Abstract

Epidemiological studies of chronic diseases began around the mid-20th century. Contrary to the infectious disease epidemiology which had prevailed at the beginning of the 20th century and which had focused on single agents causing individual diseases, the chronic disease epidemiology which emerged at the end of Word War II was a much more complex enterprise that investigated a multiplicity of risk factors for each disease. Involved in the development of chronic disease epidemi-ology were therefore fundamental discussions on the notion of causality, especially the question when causal inferences could be justified. In this paper, I shall analyze the implicit normativity of these de-bates. First, I shall give a brief overview of the historical background on which chronic disease epi-demiology emerged and describe how the pioneer studies on smoking and lung cancer became icon of the major challenge that the emerging chronic disease epidemiology was facing: the impossibility of proving that statistical associations reflected causal relations. Next, I shall describe how the develop-ment from the monocausal enterprise of infectious disease epidemiology to the multicausal enterprise of chronic disease epidemiology gave rise to intense discussions of the possible criteria by which to establish causal relationships between a given factor and a particular disease. I shall show how the necessary and sufficient conditions expressed in the so-called Henle-Koch criteria that had proved useful for the 19th century investigations of infectious diseases remained an ideal, although clearly an unobtainable one. Thus, I shall show how 20th century chronic disease epidemiologists on the one hand were searching for a new set of general principles which would provide a logical framework for their investigations, but on the other hand admitted that they would have to accept something more "pragmatic". I shall analyze the various positions in this debate, arguing that the implacability of the debate was due to unrecognized normative issues. I shall argue that many insisted on a distinction between science and application that was untenable, but that due to this distinction the values in-volved in deciding whether or not to act on the basis of a hypothesis were rarely explicitly discussed and the decision therefore continued to appear as a matter of taste rather than the result of a cogent normative analysis