CAUSAL-MODELING OF EPIDEMIOLOGIC DATA ON PSYCHIATRIC-DISORDERS

This paper reviews the logic of causal inference from epidemiological data. I maintain that the dearest causal statements can be made when t he philosophical causal principles of association, direction and isola tion are upheld in epidemiological research. After reviewing the argum ent by Holland that only experimental manipulation affords clear causa l claims, I examine the utility of structural equation models and long itudinal methods for making causal claims from non-experimental data. This examination leads to the conclusion that mental health epidemiolo gists should begin to incorporate intervention trials into the last ph ases of their research programmes when they want to make strong causal claims.