What astonished Abrahams and helped drive his effort to publicize the diet was that keto was not a new idea. It was first used as a medical treatment for epilepsy in the 1920s. The principles underlying the diet have been around since Hippocrates touched on them nearly 2,500 years ago. Starvation had long been one approach to treating epilepsy. Deny the patient food for, say, a week and often their seizures went away. But there were obvious limits on how long starvation could be used as a treatment. In the 1920s, researchers at the Mayo Clinic, looking for a way to treat diabetics, figured out that it was not fasting per se that helped control seizures. Rather, they found that it was what the body did during an extended fast that helped control them. Deprived of food, the human body starts burning body fat as fuel, and it was that process of ketosis that somehow had the antiepileptic effect. Trick the body into thinking it was starving by taking away its primary fuel of carbohydrates and forcing it to subsist on an all-fat diet, and you could create that antiepileptic effect as long as necessary.
The diet was quickly adopted and widely used through the 1930s. And then, almost as fast as it had appeared, the keto diet disappeared. When Dilantin was first used as an antiepileptic drug in 1938, its success steered medical minds toward pharmaceutical solutions. A generation later, the diet had been all but forgotten. There was no scientific evidence that it worked, after all. More important, it was incredibly difficult to administer. Even in the 1990s, Millicent Kelly, Charlie Abrahams’s dietitian at Johns Hopkins, was planning menus with a calculator and a legal pad.