Epidemiology of brain tumours in man and their relationship with chemical agents.

The 2000 or so deaths that result each year in England and Wales from brain tumours, as defined by ICD 191, represent approximately 0.3-0.4% of all deaths, with a slightly greater incidence in men than in women. Because of the low incidence of mortality from other causes in childhood, brain cancer accounts for 3.5% of all deaths in the 1-14-year age group. Mortality from brain cancer increased considerably between 1931 and 1961 but this is probably explained by changes in disease classification and improvements in diagnostic techniques. Because of the relatively low incidence of brain cancer in man, cohort studies need to be large in order to accrue sufficient cases for meaningful interpretation. Large cohort studies may run the risk of diluting a high-risk group of workers with lesser or non-exposed individuals. These considerations have encouraged the investigation of brain cancer by case-control studies and the interpretation of these must take into account possible sources of bias. At present the evidence for any chemical causation of brain cancer in man is inconclusive.