Diet and Epilepsy: The Evidence for MCT and Ketogenic Diets

The rationale is worth getting plain, because the marketing usually mangles it. The brain normally runs on glucose, and part of the thinking in epilepsy is that this glucose use is impaired, leaving neurons short of a stable energy supply. Ketone bodies, which the body makes from fat, can serve as an alternative fuel. Medium-chain triglycerides are a fat the liver converts to ketones much more readily than ordinary long-chain fats do, raising blood levels of a ketone called beta-hydroxybutyrate, or BHB (Law et al., 2015; Han et al., 2021). So the first half of the idea is simply to give the brain a steadier alternative fuel.

There is a second mechanism, and it is the part that makes MCTs genuinely interesting rather than a fad. MCT oil is mostly two fatty acids: caprylic acid (C8) and capric acid, also called decanoic acid (C10). Decanoic acid has a direct anticonvulsant action of its own, dampening excitation by blocking AMPA receptors, one of the main excitatory glutamate receptors in the brain, through a different route from ketones and from the human anti-seizure drug perampanel (Chang et al., 2016). It makes up roughly 40% of the medium-chain fat in an MCT ketogenic diet, so it is present in meaningful amounts (Chang et al., 2016). The plausible benefit, then, comes from both raised ketones and a direct calming effect on excitable neurons, not from "starving" the brain of carbohydrate. That is exactly why a sensible MCT approach is not the strict, very-high-fat ketogenic diet you may have read about in people.

The canine evidence is better than you might expect and more honest than the marketing makes it sound. The landmark study came from the Royal Veterinary College with Royal Canin: a six-month, randomised, double-blinded, placebo-controlled crossover trial in 21 dogs with idiopathic epilepsy, all already on anti-seizure medication, fed a ketogenic MCT diet (5.5% MCT) for three months and a matched placebo diet for three months, in randomised order (Law et al., 2015). That design is the gold standard, because each dog acts as its own control and neither owners nor vets knew which diet was which. Seizure frequency was significantly lower on the MCT diet: a median of 2.31 a month versus 2.67 on placebo, P = 0.020 (Law et al., 2015). The individual dogs give the honest texture: 3 became seizure-free, 7 more had at least a 50% reduction, 5 had a smaller reduction and 6 showed no response at all (Law et al., 2015). So roughly half got a meaningful benefit and a substantial minority got nothing.

A larger multicentre trial confirmed the signal and added something useful. This time the researchers tested an MCT oil supplement rather than a whole diet, added to the dog's existing food at around 9% of daily calories, roughly 7 mL a day, in the same six-month blinded crossover design, now in 36 dogs with drug-resistant idiopathic epilepsy, of which 28 completed (Berk et al., 2020). Seizure frequency was again significantly lower with the oil: a median of 2.51 a month versus 2.67 on control, P = .02, with 2 dogs becoming seizure-free and 3 more achieving a 50 to 66% reduction (Berk et al., 2020). So the effect can come from an oil added to ordinary food, not only from a special kibble.

Two further studies show why study design matters. An open-label field trial across ten European practices fed 21 dogs a 6.5% MCT diet (the commercial NeuroCare formula) as an add-on, and mean seizure frequency fell by 32% (P = 0.04) and seizure days by 42% (P < 0.001), with 9.5% becoming seizure-free and 42.9% achieving at least a 50% reduction (Molina et al., 2020). Encouraging numbers, but it was non-blinded with no placebo group, and the authors honestly note a placebo effect could account for perhaps 30% of the improvement (Molina et al., 2020). A blinded, placebo-controlled crossover of the same diet in dogs on zonisamide makes the opposite point: it produced a 35% numerical reduction (1.90 versus 2.95 seizures a month), yet because only 7 dogs completed and the study was openly underpowered, that difference was not significant (P = 0.37) (Nakatsuka et al., 2023). A small study can miss a real effect simply for lack of dogs, though it did at least confirm the diet is safe alongside zonisamide (Nakatsuka et al., 2023).

Put the four together and the honest synthesis is this: as an add-on, MCT produces a modest but reproducible reduction in seizure frequency, with roughly a third to a half of dogs reaching at least a 50% reduction in the positive trials, a minority becoming seizure-free, and a sizeable minority not responding. It is not a cure and does not replace the drugs. This is also genuinely newer evidence: the 2015 international consensus on medical treatment of canine epilepsy did not include dietary therapy at all (Bhatti et al., 2015).

There may be a bonus beyond seizures, though we should not oversell it. The same research group found an MCT diet reduced certain ADHD-like behaviours such as hyperactivity and inattention, and lowered stranger-directed fear (Packer et al., 2016). An MCT oil supplement at 9% of calories also improved spatial working memory (P = 0.008), problem-solving (P = 0.048) and owner-rated trainability (P = 0.041), with those cognitive gains linked to higher BHB after meals (Berk et al., 2021). Treat that as a possible added benefit, not a reason to feed MCTs in the first place.

Now the marketing, dismantled gently but firmly, because this is where owners get sold things that do not help and sometimes things that carry their own risk.

Strict, human-style ketogenic diets are not established in dogs. This is the most common confusion. The controlled canine evidence is all for MCT-supplemented diets and oils, which keep moderate carbohydrate and raise ketones only modestly, not for the very-high-fat classic keto used in some people. A 2024 systematic review concluded the canine epilepsy evidence centres on MCT diets, and that there is not yet evidence to recommend classic high-fat ketogenic diets in dogs or cats (Vendramini et al., 2024).

"Grain-free" is not an epilepsy therapy. There is no evidence that removing grain reduces seizures. Worse, grain-free and so-called boutique, exotic-ingredient, grain-free (BEG) diets high in peas, lentils, other legumes or potatoes have been investigated by the US Food and Drug Administration over a possible link to diet-associated dilated cardiomyopathy (DCM), a serious heart-muscle disease, in dogs not genetically prone to it (FDA, 2019). The FDA stresses that DCM is complex and multifactorial and that a causal link has not been established (FDA, 2019), so this is not cause to panic if your dog already eats one. But it is an active safety signal, and switching to grain-free hoping to help seizures could trade an imaginary benefit for a real, if unproven, risk. Choose any diet change with your vet, not on an internet claim.

No over-the-counter "epilepsy" supplement is proven. Beyond MCTs, there is not robust controlled evidence that any shop-bought supplement reliably reduces seizures. Treat such products sceptically and run anything tempting past your vet first. CBD is the big one people ask about, and it is a genuinely separate topic with its own real research and real safety catches, so it has its own article: see CBD for dogs with epilepsy. The same never-stop-your-meds rule applies there too.

If you and your vet decide an MCT trial is worth a go, here is how to do it so you actually learn something.

The most accessible MCT option in the UK is a veterinary therapeutic diet, Purina Pro Plan Veterinary Diets NC NeuroCare, which contains 6.5% MCT oil plus added omega-3s, arginine and antioxidants, and is the formula used in two of the trials above (Molina et al., 2020; Nakatsuka et al., 2023). In the spirit of honesty, much of the RVC research programme behind MCTs was funded by Nestlé Purina, formally clarified in a correction to the 2020 trial's disclosures; the correction did not change the peer-reviewed, placebo-controlled results, but the commercial interest deserves to be on the table so you can weigh it yourself (Berk et al., 2024).

A few practical rules make a trial worth running:

• It is an add-on, started with your vet. Keep the medication, dose, timing and even the brand exactly as they are while you test the diet, so any change you see is down to the diet alone. And the non-negotiable, again: never reduce or stop anti-seizure drugs to "make room" for diet (Bhatti et al., 2015).
• Introduce it gradually. Switch foods over several days to avoid stomach upset. MCT diets and oils add calories, so watch your dog's weight, and know that a minority of dogs get loose stools.
• Give it a fair, long enough trial. The realistic benefit is a frequency reduction in a subset of dogs, not an overnight cure, so the only honest way to know whether it is helping yours is a before-and-after comparison of seizure frequency over weeks to months.

That last point is where the Seizure Diary earns its keep. Log seizures for a stretch before the change, then keep logging through the trial, and you will have a real frequency comparison to review with your vet rather than a vague impression. The diary is the evidence base for the decision that matters: keep going, or stop. If you are not yet in the habit, keeping a seizure diary walks you through it. And if a fair trial does not help, that is not a failure on your part or your dog's. It is the expected outcome for some dogs, and diet is only one lever anyway, with routine, sleep and stress mattering too, which is where identifying and managing seizure triggers picks up.

Almost everything above is dog research, and it does not transfer to cats. There is even less evidence in them, effectively no controlled diet-and-epilepsy trials at all (Vendramini et al., 2024). Cats are obligate carnivores whose metabolism does not slip into ketosis as readily as a dog's, and high-fat dietary manipulation carries species-specific risks: a cat whose food intake drops can develop a dangerous liver condition called hepatic lipidosis. So do not extrapolate the dog MCT data to a cat. Any dietary change for a cat with seizures must be vet-directed and cautious, and feline seizure management remains a medical matter handled through the drug articles.

So where does that leave you and a bag of "miracle" food? With a clearer head, I hope. The genuinely supported option is MCT as an add-on to your dog's medication, with a modest but real chance of fewer seizures and a smaller chance of a bigger win. The rest, strict keto, grain-free and the supplement aisle, is unproven or running ahead of the evidence. If you want to try MCT, do it the rigorous way: talk to your vet, keep the meds exactly as they are, introduce the food slowly, and log every seizure in the diary so that in a few months you will know, from your own dog's data, whether it is worth keeping.