'Evidence and Action'
By Carla Garnett
On the Front Page...
Back in 1926, prominent NIH scientist Dr. Wilhelm Hueper wrote: "We may eliminate the inhalation of cigarette smoke as a causative factor for this increase [in lung cancer incidence and deaths]...cigarette smoke may only have a contributory influence, if at all." By the 1950's, when the issue began coming to a public head, Hueper was staunchly defending his position, even in the face of mounting scientific evidence to the contrary. A pathologist who in 1948 was appointed chief of environmental cancer research at NCI, Hueper also routinely discouraged -- in word and deed -- launching public health efforts to warn the American people about the potential dangers of smoking.
With today's hindsight, Hueper's active skepticism seems suspect. In the current climate, his motives would be questioned, his bank books examined. Most likely, the public would speculate, he was being paid off by the powerful tobacco industry. But before mentally pillorying Hueper (and a few of his esteemed colleagues in the Public Health Service who agreed with him), take a look at the beliefs that drove his actions. Reviewed in the context of a short history lesson, Hueper's position and actions may seem more honorable -- or at least more understandable, according to medical history scholar Dr. Mark Parascandola, one of two speakers at the Mar. 26 mini symposium, "Evidence and Action: How Epidemiologists Make Decisions About Science and the Public's Health."
Cosponsored by NCI's Division of Cancer Prevention and the DeWitt Stetten, Jr., Museum of Medical Research, the lectures focused on how epidemiologists progress from observation to inferences about disease causation to public health recommendations.
"There were substantial challenges to enacting a strong public health response to tobacco -- but they did not come solely from the tobacco industry," said Parascandola, who is the 1997-98 Stetten memorial fellow in the history of 20th century biomedical sciences and technology. "Differences in values and goals among members of the scientific and public health community led individuals to very different responses, even though they were working from similar evidence. Understanding their motivations can help us better understand and identify assumptions and concerns -- often hidden or unacknowledged -- that drive current public health debates."
Battle Lines Drawn
It was at the beginning of one such contentious and widely publicized public health debate, one which will rage probably well into the 21st century, that Hueper found himself coincidentally aligned with -- and not financially rewarded by -- the tobacco industry, explained Parascandola in his talk, "Cigarettes and the NIH in the 1950's." Historical documents reveal that Hueper believed a public health intervention against smoking would unnecessarily draw the nation's attention and resources away from what were, to his way of thinking, far more pernicious cancer-causing agents -- asbestos, road tar and other workplace carcinogens that thousands in the U.S. workforce were being exposed to involuntarily.
By 1953, when the powerful Tobacco Industry Research Committee was formed, a triangular debate about lung cancer and public health was already well documented: Hueper was using epidemiologic evidence to condemn occupational exposure to asbestos and environmental toxins as the more important causes of increased lung cancer morbidity and mortality; the asbestos industry was using similar population studies both to deflect criticism and to implicate smoking and the tobacco industry; and the tobacco industry was citing Hueper's testimony and journal articles to refute the asbestos industry's claims about the role of cigarettes in lung cancer.
"Hueper's strong opposition to one public health problem blinded him to another," said Parascandola, explaining further that Hueper's stridency and conflict of interest may be understandable, given that era -- before workplace hazards were reported or regulated fully, and before creation of the Environmental Protection Agency or the Occupational Safety and Health Administration.
A colleague and friend of Hueper's, Dr. Harold Stewart joined with his fellow NCI pathologist in early and vocal skepticism about the link between smoking and lung cancer. Stewart's reasons were not nearly as cut and dried as Hueper's, said Parascandola. A pipesmoker himself and an outspoken champion of basic research, Stewart -- who was instrumental in getting Hueper hired at NIH -- publicly disagreed that it should be within a federal scientist's purview to launch public health campaigns and offer medical recommendations to the citizenry at large.
"Stewart was opposed to making any sort of official statement to the public about possible dangers of smoking," noted Parascandola. "And he was opposed to the Surgeon General directing research from above. Stewart was particularly incensed by the moralistic tone of administrators and public health officials."
More Than Science
Hueper's and Stewart's stories were just two among several that Parascandola used to make his point: More than in other sciences, whenever epidemiologic evidence is considered in a public health debate, often a number of ancillary factors -- some less scientific than others -- combine to influence the argument.
Questions that factored into both NIH's and the Public Health Service's response to cigarette smoking in the 1950's include: Is a government scientist's role to tell the public how to behave (i.e., "don't smoke"), if their behavior could have an impact on their health? Should federal researchers speak out against a public health threat if speaking out could harm industry, which in turn could undermine public and congressional support of scientific research? Why target science in any particular direction, when, according to several prominent NIH officials at the time -- including Stewart and NIH director Dr. James Shannon -- true scientific breakthroughs occur chiefly as a result of the "unguided" efforts of researchers? Laboratory science, clinical research and epidemiologic analysis are all clearly acknowledged as important weapons in public health, but what is the most effective balance among them?
Two years after the Surgeon General produced the first Smoking and Health Report, when ethical, moral, and political considerations had caught up with the epidemiologic evidence in the government's tobacco and lung cancer debate, the National Cancer Institute issued a definitive public health statement warning against smoking cigarettes. The year was 1966 -- 40 years after Hueper's erroneous assertion about inhaling tobacco.
"Epidemiology presents unique challenges compared with other biomedical sciences," concluded Parascandola. "Epidemiologic evidence seems particularly contentious because, more so than any other science, its claims have direct impact on public health. The fact that such extra-scientific values play a role in the move from evidence to action is not something to be ashamed of. After all, we are talking about human life, and decisions that affect human life should consider values other than scientific truth. But we should also be aware and explicit about such influences, as the real danger comes when interested individuals attempt to disguise themselves as objective observers."
Epidemiology Alive and Well
Dr. Douglas Weed, chief of the Preventive Oncology Branch in NCI's Division of Cancer Prevention, concluded the seminar with his lecture, "End of the Era of Weak Associations: A Historical Study of Epidemiologic Discovery."
A physician and epidemiologist who established the NCI branch he now heads, Weed said that during the last decade, important questions have arisen about the perceived value and reliability of epidemiologic research. Scientists, the media and the public have wondered aloud if the field has reached its limit. As Weed noted, "These are very troubling and extraordinary claims for any science. These perceptions threaten the link between epidemiologic research and the public's health."
Weed then showed historical data that he and two NCI cancer prevention fellows, Drs. Noreen Aziz and Pam Marcus, collected on the discovery of breast cancer risk factors from the mid-1800's to the present. The NCI investigators used two different search methodologies to determine the date of publication of the first study reporting on now-established breast cancer risk factors. They subsequently graphed the magnitude of the risk associated with each factor against its date of discovery. As Weed explained, "if epidemiology has reached its limit, then the strong risk factors should have been discovered in the past, with only the weak risk factors discovered recently."
In contrast, Weed's data from both methodologies show that in the past decade, only relative strong breast cancer risk factors have been discovered. He concluded that the data refute the perception that epidemiology has reached its limit.
"We are truly in a new era of epidemiology, driven by technology, driven by a new capacity to measure genetic factors and molecular factors," he said, suggesting that epidemiology's future is alive and well.
Up to Top