Authors: Richard Kluger
But could such a small average daily dosage of ETS-derived tar realistically cause that many deaths? If Gio Gori believed that there was a level of exposure to tobacco smoke below which the death risk was negligible, or at least unmeasurable, James Replace did not; indeed, he had, arguably, just measured it, even if in the abstract. Repace believed that each individual had a different genetic code from the next person, varying in body chemistry and delicately balanced immune systems and thus in vulnerability to disease. His own family illustrated this conviction painfully: his father had contracted lung cancer, his mother ovarian cancer, a brother suffered from Hodgkin’s disease, and he himself and two of his four children had asthma. There was, to his way of thinking, a predisposition to disease among many in the population and no telling what the tipping point might be in the way of exposure to the disease-causing agents and mechanisms.
The plausibility of Repace’s projected ETS death toll was enhanced by emerging studies on ETS-derived toxins residing in the bodily fluids of non-smokers as measured by the amount of cotinine, the residue of nicotine after their bodies had metabolized it. Early cotinine studies tracking its presence in nonsmokers’ blood and urine disclosed a quantity equivalent to the yield of one or two low-tar cigarettes a day, Repace suggested. But later studies put the figure considerably lower, closer to the yield from about one-sixth of a cigarette per day, or between .75 and 1 percent of the cotinine found in the bodies of active smokers. Even at that level, though, according to Repace’s construct, the exposure was fatal for some, given the immense population at risk.
Submitted, in 1984 to the journal
Environment International
, the Repace-Lowrey article, “A Quantitative Estimate of Nonsmokers’ Lung Cancer Risk
from Passive Smoking,” withstood an unusually stringent peer-review process and a visit to the journal’s editor, A. Alan Moghisi, by a four-man delegation of Tobacco Institute lawyers and scientists in their hire who tried to label the authors as antismoking fanatics and their work as hopelessly unscientific. Moghisi himself had some reservations but thought enough of the controversial study to run it, and the Tobacco Institute feared it enough to issue a forty-three-page rebuttal pamphlet,
Tobacco Smoke in the Air, a
few months after its early 1985 publication, charging it with too many theoretical or unwarranted assumptions. On the other hand, the
American Review of Respiratory Diseases
stated the following year, “Despite the simplifying assumptions of the risk estimates and the flaws in the epidemiological data from which they are derived, Repace and Lowrey’s figures remain the best current estimates of lung cancer deaths from passive smoking.”
Repace devoted much of his spare time to speaking out on the ETS problem, so the tobacco industry did its best to get him fired or demoted at the EPA. The Tobacco Institute fed information about Repace to a conservative Tennessee congressman (and later governor), Representative Don Sundquist, who was a member of the House subcommittee that controlled the EPA’s purse strings. Sundquist addressed a four-page letter dated March 12, 1987, to the agency, objecting to Repace’s “apparently inappropriate outside activities” that he said took up government time, his appearances as a “well compensated expert witness,” and other antismoking efforts that “would seem to violate EPA guidelines.”
Such a complaint from a well-placed congressman could not be ignored, and for the next year Repace’s life became a hell as agency investigators examined every phase of his ETS studies. “Once you’re caught up in an investigation, whether you’re guilty or not has nothing to do with it,” Repace said in recounting the experience. His usefulness as a government scientist was in limbo while he defended himself, showing that his outside antismoking efforts were indeed separate and apart from his office work, that any fees he received for speaking or testifying on ETS were appropriate and within his rights as a private citizen, and that every article he wrote carried a disclaimer stating that he was not speaking in behalf of the EPA—except for one that appeared in the British journal
Lancet
, whose editors left out the disclaimer. Eventually Repace was totally exonerated, but not before suffering a loss in standing with some of his uneasy superiors.
III
OF
comparable importance to the Repace-Lowrey mathematical modeling studies in the growing effort to gauge the ETS risk—and comparably attacked
for its alleged flaws—was a very different kind of investigation reported early in 1981 by Takeshi Hirayama, chief of epidemiology of the Research Institute at Tokyo’s National Cancer Center. For fourteen years Hirayama and his associates followed 92,000 nonsmoking wives of smoking husbands to learn what their risk was of contracting lung cancer, compared to a similarly sized control group married to nonsmokers. The results of the Japanese spousal study were viewed as particularly significant because they were dose-related: nonsmoking wives married to ex-smokers or current smokers of up to fourteen cigarettes a day showed a 40 percent elevated risk of lung cancer over wives married to nonsmokers; those married to husbands smoking fifteen to nineteen cigarettes a day had a 60 percent higher risk; and those whose husbands smoked a pack or more a day had a 90 percent heightened risk. No such correlation was found for other diseases, and the data did not vary much between urban and rural areas, eliminating industrial pollutants as a confounding causal factor.
Accepted at their face value, Hirayama’s findings amounted to stating that nonsmoking wives of smokers had a risk of getting lung cancer from ETS that was 4 to 9 percent of the risk they would have run if they themselves had been pack-a-day smokers (who had a 1,000 percent higher risk of suffering the disease than nonsmokers). That elevated range of risk amounted to a considerable menace when applied to all Japanese wives, for whom cigarettes were generally taboo, married to smoking husbands, who made up about 75 percent of all married Japanese men.
Hirayama’s report did not go uncontested. Critics ripped into the Japanese spousal study for its alleged methodological and mathematical flaws, including failure to delineate its subjects’ ages adequately (other than that they were all forty or older), to exclude occasional smokers from the nonsmoking category, and to determine how long the wives were present when their husbands smoked, what the degree of ventilation was, and the amount of exposure to airborne charcoal particles from traditional hibachi cooking. These and other objections were featured in a Tobacco Institute pamphlet that savaged the Hirayama study and prominently noted that the head of Hirayama’s own institute said that Japanese husbands were not home enough to expose their wives to much cigarette smoke.
As the 1980s unfolded, other spousal studies proved to be less than fully corroborative of the Hirayama findings, and some contradicted them. Most important among the latter was a report in the June 1981 number of the
Journal of the National Cancer Institute
by the ACS’s biostatistician, Lawrence Garfinkel, who, using older data but a far larger population base than the Japanese investigators, reported “very little, if any,” elevated risk of lung cancer among nonsmoking wives of smoking husbands. In a 1984 follow-up report taking into account the workplace exposure to ETS of nonsmoking
wives—as his 1981 article had not done—Garfinkel calculated that wives married to smokers of up to a pack a day and working wives had a 27 percent higher risk of lung cancer than nonsmoking, nonworking wives of nonsmokers. Yet he found, in a puzzling inconsistency with dose-response norms, that wives whose husbands smoked over a pack a day had only a 10 percent elevated risk. Garfinkel summed up the spousal risk as “small but real.”
Evidence now emerged as well that ETS might affect breathing function and respiratory vulnerability to disease. California investigators James White and Herman Froeb followed some 2,000 nonsmokers chronically exposed at their work sites over a twenty-year period and found that their lung function was reduced by 5.5 percent and their air flow rate by 13.5 percent, or about the equivalent effects in smokers of up to half a pack a day from airways narrowed by hyperplasia. Industry critics picked apart the White-Froeb study, as they did almost every other negative finding, but in this case they were helped by other studies that did not support the California investigation. A consensus, though, formed with regard to the heightened risk of respiratory disease among infants and children in families with smoking parents and households with other smokers. A late ’Sixties study of 10,000 Israeli births had shown a 27 percent higher hospital admission rate for pneumonia and bronchitis among smokers’ children; a British study found similarly elevated rates for the same diseases for the first two years of the infants’ lives but not beyond. A 1978 study of 3,000 Finnish youngsters during their first five years disclosed a 70 percent higher chance of hospitalization for respiratory diseases among the offspring of smoking mothers. And a 1985 investigation of 1,144 American infants reported a 45 percent higher risk of bronchitis among smokers’ children. On the other hand, no consistent relationship was reported between doctor-diagnosed cases of asthma and ETS exposure.
Much uncertainty, then, attached to the unfolding exploration of how great a peril secondhand smoke truly posed. In examining the literature to date,
Consumer Reports
, one of the very few periodicals that regularly and objectively covered smoking’s effects on health, stated in its February 1985 issue that the evidence against ETS was “spare” and “often conflicting” except for the effects on the youngest children and small but measurable deficits in function in the developing lungs of fetuses in smoking mothers. Draconian social measures in the form of public smoking restrictions might stir civil contention before the effects of ETS could be authoritatively determined, the magazine suggested.
The following year, three careful and authoritative appraisals of the ETS peril were issued, two by federal government agencies and one by a blue-ribbon panel of establishment scientists. The first came from the Office of Technology Assessment (OTA), a unit that provided background information for Congress, and dwelled on smoking in the workplace. After noting that ETS
was an irritant to the eyes and mucous membranes, the OTA report said of the studies thus far made, “The case is less clear for the contribution of passive smoking to chronic diseases These studies do not have the methodological strength of studies on direct smoking and cancer… .” It stressed that the classification of nonsmoking wives in spousal studies was a tricky business and found that some of the assumptions in the Repace-Lowrey studies, like basing the risk differential for lung cancer between Seventh-Day Adventists and other nonsmokers solely on ETS, were “inappropriate”. Taken piece by piece, the results of some twenty ETS investigations were “equivocal,” but taken as a whole, they prompted “stronger conclusions,” for even if there were obvious difficulties in measuring actual dosage and exposure, these “do not invalidate the studies … . Despite the uncertainties of the evidence, the data are sufficient to warrant serious concern.”
A similarly cautionary note was struck in a second report late in the year by a select committee of the National Research Council (NRC) for the august National Academy of Sciences, made at the request of the EPA and Office on Smoking and Health. Gauging the health hazards of ETS was difficult, the NRC panel observed, since there was as yet no strictly scientific way to measure human exposure to the pollutant. Even where the evidence appeared strongest—the risk from ETS exposure in infants and children—the picture was far from clear; it was debatable whether respiratory diseases and reduced function could be ascribed to a single type of exposure, such as ETS, or whether small decreases in breathing flow noted in children with smoking parents were temporary or might lead to permanent pulmonary dysfunction. Non-smokers in general were reported by the consensus of studies to absorb from ETS the toxic equivalent of about one-fifth of a cigarette per day, which, assuming a linearity in dose-response relationships, translated into about 1 percent of the excess risk of lung cancer that active smokers faced and about 2 percent of their risk of heart disease.
But when the NRC study was reported in
The New York Times
, science writer Jane Brody did not cite this basis of comparison. In describing the heightened risk found in the consensus of the spousal studies, most of them below the levels observed in the Hirayama investigation, Brody wrote, accurately enough, that “nonsmoking spouses of smokers were 30 percent more likely to contract lung cancer than nonsmokers married to nonsmokers.” True, but non-smokers married to nonsmokers rarely contracted the disease; in fact, nonsmoking spouses married to smokers were found by the scientific consensus to have a heightened risk rate of lung cancer that was 2 to 3 percent of what it would have been if they themselves smoked a pack of cigarettes a day. Put another way, about 6 nonsmoking wives of nonsmoking husbands succumbed annually to lung cancer among every 100,000 in that category; if their husbands smoked, their death ratio from the disease rose to 8 per 100,000, or an
increased risk of 1 in 50,000. The public-health community, including even its most conscientious members in the press corps, was not above using what Peter D. Finch, in a 1992 essay entitled “Creative Statistics,” found to be distorting relative risk data rather than absolute figures. “Health activists often present their arguments in terms of relative risk to evoke fear … that will activate us to adopt a lifestyle which they have deemed to be desirable,” Finch wrote as part of a booklet issued by the conservative Manhattan Institute. At any rate, the NRC study concluded that much more research was needed “if [public-health] policies are to be based on possible adverse health effects and not solely on the discomfort that passive smoking causes among the two-thirds of Americans who don’t smoke.”
The last, and most ambiguous, word on the subject came at the end of the year with publication of the Surgeon General’s report on “The Health Consequences of Involuntary Smoking.” Dr. Everett Koop had by then established himself, much to the amazement of the liberals who had opposed him and the conservative-fundamentalist camp that had championed his nomination, as a humane and powerful voice, calling for society-wide crusades against such scourges as AIDS (after having once referred to homosexuals as “anti-family”) and cigarettes, whose makers he routinely termed purveyors of “death, disease, and deception.” A barrel-chested thunderer with his grizzled, square-cut beard and steel-framed glasses, Surgeon General Koop was not easy to ignore when he wrote in the preface to his 1986 report: