IN 1965, the American scientists Clint Fuller, Andrew Benson, and Harlan Wood unexpectedly received mail from Germany. Inside each envelope was something even more unexpected: a personal check from Otto Warburg for 4,000 deutsch marks (about $1,000 at the time). As the accompanying letter explained, Warburg wanted the three scientists to attend a symposium in Strasbourg, a scenic city on the French-German border, where they would have the honor of responding to his latest photosynthesis papers.
The Americans’ surprise quickly gave way to puzzlement. Warburg, by then, was embroiled in yet another debate about photosynthesis—this time over the question of whether the oxygen given off by photosynthesis is derived from water or carbon dioxide. Warburg was sure the answer was carbon dioxide. The men he had invited to the symposium—along with most others in the field—were sure it was water. Still, 4,000 deutsch marks was a lot of money and Strasbourg was a beautiful city. And so all three Americans flew to Europe, a little confused but happy to tell Warburg that he was dead wrong, if that’s what he wanted.
The day before the symposium, Fuller, Benson, and Wood met a German professor who was to chair the event and asked what they should anticipate. Warburg, the professor explained, had planned a “hanging party.” The three Americans were there only so that they could “surrender publicly” and acknowledge that Warburg had been right all along.
Needless to say, the visitors were taken aback. They had no intention of surrendering to Warburg, let alone publicly. The next day, the symposium attendees gathered on the steps of the convention hall to greet Warburg, as though he were a foreign dignitary gracing them with his presence. When he finally arrived, he did so in style. As Fuller later recalled, a long black Mercedes pulled up with the license plate that read “MP-B1” (Max Planck Society–Berlin One). The chauffeur emerged from the car first and opened a back seat door. As the crowd looked on with anticipation, Warburg stepped out of the vehicle in a beige suit with what Fuller described as a “10-gallon Cowboy hat” on his head. At age 81, Otto Warburg had finally gone full diva.
Warburg took a moment to look Fuller up and down before shaking his hand: “Herr Fuller, you are much too young!” Warburg said. Fuller thought it was a compliment, though he wasn’t entirely sure.
The symposium began and Warburg presented first. Heiss accompanied him onto the stage carrying two rolled-up charts. Warburg and Heiss unrolled the first together and held it up for the audience. There was not much to see. It was merely a picture of one of the glass vessels that Warburg used to make his measurements. Warburg explained that he could accurately measure photosynthetic reactions with this simple vessel in a matter of hours.
With that, it was time to unroll the second chart. This one was different. It featured a long series of complicated equations based on photosynthesis experiments carried out with radioactive carbon—the same technique Weinhouse had used to supposedly “overthrow” Warburg’s cancer research. Such studies, Warburg explained, could take days to complete and would achieve much less precise measurements.
When his profoundly biased presentation of the two different approaches was complete, Warburg looked out at the audience: “It’s up to you to make your decision.”1
When their own turns came, Fuller, Benson, and Wood ignored Warburg’s talk and presented their own, entirely contradictory data. After the symposium, Warburg showed no signs of displeasure. Together with the mayor of Strasbourg, he hosted an elaborate dinner and said nothing of what had transpired that afternoon. He even invited the three Americans who had just contradicted his every word to visit his institute.
It’s possible that Warburg was in an unusually generous mood that evening. More likely, he had achieved a state of self-assuredness so complete and all-consuming that he was no longer even capable of considering that he might be wrong. The strong reaction of his colleagues was itself evidence that he was right. The greater the discovery, Warburg wrote to Burk, “the greater is the resistance.”
Shortly before his death in 1959, Emerson discovered that there are two light reactions in photosynthesis, meaning that a measurement of more than four photons did not contradict Einstein’s explanation of how photons and atoms interact. It was as clear a sign as possible that Warburg had been mistaken all along, but when a German biochemist later came across one of Warburg’s old notebooks, he discovered that Warburg had a different response to the news. “Finally, Emerson has confirmed my results,” Warburg had written.2
THE YEAR AFTER THE “hanging party,” Otto Warburg spoke at a gathering of Nobel laureates in the Bavarian town of Lindau. He spent most of the nearly hour-long lecture making the same exaggerated claims about the role of damaged respiration in cancer that he had been making ever since he had found the exceptionally high rates of fermentation in cancer cells of abdominal fluid. Citing the widely accepted estimates of the era, he pointed out that about 80 percent of cancers could be prevented. From Warburg’s vantage, the lack of emphasis on prevention—as opposed to treatment—was especially galling, considering that, as he put it in his talk, there is “no disease whose prime cause is better known.”3
The “Lindau Lecture,” as the talk became known, was Warburg’s final performance, and it received mixed reviews. His broader message about prevention and metabolism was as important as ever, yet the world had grown tired of his outrageous antics. In a dismissive article about the lecture, the German magazine Der Spiegel suggested that Warburg was now making false claims about cancer on an almost annual basis. Warburg had once told Hans Krebs that the secret to winning academic battles was to outlive one’s scientific opponents. He hadn’t considered that it may be possible to live too long and to see your once dominant position slip away and your theories forgotten.4
As the Nobel laureates in the audience in Lindau surely knew, Warburg’s notion of a “prime cause” of cancer was misleading. Warburg was using the language of Robert Koch, but he had subtly changed the terminology. Even if Warburg had been entirely correct that cancer arose when cells replaced respiration with fermentation, the “prime cause” wouldn’t be the transition to fermentation itself but whatever caused that transition to take place.
The researcher most qualified to identify the prime causes of cancer in the 1960s was not Otto Warburg but Richard Doll, the celebrated British epidemiologist who had linked smoking to cancer in the 1950s. Doll’s path to cancer research had been fortuitous. At 17, Doll had decided to become a mathematician, only to ruin his chance of winning a scholarship to Cambridge by drinking 3 pints of 8 percent alcohol ale the night before an exam. Doll elected to become a physician instead. While training at St Thomas’ Hospital in London in the 1930s, he and his fellow students took annual trips abroad to observe how medicine was taught in other countries. In 1936, they traveled to Frankfurt, Germany. One evening, while drinking with a group of German medical students, Doll shared his concerns about anti-Semitism. His German companions insisted that only a Jew would say such things and forced Doll to stand atop the table so that they could measure his ankles—thick ankles and flat-footedness were said to be typically Jewish traits.
It wasn’t even the most shocking moment of Doll’s trip. Before attending a talk on radiology, Doll had been warned that the lecturer was “a keen Nazi” and would expect everyone in the room to stand up and say “Heil Hitler” when he entered. Doll refused to stand and salute. The presentation proved as horrific as he could have expected. The lecturer showed a drawing of X-rays attacking cancer cells, only it wasn’t a typical medical illustration. The X-rays were depicted as Nazi troops. Their target: cancer cells adorned with Jewish stars. It “didn’t require many experiences of that sort to realize that there was something evil that had to be eliminated from the world,” Doll said.
Though known for his patrician bearing, Doll was a radical. In 1941, while serving in the war as an English medical officer, he applied to become a paratrooper so that he could confront the Nazi evil head-on. He was rejected and ended up serving in an infectious disease ward in Cairo. He had every reason to think he had missed his moment, but only three years after the war, history would offer Richard Doll another chance. England, having lost more than 300,000 soldiers and civilians during the war, was now under attack from a different sort of enemy that was killing its citizens by the tens of thousands. Lung cancer, a relatively rare diagnosis prior to the 1920s, was turning into a full-fledged epidemic. Between 1930 and 1944 alone, lung cancer deaths among English men had increased sixfold.5
That no one knew why lung cancer deaths were becoming so much more common only made the trend all the more terrifying—or at least terrifying to many, if not most, people. Some medical authorities at the time made the same argument long used to deny that cancer was on the rise in the late nineteenth and early twentieth centuries. The entire scare was said to be an illusion created by new diagnostic techniques. Chest X-rays, in particular, were thought to be revealing lung cancers that had always occurred at contemporary rates but had previously gone undiagnosed.
Smoking had become far more common in the decades before the increase in lung cancer rates, and, among those who believed the epidemic was real, smoking was a leading suspect. But that smoking caused lung cancer was not considered obvious. The medical attitude of the 1930s and 1940s was famously revealed in a sarcastic comment made by the American surgeon Evarts Graham, who pointed out that nylon stocking sales had increased during the same years that smoking had increased. Graham’s skepticism turned out to have been wrong, as his own research would later show, but he was making an important point. That two trends coincide—in this case cigarette smoking and lung cancer—does not mean that one causes the other. And as the skeptics pointed out, cigarette smoking was only one of a number of possible carcinogens that had grown more common in tandem with lung cancer. There were more pesticides on plants than ever before. Driving (and thus inhalation of car exhaust) had also increased dramatically during the same period. More cars meant more roads paved with carcinogenic tar. If smoking was a suspect for the crime of lung cancer in the late 1940s, it was standing in a police lineup together with a lot of other dangerous-looking suspects.
In the late 1940s, the British government first turned to Austin Bradford Hill, a highly respected epidemiologist and statistician, for help in solving the lung cancer mystery. Hill wanted a medical doctor on his team and knew that Doll, though not yet 40, was adept with numbers. Had it been feasible to do so, Doll and Hill might have run a true experiment, randomly assigning subjects with similar backgrounds to different groups. One group would inhale more car exhaust; another would smoke more; still another might consume more chemicals. The “control” group would carry on with their lives as normal. Such studies, known as randomized controlled trials, are the only way to definitively show cause and effect, and Hill himself had helped to design the method. But such a study, in addition to being profoundly unethical in this case because it could have caused great harm to the participants, would have to be conducted over many years, perhaps decades, before it could provide meaningful data on the causes of cancer. Doll and Hill, in need of a more realistic approach, turned to epidemiology, the study of health and disease patterns among populations.
For all the successes of Pasteur and Koch in tracing infectious diseases to specific microbes, it was epidemiology that did the most to ease the burden of the diseases in the nineteenth and early twentieth centuries. Doll’s scientific idol was John Snow, the English physician who had founded modern epidemiology in 1854 by locating the source of a deadly cholera outbreak in London. Snow solved the puzzle by mapping where the cholera victims lived and where they had obtained their drinking water. The dead, it turned out, had all been drinking from the same water pump. The question for Doll and Hill, and for England, was whether a similar technique could work for a chronic disease like cancer, which arises over many years and can have many different causes.6
As a first step, Doll and Hill selected two groups of hospital patients in and around London. One group consisted of 649 people who had been diagnosed with lung cancer. The other, the control group, was made up of hospital patients of similar ages who did not have cancer. Both groups were then asked to respond to surveys about their lives. The questions about smoking accounted for only one of the nine sections of the survey. Doll, a cigarette smoker himself, favored the car exhaust and road tar hypotheses.
The surveys were carried out by social workers. When Doll obtained the results, he recorded them with his fountain pen and did the calculations himself, like an accountant filling out a ledger of profits and losses. By the time he was done, he had given up his own cigarette habit. The study had found that those who smoked 25 or more cigarettes a day were 50 times more likely than nonsmokers to get lung cancer.
It took many more studies and many more years to convince the world that cigarettes were responsible for the increase in lung cancer deaths. Wilhelm Hueper, who continued to insist that artificial chemicals were the true cause of cancer, would emerge as one of the most outspoken skeptics of the smoking hypothesis.
Doll and Hill, as well as the American researchers Ernst Wynder and Evarts Graham, who carried out a similar study at approximately the same time, are now celebrated as the first to definitively link smoking to cancer by means of a modern scientific study. Yet, as was so often the case when it came to cancer, the Germans were more than a decade ahead. In 1939, the German doctor Franz H. Müller had published his own sophisticated epidemiological study and arrived at an even stronger conclusion. In 1950, Hill and Doll wrote only that smoking was “an important factor” behind the rise in lung cancer deaths; Müller had concluded that it was “the single most important cause.”
Müller’s paper was only one of a number of groundbreaking studies of smoking and cancer published by Germans during the Nazi era. And though their efforts largely failed to curb smoking, the Nazis launched the world’s most aggressive anti-smoking campaigns, complete with workplace bans, counseling centers for addicts, and nicotine-free cigarettes. In typical fashion, the Nazis would find a way to blend this initiative to protect life with beliefs designed to destroy it. Smoking was said to be the habit of depraved Jews and Communists, as well as Blacks and the Romani people. Jewish capitalists were blamed for the spread of tobacco use across Europe. Like the people who smoked it, nicotine itself was considered a threat to Aryan genetic material.
Hitler had smoked as many 40 cigarettes a day as a young man. He claimed he had given up smoking and “tossed his cigarettes into the Danube” upon realizing how much money he was wasting. It’s more likely that he was too poor to support his habit. In 1942, he said his decision might be credited for “the salvation of the German people.” Had he continued to smoke, Hitler reasoned, he would have died years before.7
IN THE DECADES AFTER Hill and Doll published their landmark study of smoking and cancer in 1950, Doll conducted a series of innovative cancer epidemiology studies and helped to identify a number of new carcinogens, including asbestos. And so when, in the early 1980s, the US government wanted a better understanding of why so many Americans got cancer, Doll was the obvious person to turn to. Together with his colleague Richard Peto, he carried out an extensive review of all of the available literature. The result was a 112-page report, “The Causes of Cancer,” published in 1981. It would stand as the definitive word on the subject for years.
Like Warburg, Doll and Peto believed that 75 to 80 percent of cancers were avoidable. That was not a surprise. Nor was it surprising that they attributed 30 percent of American cancer deaths to smoking. The truly startling finding was not what caused cancer but what did not. By Doll and Peto’s estimate, the artificial chemicals in our air and food accounted for only 2 percent of all cancer deaths. They attributed another 4 percent of cancers to toxic chemicals in the workplace.
Doll has been accused of being a shill for big business, but he wasn’t the only cancer authority pushing back against the chemical hypothesis at the time. In the 1970s, Bruce Ames’s test for carcinogens made him a hero to environmentalists. But in the 1980s, Ames and others began to test natural chemicals as well. The natural chemicals, it turned out, could also cause the bacteria to mutate and grow. There were even more carcinogenic chemicals in plants than in the pesticides sprayed on the plants. The findings could sometimes make a mockery of the entire field: a single cup of coffee had 19 distinct carcinogens in it. After running one test after another, Ames concluded that 99.9 percent of the “toxic chemicals” we are exposed to are “completely natural.” “Pollution,” Ames admitted, “seems to me to be mostly a red herring as a cause of cancer.”
Ames, too, would be accused of siding with corporate interests over ordinary Americans. But there was always a much better argument to be made in defense of the chemical carcinogen hypothesis. When Ames found that the natural world was full of toxic chemicals, his point wasn’t that plants cause cancer, but that his own eponymous test couldn’t tell us what did cause the disease. Doll and Peto, likewise, knew that their survey-based epidemiological methods were far from perfect—after all, people don’t necessarily know what chemicals they might have been exposed to over many years. Doll and Peto looked at animal studies in their analysis as well, but whether the high doses of a chemical that could cause cancer in a rat could also cause cancer in humans at much lower levels of exposure was far from certain.8
Doll and Peto concluded that only a small percentage of American cancers were caused by artificial chemicals because they had no convincing evidence to the contrary. But in “The Causes of Cancer,” they attributed less than a third of cancer deaths in America to smoking. That left most of the hundreds of thousands of cancer deaths each year still in need of an explanation.