Authors: Richard Kluger
To his dismay, Oscar Auerbach was not greatly embraced when he reported his findings at professional gatherings, “where I had to wade through clouds of smoke—it seemed as if 90 percent of doctors were smoking then.” After a typical presentation, he would be assaulted by thoracic surgeons asking him, “How come my pathologists don’t find what you do?” The answer, though he was too polite to say it, was probably that the smoke kept getting in their eyes.
IV
FINAL
results of three major epidemiological studies on the relationship between smoking and health now showed not only a statistically significant linkage but, in the case of heavy users of cigarettes, a ratio of peril so large as to indict the habit as causally connected, even if nobody could yet say for certain just how the damage was inflicted.
After fifty-three months, the study of British doctors by investigators Doll and Hill recorded 81 lung cancer deaths among the 1,714 subjects who had succumbed, with a mortality rate for that disease among heavy smokers of twenty times that for nonsmokers. For those who had quit the habit, the rates declined in proportion to the length of time they had been off tobacco. The 1956 Doll-Hill findings were dwarfed the following year by the final report from American Cancer Society investigators Hammond and Horn, who, after close analysis of the 11,870 deaths among nearly 200,000 subjects during a forty-four-month study, found that cigarette smokers had a 68 percent higher
mortality rate than nonsmokers, while cigar smokers had an elevated rate of 22 percent and pipe smokers a 12 percent higher rate. Among the most persuasive ACS data were the risk figures for quitters. For those who had smoked up to a pack of cigarettes daily, the death rate was 104 percent higher than for non-smokers within the first year after they had quit, but dropped to only 30 percent higher between one and ten years off the habit and to just 8 percent higher after ten years of not smoking. For those who had smoked more than a pack, the contrast was equally dramatic, but even after having quit for a decade, heavy users had a mortality rate half again as high as those who had never smoked. Of the 448 lung cancer deaths recorded during the ACS study (80 percent of them confirmed by microscope diagnosis), only 15 victims had never smoked. Regular cigarette users had a lung cancer death rate eleven times higher than nonsmokers, while those who smoked two packs a day were struck down by the disease sixty-four times as frequently as those who did not smoke. So conscientious had the ACS’s volunteer interviewers been that almost 99 percent of those who had been enrolled in the study were accounted for at the end, prompting chief investigator Cuyler Hammond’s wife, Marian, to handwrite a note of gratitude to all 22,000 interviewers.
A still larger population, using federal government data, was investigated for the toll smoking took upon it by biostatistician Harold F. Dorn of the National Cancer Institute (NCI). Dorn had hit upon a relatively easy way to accumulate a massive data base through readily accessible government records: he enlisted a quarter of a million veterans of American military service who had signed up for attractively priced life insurance policies that the U.S. had offered its soldiers between 1917 and 1940 and, at the outset of the study, asked them to report their smoking habits (or absence of same). During the thirty-month tracking period, Dorn learned of the deaths of his subjects through the insurance benefit claims filed on their policies with the Veterans Administration and obtained details on the causes of death from forms that were required to be filed with the U.S. Public Health Service before the claims could be honored. Dorn’s findings, reported in 1957–58, tallied closely with the Hammond-Horn and Doll-Hill results. Smokers suffered a 58 percent higher death rate than did those who had never smoked, while heavy smokers had an excess rate twice that high, and regular smokers had a chance of contracting lung cancer 9.35 times higher than nonsmokers in the Dorn study. The earlier you started smoking, Dorn also found, the poorer your health prognosis.
On many fronts, and in many laboratories, the case against smoking grew in the late ’Fifties. Half a dozen labs, for example, reported confirming Ernst Wynder’s success in producing cancerous tumors on the backs of mice painted with tobacco tar condensate, though most were less numerous than in Wynder’s study, while other labs reported that they had not been able to replicate Wynder’s results after following his method. One confirming laboratory, the
Arthur D. Little program for Liggett & Myers, had consulted closely with Wynder and succeeded in 1955 in its own skin-painting effort, albeit at a lower tumor level than Wynder’s, but did not report the results publicly until 1962—and did so then in an offhand manner. A timely report of its findings by a research institution as well regarded as ADL might have contributed significantly to the ongoing health debate.
Wynder himself continued to turn up important new evidence. By way of assessing the possible role of air pollution in causing lung cancer, the Sloan-Kettering epidemiologist arranged a four-year study of some 8,700 patients admitted to hospitals run by the Seventh-Day Adventists in highly polluted Southern California. About 8 percent of the subject patients were adherents of the Adventist faith, which forbids smoking. If air pollution and not smoking were a major causative factor in lung cancer, the nonsmoking Adventist patients could have been expected to contract the disease in about the same proportion as the other patients since they all had been breathing the same fouled air. A proportionate death toll from lung cancer during the course of Wynder’s study would have taken the lives of 10 or 11 Adventists; in fact, only one died from the disease, and that victim had been a pack-a-day smoker for twenty-five years before joining the Adventist church thirteen years earlier. For other body sites, the Adventists fell victim to cancer in comparable proportion to the rest of the subject group, according to Wynder’s 1958 report.
Gifted at attracting publicity for his findings, Wynder further demeaned the air pollution argument a few years later by testing the level of airborne toxic substances in New York City’s thronged Herald Square and finding it to be the equivalent of the intake from seven cigarettes, not the two packs that skeptics of the smoking and health relationship had suggested. Nasal breathing of the city air, moreover, Wynder noted, allowed the body to filter out a sizable portion of the contaminants, whereas inhaled cigarette smoke bypassed the nose’s natural filtration system.
By the late ’Fifties, Wynder had taken on an invaluable ally in his smoking studies. Where Wynder had fled from Nazi anti-Semitism, fellow German Dietrich Hoffmann had fought for Hiker’s Wehrmacht before beginning his career as a gifted bench chemist. Despite their deep-seated differences, there was mutual respect between the flamboyant Wynder and the methodical Hoffmann, and the pair would collectively publish more papers of a substantive nature than any other investigators of the smoking peril. With Hoffmann doing most of the laboratory work on the fractionation of smoke, breaking down the immensely complex substance into its component parts, the team had by 1959 isolated no fewer than seventeen carcinogenic compounds in cigarettes. But these were present only in trace quantities, unlikely in a given individual to be adequate to trigger carcinogenesis, though nobody knew if there might indeed be a minimum threshold dosage for any cancer-inducing substance. The
Wynder-Hoffmann chemical analysis fed speculation that the true carcinogen in cigarette smoke still eluded investigators or, alternatively, that the various carcinogens might behave synergistically
(i.e.
, combining their potency), or might become far more carcinogenic in the presence of such suspected promoters of the disease as phenol or carbolic acid, a crystalline distillate found in tobacco smoke.
V
WHY
people smoke became a legitimate topic of social research as the evidence piled up against the prudence of the custom. Psychologists were naturally reluctant to label as abnormal behavior or a personality disorder a practice indulged in by nearly half of the adult population. The most that could plausibly be advanced by way of a rational theory to explain the popularity of smoking was that the stresses of modern life induced a form of self-destructive behavior that seemed to stabilize the emotions and prevent many forms of dysfunction.
A mock-Freudian analysis of the habit, appearing in the April 1955 issue of the
Harvard Medical Alumni Bulletin
under the title “Smoke Gets In Your Id” by Boston University psychiatrist Peter Knapp, offered a tongue-in-cheek analysis of the mental state of smokers. Four percent of them, Knapp suggested, confecting the numbers, were social smokers, most of them women, from whom “[l]arge puffs of smoke are emitted rapidly, sometimes with an audible squeal;” 26 percent were “neurotic or reactive” smokers, responding to stresses of a usually minimal nature, like “the strain of waiting for a cocktail to chill in a shaker;” and 50 percent were addictive, for whom the main motive to light up was to relieve “the stress of finishing a prior cigarette.” Smoking was a highly useful behavior, he noted, because it allowed the practitioner to express a variety of drives and resolve conflicts as part of “an individual’s web of ritualized activity”—solving, for example, the pressing problem of what to do with one’s hands. The habit was rooted in “orality,” Knapp hypothesized, meaning that it was no mere case of emotional deprivation from insufficient breast-feeding but a kind of respiratory eroticism linked to “the biphasic nature of the breathing act,”
i.e.
, the luxuriating nature of inhalation followed closely by the satisfying relief of exhalation.
Among the first to posit that smoking might provoke rather than allay stress was researcher Charles MacArthur, who followed 250 Harvard students for many years after enrolling them in his study during their sophomore year. After finding that anxiety and heavy smoking “are not as clearly related in our data as we had expected,” MacArthur added: “It seems likely that heavy smoking is best used as a tension reducer after it is already established high in one’s
habit hierarchy”—in other words, as a way to counter the stressful symptoms of withdrawal.
The ambivalence of heavy smokers toward their habit was well and variously reflected in a 1957 motivational research study of 750 subjects, each of whom submitted to two-hour interviews with the Market Planning Corporation at the behest of Liggett & Myers. Many of the subjects believed that smoking did them no good, caused coughing and shortness of breath, made them feel awful upon awakening, stained their teeth, and burned their clothing; indeed, “a majority of those we interviewed said they wished they could stop smoking” (underscoring in original text) and earnestly hoped that their children would not begin. Yet most subjects also dreaded the supposed adverse effects of quitting, fearing that they “would not be able to cope with life” without their cigarettes. Among the cited gratifications of the practice were “pleasure at will,” the draining of tensions and aggressions, the company it provided in times of loneliness, relief from boredom and social pressures, and a sense of adulthood, independence, vigor, and the thrill of risk-taking, so that the prospect of living dangerously was thus an inducement, not a deterrent, to the smoking habit.
A leading epidemiologist at Johns Hopkins—Abraham Lilienfeld—studied some 4,500 adults in the Buffalo, New York, region in an effort to isolate emotional and behavioral characteristics that might explain why so high a proportion of smokers contracted lung cancer. Reporting in 1959 in the
Journal of the National Cancer Institute
, Lilienfeld found that smokers tended to marry, change jobs, move their home, play sports, and enter the hospital more than nonsmokers and were more “neurotic,” though he did not define that condition, but these differences were too marginal, in his judgment, to matter much. At about the same time, a leading British psychologist, Hans Jurgen Eysenck, began publishing in what would amount to a three-decade campaign of denial that cigarette smoking was a prime cause of disease. Eysenck argued instead that personality configuration held the key to disease formation; repressed emotions and hopelessness, for example, promoted cancer just as hostility and aggression did heart disease—tendencies that he believed, not surprisingly, could be reversed by psychotherapy. In a 1960 study of 2,300 males, he found that smokers were decidedly more extroverted than nonsmokers—the act itself, with its great billowy clouds, was essentially a form of public display and extremely sensual in nature—but were neither less rigid nor more neurotic than those who refrained from the habit.
Perhaps the most telling behavioral study of smoking habits in this period was a survey directed by ACS psychologist Daniel Horn, who in association with Cuyler Hammond had run the cancer society’s big prospective study. Horn, at least Hammond’s mental equal but lacking his looks, social graces, organizing ability, and drive, had split with him after coming to feel that he
had been relegated to a decidedly secondary position behind the epidemiologist. Now he applied his talents to a study of 22,000 high school students in Portland, Oregon, in a quest for clues as to why and how the smoking habit took hold. In findings published in 1958, Horn reported that three out of four smokers had begun the practice before they finished secondary school—in the most suggestible years of adolescence. One-third of all the Portland high school students smoked—38 percent of the boys and 29 percent of the girls. Said one youngster in the Oregon contingent, cutting to the core of smoking’s perilous attraction: “Nobody young worries much about some disease they might get forty or fifty years from now.” Family role models were deemed the most influential factor in starting them on the custom: 40 percent of the students both of whose parents smoked became smokers themselves, as did 36 percent of those with an older sibling who smoked and 33 percent with one parent who did. By contrast, only 8 percent of the surveyed high schoolers smoked against the wishes of both parents. Horn estimated that for about 25 percent of smoking high schoolers, the practice was “a compensatory form of behavior” that reflected “failure to achieve peer-group status or satisfactions” whether academically, socially, or extracurricularly. The third main reason youngsters smoked, said Horn, was rebellion against authority or discipline, a motive particularly common at Portland’s five Catholic high schools.