What the work claims
Smoking increases risk of lung cancer, myocardial infarction (then called coronary thrombosis), respiratory disease, and other smoking-related illnesses. Smoking reduces lifespan up to 10 years. More than half of all smokers die from a smoking-related disease.
How it was done
Researchers contacted all UK-registered physicians in October 1951 and enrolled 40,701 respondents — two-thirds of those approached. No further cohorts were recruited. Respondents were stratified by decade of birth, sex, cause-specific mortality, physical health, and smoking habits. Follow-up occurred via questionnaires in 1957, 1966, 1971, 1978, 1991, and 2001. Cause-specific mortality was monitored until 2001.
What holds up
The 1956 finding — that smoking increases lung cancer risk — holds up. So does the 50-year analysis showing markedly higher rates of lung cancer and coronary thrombosis among smokers. The conclusion that smoking reduces lifespan up to 10 years, with excess mortality dependent on duration of smoking, is directly supported by the data.
What does not
It does not generalise to women: females were excluded from most analyses due to limited sample size. It does not quantify risk for non-fatal disease or intermediate biomarkers. It does not establish biological mechanisms. It does not assess cessation timing beyond age-based lifespan loss estimates.
Why it matters beyond the lab
It shifted public health policy from speculation to evidence-based intervention. It proved that a single modifiable behaviour could dominate population-level mortality — a template later applied to obesity, alcohol, and air pollution.
Is it worth your time
Yes. It remains the longest-running prospective cohort study on smoking and mortality, with direct, repeated measurements across five decades. Its design eliminated recall bias for baseline exposure and linked behaviour to cause-specific death with high precision — rare for mid-20th-century epidemiology.