MIETTINEN os. Causal and preventive interdependence: Elementary principles. Scand j work environ health 8 (1982) 159-168. "Synergism" of two factors in the causation or prevention of an all-or-none event means the existence of instances in which both factors are needed for the effect, while "antagonism" means that at least one can block the solo effect of the other. The manifestation of such interdependences -or of their complement, independence -in terms of event rates is complicated by the correlation of the susceptibilities to the two factors. Thus, given the risk difference (RD) values representing the solo effects, the RD corresponding to the joint exposure has a range consistent with independence so that independence cannot be inferred even from very ample data without knowledge of the degree of correlatedness of the susceptibilities. The definition of this range is closely analogous for causal and preventive factors, respectively. However, when knowledge about interdependence is used in inference about the factors' interrelation in the mechanisms for the effect, sharp distinctions may have to be made between causal and preventive factors. In each case, the interdependence is a result of the interrelation of the actions of the two factors and!or interaction between them. Operational decisions, having to do with the wisdom of the joint exposure, can be guided by knowledge about the interdependence of the factors; however, knowledge of the risks corresponding to the various exposures is a sufficient guide, without any need for inferences about causal or preventive interdependence.

This article in PubMed: www.ncbi.nlm.nih.gov/pubmed/10884143 Scand J Work Environ Health 1999;25 (6, Even though etiologic research has been the central concern in academic epidemiology, its concepts have remained confused or malformed, starting from that of etiology itself; and the same applies to its principles, starting from the notion that the principal variants of an etiologic study are the 'cohort' study and the 'case-control' study. This article suggests revisions of some central concepts pertaining to the object (and objective) of an etiologic study, and it posits an updated conception of the essence -singular-the study itself. This is supplemented by some novel, yet merely orientational, propositions in respect to quality-assurance in etiologic research.

No abstract