November 26, 2004

Letter: Risks of Smoking

Sir: Because so many smokers have given up, the number of UK deaths from smoking is decreasing: it was 140,000 in 1990, and is now "only" about 100,000 a year. Still, the risks are so great that those who smoke should really understand them (about half of all persistent cigarette smokers are eventually killed by their habit, although half are not) when choosing whether to continue. Journalists such as Tim Luckhurst ("Smoke screen", 16 November) do not help consolidate this understanding when they misunderstand, and hence deny, the main evidence.

Our estimates are that smokers die, on average, 10 years earlier than non-smokers, and that in 2000 smoking caused about 30,000 of the 34,000 UK lung cancer deaths; 13,000 of 117,000 other UK cancer deaths; 31,000 of 237,000 UK deaths from heart attack, stroke and other vascular diseases; and another 40,000 deaths from other conditions: total 114,000 (18 per cent of all deaths in England and Wales plus 24 per cent of those in Scotland).

Lung cancer is a "smoking-related" disease, and so are heart attack and stroke, but we and others, including the health minister he quoted, attribute only some of the deaths from these smoking- related diseases to smoking, not all of them. More than 10 per cent of persistent cigarette smokers die from lung cancer; Luckhurst quotes somebody saying that under 1 per cent do, because he mixed up the risk per year with the lifetime risk.

We and the Government claim that half of all smokers are killed by it, not that "death is always the result", and we do not include among the deaths from smoking "people who should have died when they did without smoking a single cigarette".


Professor of Medical Statistics & Epidemiology

University of Oxford