These missionary and colonial physicians did often diagnose tumors and other diseases of civilization in local whites, and among natives who were working for European households and industries. In August 1923, for instance, A. J. Orenstein reported in the British Medical Journal on his experience as a superintendent of sanitation for the Rand mines in South Africa: “In a series of one hundred consecutive necropsies on native mine laborers conducted by myself in the latter part of 1922 and the first two months of 1923, two cases of carcinoma were observed—one was carcinoma of the pancreas and glands of the neck in a native male of the Shangaan race, age about 40, the other was a case of carcinoma involving practically the whole of the liver, in a native male of the same race, age about 25.” The reports from these physicians were a reminder of how dramatic the course of the disease could be, and evidence against the argument that sophisticated diagnostic technology, unavailable in these outposts, was required to diagnose cancer. In 1923, George Prentice, who worked in Nyasaland, in southern central Africa, described one native patient with an inoperable breast tumor in the British Medical Journal: “It ran an uninterrupted course,” Prentice wrote, “completely destroyed the breast, then the soft structure of the chest wall, and then ate through the ribs; when I last saw the negress in her village, I could see the heart pulsating. That was just before her death.”

The absence of malignant cancer in isolated populations prompted questions about why cancer did develop elsewhere. One early hypothesis was that meat-eating was the problem, and that primitive populations were protected from cancer by eating mostly vegetarian diets. But this failed to explain why malignancies were prevalent among Hindus in India—“to whom the fleshpot is an abomination”—and rare to absent in the Inuit, Masai, and other decidedly carnivorous populations. (This hypothesis “hardly holds good in regard to the [American] Indians,” as Isaac Levin wrote in 1910. “They consume a great deal of food [rich in nitrogen—i.e., meat], frequently to excess.”)

By the late 1920s, the meat-eating hypothesis had given way to the notion that it was overnutrition in general, in conjunction with modern processed foods, lacking the vital elements necessary for health, that were to blame. These were those foods, as Hoffman put it, “demanding conservation or refrigeration, artificial preservation and coloring, or processing otherwise to an astonishing degree.” As a result of these modern processed foods, noted Hoffman, “far-reaching changes in bodily functioning and metabolism are introduced which, extending over many years, are the causes or conditions predisposing to the development of malignant new growths, and in part at least explain the observed increase in the cancer death rate of practically all civilized and highly urbanized countries.”