Another example of this thinking also dates to the 1970s and comes from George Blackburn and Bruce Bistrian at Harvard Medical School. Bistrian and Blackburn developed what they called a “protein-sparing modified fast” to treat patients with obesity: 650 to 800 calories a day of nothing but lean fish, meat, and fowl. It had effectively no carbohydrates, making it a ketogenic diet, albeit a very low-calorie version. Bistrian and Blackburn prescribed the diet to thousands of patients, as Bistrian told me when I interviewed him in January 2003, and half of them lost at least forty pounds. In one 1985 publication reporting on almost seven hundred patients, the average weight loss was nearly fifty pounds in four months. The patients felt little hunger while on the diet. “They loved it,” Bistrian said. “It was an extraordinarily safe way to get large amounts of weight loss.” 

But one point that Bistrian made in our interview was critically important: If he and Blackburn had tried to balance these diets by adding, say, vegetables, whole grains, and legumes, meaning the patients would obviously be eating more calories and more carbohydrates, the diets would have failed. You’d think more calories would mean less hunger, but it would cause more. Bistrian was the first one who pointed out to me the different responses between Ancel Keys’s starvation experiment subjects—eating 1,600 calories a day and, well, starving—and the experiences of the patients he and Blackburn were working with, or Sidbury was treating, who were perfectly content consuming far fewer than 1,000 calories a day. “The proof of the pudding,” he said to me, “was in the eating.” 

Alas, Bistrian and Blackburn’s thinking, and Sidbury’s, was flawed. They were telling their patients—Sidbury’s were kids; Bistrian’s and Blackburn’s were adults—to severely restrict calories because that was what they still thought was necessary. Despite everything they knew about insulin and fat metabolism, they too couldn’t escape the trap of energy balance thinking. Because Bistrian and Blackburn were feeding their patients so few calories, this created a problem that appeared to the two researchers to be insurmountable. It had to do with maintaining the weight loss. 

For a diet to work for a lifetime, it has to be maintained for a lifetime, and for a diet to work—to get us lean, or relatively so—it has to remove or limit the cause of why we get fat. If the cause is too many calories, then a lifetime of calorie restriction at some level is necessary. If the cause is elevated insulin levels and too many carbs, then a diet that maintains insulin at a low threshold for a lifetime—carb-restricted, high in fat—is necessary. There seems no getting around it. Bistrian and Blackburn were perfectly aware of this problem. They knew that if their patients went back to eating the way they did, they’d gain the weight back. If they ate more calories but still worked to keep insulin low, they’d be restricting carbohydrates and replacing them with fat. They’d be eating what Bistrian and Blackburn thought of as an Atkins diet. And unless you believed that eating all that fat was benign—as many physicians finally do today—that was unacceptable. Bistrian said this was a primary reason he and Blackburn left the field. They had two choices, they thought: Tell their formerly fat patients to take appetite-suppressing drugs so they could spend a lifetime battling hunger on a calorie-restricted but balanced diet, or tell them to eat the Atkins/ketogenic way, to satiety of fat and protein. They considered neither to be a safe option. “All that saturated fat,” Bistrian said to me. He and Blackburn turned their attention to other things. We don’t have that luxury.

Gary Taubes. The Case for Keto: Rethinking Weight Control and the Science and Practice of Low-Carb/High-Fat Eating (Kindle Locations 1733-1734). Knopf. Kindle Edition.