Authors: Richard Kluger
Among these early cancer physician-detectives was Johns Hopkins-trained public-health specialist Morton L. Levin, working on the epidemiology of the scourge’s elusive causes at the State Institute for the Study of Malignant Diseases (later called Roswell Park Memorial Institute) at Buffalo during the 1930s and serving jointly as assistant director of New York’s new state bureau of cancer control, a unit of the health department at Albany. Levin’s discipline was so new, actually, that he was the sole investigator in the area at a time when only lip and tongue cancer had been definitely linked to smoking. “They didn’t know what to do with me,” Levin recalled, and much of his time was devoted to lecturing women’s groups on the need for early detection of cervical cancer.
But the data on smokers’ shorter life span reported in 1938 by Dr. Levin’s
former Hopkins mentor, biostatistician Raymond Pearl, and Franz Hermann Müller’s still more troubling study in Germany set the epidemiologist thinking, even though as a longtime pack-a-day smoker of Lucky Strike he was reluctant to probe the perils of the cigarette habit. He ordered the staff at Roswell Park to begin the systematic collection of data on the smoking habits of every entering patient, regardless of complaint and before any ailment had been diagnosed—in order to prevent unintended bias. Levin accumulated his data over a ten-year period, during which he encountered heavy skepticism from fellow physicians who were used to dealing with individual cases rather than representative samples of people and comparing multiple factors related to their pathology. Besides, many of his associates in the high-stress medical profession liked to smoke as much as he did and were not eager to be disabused about the effects of a favorite pleasure. But his findings were hard to ignore, if hardly conclusive. Among 236 cancer hospital patients who had smoked for twenty-five years or longer, the lung cancer rate was nearly 21 percent compared to an 8.6 percent incidence among patients who were nonsmokers; among those smoking a pack of cigarettes a day, the relative risk of lung cancer was ten times as high, Levin found, as among those who did not smoke.
He wrote up his painstakingly gathered data and, while serving as director of a temporary national Commission on Chronic Diseases installed at the American Medical Association’s national headquarters in Chicago, submitted his paper on smoking and cancer to the AMA’s
Journal
. Rejected without explanation and returned to his regular office in Albany, the paper caught up with the highly regarded epidemiologist in Chicago, where he traveled a few floors within the same building to ask the editor of
JAMA
why his paper had been turned down. “Oh,” said the red-faced editor, “the statistician did that,” and the latter was consulted. Surprised that the paper’s author was the same eminent Dr. Levin heading a national commission installed at the AMA, the statistician explained that he had trouble with the concept of “smokers” versus “nonsmokers” since the two could trade places at any moment. Levin satisfied the balky editor that this problem had been addressed by careful interrogation to determine the subjects’ smoking histories, but he was then presented with a second paper on the same subject that was about to meet the same fate as his own. Levin read it, recognized its value at once, reassured the statistician that it had been properly adjusted to account for the ages of the subjects, and urged that it be run in the same issue as his own study. It would be this second paper that led off the May 27, 1950, number of
JAMA
and gained far more attention than Levin’s useful study.
Ernst L. Wynder was the son of a Jewish physician in the Westphalian town of Herford, where his father, though viewed as somewhat leftist politically, had been socially prominent before Hitler came to power. The family fled Nazi persecution and settled in suburban New Jersey, where the senior Wynder resumed
his practice while young Ernst, now Ernest—he would alternate between the two spellings as part of his lifelong ambivalence toward his native Germany—graduated from New York University in 1943 and served in the U.S. Army intelligence section. As a medical student at Washington University in St. Louis, he wore his dark hair long, and his angular face often bore a dour, world-weary, even distasteful expression that looked to some as if he had just detected an unpleasant aroma in the vicinity. If his high opinion of himself was not entirely endearing, there was no gainsaying Ernie Wynder’s brains, energy, and ambition.
In the course of a summer internship at NYU after his second year at medical school, Wynder attended an autopsy of a lung cancer victim and was struck by the absence of any report of smoking in the deceased’s medical history. Inquiring of the widow, he discovered that the man had been a two-pack-a-day smoker, and Wynder began to suspect that the relationship between smoking and lung cancer had been underestimated due to such faulty reporting. With Bellevue Hospital right across the street, he approached the head of the chest service there and obtained permission to conduct in-depth interviews with those stricken by the disease, seeking not only his subjects’ smoking history in detail but information on such other possibly related factors as occupational exposures and the cause of parents’ death. “After twenty or so interviews,” Wynder recalled, “I knew I had something.”
Returning to medical school, Wynder asked his anatomy professor to approach famed thoracic surgeon Evarts Graham, head of chest service at Barnes Hospital, which was closely affiliated with the Washington University medical program, to inquire whether the precocious investigator could continue his data-gathering with the patients in Graham’s charge. The surgeon, renowned for having performed one of the first successful removals of an entire cancerous lung, was primarily a clinician with what struck young Wynder as somewhat limited curiosity about the causes of the diseases he sought to check with his scalpel. Besides, Graham was a heavy smoker and dubious about the cigarette-lung cancer connection. Other faculty members thought Wynder’s questionnaires a waste of time, but Graham saw no harm in the undertaking. Once the results began to accumulate, Wynder remembered, “Graham became quite supportive. It was a great break in my life.”
During his last two years at medical school, Wynder kept collecting his data, traveling to widely separated sites whenever possible. Among them was California, where he showed up one day in the state’s Department of Health office occupied by Dr. Lester Breslow, an alert young epidemiologist who believed the smoking-cancer link was more anecdotal than clinically demonstrable. He had his doubts, furthermore, that an investigator as brash and inexperienced as Wynder could perform reliable field work. But Breslow gave him access to state hospital patients and by way of monitoring his technique sent
along a health department examiner. “Our man reported back that Wynder’s work was just awful—the questioning was crude and the record-keeping inconsistent,” Breslow recounted. But Wynder apparently learned some lessons from his first serious brush with public health professionals, and so did Breslow, who as a result ordered his department to undertake its own investigation into the subject, the findings of which eventually proved almost identical to those in Wynder’s 1950
JAMA
report.
That article, “Tobacco Smoking as a Possible Etiologic Factor in Bronchiogenic Carcinoma: A Study of 684 Proved Cases,” with Dr. Graham tactfully listed as the second author, came to be regarded as a classic in epidemiology due to the size and geographic spread of its sample, the fact that its data derived from personal interviews and not from hospital records or other secondhand sources, and its breakdown of subjects into five clearly calibrated categories of smoking dosage, ranging from none (meaning less than one cigarette daily over the previous twenty years) to chain-smoking. Among its chief findings were that 96.5 percent of the lung cancer patients interviewed were moderately heavy to chain-smokers (compared with 73 percent of the 780 patients with diagnoses other than lung cancer), that almost all those stricken had smoked for twenty or more years, and 98.7 percent of the cigarette smokers were conscious inhalers (compared with 62.5 percent of the cigar smokers and only 18 percent of pipe smokers). There were no nonsmokers among the lung cancer victims in his collective sample—but then only 14 percent of the non-cancer hospital patients covered in the study were nonsmokers, suggesting that smoking was merely an incidental factor in the formation of the disease or that it had a major impact as well on diseases other than lung cancer. All of this invited far deeper suspicion about the cigarette habit.
Four months after the Wynder-Graham study appeared, the
British Medical Journal
carried the first preliminary report on the same subject by an Englishman who would in time rank with Wynder and one other scientist as among the world’s foremost investigators on the relationship of smoking to health. Dr. Richard Doll, ten years Wynder’s senior and a far more reserved personality, was a well-regarded scholar-biostatistician on Britain’s Medical Research Council when he and a colleague, A. Bradford Hill, made a study of 1,732 cancer patients at twenty London hospitals and compared them with 743 non-cancer patients. While somewhat more ambiguous than Wynder’s findings, the Doll-Hill results were still more incriminating when examined closely for the dose-response relationship: heavy smokers (twenty-five cigarettes or more a day) in the British study were found fifty times as likely as nonsmokers to contract lung cancer.
VII
HARDLY
noticed at the time but of major consequence to the future of the cigarette industry was the 1940s entry into the field by Joseph F. Cullman, Jr., a short, plump, high-spirited tobacco grower, broker, and investor, already in his sixties. Possessed of an old-shoe ease of manner and reputation for straight dealing, Joe Junior, as he was widely called, was among the best-loved men in his business, though his dealings had been mainly in the cigar end of the trade. It did not hurt that one of his family enterprises, Tobacco and Allied Stocks, a publicly held company but controlled by the Cullmans, had sizable holdings in many of the big cigarette manufacturers, including Reynolds.
Joe Junior traced the family’s interest in tobacco to his grandfather, Ferdinand Kullman, who in the wake of the social revolutions sweeping Europe in 1848 left his German Jewish ancestors in a Rhine River town and voyaged to America on the same ship with countryman and social reformer Carl Schurz. Kullman had less exalted business to conduct, trying to continue in the family wine trade but having better luck in the new and booming U.S. cigar trade. The powerful smokes the first Kullman manufactured and purveyed, family lore held, did more damage to the Union troops during the Civil War than Rebel bullets and bayonets. A mediocre businessman, he was kept afloat by cash infusions from the old country, but his son Joseph, a more vigorous and personable sort, built speedily and well on his father’s beginnings.
Starting out at fifteen in a New York tobacco merchant’s Water Street offices, he quickly absorbed the dynamics of the business, became an expert cigar roller in the bargain, and proved an accomplished salesman, known as “Piano Joe” for his aptitude on the ivories. Taken into partnership while in his early twenties, the first Joe also had the good fortune to marry into money, brought his relatives into the business, outlived his original partner, and soon commanded a sizable leaf brokerage enterprise. Business travel broadened his horizons, and he was a bold enough operator to buy up almost all the Dutch East Indian leaf offered at auction in Holland in 1890, just before the U.S. erected a steep tariff wall against such imports, and thereby made a killing that added to his hefty holdings.
His firstborn and namesake was given the best of everything, including an education at New York’s elite Collegiate School, where Joe Junior was captain of the football team; a horse of his own, stabled near the Cullmans’ West Seventy-first Street home a block from Central Park, where he exercised the animal; and membership in Yale’s class of 1904. Though physically unprepossessing—he was a bull of a man, with a large head and neck, bushy brows, powerful arms, and short legs—he was popular in college, jocular, and little
inhibited socially by his religious affiliation. Indeed, his father had turned from Judaism to the Ethical Culture Society, and Joe Junior was very much the assimilationist, at ease in almost any social sphere. Indoctrinated early in the tobacco business and as a youngster attending leaf auctions abroad with his father, Joe Junior worked for a while after college in the tobacco fields of Wisconsin and on Cuban plantations before sharing a rolltop desk with his father in the family office. Over time, he would master the horticulture of cigar tobacco to the point where he could tell by the look, feel, taste, and smell of a leaf just where it grew, the nature of the soil that had nourished it, the density of the yield per acre, and what kind of fertilizer had been applied to it. He also moved the family beyond the business of buying and reselling Sumatran leaf for wrapping cigars and Cuban leaf for filling them by pushing his father in 1910 into buying land in the Connecticut River Valley to raise their own “shade” tobacco, grown under a veil of protective cheesecloth, for sale to cigar makers. With that effort blooming, the family began investing in the stock of other, larger tobacco companies, among the nation’s most prosperous businesses.
Throughout the ’Twenties, Joe Junior worked and played hard, enjoying the great loves of his life: a growing family (four sons and a daughter), food (mostly heavy German, which did nothing to improve his stocky shape), liquor (speakeasies were part of his regular cigar-sales territory), dance (including the conga, learned at Havana night spots), sports (he regularly hunted, fished, swam, sailed, played tennis, and most of all rode horseback—the friskier the mount the better he liked it), gambling (for fun), and the ladies (unflagrantly but an open secret, even to his wife, who was not robust). At Cedar Lodge, the Cullmans’ substantial though unpretentious country place in Stamford, Connecticut, Joe was an expansive host in his jodhpurs, loud jackets, and outlandish hats. Among those regularly entertained was fellow tobacco industry leader Al Lyon of Philip Morris, whose own country place was only a few miles away in Greenwich; the pair often rode together.