A comment on the reference series with regard to multiple exposure evaluations in a case-referent study.

Case-referent (case-control) studies permit multiple exposures to be considered within the same series of cases and referents (controls) with regard to the etiologic contribution to a certain disease. Specific aspects of validity should be considered however, one being the utilization of an adequate reference series. Should the exposure cause, prevent, or otherwise have a certain relationship to disorders other than the one being studied, individuals with such other disorders cannot be uncritically included in the reference series if obtained from a hospital register, a register of causes of deaths, or some other source of subjects with relations to the individuals' health status. If a general population register is utilized, this problem is avoided since individuals obtained from such a register would adequately reflect the exposure frequency of the source population of the cases. These aspects are well illustrated by a consideration of the use of drugs among cases and, on one hand, referents with some sort of disorder as compared to referents (whether healthy or not) obtained from a population register on the other hand. Studies of industrial exposures seem to be somewhat less sensitive to bias of this type however, since industrial exposures are not primarily disorder-related, whereas drug exposures are.