Abstract
Trends in cancer mortality in England and Wales are dominated by a slowly evolving epidemic of lung cancer attributable to smoking. When the substantial effects of tobacco are discounted there is no evidence that the overall incidence of cancer is rising, but striking trends are apparent for several specific tumours. These may offer important clues to aetiology. * This is the first in a series of articles examining developments in cancer and updating what we know about the disease. Among members of the general public the incidence of cancer is widely perceived to be rising. It is true that many more people die from the disease now than 100 years ago, but this is largely attributable to reductions in mortality from other causes. Cancer is for the most part an affliction of old age, and as longevity increases so will the number of cancer cases. To get a clearer picture we need to look at trends in death rates at different ages. Not all cancers are fatal, and for those in which treatment has improved (for example, lymphoma and testicular cancer) mortality statistics must be interpreted with care. Also, allowance must be made for advances in diagnostic methods. For example, new imaging techniques may have led to better recognition of pancreatic tumours and thereby contributed to an apparent increase in mortality. Nevertheless, death rates are a good starting point for examining whether we really face a cancer epidemic. ### Trends in death rates at different ages Figure 1 shows the trends in overall cancer mortality in England and Wales during 1950-89 for five year age groups. In both males and females rates increased in elderly people but declined at younger ages. It turns out that this divergent pattern was determined mainly by changes in mortality from lung cancer, which contributes more deaths than any other tumour in men, …