Abstract
There is an increasing interest in the possible role of psychosocial environment or life stress as an etiological factor of chronic diseases. When the relationship between psychological stress and a disease is to be studied in human populations, epidemiologic methods are most suitable. A case-control study is one of the most useful methods for such studies. However, the study might lead to a false conclusion, unless confounding variables are taken into consideration. In this paper, the natural history of an ill person who visits a medical facility with a disease caused by stress from psychosocial environment was discussed and two confounding variables were successfully identified, which might cause spurious association between stress and disease when a case-control study was conducted using these hospital patients. First, susceptibility to psychological stress might cause a spurious association if the difference in the psychological process of seeking-medical-aid behavior between cases and controls was not taken into consideration. Second, the study also leads to a false conclusion when psychological stress plays a role not in the biological process of the disease but in the psychological process of the diseased people, through which they decide to seek medical care. For these reasons, patients selected from the general population through health examinations are preferable to hospital patients. If the hospital patients are the only available source for the study, controls should be selected with great care so as to avoid such spurious associations.
