British Doctors Study (Doll and Hill prospective cohort)

In 1950, Richard Doll and Austin Bradford Hill published a case-control study showing that lung cancer patients reported heavier smoking histories than controls. It was suggestive but methodologically vulnerable: patients with a serious illness might recall their habits differently, and the design could not measure mortality rates directly. Hill, one of the leading biostatisticians of the era, knew that a prospective cohort study would answer the objection. In 1951 they mailed questionnaires about smoking to every doctor on the British Medical Register.

About 40,000 doctors responded, providing detailed smoking histories. Doll and Hill then followed their deaths, a straightforward task because British doctors had to keep their names on the Medical Register to practise and their deaths were well recorded. The 1954 report in the BMJ presented findings from just the first years of follow-up. Even in that short interval, lung cancer mortality rose sharply with the number of cigarettes smoked daily, with heavy smokers carrying substantially higher risk than lifelong nonsmokers.

The prospective design neutralized the main objection to the earlier case-control work. Smoking histories were recorded before any participant developed lung cancer, so there was no possibility that diagnosis had colored self-reporting. Death rates could be calculated directly, which the case-control approach could not do. The study also separated lung cancer deaths from deaths by other causes, addressing the concern that smokers might simply die of other diseases before lung cancer appeared.

Doll followed the 34,439 men in the cohort for fifty years, from 1951 to 2001. The report published in 2004 showed that men born between 1900 and 1930 who smoked only cigarettes and kept smoking died on average about 10 years younger than lifelong nonsmokers. Stopping at age 30 gained about 10 years of life expectancy, at 40 about 9, at 50 about 6, and at 60 about 3. Among men born around 1920, prolonged cigarette smoking from early adult life tripled death rates at each age.

The British Doctors Study ran in parallel with the U.S. evidence accumulating through the 1950s, and both fed into the 1964 U.S. Surgeon General's report, which concluded that cigarette smoking is causally related to lung cancer in men. In 1965 Hill published "The Environment and Disease: Association or Causation?", on how to judge whether an association between an exposure and a disease is causal.