Some cautions for research on personality and health.

This article comments on three conceptualizations of the relationship between personality and disease as they relate to the papers in this issue. Despite the sophistication and promise of recent research in this area, conceptual and methodological difficulties present in earlier psychosomatic research have persisted. In this area, robust and convincing associations between traits and disease remain largely elusive, and stable, valid measures of psychological factors are relatively rare. In addition, the mechanisms relating behavior to health are sometimes sketchy or overstated. We suggest that the most productive research strategies will focus on mechanisms linking behavior to health, rather than merely correlating traits with measures of disease. Needed are continued close attention to these pitfalls by researchers as well as the maintenance of a critical attitude by readers.