ETIOLOGIC FRACTION ANALYSIS FOR CONTINUOUSLY DISTRIBUTED OUTCOME VARIABLES AND EMPIRICAL ANALOGY WITH DICHOTOMIZED OUTCOME VARIABLES

Background. It is not clear from the published literature whether R(2) estimated from linear regression models for continuously distributed outcome variables is analogous to the etiologic fraction for dichotomi zed outcome variables. This article attempts to address this issue. Me thod. Continuous and dichotomous outcomes of the same underlying attri butes (gestational age and fetal growth) were compared using data from a recent study of birthweight distributions in ethnic Caucasian infan ts. Results. The relative magnitudes of the etiologic fraction and R(2 ) were quite similar for the same underlying attributes. For example, R(2) and etiologic fraction for weight gain rate ranked 2 and 3, respe ctively, for fetal growth and ranked 4.5 and 5, respectively, for gest ational duration. Conclusions. R(2) estimated from linear regression m odels for continuously distributed outcome variables appears analogous to the etiologic fraction for dichotomized outcome variables. If due consideration is given to the underlying biological mechanisms of the studied attributes, R(2) can be used as a measure of public health imp act.