The Missing Facts About My Dog’s Epilepsy

First, a bit of newsletter business: This is the first edition of Nine Lives since I split it into two separate newsletters. If you subscribed to Nine Lives before April 2026, you will also be receiving News & Notes (more casual, more personal, more frequent, not archived, and — usually — shorter) unless you separately unsubscribe.

Longtime Nine Lives readers will remember our dog Tashi, whom Jak and I found as a starving, half-grown puppy in 2023. A quick recap for folks who haven’t been following my dog news over the past three years: I introduced Tashi to “talking pet buttons” and he took to them right away, eventually topping out at a vocabulary of around 150 words. With this extra level of communication, we were able to negotiate our way out of his (fairly intense) early problems with self-control and emotional regulation, and he became a very well-behaved, loving dog. In large part because of our shared language, my bond with him now is very deep. He is my heart on four paws, in a way no creature has been since my long-departed, also extraordinary cat Misha. I would move the moon for him.

And that’s just about what it’s come to, because at around a year and a half, Tashi developed idiopathic epilepsy. This was on top of the severe hip dysplasia that he’d already had one surgery for, and which continues to cause him pain and reduce his mobility. (In fact, as I write this, Tashi is at the vet hospital recovering from surgery to remove the joint of his other hip.)

The one and only veterinary response to Tashi’s seizures has been to add more and more anticonvulsant drugs, which have only resulted in partial seizure control, at the price of terrible side effects. Only when I went off-script with my own research and experimentation did I discover both a cause and a treatment that no one else had mentioned.

I desperately want to boost awareness of these little-known facts among dog and cat owners. Life expectancy for an epileptic pet is inversely correlated with frequency and duration of seizures — or in plainer language: every seizure event brings a pet that much closer to death.

By the most conservative estimate,¹ almost half a million dogs in the United States alone have idiopathic epilepsy. Having a better method of seizure control is literally a life-and-death issue for millions of pets worldwide — as is the knowledge to avoid triggering epilepsy in the first place.

Jak and I first observed Tashi’s tonic-clonic seizing in January of 2025, although in hindsight I realized that he’d had two other seizures several months earlier, unobserved. (He always hypersalivates during his grand mal seizures; twice before we had found him covered in drool without at the time knowing why.)

In the US there are four drugs commonly used for control of canine epilepsy; only three are available in Mexico, where we live. Two of those three — potassium bromide (KBr) and phenobarbital — have significant negative side effects. (Levetiracetam, the generic of Keppra, is the third; its only downside has been the short half-life, requiring dosage every eight hours.)

Under the guidance of our vet, we started Tashi on a moderate dose of potassium bromide; when the seizures continued to escalate, we added and then maxed out levetiracetam. By May he was having tonic-clonic seizures every ten days, so we raised the KBr again, and then finally took it all the way up to the max dose (900 mg per day, for a 22 kg dog) in October. At that point we finally got some relief — his next two seizures, about five weeks apart, were the gentler, focal sort instead of violent, full-body convulsions.

Unfortunately, once we passed 450 mg KBr, Tashi began to exhibit ataxia — a lack of muscular control — in his hind legs, a known side effect of bromides. By 900 mg, his legs collapsed under him every time he tried to run, and sometimes even at a walk. Since he was already semi-disabled in the hips due to dysplasia, this was particularly dangerous.

According to our best local vets, as well as the veterinary neurological specialist we consulted in Guadalajara, the only remaining treatment would be adding phenobarbital to the bromide and the lev. Phenobarb is a drug of last resort, because it does severe damage to the liver, eventually resulting in complete liver failure … and in any case, it also causes ataxia like the KBr, so it wouldn’t help with the problem we were having.

Coincidentally, in early December I happened to be reading a (fascinating) nonfiction book about food science, wherein I was reminded that humans with epilepsy are frequently placed on a ketogenic diet: high fat, low carbohydrates, prompting the body to use ketones (from fat) instead of glucose (from carbs) for energy. For reasons that are not entirely understood, epileptics suffer fewer seizures under ketosis.

It occurred to me to wonder whether that might possibly be a useful option for dogs as well, and so I began specifically searching for scientific literature on the subject.

While I was engaged in that research, I stumbled across an FDA alert from 2018 about a class of flea control treatments that sometimes caused seizures in dogs and cats. To be clear, these are FDA-approved medications, widely sold in the US and around the world for both dogs and cats without any kind of neurotoxicity warning on their exterior packages.

The flea treatment that Tashi was on at the time, Bravecto (fluralaner), was on the warning list. His first seizures — as indicated by the hypersalivation that we noticed without immediately understanding what it signified — had occurred soon after I switched Tashi from Advantix II to Bravecto in fall 2024.

In fact, an ongoing class action lawsuit against Merck, the manufacturer of Bravecto, states that “the FDA has received three times as many reports for neurological adverse events after animals ingested Bravecto than any other flea and tick product containing isoxazoline”. I suspect that the fact that Bravecto is 1) ingestable rather than topical and 2) intended to last for twelve weeks instead of four are what makes it more frequently toxic, but all isoxazolines are a neurological risk.

To make things worse, I then discovered that literally every single flea preventative sold in Mexico is on that warning list — they’re all from the same isoxazoline class of drug. So, you know, fuck me.

Meanwhile I found some tantalizing indications that ketogenic diets could help control seizures in dogs. There hasn’t been nearly as much scientific study as I would like, but what I read convinced me that it was worth a shot. At least there is a century’s worth of data and information on various ketogenic diets for epilepsy management in humans, which follows similar principles. I hoovered up all of it, everything I could find.

I was already making homemade BARF-style raw food for Tashi — an original diet I’d concocted from locally available ingredients with the aid of the USDA nutrition database,² dog nutrition requirements from both NRC and AAFCO, and a gigantic and complex spreadsheet that I created to track it all. During December I began working on a new recipe, trying — with the basic whole human-grade foods available to me here in Mexico (i.e. no specialized pet supplements) — to hit ketogenic percentages of fat and carbs, while still generously meeting all the nutritional requirements, and not breaking our budget. It took a truly epic number of hours, and I ultimately had to resort to importing human supplements for a few specific vitamins and minerals, but I got there. I planned to start Tashi on the new diet in early January, after our end-of-year trip to the States.

For the last several years, Jak and I have been making an annual trip up to Texas to visit his brother's family — driving, rather than flying, so we could load up on a variety of items that are either prohibitively expensive or completely unavailable in Mexico. Jak’s brother has two Great Pyrenees dogs who are Tashi’s age, and so we had brought him with us on two previous visits. The canine cousins were a huge hit with Tashi, and while the long car ride was not his favorite, he much preferred being with us instead of stuck alone at home.

But maintaining Tashi's complex homemade raw diet during the ten-day trip was just logistically impossible. My plan had been to temporarily swap him to a high-quality kibble, already purchased and waiting for us in Austin. As I read more about diet and epilepsy in both dogs and humans, I began to worry that a sudden change in Tashi's diet might actually trigger a seizure. However, the trip was imminent and there was no time to change the plan — I couldn’t produce a petsitter capable of looking after an epileptic dog, on short notice, over a major holiday.

Sure enough, Tashi had a seizure on the way home, a full grand mal — worse than the focal seizure he’d had just a week and a half earlier, before we left. I felt bad, knowing that the trip had surely caused that one, but at least now I was equipped to get him on different flea meds and try a dietary therapy. From the States I brought back a year’s supply of Advantage II (the safest flea preventative for seizure-prone dogs and cats, sadly not sold in Mexico), and more than a year's supply of lab-certified MCT oil, as my research had convinced me that it was worth incorporating in a ketogenic diet.

January fourth was the date of the seizure he had while we were stopped for the night in Monterrey. Within one week of our return, I had Tashi fully switched over to the high-fat keto diet; I graduated introduction of MCT oil across four weeks, because I was concerned about stomach upset. (This was probably overly cautious; I now think I could have moved much faster on that.)

January and February passed with no seizure activity. In early March I reduced Tashi's KBr by one-quarter (900 mg to 675 mg). Waited a cautious four weeks more (KBr has a very long half-life, especially in dogs). Still no seizures, so I dropped it again in April to one-half of the max dose. His ataxia slowly diminished and his mobility improved.

And still, no seizures. In early May I took the KBr down by another quarter, to 225 mg. By this point, four months seizure-free and counting, I began to have wild hope. What if the pesticides were the cause of his seizures, and the damage was reversible? I could not find any information on what happened to the dogs who developed epilepsy from a flea pesticide after the pesticide was stopped. Could I possibly have just cured him, by taking him off the Bravecto? That would be beyond amazing.

Three weeks later, on May 21, Tashi finally had a short tonic-clonic seizure, dashing my hope that I had managed to solve the whole problem in one fell swoop. But he had made it just shy of twenty weeks with no seizures.

The timing initially made me suspect that the lower levels of bromide were what allowed the breakthrough, so I immediately took him back up to 450 mg. However, the same weekend that I lowered his KBr to 225 mg, I also accidentally dosed Tashi with leftover Advantix II flea treatment instead of Advantage II. Although Advantix II does not contain an isoxazoline-class drug, it does contain permethrin, which causes severe neurological damage to cats and — I later confirmed — is contraindicated for seizure-prone dogs.³ So Tashi’s one and only seizure in the six months since conversion to a ketogenic diet might also have been a reaction to a flea treatment.

Between the suspicious timing, and the preponderance of other cases of this type, I believe that Bravecto is almost certainly the proximate cause of Tashi’s “idiopathic” epilepsy. The pharmaceutical companies, eager to keep raking in the profits, have maintained in court that isoxazaline-class drugs are simply triggering a pre-existing epileptic tendency … which isn’t much of a defense, in my opinion. But proving sole causation in such a way that would revoke the FDA’s approval and take isoxazoline treatments off the market — which is what I honestly believe should happen — is functionally impossible; it’s not even clear that Merck won’t end up beating the class action lawsuit on the lesser issue of adequate disclosure.

In either case, it is appalling to me that even after Tashi began having seizures, none of my vets knew to suggest that I take him off the two classes of pesticides (isoxazolines and pyrethroids) that are known to cause seizures in mammals.

The good news is that the ketogenic diet is helping tremendously with stopping Tashi’s seizures. Right now (June 2026) he’s on half the previous amount of KBr, and doing far better than he was before the diet, without the debilitating and dangerous hind-limb paralysis. If the best we ever get is stabilizing at, say, two seizures per year — that’s eighteen times longer than we were managing on the same drugs alone, without the ketogenic diet. In terms of both quality of life and life expectancy, that is huge.

And it might get better still. One of the published case studies I referenced followed a dog with refractory (medication-resistant) epilepsy for more than five years. The owners were able to stop both phenobarbital and potassium bromide completely during the first year on the keto diet, leaving only levetiracetam (which Tashi is also taking) and zonisamide (which I cannot get in Mexico). It took about two years in ketosis — during which the dog’s seizures slowly got further and further apart — but she eventually reached full remission. When the case study was published, the dog had been on a keto diet for five years and totally seizure-free for the last three of them — down from a starting point of one seizure a week.

The handful of canine ketogenic diet double-blind studies have tested only short-term effects, over a few weeks or months, among other limitations. But from the tiny amount of long-term data I could find, it sure looks like ketosis has a cumulative protective effect across years. So I’m still hoping that over the next year or two Tashi’s situation will continue to improve beyond where we are after the first six months — especially as I continue to keep him away from the neurotoxic flea meds.

Tashi has gained a bit of extra weight on the new high-fat diet, so I’m still trying to zero in on the correct caloric intake for him. Also I’ve been tweaking the recipe to be as uncomplicated for me as possible, especially because I now have a second, non-epileptic dog to feed. I still spend at least three hours a week just prepping dog food; people with more money and the resources of a first-world country have easier options,⁴ but this is what I can do, and it’s absolutely worth it to me.

Regardless of whether the diet change brings Tashi to full remission or merely tremendously improved seizure control, I think the idea of ketogenic diets for epileptic pets is worth a lot more attention. Few veterinarians seem to be aware of diet as an effective seizure treatment, or ever broach it with clients as a possibility. And when I was reading up on canine epilepsy in the wake of Tashi’s first seizures, the idea of ketosis for seizure control was never mentioned in any of the resources I found … even though it’s been used with humans for over a century.

Also, I really want to sound the alarm about the potential for isoxazoline-class flea treatments (including Bravecto, Bravecto Plus, Credelio, Nexgard, Nexgard Plus, Nexgard Combo, Symparica, Symparica Trio, and Revolution Plus) to cause seizures in dogs and cats. Again, none of the vets I consulted had any awareness of this connection, not even the neurological specialist.

Perhaps US veterinarians tend to be better-informed, but with the way private equity has taken over the entire veterinary industry, turning the focus from “helping pets” to “making a giant profit”, I’d be surprised — after all, flea medication is a major revenue stream for veterinarians. (And the leading veterinary manual is published by Merck, the pharmeceutical company that makes Bravecto. It’s conflicts of interest all the way down.) As with so many things now, the burden of research and education is on individuals.

If you know anyone with a dog or cat, please share the FDA warning about flea treatments and seizures. If you know someone with an epileptic dog or cat, please share this article with them. I am happy to discuss ketogenic diets and share the information I have sourced and compiled with anyone who is interested in pursuing dietary therapy for their epileptic pet.

And just to be clear: I am not a MAHA conspiracist. I am pro-science; I believe vaccines are generally an amazing thing; I like having access to new and better pharmaceuticals. But I really do think the drug companies are protecting their profits here by covering up the fact that their flea treatments are causing major neurological damage to a significant percentage of beloved pets. And that’s just fucking evil.