Many people say they hardly get headaches any more. The studies point the same way, with less certainty than one would like.
This page follows on from What changes in the body when you cut carbs, which explains why keto often starts by bringing headaches.
Two or three headaches a month before, hardly any after: the story comes up often among people eating keto, and even among those who simply stay low-carb.
It is the opposite of what happens in the first week, when keto often brings headaches while the body regains its salt and water. That paradox is explained in What changes in the body when you cut carbs.
Research has looked at a single kind of headache: migraine. Ordinary headaches, known as tension headaches, have barely been studied from this angle.
It dates from 2019. Thirty-five overweight people with migraine followed two very low-calorie diets for a month each, in an order drawn at random: one ketogenic, the other not. Neither the patients nor the doctors counting the attacks knew which one was under way.
During the ketogenic month, participants had on average 3.7 fewer migraine days than during the non-ketogenic month, and 3 fewer attacks. Three in four (74%) saw their migraine days fall by at least half, against fewer than one in ten (9%) on the other diet.
Its value lies in its design: at equal calories, the only major difference between the two months is ketosis. That is what the same team's earlier study, in 2015, lacked: 96 overweight women, of whom those who went through a month of ketosis were better from that first month, while those on a standard diet only improved from the third. But the groups there were not drawn at random.
A single randomized trial, small and short, but one that isolates ketosis. Thirty-five people for a month: enough to take the lead seriously, not enough to take it for granted.
A 2023 meta-analysis pooled ten studies, about 270 adults, with several forms of ketosis: very low-calorie diets, the modified Atkins diet, classic keto. All reduce attack frequency, and depending on the study, 58 to 83% of participants see their attacks fall by at least half.
In people with chronic migraine resistant to treatment, a 2021 Italian study of 38 patients saw headache days drop from 30 to 7.5 a month over twelve weeks. The figure is striking, but there was no comparison group.
The limits are the same everywhere. Most studies have no control group, almost all come from Italy and often from the same team, they last one to three months, and ketosis is not always measured. A 2024 Danish review, covering every diet tested against migraine, rates the evidence as a whole as still weak.
If ketones protected on their own, swallowing them would be enough. A 2022 Swiss trial tested this in 41 people with migraine: twelve weeks of beta-hydroxybutyrate as a supplement, twelve weeks of placebo, without changing their diet.
Result: barely one fewer migraine day a month, a difference too small to count. The authors conclude it was not effective, at the dose tested.
Two readings remain possible: the dose was too low, or what acts is not only the ketone, but everything the diet changes around it.
Several mechanisms have been proposed. None is demonstrated in humans, but all of them hold up.
A brain short of energy. Brain scans of people with migraine, between attacks, show lower energy reserves than normal. Ketones offer it a backup fuel, which the brain knows how to use, as Keto and focus shows.
Less excitable neurons. Migraine comes with a cortex that overreacts. Beta-hydroxybutyrate calms that excitability, the same action that has earned the ketogenic diet its century-old use against epilepsy.
Less inflammation. In the lab, beta-hydroxybutyrate dampens an inflammatory pathway involved in pain. In people with migraine, the effect has yet to be measured.
Stable blood sugar. Skipped meals and fasting are well-known attack triggers. With no sugar spike, there is no crash an hour later, as described in What happens after a meal. This mechanism needs no ketosis, and could explain feeling better on a simple carb reduction. The only study that tested this case did not find a clear effect, though.
Everything that changes around it. Less alcohol, sugar and processed food, more regular meals, and often lost weight, which on its own reduces migraine. In the 2015 study, the group on the standard diet also ended up better, as it lost weight.
The first week. Keto often brings headaches, from a lack of salt and water. It passes within a few weeks.
After that. For migraine, one randomized trial and around ten studies point the same way: fewer attack days, often by half.
The evidence. Small, short, mostly Italian, in overweight people. A serious lead, not yet a certainty.
The mechanism. Plausible, not demonstrated. Ketone supplements alone did not work.
Migraine is treated with a doctor. In these studies the diet was supervised, and it can interact with some treatments. A new, unusual or sudden headache should be shown to a doctor, whatever you eat.
Randomized trial: Di Lorenzo et al., A Randomized Double-Blind, Cross-Over Trial of very Low-Calorie Diet in Overweight Migraine Patients: A Possible Role for Ketones?, Nutrients 2019.
2015 study: Di Lorenzo et al., Migraine improvement during short lasting ketogenesis: a proof-of-concept study, European Journal of Neurology 2015.
Meta-analysis: Neri et al., Ketosis and migraine: a systematic review of the literature and meta-analysis, Frontiers in Nutrition 2023.
Review of diets, including the 2021 study and the plain low-carb case: Nguyen and Schytz, The Evidence for Diet as a Treatment in Migraine, Nutrients 2024.
Ketone supplements: Putananickal et al., Efficacy and safety of exogenous beta-hydroxybutyrate for preventive treatment in episodic migraine, Cephalalgia 2022.
Mechanisms: Gross et al., Potential Protective Mechanisms of Ketone Bodies in Migraine Prevention, Nutrients 2019.