Step 5 of 8
What Ketosis Actually Is: Brain Fuel Reimagined
By the end of this lesson you will understand what nutritional ketosis actually is at the level of hormones, fuel and cell signaling — and why the brain's response to ketones is not the same thing as fat loss.
Ketosis has been marketed so heavily as a weight-loss strategy that its more interesting property has been almost buried: it changes what the brain runs on, and how the brain's power plants behave. This lesson takes the concept apart mechanically — insulin down, ketones up, and what happens between those two events. We will also be clear about where the evidence is solid, where it is preliminary, and who should not attempt this without medical supervision.
Think about a hybrid car. It has two power systems: a gasoline engine and an electric motor. The whole point of the design is that it can switch between them depending on conditions. Now imagine that someone has jammed the switch so that the car can only ever run on gasoline. The battery is still there. It is fully charged. But the vehicle has no access to it, and every time the tank runs low the driver feels the car falter and pulls into a filling station. That is a reasonable picture of what happens in a body running on frequent carbohydrate intake: the alternative fuel system exists, it is well stocked, and something is holding the switch down.
The thing holding the switch is insulin. When you eat carbohydrate, blood glucose rises, and the pancreas releases insulin to move that glucose into cells and to store surplus energy as fat. Insulin also does something less often discussed: while it is elevated, it actively suppresses the release of fatty acids from fat tissue. This is elegant design. There is no sense burning your reserves while the food is still arriving. But if eating is frequent — three meals, two snacks, a coffee with sugar, something in the evening — insulin rarely falls far enough or long enough for the second system to come online. Over years, this pattern can contribute to insulin resistance, in which cells become less responsive to the signal and the body compensates by producing still more insulin.
When carbohydrate intake falls and insulin drops, the sequence reverses. Fatty acids are released, travel to the liver, and are converted through beta-oxidation into ketone bodies — chiefly beta-hydroxybutyrate (BHB) and acetoacetate. Unlike fat itself, ketones are water-soluble and cross the blood–brain barrier readily. This matters enormously, because the brain cannot burn fatty acids directly. Ketones are the brain's only route to the body's largest energy store. That state of elevated circulating ketones is nutritional ketosis.
Three details make this more than a fuel swap. First, mitochondria appear to extract slightly more ATP per unit of oxygen consumed from ketones than from glucose — a small efficiency gain that may matter most in tissue already under metabolic strain. Second, BHB is not only fuel; it is a signaling molecule. It influences gene expression, interacts with cellular sensors such as the NLRP3 inflammasome and the sirtuins, and is associated in laboratory work with dampened inflammatory signaling. Animal studies also show increased brain-derived neurotrophic factor (BDNF), a protein involved in synaptic maintenance — a promising finding, but one not yet firmly established in humans. Third, ketosis shifts the balance between GABA, the brain's main inhibitory neurotransmitter, and glutamate, its main excitatory one, generally in the direction of more GABA. That is a plausible mechanistic bridge to the calming effects some people report, though plausible is not the same as proven.
In working life, the relevant experience is rarely dramatic. It is the two o'clock collapse after a carbohydrate-heavy lunch, when a routine spreadsheet suddenly requires effort. It is the sense that concentration is available in narrow windows and must be refuelled constantly. It is irritability that arrives at predictable times of day rather than in response to anything that happened. None of this is diagnostic of anything, and stress, sleep debt and workload explain a great deal of it. But a brain that can only run on one fuel, delivered in pulses, is a brain with a narrow margin.
Several misconceptions deserve unpicking. The first is the confusion between nutritional ketosis and diabetic ketoacidosis, a dangerous condition seen mainly in type 1 diabetes where ketone levels rise many times higher in the absence of insulin. In an adult with intact pancreatic function, insulin and glucagon keep nutritional ketosis within a self-limiting range. The second is treating ketosis as a switch rather than a spectrum: there is a continuum from mild elevation to sustained deeper ketosis, and the early transition may bring temporary fatigue, headache or poor sleep — the so-called keto flu — largely reflecting fluid and electrolyte shifts rather than failure. The third is assuming the benefit is the weight loss. Some people report cognitive changes with no meaningful change in body composition at all, which points toward mechanisms other than fat mass.
Here is the counterintuitive part. A person can be metabolically overfed and neurologically underfed at the same time. Energy can be abundant in the bloodstream and stored in enormous quantities in adipose tissue while individual neurons operate under a supply–demand mismatch. Famine in the middle of a full pantry. That reframing — not more calories, but better access to the ones already present — is the core idea of metabolic psychiatry, and it is why this approach is unsuitable to attempt unsupervised for anyone with type 1 diabetes, anyone taking insulin or sulfonylureas, anyone on lithium or certain antipsychotics, during pregnancy, or with a history of eating disorders. Those conversations belong with a clinician.
Key points
- Insulin does not only move glucose into cells; while elevated, it blocks the release of stored fat, keeping the body locked into a single fuel system.
- When insulin falls, the liver converts fatty acids into ketone bodies — mainly beta-hydroxybutyrate and acetoacetate — which cross the blood–brain barrier and give the brain access to fat-derived energy it cannot otherwise use.
- Beta-hydroxybutyrate acts both as fuel and as a signaling molecule, influencing gene expression, inflammatory pathways and mitochondrial function independently of calories.
- Nutritional ketosis is a self-limiting, reversible state and is physiologically distinct from diabetic ketoacidosis, which occurs mainly in type 1 diabetes.
- Human evidence for psychiatric benefit currently comes from small pilot studies and case series; mechanisms are well described, but who benefits and for how long remains unresolved.
- Carbohydrate restriction is not appropriate for everyone, and interactions with psychiatric and diabetes medications make clinical oversight necessary rather than optional.
Do this before the next step
Spend one week keeping a simple note of how alert you feel roughly ninety minutes after each meal, alongside what that meal mostly consisted of. You are not dieting or restricting anything — you are collecting your own data on whether energy dips track with particular meal patterns. Most people have never actually checked, and the pattern is often more consistent than expected.
Count how many separate times you eat in a typical day, including drinks with sugar or milk. Frequency, independent of what the food is, drives how much of your day is spent with insulin elevated. Simply noticing that the number is nine rather than three is more informative than any change you might make in the same week.
If anything in this lesson makes you want to try carbohydrate restriction seriously, write down your current medications and any diagnosed conditions first, and take that list to your doctor or psychiatrist before changing your diet. Blood glucose medications and some psychiatric medications need active monitoring when fuel intake shifts, and that is a conversation that requires your individual clinical picture, not a general article.
Answer these honestly
- Looking honestly at a typical working day, how many hours pass between your first intake of the morning and your last of the evening during which you are genuinely not eating anything?
- When your concentration fails in the afternoon, what explanation do you usually reach for — and have you ever tested whether that explanation is the right one?
- If a change in how your brain is fueled produced no change at all in your body weight, would you still consider it worthwhile, and what does your answer reveal about how you have been framing food until now?