Abstract
Stress-related research has thus far failed to provide an adequate understanding of hypertension and other psychosomatic ailments for three main reasons: First, there has been a continual failure to view stress as a relational phenomenon, that is, as a particular kind of transaction between person and environment. Second, there has been much confusion about the social, psychological and physiological levels of stress analysis; each is to some extent independent of the other, so that what happens at one level cannot stand for what happens at another. Third, the predominant research model has been structural and static. That is, the researcher looks at some environmental or personality characteristic, treating it as a stable property, and attempts to relate it to the risk of hypertension across persons or groups. Such an approach overlooks the key social, psychological and physiological mediating processes (e.g., social supports, cognitive appraisals, and coping) that are concurrent with and have causal significance in blood pressure elevation or change. Structural research models need to be supplemented with process-oriented ones in which the same persons are observed across various adaptational encounters and over time.