After cardiovascular disease, cancer is the second most frequent cause of death in developed countries; accounting for 21% of all deaths. In developing countries, where it ranks well below infectious and parasitic diseases, it accounts for 7% of all deaths. As the incidence of cancer is strongly dependent on age, the different age structure of the population in the two groups of countries may be largely responsible for the difference in mortality between the two. The disparity may also be due to the different lifestyles and environmental exposure in the two groups, giving rise to different types of health risks. In addition, the vulnerability of cancer patients to infectious diseases may conceal cancer mortality in developing countries by masking the attribution of the ultimate cause of death.

In both groups of countries, cancers of different sites make different contributions to total cancer mortality. Here again differences in health care standards and dietary and other lifestyle habits are likely to play a major role. Thus, screening and therapy contribute to the lower mortality from cervical cancer in developed countries, and high exposure to hepatitis B and to aflatoxins to increased liver cancer in developing countries. A substantial fraction of excess mortality from cancer of the colon and rectum and of the breast in developed countries is probably related to dietary factors (such as high fat and low fibre content), while the very high mortality from lung cancer, which is now declining in some of these countries, at least among middle-aged men, reflects the high frequency of smoking a few decades ago. If spread of the use of cigarettes continues at the present rate in developing countries, mortality from lung cancer may be expected to reach epidemic proportions in the future, as development brings about a lengthening of the life span and a reduction in deaths from infectious diseases.

Time trends in cancer mortality based on historical data from the United States have been studied in detail. The figure overleaf shows the changes in mortality from cancers of various sites over a 55-year period. The sharp increase of lung cancer and the steady decrease of stomach cancer stand out as the most significant, the former mostly due to past trends in tobacco smoking, the other possibly to changes in food preservation and the amount of fruit and vegetables in the diet. These two cancers show broadly similar trends in all countries for which data are available. The other cancers recorded in the figure have undergone little variation since the 1950s, after rising until that time.

The list of exposures that have proved or are likely to be carcinogenic in humans is very long. It is not possible, however, to assess the burden of cancer attributable to most of them in either developed or developing countries, except in such cases as occupational exposure (e.g. to asbestos, vinyl chloride, benzene), urinary schistosomiasis which carries a high risk of cancer of the bladder, or hepatitis B virus, associated with a high incidence of liver cancer. Indeed, the most significant factors for the general population, beside tobacco smoking and alcohol consumption, appear to be dietary composition, probably responsible for a large fraction of some of the most common cancers (stomach, colon, breast, endometrium), and reproductive and associated behaviour (breast, ovary, cervix). The role of ionizing radiation in inducing cancers of various types is also well established but forms only a small percentage of the total. Exposure to the ultraviolet component of sunlight is responsible for a significant number of skin cancers.

Chemicals in the environment make a contribution to the occurrence of cancers in the general population, but the extent cannot be assessed with any certainty. If the factors mentioned above play the major role they are assumed to in the causation of cancer, a fraction of the order of 5% of all cancers in the general population is likely to be due to environmental chemicals. This gives no cause for complacency. The variety of chemicals produced used in everyday life, and released to the environment is increasing even In developing countries and exposure to them often involves more than the possible occurrence of cancer. Efforts should therefore be made to limit exposure so that the benefits deriving from their use are not offset by the damage they may cause to health. But this should not divert attention from the other factors, avoidance of which will achieve a significantly greater measure of prevention of cancers and other diseases.

Sources: Doll, R & Peto, R. The causes of cancer quantitative estimates of avoidable risks of cancer in the United Stares today Oxford, Oxford University Press, 1981; Kurihara, M. et al. Cancer mortality statistics in the world, 1950-1985. Nagoya, University of Nagoya Press, 1989; Silverberg, E. & Lubera J.A. Cancer statistics, 1988. CA-A cancer journal for clinicians, 38: 5-22 (1988) (cited in Tomatis, L., ed. Cancer causes, occurrence and control Lyon, IARC, 1990)