Only 60 years ago, smoking was generally regarded as innocuous. Now it is recognized as the main cause of lung cancer, the world's commonest cancer. More than any other person, Richard Doll was responsible for this remarkable transformation in attitude, as well as for the now widely accepted view that much cancer (not just of the lung) is, at least in principle, avoidable. Doll, who died on 24 July 2005 aged 92, was the world's foremost epidemiologist and had a wide influence on the progress of the subject.

Doll was born on 28 October 1912, the son of a London general practitioner. He might have pursued his strong interest in mathematics, but at the last moment decided to follow his father into medicine, qualifying at St Thomas' Hospital in 1937. His epidemiological career began after war service, with work on the causes of peptic ulcers that attracted the attention of Tony Bradford Hill. At this time many considered that the marked rise in lung cancer deaths between the two world wars was simply due to improved diagnosis. In 1948, however, Percy Stocks and Ernest Kennaway succeeded in persuading the Medical Research Council (MRC) to investigate the issue.

For this, the MRC turned to Hill, who recruited Doll, and together they obtained detailed smoking histories from a large number of patients with and without the disease. In 1950 they reported the findings of this ‘case-control’ study. The rarity of non-smokers and the relative excess of heavy smokers among the lung-cancer cases, compared with the controls, convinced them that cigarette smoking was “a cause, and an important cause” of the disease.

This was not, in fact, the first such study, for earlier in the same year similar findings were reported in the United States by Ernst Wynder and Ewart Graham. Doll and Hill, however, went on to collect details of smoking habits from 40,000 British doctors, and to ascertain the causes of death of those who died. Their earlier findings were confirmed, including the dose–response relationship (the more cigarettes smoked, the greater the lung cancer risk). But this ‘cohort’ approach was also able to reveal the wide range of other smoking-related diseases; in the recent 50-year follow-up, published with Richard Peto, he showed that one-half of persistent smokers were estimated to die as a result of the habit. The long interval from starting smoking before its major effects on mortality appear was another notable finding.

This study has continued and is unique in its regular updating of the smoking habits of participants that has allowed recognition of the (unexpected) benefits of stopping smoking. The strong dose–response relation between lung cancer and cigarette smoking, the high standard of the design and conduct of the study, and the balanced assessment of its findings in many papers — all of these played a part in making a convincing case about the perils of smoking. In consequence, habits have changed. For example, in Britain the proportion of men who smoke has fallen from 80% a few decades ago to less than 30%.

Doll's work went far beyond smoking, however. He was involved in early investigations of the carcinogenic effects of ionizing radiation; in 1957 with Michael Court Brown he followed the fates of 14,000 patients with ankylosing spondylitis who were treated with radiation. This cohort study brought independent confirmation of studies of Japanese A-bomb survivors that radiation could cause leukaemia, and it has been a major source of data on the dose–response relation of radiation and cancer.

In another project, British radiologists were investigated because of their repeated exposure to low doses of radiation, the only group with excess cancer being those with long practice in the early twentieth century when high cumulative doses were likely as a result of the relative lack of protective procedures. Later Doll studied many aspects of exposure to low levels of radiation, both ionizing (for example radon and its relationship to lung cancer) and non-ionizing (electromagnetic fields from power lines and childhood leukaemia).

Doll was an author of the first compendium of worldwide data on cancer incidence, and was the first to recognize that each cancer that was common in one part of the world was rare in another — and for reasons that are not primarily genetic, but largely extrinsic. In 1954 with Peter Armitage, long before the relevant advances in molecular biology, he adduced evidence for the multi-stage nature of carcinogenesis. In 1955 he was also the first to show a significant excess of lung cancer among asbestos workers and later, with MartinVessey and others, he studied the side-effects of oral contraceptives. He continued a clinical attachment at the Central Middlesex Hospital until 1969, conducting therapeutic trials on peptic ulcers, including the first demonstration of the efficacy of both liquorice extract and carbenoxelone.

Credit: R. JUDGES/REX FEATURES

In 1969 Doll left the MRC Statistical Research Unit in London (of which he had been director since 1961) to become Regius Professor of Medicine at Oxford, an unexpected appointment that brought added attention to the subject of epidemiology. The range of his epidemiological activities was extraordinarily varied. In his career he published over 500 papers, reputedly without the irritation of writing a single grant application (he was, incidentally, saddened that new government rules will prevent certain harmless but valuable research into public health, like some of his own). Late in life he participated in randomized trials, and meta-analyses of trials, of treatments for breast cancer and vascular disease, as well as in collaborative meta-analyses of risk factors for breast cancer.

Doll's approach to his work was exemplary. He examined data with extraordinary detachment, even when surrounded by intense preconceptions. The more closely people worked with him, the more respect they had for him. Highly (but unobtrusively) organized, his concentration, efficiency and memory were formidable — as was his capacity for work, which was unaffected by long-haul flights or age.

He is commemorated in Oxford by Green College, which he established, and by the two research units he helped to found. Both of these are now housed in the new Richard Doll Building.