbauer, Seyfried) proposed GKI — glucose divided by ketones — to monitor metabolic therapy in brain cancer, with values near 1.0 as the therapeutic target. It was not developed for weight-loss keto.
More in Keto Research
Educational information only — not medical advice, and not a recommendation to start, stop, or change any diet, supplement, or treatment. Talk to a qualified healthcare professional before making changes. Copper Keto Companion and Copper Sun Content and Creative, LLC are not medical providers.