ECONOMETRIC APPROACHES TO EPIDEMIOLOGIC DATA - RELATING ENDOGENEITY AND UNOBSERVED HETEROGENEITY TO CONFOUNDING

The concepts of endogeneity and unobserved heterogeneity are well-know n among econometricians. However, these issues are rarely addressed in epidemiologic studies. This paper explores these two concepts, their relationship to each other, and the implications for analysis in epide miologic studies. An endogenous variable is defined as a predictor var iable which is partly determined by factors within the model itself, w hile unobserved heterogeneity is conceptualized as a vector of missing variables acting through the error term. Under certain assumptions, t he simultaneous existence of an endogenous variable and unobserved het erogeneity is shown to act in a manner analogous to confounding. Speci fically, this occurs due to an association between the error term in t he equation and che endogenous predictor variable. The accepted econom etric solution to this problem is to replace the endogenous variable w ith an 'instrumental variable' which is not correlated with the error term and thus not susceptible to confounding. The validity of these co ncepts and of the proposed solution are discussed. (C) 1997 by Elsevie r Science Inc.