The Development of Suspicion
Ods me, I marvel what pleasure or felicity they have in taking their roguish tobacco. It is good for nothing but to choke a man, and fill him full of smoke and embers.
THE MAIN TASK of the epidemiologist is to discover variables which are always or nearly always found in conjunction with a disease in which he is interested. Quite a curious selection of variables has been found to be associated with lung cancer. Among them are sex, age, marital status, country of birth and residence, population density, social class, occupation, air pollution, smoking, consumption of alcohol, consumption of coffee, consumption of cooked shell fish and crustacea, hairiness of the second phalanges of the fingers and toes, number of teeth lost, familial factors, past history of respiratory disease, being gassed in the 1914-18 War, the number of doctors per square mile, the purchase of petrol lighters and the possession of an extraverted temperament. Of all these possible causes smoking has from the beginning been singled out. Why ?
Interest in the problem began probably from a consideration of the vast increase in the number of deaths from cancer during the last sixty years. Figure 3 shows the growth of the number of cancer deaths in thousands from 1900 to 1960 in the United States. Some of this increase is simply due to a growth of the total population. Some of it is due to an increase in the age of the population. But the major part of the increase, as will be seen, is not due to any such external factors. Clearly we