UNSURPRISINGLY, Warburg saw his “Lindau Lecture” as a triumph. “The path proposed in Lindau seems to be succeeding,” he wrote to a colleague. “If it turns out to be so, the opposition to the Dahlem work will be silenced; it will be similar to what happened to Pasteur. After he had healed the dogs, everyone agreed with him.”1
The following year, with Dean Burk’s help, Warburg published a revised version of his lecture under the title The Prime Cause and Prevention of Cancer. The published edition included a preface in which Warburg stressed the importance of vitamin supplements for cancer prevention. Warburg’s thinking on vitamins was related to his concerns about chemical poisons. Artificial chemicals, he believed, caused cancer by interfering with the enzymes that allow cells to use oxygen. Vitamins provide critical components of those same enzymes, meaning that a lack of vitamins could cause the enzymes to malfunction just as a poison might.
Warburg’s interest in supplements wasn’t new. In the 1920s, he had developed a mineral tablet that was sold by a German pharmaceutical company. And he had been researching vitamin-based treatments for cancer since the 1930s, when he made his major discoveries about coenzymes. But in his last years, Warburg was newly convinced that consuming large quantities of vitamins could prevent cancer. In 1968, a colleague sent Warburg a letter asking for his thoughts on one particular vitamin-based therapy. Warburg suggested that the formula needed 10 times as much vitamin B2.2
Warburg might have been making extreme and unsupported claims, but the idea that a lack of particular vitamins could lead to cancer wasn’t unreasonable. Richard Doll and Richard Peto considered the relationship between vitamins and cancer in the “Diet” section of “The Causes of Cancer”—which included the natural ingredients of food, as opposed to artificial chemicals added to food. Doll and Peto found the data on vitamins and cancer inconclusive, but unlike Warburg, they also considered other possible ways that people’s diets might cause cancer. They went through the possibilities one by one, from the fat and fiber content of food, to natural chemicals that might act as carcinogens, to how cooking might turn an otherwise safe ingredient into a potential hazard.
It was a well-intentioned effort, but figuring out how many cancers could be linked to our diets turned out to be far harder even than determining the hazards of artificial chemicals. Because it is nearly impossible to run an experiment that controls what individuals eat and drink over many years, Doll and Peto relied primarily on epidemiological studies, which merely track what people choose to eat and whether they develop health conditions. Doll had used the same approach to study smoking and cancer, but such “observational” studies depend on study subjects accurately recording what they have consumed each day, and a separate body of literature suggests that the vast majority of people are unable to do this.
An additional obstacle, Doll once pointed out, is that “we don’t eat individual things.” Most of our meals include many different foods and varying mixes of fats, proteins, and carbohydrates. The black-and-white world of smokers and nonsmokers is impossible to reproduce in studies of nutrition and cancer. This is one reason why so many different foods have been shown to cause cancer in one study and to prevent it in the next.
Even when the data are accurate and one food or drink can be successfully isolated from the rest of the diet, it remains difficult to make definitive statements about diet and cancer. That people who eat a lot of vegetables are found to be less likely to get cancer in a study might mean that eating vegetables lowers one’s cancer risk. It might also mean that people who eat lots of vegetables tend to eat less of another food or exercise more and that it’s the absence of the other food or the additional exercise that is responsible for the protection against cancer. Researchers make statistical adjustments to account for these possibilities, but no amount of mathematical maneuvering can eliminate all of the possible confounding factors. As UCLA epidemiologist Sander Greenland once put it, identifying smoking as a cause of cancer was a “turkey shoot.” Assessing most other environmental causes is more like shooting without aiming.3
The final part of the “Diet” section of “The Causes of Cancer” appeared under the heading “Overnutrition.” Amid the incredibly complex scientific discussion in the report, the idea that cancer could be a problem of eating too much sounded almost laughably simplistic, but Doll and Peto took the idea seriously, writing that overnutrition should perhaps have been the first part of their diet discussion given some of the striking findings.
The idea that eating too much or carrying too much weight made someone prone to cancer wasn’t new. In his 1810 book Observations on the Cure of Cancer, the Scottish surgeon Thomas Denman wrote that it was “thought, if not proved, that those who become corpulent or fat” are more susceptible to cancer and also that cancer was “a more rapid and intractable disease” among the overweight. Over the next century, many other cancer authorities arrived at the same conclusion. “Probably no single factor is more potent in determining the outbreak of cancer in the predisposed, than excessive feeding,” the British cancer expert W. Roger Williams wrote in 1908.4
The belief that cancer had something to do with eating too much and excess weight gained traction in the early twentieth century via the rodent feeding studies done by Peyton Rous and others, which found that low-calorie diets were protective against cancer. These studies faded to the periphery of mainstream science after the First World War, but overeating emerged as the most popular of the many food-related cancer theories in Germany in the 1920s and 1930s. Much of the German public’s interest in the topic can be traced to a slew of unscientific books touting “hunger cures.” But more serious work also considered the quantity of food one ate to be an essential part of the modern cancer story.5 Frederick Hoffman, the insurance actuary who became the world’s authority on cancer statistics, considered dozens of different ways that our diets might contribute to cancer in his 1937 book Cancer and Diet. And though Hoffman typically acknowledged his doubts about each hypothesis, he ends the book with a surprisingly definitive statement on overeating and cancer:
I consider my own duty discharged in presenting the facts as I have found them, which lead to the conclusion that overnutrition is common in the case of cancer patients to a remarkable and exceptional degree, and that overabundant food consumption unquestionably is the underlying cause of the root condition of cancer in modern life.6
In their discussion of overnutrition, Doll and Peto singled out Albert Tannenbaum, a cancer researcher in Chicago who had studied the hazards of uranium as part of the Manhattan Project. While working on cancer experiments that were unrelated to diet, Tannenbaum noticed that mice that weighed less appeared to get cancer less often. He spent much of his career studying the phenomenon. In one experiment, Tannenbaum gave a group of 50 female mice a diet of 2 grams of standard rations each day. A second group of 50 female mice received the same 2 grams of food plus as much cornstarch—a carbohydrate formed from glucose—as they cared for. After 100 weeks, the mice that had binged on cornstarch had 26 mammary tumors. None of the mice on the restricted diet had a tumor.
Early twentieth-century researchers had performed similar experiments and arrived at similar results, but in Tannenbaum’s next experiments, he did something that earlier researchers had not been able to do. Before running the studies, he first exposed two groups of mice to a powerful chemical carcinogen. After 60 weeks, mice on the restricted diet had developed 11 skin tumors. The mice who also ate cornstarch had 32 skin tumors. Chemical carcinogens alone could cause a mouse to develop cancer, Tannenbaum saw, but they appeared to do so more readily when animals were well fed.7
Tannenbaum found that underfeeding mice prevented or delayed the growth of every type of cancer he studied. The less Tannenbaum fed his mice, the greater the effect. With each new experiment, he grew more convinced he was onto something significant—his colleagues joked about how often he could be found weighing his mice. But though confirmed by other scientists, Tannenbaum’s results failed to bring nutrition studies into the mainstream of cancer research. Such studies didn’t exactly seem like cutting-edge science in the early twentieth century, let alone in the postwar period. And the skepticism of feeding studies was reasonable. No one could say whether the low-calorie diets that benefited mice would also benefit people.
In an effort to see if his findings were in fact relevant to humans, Tannenbaum searched through the records of insurance companies in the late 1940s. What he found likely didn’t surprise him: the more someone weighed when past middle age, the more likely the person was to die of cancer. Avoiding excess weight “might result in the prevention of a considerable number of cancers in humans,” Tannenbaum wrote, “or, at least, the cancer process may be delayed in time of appearance.”8
THAT EATING TOO MUCH appeared connected to cancer was a perplexing plot twist for Germany, a nation that began the twentieth century panicked about both cancer and having enough to eat. In fact, the abundance of nutrients in many Western societies in the second part of the twentieth century was not unrelated to Germany’s long-standing anxieties about its food supply.
Warburg had always wanted to grow more food by making photosynthesis more efficient, yet the scientist who found a way for Germany (and the world) to have far more food was not Warburg but Fritz Haber, the Kaiser Wilhelm Society chemist who had pioneered the use of gas weapons during the First World War. Haber’s plan was to secure the German food supply by solving the nation’s nitrogen problem.
Outside of scientific circles, nitrogen may be less appreciated than the three other primary components of the organic world—carbon, oxygen, and hydrogen—but we cannot live without it. Animals get their nitrogen by eating plants or by eating animals that have eaten plants, and almost all plants take their nitrogen from the earth. Without enough nitrogen, crops will not grow and humans will not eat. As the world’s population grew in the nineteenth century, agricultural production increased and the nitrogen in the earth’s soil rapidly dwindled. For the farsighted scientists waking up to the problem of nitrogen depletion, it was an existential crisis, not unlike the threat of global warming today. In the early twentieth century, nitrogen “was the weakest link in the chain of life,” in the words of Haber’s biographer Daniel Charles, “a substance more scarce than water, sunlight, or any other nutrient.”
Germany did what it could. Manure is excellent fertilizer because it contains nitrogen, and in the nineteenth century, Europeans were transporting bird excrement from Pacific islands by the boatful. But birds could defecate only so often. A new answer was needed, and it was found in the mountains of Chile, where nitrogen (bonded to oxygen in the form of nitrate) could be mined. It was a lifesaving, but temporary, solution. By the end of the century, Europe was importing more than a million tons of Chilean nitrate each year. It was only a matter of time before the supply ran out.
For European scientists of the era, the lack of nitrogen in the soil was a maddening crisis. Nitrogen, after all, is everywhere. It makes up almost 80 percent of the air we breathe. Yet this nitrogen is a gas formed from pairs of nitrogen atoms locked together in a tight embrace. Scientists had no way to separate the 2 nitrogen atoms from each other and thus no way to make them bond with oxygen and form the molecules plants rely on to grow. The nitrogen in the air, Charles wrote, was “as nourishing as the seawater surrounding a thirsting sailor.”9
In 1908—the year Warburg first measured the breathing of sea urchin eggs—Haber created a device that placed nitrogen gas in the air under so much heat and pressure that the atoms finally gave in and released their twins. Inside the device, hydrogen molecules stood by, waiting to bond. The resulting nitrogen-hydrogen compound, ammonia, naturally reacts with oxygen to form the nourishing molecules plants need.
The breakthrough didn’t immediately revolutionize worldwide agricultural production. By the time German industry had managed to mass-produce ammonia, World War I had started. Rather than going into German fields in the form of artificial fertilizers, the ammonia was used to make explosives for the military. But in time, more and more of the nitrogen in Haber’s ammonia did find its way into the earth, and the impact could hardly have been greater. In the opinion of the Czech Canadian scientist and author Václav Smil, Haber’s ammonia-producing process was “the most important invention of the twentieth century.” Without it, Smil argues, the world population could never have grown from 1.6 billion in 1900 to 6 billion by the end of the twentieth century. In 1997, Smil estimated that at least 2 billion people were then alive thanks to Fritz Haber.10
Hitler had planned to solve Germany’s food crisis by conquering Eastern Europe. Had he not been longing for war, perhaps he would have noticed that one of the Jewish scientists he had chased out of Germany had already found a way to provide Germany with far more bread than all the grain fields of the Ukraine ever could.
Warburg, too, might have taken a lesson from Haber—a lesson that extended beyond food production. Though Warburg thought that diet was hugely important, his interest was limited to the hazards of artificial chemicals and the protective effects of vitamins. He had no interest in the connection between overnutrition and cancer. Haber’s success with nitrogen and the subsequent increase in agricultural production was a reminder of the fundamental relationship between nutrients and growth. And if ignored by Warburg and almost all other cancer scientists during his lifetime, the lesson was not lost on everyone.
In an 1889 article in The Lancet, the English surgeon Stephen Paget pondered why a cancer might spread to one tissue of the body rather than another. “When a plant goes to seed,” Paget wrote, “its seeds are carried in all directions; but they can only live and grow if they fall on congenial soil.” The “seed,” in Paget’s construction, is the cancer cell. The “soil” is the body—the environment in which the seed grows.
The first researcher to draw a clear connection between glucose in the diet (in the form of carbohydrates) and Warburg’s glucose-hungry cancer cells may have been the pioneering cancer researcher Anna Goldfeder, who carried out her own series of feeding studies in the late 1920s. Goldfeder reported that tumors grew more quickly when animals ate high-carbohydrate diets and mentioned that Warburg’s research might explain the phenomenon. “The organism is to be regarded as a soil for tumor growth,” Goldfeder wrote.11
If Warburg missed Goldfeder’s paper, he had plenty of other chances to think about aspects of a cancer’s soil other than oxygen. In 1926, on Warburg’s recommendation, the Kaiser Wilhelm Society had invited the Danish cancer scientist Albert Fischer (no relation to Emil) to come to the Kaiser Wilhelm Institute for Biology as a visiting scientist for three years. Warburg worked in the same building as Fischer and knew him well. In 1929, Fischer suggested that the cells of a body could be analogized to plants. Cancer cells are like “weeds,” Fischer said. They outgrow their neighbors because they are better at taking up the food in their environments.12
Ernst Freund, the young Austrian doctor who had found a connection between elevated blood glucose and cancer in the 1880s and then attempted to cure Crown Prince Friedrich, also thought in terms of seeds and soil. In 1938, the year Germany invaded Austria, Freund fled to England, where, at age 75, he obtained funding for a new cancer research program. By then, Freund had moved on from his single-minded focus on the role of elevated glucose in cancer, but he remained convinced that cancer was a systemic problem that could only be understood in the context of the metabolism of the entire body.
As Freund saw it, the microscope itself had led scientists astray by allowing them to peer into the cancer cell and thus to lose sight of the larger picture. “[I]nstead of confining ourselves to an investigation of the action of the malignant cell on the organism,” Freund wrote, “we ought to investigate the action of the organism on the malignant cell.”
Shortly before his death in 1946, Freund published a book, Metabolic Therapy of Cancer, in which he urged cancer scientists to think beyond the cancer cell and to consider the soil of the body in which it grows. “[I]t is only natural to link growth of any kind with nutrition,” Freund wrote. “The farmer ascribes a good or bad crop primarily to the nature and composition of the soil, i.e., the nutrient, though he must make allowances for the quality of the seed; and every layman realizes the value of diet to growth.”
“Curiously enough,” Freund continued, “medical science takes a different attitude.”13
IN “THE CAUSES OF CANCER,” Doll and Peto paused to consider medical science’s attitude toward overnutrition. Why, they wondered, had Albert Tannenbaum’s feeding studies made “little impact on cancer research,” given the evidence showing that the quantity of food an animal consumes can have “profound effects” on cancer risks? Part of the problem, they suspected, was that low-calorie diets were impractical for most people. But there was a deeper issue as well: “[W]e still have no clear idea of the mechanisms whereby dietary restriction protects laboratory animals,” Doll and Peto wrote. Tannenbaum had struggled with the same issue. Understanding how a low-calorie diet prevented cancer “might really give an insight into carcinogenesis itself,” he noted in 1959. And yet there remained “no explanation.”14
The lack of certainty on the role of diet is reflected in the conclusions of “The Causes of Cancer.” Doll and Peto estimated that diet accounted for 35 percent of American cancer deaths. That figure made diet the single greatest cause of cancer, and a 2015 review found that Doll and Peto’s findings have held up remarkably well. But it was little more than an educated guess. The actual percentage of American cancer deaths caused by diet, they stated, might be as low as 10 or as high 70. “Diet is a chronic source of both excitement and frustration to epidemiologists,” they wrote.15
Neither Tannenbaum nor Doll and Peto considered that Warburg’s discovery of fermentation in cancer might offer a way to explain the fundamental connection between seed and soil, between overeating cancer cells and overfed bodies. The oversight is perhaps understandable. Warburg himself could not see the link, and though the two phenomena might have seemed related, no one could fully explain the relationship. What was missing at the end of the postwar period was a metabolism synthesis, a way to combine Warburg’s discovery of fermentation in cancer cells with a long history of observations on food consumption, body weight, and cancer.
That synthesis would eventually arrive, and with it, perhaps, the best answer of all to the question of why cancer became so common in the Western world in the nineteenth and twentieth centuries. But before Warburg’s discovery of how cancer cells eat could help solve a long-standing cancer mystery, something odd happened: everyone forgot about it.