Weighing the epidemiological evidence for environmental association and causality

摘要

Methodological explanations, such as bias or (genetic) confounding, must be ruled out.rnBias may arise, for example, when cases or controls are selected, followed overtime, or interviewed in such a way that they are more or less likely to report environmental exposure in the case-control comparison (for example, a case-control study of psychotic syndrome and cannabis use recruiting controls from a selected community with very strict lifestyle rules).