The third possible explanation for the relationship between psychological factors and CHD events is that psychological factors may be associated with confounding variables that lead to CHD events. Most, although not all, studies have demonstrated that the increased risk for CHD events associated with psychological disorders is independent of age, diabetes, smoking, lipids, obesity, physical activity, and measures of baseline CHD (5–6). Thus, it seems unlikely that traditional risk factors explain the association of psychological factors with CHD events. However, patients with psychological disorders may have a higher frequency of unmeasured characteristics that predispose them to CHD events. To exclude confounding, a randomized trial would need to demonstrate that treatment of psychological disorders decreases the incidence of CHD events without affecting traditional risk factors. Showing that changes in psychological factors precede a decreased incidence of CHD events would eliminate residual confounding and strengthen the causal hypothesis.