GKI Chart: Glucose Ketone Index Ranges
What each band on the glucose ketone index scale means, where the numbers came from, and the one situation where a “good” GKI is a warning sign rather than a result.
The GKI scale
A lower number means ketones are high relative to glucose. Because it is a ratio, the same GKI can come from very different pairs of readings — which is both its strength and the source of its main pitfall.
| GKI | Zone | What it generally reflects |
|---|---|---|
| Below 1.0 | Peak therapeutic | The band used in oncology-focused protocols. Seyfried’s stated target for metabolic management of brain tumours is roughly 0.7–2.0, ideally around 1.0. |
| 1.0 – 3.0 | High therapeutic | Deep, sustained ketosis. This is the range associated in the literature with extended fasting and clinical therapeutic protocols. |
| 3.0 – 6.0 | Moderate therapeutic | The band typically targeted in ketogenic-diet management of epilepsy. |
| 6.0 – 9.0 | Light ketosis | Where most people following an ordinary low-carbohydrate diet sit day to day. |
| Above 9.0 | Not in significant ketosis | Glucose is high relative to ketones. Common after a higher-carbohydrate meal or on a standard diet. |
Want the number rather than the band? The GKI calculator works it out from your glucose and ketone readings in either unit system.
The formula, and the unit trap
Both inputs must be in mmol/L. Blood ketone meters already report beta-hydroxybutyrate in mmol/L everywhere, so the ketone side is rarely a problem. Glucose is where people go wrong: if your meter reads in mg/dL, you have to divide by 18.0182 before dividing by ketones.
Skipping that step inflates the result by roughly eighteenfold. A genuine GKI of 1.7 comes out as 30 and looks like no ketosis at all. If your number seems implausibly high, this is almost always why — see the glucose unit converter if you need to switch between the two.
| Glucose | Ketones | Glucose in mmol/L | GKI |
|---|---|---|---|
| 90 mg/dL | 3.0 mmol/L | 5.0 | 1.7 |
| 100 mg/dL | 1.5 mmol/L | 5.5 | 3.7 |
| 110 mg/dL | 0.8 mmol/L | 6.1 | 7.6 |
| 140 mg/dL | 0.3 mmol/L | 7.8 | 25.9 |
Where the scale came from
The glucose ketone index was published in 2015 by Meidenbauer, Mukherjee and Seyfried at Boston College. It was not designed as a wellness metric. The purpose was to give researchers a single number for tracking how effectively a ketogenic protocol was shifting a patient’s metabolism during metabolic management of brain cancer — specifically glioblastoma.
That origin explains the shape of the scale. The sub-1.0 band that circulates widely in keto communities is the oncology research target, not a general recommendation, and the paper makes no claim about it being desirable for healthy people. Treating it as a wellness goal reads more into the number than its authors did.
It also explains what the index does not have: no diabetes or endocrine society has adopted it, so there is no ratified normal range, no diagnostic threshold, and no outcome data outside the research context it was built for.
When a low GKI is a warning, not an achievement
This is the part most GKI charts leave out, and it matters more than everything above it.
The index cannot tell the difference between nutritional ketosis and diabetic ketoacidosis, because arithmetically they can look alike. Consider someone with type 1 diabetes whose insulin has been interrupted: glucose 300 mg/dL (16.6 mmol/L), ketones 5.0 mmol/L. That is a GKI of 3.3 — squarely inside the “moderate therapeutic” band on every chart online. It is also a picture of developing DKA.
The difference is not in the ratio. It is in the absolute numbers. Nutritional ketosis pairs normal or low glucose with raised ketones. Ketoacidosis pairs high glucose with raised ketones, and the ketone level climbs far beyond what dietary ketosis produces.
Our DKA assessment tool is built for exactly this scenario, and reads the absolute values rather than the ratio.
What actually moves the number
Because GKI is a ratio, anything that shifts either reading shifts the index — and glucose is usually the more volatile of the two.
Carbohydrate intake is the obvious lever, but it is far from the only one. The pre-dawn rise in cortisol and growth hormone lifts fasting glucose, which is why a morning reading often looks worse than an evening one on identical eating. Poor sleep, acute stress, illness, and hard exercise all raise glucose temporarily. Coffee does it in some people. Any of these can move a GKI a full band without a single gram of carbohydrate changing.
On the ketone side, readings drift with hydration, time since the last meal, and how keto-adapted someone is — long-term adapters often run lower blood ketones while using them more efficiently, so a rising GKI over months is not necessarily a step backwards.
The practical implication is to compare like with like: same time of day, same fasted state, and read the trend across a week rather than reacting to a single morning.
Limitations & Safety Notes
- The bands on this page are research-derived reference points, not clinical thresholds. No diabetes or endocrine society has ratified a GKI target.
- The index requires blood beta-hydroxybutyrate. Urine strips measure a different ketone body and breath meters measure acetone; neither substitutes into this formula.
- It cannot distinguish nutritional ketosis from diabetic ketoacidosis. Anyone with type 1 diabetes should read absolute glucose and ketone values, not the ratio.
- Single readings carry little information. Time of day, hydration, sleep, illness and stress all move the number independently of diet.
- Therapeutic ketogenic protocols for epilepsy or oncology are managed by specialist teams. This page describes what the scale means; it is not a protocol to follow unsupervised.
Frequently Asked Questions
What is a good GKI?
It depends entirely on why you are tracking it, and there is no officially endorsed target. Ordinary low-carbohydrate eating usually produces 6–9. Ketogenic-diet protocols for epilepsy commonly aim at 3–6. The sub-1.0 figure quoted online comes from oncology research and is not a general wellness goal.
How do I calculate GKI?
Divide blood glucose in mmol/L by blood ketones (beta-hydroxybutyrate) in mmol/L. If your meter reads glucose in mg/dL, convert first by dividing by 18.0182. A reading of 90 mg/dL with ketones of 3.0 mmol/L is 90 ÷ 18.0182 = 5.0, then 5.0 ÷ 3.0 = a GKI of 1.7.
Why is my GKI so high?
A high GKI means glucose is high relative to ketones. The usual causes are carbohydrate intake, a recent meal, the dawn rise in cortisol, stress, illness, poor sleep, or simply measuring at a time of day when glucose naturally runs higher. It is a ratio, so it moves when either number moves.
Can I use urine ketone strips for GKI?
No. The index is defined on blood beta-hydroxybutyrate. Urine strips measure acetoacetate, and how much of it spills into urine changes with hydration and with how keto-adapted you are, so the two are not interchangeable. Breath acetone is likewise a different measurement.
Is a low GKI always good?
No, and this is the most important caveat on the page. In someone with type 1 diabetes, high glucose together with high ketones can produce a GKI in the “therapeutic” range while the person is actually developing diabetic ketoacidosis. The index cannot distinguish nutritional ketosis from DKA, because the arithmetic looks similar. Ketones above roughly 1.5 mmol/L alongside high glucose is a medical emergency, whatever the ratio says.
Does GKI have official clinical guidelines?
No. It was published in 2015 as a research tool for monitoring metabolic therapy in brain cancer. No diabetes or endocrine society has adopted it or set a target, so treat published bands as research-derived reference points rather than clinical thresholds.
When should I measure it?
Consistency matters more than timing. Fasted, first thing in the morning, before any food or coffee, is the usual choice because it removes meal effects. Comparing a fasted reading one day with an afternoon reading the next tells you almost nothing.
Sources
- Meidenbauer JJ, Mukherjee P, Seyfried TN. The glucose ketone index calculator: a simple tool to monitor therapeutic efficacy for metabolic management of brain cancer. Nutrition & Metabolism. 2015;12:12. Link — the paper that defined the index and its therapeutic bands.
- American Diabetes Association. Standards of Care in Diabetes — 2026. Diabetes Care. 2026;49(Suppl 1). Link — ketone monitoring and hyperglycemic crisis thresholds.
- Kitabchi AE, et al. Hyperglycemic Crises in Adult Patients With Diabetes. Diabetes Care. 2009;32(7):1335–1343. Link — the DKA picture this page warns about.
Last reviewed: July 2026