The Evolution of Public Health Research: Sir Austin Bradford Hill and the Cigarette-Rolling Machine

Posted by Jaydeb Mukherjee on Apr 7, 2016 11:00:00 AM


In 1923, a general practitioner in rural Virginia received a phone call from a friend, a surgeon at a hospital in Richmond. A middle-aged man with an uncommon form of cancer had been admitted to the hospital - would he like to come and see the patient, the friend asked. It was a rare opportunity to examine and diagnose an unusual illness first-hand. The country doctor was happy to visit the hospital and further his education, especially since he'd doubted seeing more than one or two similar patients over the rest of his career.

Lung_Cancer_CellsHowever, to the doctor’s dismay, within a few years he unexpectedly saw dozens of cases of this “rare” cancer, and so did the hospital in Richmond and most other medical practices across the country. By the 1950s, in a span of only 20 years, a rare disease known as adenocarcinoma of the lung, along with other types of lung cancer, had gone from “rare” to the leading cause of death in men in the US. Even worse, the tidal wave of lung cancer cases was accompanied by a similar flood of patients with cardiovascular disease.

The middle of the 20th century marks the shift from infectious disease to chronic illness as the leading cause of morbidity and mortality in the US, and likewise marks a change in the way epidemiological and public health research is conducted. The introduction of penicillin in 1942 dramatically reduced death from infection and also displaced the role of bacteriology in public health research. Diseases that could not be addressed with a vaccine or antibiotic, and had an obscure cause and long latency period, necessitated a new approach. Lung cancer led the way.

Mass Production and Mass Disbelief

The statistical methods and epidemiological tools we use today are to some extent a result of the invention of the cigarette-rolling machine. Up until the late 1800s, cigars and cigarettes were hand-made luxury items and consumption was relatively low. However, in 1880, the Allen and Ginter Company of Virginia offered a prize of $75,000 for a machine to replace their employees, who clocked in at an unprofitable 4 cigarettes a minute. The prize was won by James Albert Bonsack, whose invention produced 200 cigarettes per minute, which obviously resulted in a great cost reduction. By the start of World War I, about half of all males in Europe and the US were smokers, and women were quickly acquiring the habit as well.

Cigarette_Rolling_MachineThe link between cigarette smoking and lung cancer was first identified by German doctors around 1930. However, the relationship continued to be debated in the US, despite the overwhelming statistical associations, for a number of reasons. For instance, researchers tried but failed to induce lung cancer in laboratory animals with tobacco smoke. Cigarette ads continued to claim that smoking had health benefits and depicted physicians smoking cigarettes. Although the public dread of cancer resulted in the creation of the National Cancer Institute (NCI) in 1937, the US military continued to supply soldiers with cigarettes in WWII. And while the Nazis instituted the first anti-tobacco campaign (given that German researchers first discovered the link), wartime nationalism helped cement Allies' affection for smoking.

Establishing Cause

The revolution in epidemiology that established a causal relationship between cigarettes and lungs was led by British statistician Sir Austin Bradford Hill. Hill and his colleague, Richard Doll, published their research on this link in 1950, as did US researchers Evarts Graham and Ernst Wynder. However, Hill and Doll’s contributions to public health and epidemiology went well beyond linking cigarette smoking and related illnesses.

It was Hill who pioneered the Hill Criteria of Causality, which still form the foundation of epidemiology, and although the list has varied over the decades, his criteria included:

  • Strength (effect size): The larger the association, the more likely that it is causal
  • Consistency (reproducibility): Consistent findings observed by different persons in different places with different samples strengthens the likelihood of a causative
  • Specificity: Causation is likely if there is a specific population at a specific site and disease with no other likely explanation
  • Temporality: The effect has to occur after the cause and after any expected delay between the cause and expected effect
  • Biological gradient: Greater exposure should generally lead to greater/lower incidence of the causative/protective effect
  • Plausibility: A plausible mechanism between cause and effect is helpful, but may not be available through current knowledge
  • Coherence: Coherence between epidemiological and laboratory findings increases the likelihood of an effect. However, lack of such evidence cannot nullify the association
  • Experiment: It may be possible to appeal to experimental evidence
  • Analogy: The effect of similar factors may be considered

Besides the Hill Criteria of Causality, the two are credited with originating the case-control study, and the first long-term prospective study, both of which focused on smoking-related illnesses. Hill also served as the statistician on a study of the use of streptomycin to treat tuberculosis, which is considered the first randomized controlled trial.

New Studies for New Methods

Even though cancer got the nation’s public attention, it was heart disease that pulled epidemiology into the limelight. As in the case of cancer, the cause of cardiovascular disease could not be found under a microscope, and in 1948 several organizations in Massachusetts formed the Framingham Heart Study, which is considered to be the first premier longitudinal cohort study. Soon after it was initiated, the Framingham Study was transferred to the National Heart Institute (now called the National Heart, Lung, and Blood Institute, or NHLBI), which had been formed in 1938. In 1961, data from the Framingham study was published that established smoking as a major factor in cardiovascular disease.

A noteworthy aspect of the Framingham study was that it included collection of blood samples and establishment of a biobank. Up to that point, the research linking cancer and heart disease to cigarette smoking was statistical. Through the 1950s and into the 1960s, collection of blood samples for testing served to identify “what the etiologic agent wasn’t”. The change occurred in 1961, when research from the Framingham Heart Study not only established smoking as a factor in cardiovascular disease, but also noted correlations between the disease and lipids in the bloodstream.

The NHLBI didn't miss a beat -  in the very same year, the Institute began collaborating with the Center for Disease Control and Prevention (CDC) to standardize cholesterol measurements. Since that time, the importance of blood and other biospecimens, and correspondingly, the development of laboratory methods to interrogate the biospecimens, has become an integral part of epidemiology and public health research.

So you could say that the automatic cigarette rolling machine is also, albeit indirectly, responsible for the creation of large-scale biobanking.

As medical discovery continues to evolve, the need for biospecimens and proper storage will remain a focal point of our industry. To learn more about unique sample storage and how biobanking and epidemiological tools are aiding veterinary research, check out our eBook Biobanking for Animal Health: Morris Animal Foundation is Taking Veterinary Research to a New Level.



Download eBook