
In 1953, a young hospital doctor called George Cahill was treating a young patient with a dangerous complication of diabetes. At the time, one of the molecules in that patient's blood — beta-hydroxybutyrate — was seen by many as a metabolic poison. Cahill went on to help show it was something else entirely: the fuel that lets a human brain survive weeks without food (Cahill, Annual Review of Nutrition, 2006).
That molecule is a ketone. And the story of how we got from "poison" to "survival fuel" is the best way to understand what ketosis is — and what it isn't.
The fuel switch
Your body keeps a small store of sugar, called glycogen, mainly in the liver and muscles. When you stop eating, it works through roughly three overlapping stages: first it uses that stored sugar, then it makes new sugar, and then it shifts to burning fat (Fazeli & Steinhauser, Endocrine Reviews, 2025).
Here's the catch. Most of your body can run on fat directly, but your brain can't. So the liver turns fat into ketones, which can cross into the brain and power it. That's ketosis: not a hack, just a backup generator that kicks in when the main supply runs low. With fasting, ketone production starts within about a day, and in humans this fat-burning stage can last for months (Fazeli & Steinhauser, 2025).
What the evidence shows
The history is in people, not mice. In the 1960s, researchers studied people with obesity who fasted for 38 to 41 days. The maths said their brains should have run out of sugar and protein to burn. They didn't — because ketones had largely replaced sugar as the brain's main fuel (Rebello et al., International Journal of Obesity, 2025). In those studies, ketone levels in the blood reached 2-3 mmol/L after three days and 6-8 mmol/L after several weeks.
This is the part of fasting science that nobody seriously disputes. Ketosis is real, it's measurable, and your body is built for it.
The extras are a different story. Ketones aren't only fuel; they also seem to act as signals that can change how genes are switched on and off, which might matter for disease and ageing (Newman & Verdin, Annual Review of Nutrition, 2017). One much-cited study found that the main ketone can block an "alarm system" of the immune system involved in inflammation — in human immune cells in the lab, and in mice (Youm et al., Nature Medicine, 2015). Fascinating, and worth following. But that's cells and mice. When researchers looked at inflammation in people doing long fasts, the data were limited and inconsistent, and some markers actually went up (de Ciutiis et al., Ageing Research Reviews, 2025).
Ketosis is not ketoacidosis
This is the line worth knowing, because the names are so similar.
Diabetic ketoacidosis (DKA) is a medical emergency, mostly seen in people with diabetes. The problem isn't ketones as such; it's a lack of insulin. Without enough insulin, fat floods out of storage and the liver turns it into ketones far faster than the body can use them (Huang et al., Journal of Diabetes Science and Technology, 2024; Laffel, Diabetes/Metabolism Research and Reviews, 1999). Blood sugar goes very high, ketones can climb past 20 mmol/L, and the blood becomes acidic.
Nutritional ketosis — from fasting or a very low-carb diet — is a different situation. Blood sugar dips but stays in the normal range. In one study of people on a ketogenic diet for six weeks, ketones peaked at 0.7 mmol/L, well below the 3.8 mmol/L level used to diagnose DKA in adults, and their blood stayed at a normal acidity (Rebello et al., 2025). In people whose bodies make insulin, ketones themselves nudge insulin up, which acts as a brake.
Same molecule, very different story. The difference is insulin.
Why scientists still disagree
Nobody argues about whether ketosis happens. The disagreement is about whether the extra benefits people attach to it are real in humans.
We rarely measure it. In human trials of intermittent fasting, ketones often aren't even measured; when they are, they rise (Rebello et al., 2025). One small study of 11 adults eating between 8 am and 2 pm for four days found higher morning ketones (Jamshed et al., Nutrients, 2019). Whether those modest rises do anything for health hasn't been tested.
Theory runs ahead of trials. Prof Mark Mattson, a neuroscientist and one of the leading fasting researchers, has argued that cycling in and out of ketosis may strengthen the brain (Mattson et al., Nature Reviews Neuroscience, 2018). In 2025 he set this out formally — and called it what it is: a theory (Mattson, Nature Metabolism, 2025). It's a serious, testable idea. It hasn't been tested yet.
What you'll hear online
"Ketosis burns fat." True — that's literally what it is. Whether it burns more fat than eating fewer calories some other way is a separate question this series covers elsewhere.
"Ketones are anti-inflammatory." In lab cells and mice, one ketone dampens an inflammation pathway. In fasting humans, the picture is messier.
"Ketones are dangerous." For most people, ketosis from fasting is normal physiology. The danger zone is ketoacidosis, which is about missing insulin.
Before believing any fasting claim, it helps to ask five quick questions:
Five questions before you believe a fasting claim
- Was this seen in people, or in mice?
- How many people, and for how long?
- Compared to what?
- Did they measure a real outcome, or a number in the blood?
- Does the person telling me this sell something related? (Not an accusation — just useful to know.)
What would actually settle it
Trials that measure ketones properly in people doing intermittent fasting, compare them with people eating the same calories on a normal schedule, and follow real outcomes — memory, mood, blood sugar control over years. Mattson's theory makes predictions that could be checked. That's a good thing, and worth funding.
What you can do
If you have diabetes — especially if you use insulin or take a type of tablet called an SGLT2 inhibitor, which can raise the risk of ketoacidosis (Huang et al., 2024) — talk to your doctor before you try any kind of fasting. Your doctor will know what applies to you. For a fuller checklist, see Before you fast on medication: the questions to ask your doctor.
Ketosis is one of the quiet marvels of human biology: the reason a person can think clearly after weeks without food. It deserves to be understood, not oversold.
Curiosity first.
— Dr. Brugal
Voices in the debate
If you'd like to hear the case for fasting and ketosis from people who champion it, these are some of the most influential voices. Their views often go beyond what published human trials have shown so far — which is exactly why it's worth hearing them and checking the studies side by side.
- Dr Pradip Jamnadas — MD, cardiologist (Florida), clinical assistant professor. Talks about fasting for ketosis, autophagy, brain health and heart and metabolic health. Feel Better, Live More (Bitesize)
- Dr Jason Fung — MD, nephrologist (Canada). Sees fasting as a central tool against obesity and type 2 diabetes, with insulin as the key hormone. Books: The Obesity Code, The Complete Guide to Fasting.
- Prof Mark Mattson — researcher (PhD, Johns Hopkins). Proposes that "metabolic switching" between fuels benefits the brain and body; author of the studies cited above. Nature Metabolism, 2025
For the wider debate, see our series guide: Why scientists still can't agree on fasting — and what would settle it. Related: Autophagy: the real science behind fasting's favourite buzzword and Does fasting calm inflammation? The answer is messier than you've heard (both coming soon).
References
All studies retrieved via PubMed.
- Cahill GF. Fuel metabolism in starvation. Annual Review of Nutrition. 2006;26:1-22. Personal and scientific review of human starvation research. doi:10.1146/annurev.nutr.26.061505.111258 (PMID 16848698)
- Fazeli PK, Steinhauser ML. A critical assessment of fasting to promote metabolic health and longevity. Endocrine Reviews. 2025;46(6):856-876. Review of animal and human studies. doi:10.1210/endrev/bnaf021 (PMID 40700575)
- Rebello CJ, Zhang D, Anderson JC, et al. From starvation to time-restricted eating: a review of fasting physiology. International Journal of Obesity. 2025;49(1):43-48. Review of human studies. doi:10.1038/s41366-024-01641-0 (PMID 39369112)
- Newman JC, Verdin E. β-Hydroxybutyrate: a signaling metabolite. Annual Review of Nutrition. 2017;37:51-76. Review of animal and human studies. doi:10.1146/annurev-nutr-071816-064916 (PMID 28826372)
- Youm YH, Nguyen KY, Grant RW, et al. The ketone metabolite β-hydroxybutyrate blocks NLRP3 inflammasome-mediated inflammatory disease. Nature Medicine. 2015;21(3):263-9. Laboratory study in human immune cells and in mice. doi:10.1038/nm.3804 (PMID 25686106)
- de Ciutiis I, Djakovic S, Cagigas ML, et al. Long-term fasting and its influence on inflammatory biomarkers: a comprehensive scoping review. Ageing Research Reviews. 2025;110:102797. Scoping review of 14 human studies of fasts lasting 48 hours or more. doi:10.1016/j.arr.2025.102797 (PMID 40484176)
- Huang J, Yeung AM, Bergenstal RM, et al. Update on measuring ketones. Journal of Diabetes Science and Technology. 2024;18(3):714-726 (first published online 2023). Review. doi:10.1177/19322968231152236 (PMID 36794812)
- Laffel L. Ketone bodies: a review of physiology, pathophysiology and application of monitoring to diabetes. Diabetes/Metabolism Research and Reviews. 1999;15(6):412-26. Review. doi:10.1002/(SICI)1520-7560(199911/12)15:6<412::AID-DMRR72>3.0.CO;2-8 (PMID 10634967)
- Jamshed H, Beyl RA, Della Manna DL, et al. Early time-restricted feeding improves 24-hour glucose levels and affects markers of the circadian clock, aging, and autophagy in humans. Nutrients. 2019;11(6):1234. Randomised crossover trial, 11 adults, 4 days. doi:10.3390/nu11061234 (PMID 31151228)
- Mattson MP, Moehl K, Ghena N, et al. Intermittent metabolic switching, neuroplasticity and brain health. Nature Reviews Neuroscience. 2018;19(2):63-80. Review of animal and human studies. doi:10.1038/nrn.2017.156 (PMID 29321682)
- Mattson MP. The cyclic metabolic switching theory of intermittent fasting. Nature Metabolism. 2025;7(4):665-678. Perspective proposing a theory. doi:10.1038/s42255-025-01254-5 (PMID 40087409)
Awareness content — not medical advice. Always talk to your doctor or pharmacist before making changes.