Personality and Health in the 1980s: Psychosomatic Medicine Revisited?


  • Special thanks to Anita Delongis, John Marquez, Kent Sheets, and Carolyn Heaton for their invaluable consultation in the preparation of this commentary

Requests for reprints should be sent to Kenneth A Holroyd, Department of Psychology, Ohio University, Athens, Ohio 45701-2979


ABSTRACT This special issue is a sign of a resurgence of interest in the role of personality in health not seen since the 1940s and early 1950s when the promise of the psychosomatic approach to health and illness appeared to be the greatest This new look at personality and health represented by contributions to this special issue attempts to address the limitations of earlier work in psychosomatic medicine by making more explicit efforts to define personality variables precisely, to distinguish these variables from conceptually related psychological constructs, and to embed them in a body of theory and empirical research This new work also attempts to remedy methodological limitations of earlier work by placing greater emphasis on prospective research and highlighting distinctions between symptom reports, illness behavior, and actual illness However, the new work and earlier work in psychosomatic medicine share certain working assumptions, for example, a primary emphasis on the relatively direct impact of personality on disease onset, an assumption that personality variables operate in interaction with stressful events, and a frequent emphasis on general susceptibility to disease Moreover, this new work frequently risks the same methodological pitfalls that limited scientific progress in psychosomatic medicine We argue that the rapid rise and decline of psychosomatic medicine is most likely to be repeated in research on personality and health in the 1980s if reasonable criteria for considering personality variables a risk factor for disease are not precisely defined, disease endpoints (the dependent variable) are not assessed precisely, personality variables of interest (the independent variable) are not empirically distinguished from other related psychological variables, and complex relationships among risk factors are not taken account It is emphasized that models drawn from personality research cannot be transferred unchanged to the health arena without risking false inferences about the role of personality in health