by Allan Brandt
But causal relationships remained unclear. “School records indicate,” explained the American Journal of Public Health in 1923, “that when a pupil begins to use tobacco his intellectual work is apt to decline. While this is not always true the relationship between the use of tobacco and low scholarship is so frequent and well marked as to warrant the belief that we have cause and effect exemplified.”31 Another researcher complained that “our methods of getting the cumulative results are less exact, owing to the objection to subjecting sufficient numbers of human beings to the use of tobacco for sufficient periods of time under proper control.”32
In the end, even the most carefully collected data inevitably demonstrated the researcher’s a priori assumptions: smoking was a peril. Such findings, Good Housekeeping suggested, should shift the debate about smoking and youth away from questions of “personal liberty.” “Doesn’t a college enter into a compact with society to do its utmost to make real men and real women of the boys and girls who are sent to it?” the magazine huffed. “Why permit any one accepted to fall into habits that almost inevitably lower scholarship and either result in dismissal or an inferior quality of finished product?”33 Unlike many journals reporting similar findings, Good Housekeeping took its social responsibilities seriously, refusing to accept tobacco ads.34
As they began to see the weakness of mere moral anecdotes, researchers increasingly applied statistical techniques. By the late 1920s, their investigations had become more sophisticated.35 A particularly detailed and impressive assessment of a group of smokers and nonsmokers was done with University of Minnesota students in the late 1920s by Dr. H. S. Diehl. Determined to conduct a truly scientific study, free of the bias that typically characterized the tobacco debate, Diehl found no significant differences between the two groups, dispelling long-standing concerns about the impact of smoking on development. Nonetheless, he noted that students at college had only recently taken up smoking: “[In] persons of 19 years of age the habit has not been practiced sufficiently long for degenerative effects, if there are any, to have proceeded very far.”36 This was a prophetic insight that pointed out a critical obstacle in demonstrating the harms of tobacco: the long latency period between the start of smoking and the onset of disease.
Early critics of the cigarette focused their attention on the impact on boys and adolescent men, but as women took up smoking, medical investigation quickly followed. The opposition to women smoking drew strong medical allies. Physicians prominently joined the antitobacco campaigns, and many pointed to particular harms associated with smoking for women.
Their warnings typically emphasized the vulnerabilities of the “weaker sex.” Dr. Samuel Lambert, a prominent New York physician, explained:Intemperate smoking causes nervousness and may lead to something worse. . . . Women who use cigarettes cannot be temperate. At best it is a horrible weed and should be let alone. It fouls the breath and makes woman unwomanly.
“Women smoke nervously,” agreed Dr. Samuel A. Brown, dean of Bellevue Hospital. “They cannot smoke moderately.” The cigarette became a symbol for the loss of feminine control as well as for the changing roles of women in the early twentieth century.37
Critics often revealed powerful gender expectations as they elucidated the particular degenerative effects of cigarettes on women. As Charles B. Towns explained in 1916, “It degrades everything in a woman that is worth while,” making “the lovely, devoted, clean wife and mother . . . negligent of all her womanly duties and responsibilities.”38 Early opposition to women smoking was connected to broader concerns about eugenics, degeneration, and motherhood.39 The cigarette became another example of the perceived failure of white, middle-class women to act responsibly as “mothers of the race.”40 In this respect, the rise of women smoking marked a deeper erosion of the traditional separate spheres of gender. One physician considered it fortunate that women who smoked failed to reproduce, explaining, “No more pitiful sight on earth could possibly be imagined than the spectacle of some mother who is a cigarette smoker bringing into the world a poor, pitiful physically and mentally defective child.” 41
The prominent Michigan physician and surgeon Bertha Van Hoosen argued that smoking had particularly dire consequence for prospective mothers. “Motherhood and tobacco,” she wrote, “are as antagonistic as water and fire. Motherhood is the greatest thing in life; in fact the essence of all life and the perpetuation of life. Motherhood is too complex to tamper with tobacco or any other drug-forming habit.”42 Some medical critics noted that nicotine could lead to spontaneous abortions. Others claimed that nicotine was present in amniotic fluid and breast milk.43 Already concerned about declining birth rates among white, middle-class women, physicians now suggested that smoking “may have a deleterious effect upon female fecundity.”44
By the 1930s, such eugenic hyperbole gave way to studies of the impact of smoking on fertility and lactation. Systematic assessment of smoking’s effect on breast milk began by the late 1920s. One group of researchers concluded that nicotine suppressed secretion of breast milk in several animals, but that “they had never observed any diminution in the secretion of milk, or any effect on the child, that could be attributed to the smoking of cigarettes by the [human] mother.” Nonetheless, hospitals began to prohibit smoking immediately after labor.45 Another study, conducted in the early 1940s, concluded that although nicotine could be found in both the breast milk and urine of nursing mothers, milk production was affected little. The researchers hypothesized that the mother’s tolerance of nicotine might moderate its effect on infants and lactation.46
The relation of nicotine to lactation again raised questions of causality. Did anxiety cause insufficient production of milk (anxious women tended to smoke more), or was nicotine the culprit? “It is not my belief that the effect of nicotine is the sole or even the chief factor involved in diminished lactation,” explained one doctor. “Usually it is the nervous, excitable woman who, whether a smoker or abstained, has a deficient milk output. . . . There is insufficient evidence to conclude that the one is cause, the other effect. . . . One is sorely tempted to conclude that excessive smoking does influence milk production adversely,” this investigator concluded. But unambiguous data was hard to come by. Some individuals would take a few shallow puffs of a cigarette and discard it, or not inhale at all. Others inhaled deeply all the way to the end. Further, the different brands’ nicotine content varied considerably. “These and other factors,” explained this physician, “make it extremely difficult to formulate safe and sane standards.”47
Like much smoking research, these studies often reached no definitive conclusion. Given the repeated inability to prove that cigarettes constituted a clear danger to mother and child, clinical recommendations typically reverted to the default position: mothers who smoked should practice moderation. But prospective mothers, more of whom smoked, sought more reassurance than this. Hygeia, the AMA’s magazine for the general public, concluded in 1934 that “smoking by mothers is in all probability, not an important factor” in infant mortality.48 Many physicians preferred—given the uncertainty of the data—to be risk averse in their recommendations to women patients. “Until we can prove that excessive cigarette smoking is not harmful in pregnancy” argued one doctor, “we should caution against it.”49 Nonetheless, many smokers dismissed such warnings.
Not every piece of commentary was loaded with assumptions about women, their social roles, and biological dispositions. Dr. S. Josephine Baker, a public health leader in New York, found no reason to assume that cigarettes posed a special risk for women. She considered smoking “more an individual than a race or sex problem.”50 “I have been unable to trace any valid reason for the prevailing impression that the health of women is more seriously affected by smoking than is the case with men,” she noted in the Ladies’ Home Journal.51 Harvey Wiley, who directed the earliest federal investigations of the harms of smoking, offered no disagreement. “Women have just as much right to smoke cigarettes as men,” he
wrote in 1928. But he added, with considerable prescience, “They are likely to suffer the same penalty as men. . . . I am inclined to believe that cancer . . . will increase among women exactly in proportion to the number that acquire the smoking habit.”52 Wiley anticipated a conclusion that would take nearly a quarter century to categorically demonstrate.
By the 1920s, moral claims against the cigarette began to diminish in the face of the product’s popularity. As the percentage of regular smokers grew, it became increasingly clear that not all of them would suffer the cigarette’s purported harms. Too many smokers used tobacco without any apparent consequences to sustain the reformers’ claims of incipient moral and physical decay. As the New Republic wryly noted, “Moderate cigarette smoking can scarcely be considered disastrous, as many octogenarians or nonagenarians will testify.”53
The very popularity of the cigarette typically was cited as medical reassurance. How could the cigarette be dangerous if so many millions of Americans used it regularly without any apparent consequences? “Any substance so widely and commonly used as the cigarette cannot be as dangerous and deleterious as the propaganda of the more fanatical ‘no-tobacco’ advocates might lead one to infer,” argued Emil Bogen, a tobacco researcher.54
Conventional medical wisdom settled on the idea of smoking in “moderation.” Although many agreed that excessive smoking could be harmful, moderate smoking was now deemed medically acceptable. By 1929, Hygeia declared that the “general opinion of those who have studied the subject is that a person in sound health may bear what are for him moderate doses without injury.”55 But the actual distinction between “moderate” and “excessive” smoking was rarely defined. This was, no doubt, partly the result of problems of self-reporting, and of the wide disparities in precisely how cigarettes were consumed.
Physicians increasingly viewed cigarette smoking as a behavior “tolerated” by most individuals and embraced by the public. It remained dangerous only for some people. As one physician summed it up in 1920:There are some individuals who cannot use tobacco, there are some who should not use it, there are some who use it to excess and who suffer in consequence. There is on the other hand a large army of moderate tobacco users, who indulge for years without appreciably bad physical effect and good mental effect.56
This approach protected the often contested sphere of clinical authority; it was a matter of professional discretion to identify individuals who should reduce or eliminate smoking. Individual variation became the theater of clinical judgment: some smokers seemed completely unaffected by their habit; others appeared particularly sensitive to the complex constituents of cigarette smoke. It was precisely the exercise of such discernment that distinguished physicians and made their recommendations authoritative and important. The historic aphorism, “Ask your doctor,” was predicated on these individual judgments.57
According to a number of accounts, physicians—having prominently joined the ranks of inveterate smokers—lost interest in the connection between smoking and disease after 1930. But in reality, clinical medicine claimed the issue as a matter for individual assessment. During this era, there was a strong tendency to avoid causal hypotheses in matters so clearly complex. There was—and would remain—a powerful notion that risk is variable and, thus, most appropriately evaluated and monitored at the individual, clinical level. As cigarette smoking became increasingly popular, medicine offered no new insight into how best to evaluate such variability other than after the fact. If and when an individual developed symptoms, a physician might appropriately advise restricting or eliminating tobacco. This approach kept cigarette use firmly outside the sphere of public health.
Through the first half of the twentieth century, it proved impossible to categorically substantiate the claims of the harmfulness of smoking. “It is only too true that in matters affecting human behavior, as well as in many studies that are carefully made and aim to be truly scientific, there is too great a tendency to assume that because two factors exist side by side one is necessarily either the cause or the effect of the other,” explained the unusually clear-headed S. Josephine Baker. “It may truly be that smoking lowers academic grading but it may be equally true that lowered mental status is the factor that leads to habitual smoking.”58
Many observers noted the difficulty of “impartiality.” “A large amount of partly scientific or pseudo-scientific work has been done,” commented the noted psychologist William H. Burnham of Clark University, “and not even much of the painstaking experimentation of scientific men has sufficiently considered the mental factors involved.” In a careful assessment of the available studies, Burnham concluded that “as regards causal relations the evidence justifies no sweeping assertions.” He concluded nonetheless that “moral, social, and economic considerations, perhaps, quite as much as hygiene, should determine the desirability or undesirability of the use of tobacco. . . . In the words of the older moralists, how far is it desirable to become the slave of a habit?”59 No behavior so deeply entwined with social and cultural mores could ever be evaluated exclusively on a scientific metric.
As the cigarette triumphed over its moral opponents, questions about its health effects would nonetheless persist. After 1930, researchers investigating the health effects of smoking took care to isolate their claims from moral concerns. The status and legitimacy of their work now would depend on its reproducibility and its independence from Victorian prescriptions for health and good living. This did not mean that scientific investigation shook off all vestiges of moral and cultural assumptions. “Objective science” could offer powerful prescriptions of its own—but they would now be sustained by a new, historically specific logic premised on new forms of scientific investigation and argumentation.60 Any evaluation of the harms of smoking—to be persuasive and authoritative—would henceforth require an approach dissociated from traditional moral rhetoric. But this was no simple process.
During these years, theories of carcinogenesis tended to focus on hereditary vulnerabilities. Chronic diseases, in general, were attributed to multiple causes, from genetic predispositions to environmental and behavioral exposures. This underscored the “clinical” approach to smoking. Perhaps, some argued, individuals with a personal or familial history of cancer should avoid cigarettes. Diseases of rising frequency like cancer, heart disease, and stroke were also typically labeled degenerative diseases of the aging organism, often revealed by the rise of life expectancy.
Even in the early twentieth century, physicians catalogued a wide range of chronic diseases that they associated with smoking. By the 1930s and 1940s, clinical anecdote carried considerable authority with physicians, who carefully recorded their observations of the effects of tobacco upon their patients.61 Many investigations focused on the cigarette’s impact on the heart and circulation. 62 “Tobacco heart,” a well-known syndrome, included arrhythmias, angina, and sometimes cardiac arrest.63 Physicians also commonly attributed oral cancers to smoking, especially among cigar and pipe smokers.64
Additionally, by the early 1940s, investigation into the physiologic effects of smoking had grown in sophistication. Doctors at the Mayo Clinic, for example, conducted a detailed series of studies concerning the impact of cigarettes on circulation under controlled experimental conditions. New diagnostic tests were applied to evaluate blood pressure, basal metabolism, and electrocardiographic changes. New standards of experimental design required investigators to reduce confounding variables and bias. Using these new methods, researchers found evidence that smoking constricts the blood vessels—an effect they attributed to nicotine. They also concluded that smoking increased basal metabolism, as well as heart rate and blood pressure. And yet the clinical significance of these findings remained unclear. The researchers suggested that patients with coronary artery disease or high blood pressure should abstain, but “when the heart is healthy, no harm is likely to result from smoking.”65 Smoking might exacerbate a “pre-existing” condition or “weakness,” but it was not seen a
s causing disease.
By the 1940s, researchers were using animal models to assess tobacco’s effects. Experimenting with rats, pharmacologists concluded that repeated dosing with nicotine led to tolerance.66 Other researchers demonstrated that continued nicotine ingestion by rats led to a disruption in the “estrus” cycle and aberrations in growth and development.67 Scientists experimented with new techniques to expose rats to cigarette smoke. In a 1940 study of pregnant rats, the authors found that exposure to smoke lowered birth weight and otherwise hindered growth and development; nonetheless, they noted, “individual variation is much in evidence.”68 While some rats apparently suffered harm, others “receiving the same treatment as the rest, stood the exposure to tobacco products with no apparent detriment.” They easily made the comparison with humans: “Likewise there are women who smoke continually and rather heavily and yet they and their offspring remain in what appears to be perfect health.”69
An Argentinean scientist, A. H. Roffo, developed techniques in the 1930s for distilling the residues in burning tobacco. Employing the relatively new technology of chemical spectroscopy to identify the constituents in these tobacco “tars,” Roffo discovered the presence of polycyclic aromatic hydrocarbons, well-known carcinogens. Applying these distillates to rabbits, Roffo, a vigorous critic of cigarette smoking, was able to produce tumors, confirming his hypothesis that the tars were carcinogenic. Nonetheless, observers would continue to question the applicability of these findings to humans.70
The jump from animal models to humans raised a host of scientific problems. In experiments on rats, for example, it proved especially difficult to develop a means for estimating equivalent doses for exposure to cigarette smoke and nicotine. “Under no circumstances should it be assumed that chronic effects of nicotine in the human subject are similar to acute effects of nicotine in animals,” lectured one researcher. “To date [1948] there have been no animal experiments which fully simulate smoking by the human being. Hence, no analogy can be drawn from the experimental data at hand.”71