Linear Regression Models in Epidemiology
share
The paper proposes to analyze epidemiological data using regression models which enable subject-matter (epidemiological) interpretation of such data whether with uncorrelated or correlated predictors. To this end, response functions should include not only terms linear in predictors but also higher order ones (e.g. quadratic and cross terms). For epidemiological interpretation of a regression model, the suggestion is to construct conditional functions derived from the general regression function with the values of all predictor variables held fixed excepting one predictor. Unlike the conventional techniques based on linear-predictor models in which the coefficient at any variable is interpreted, our approach proposes to interpret this conditional function, which is multivariate for any predictor being dependent on the values of all the other predictors. It is such functions that can describe relationships between Y and a predictor that have different forms in different predictor domains. The paper discusses differences in the interpretation of the proposed conditional functions between cases involving correlated and uncorrelated predictor variables. The construction and analysis of regression models for epidemiological and environmental data are illustrated with examples.
READ FULL TEXT
