Editorial | Editorial

The making of an epidemiological theory of bias and confounding We have chosen to discuss the history of epidemiology, as the history of the emergence and formalisation of its methods. In this issue and the following, Vineis (2002) and Vandenbroucke (2002) will be looking into the history of bias and confounding in epidemiology. These authors demonstrate convincingly how much epidemiology has borrowed from the social sciences and adapted methods and concepts to its own needs. We would like to point to the fact that these developments have often occurred while trying to solve particular prob-lems. In particular the controversy for the recognition of the harmful effects of tobacco on health has been an important crucible for the concept of confounding and the theory of bias