The epidemiology of life stress.

Different contemporary societies, different historical periods in the same societies, and different strata within the same societies, produce different patterns of health disorders. At the ecological level, there can be no dispute that particular social structures generate particular patterns of health disorder. To put it another way, the distribution of health in a society is one dimension of the social structure and an indicator of its nature and impact on people's lives (Susser & Watson, 1971). When we turn to the individual level, however, and to the question of how the impact of the social structure is translated into pathology, much that is said must rest on assertions of authority and faith. It is about 40 or 50 years since the'stress' hypothesis gathered momentum and began to displace the fashionable hypothesis of that time, which attributed a gallimaufry of obscure ills to foci of infection. Many excellent sets of teeth and many pairs of tonsils were sacrificed to that hypothesis. We can count ourselves fortunate that the stress hypothesis does not require surgical intervention. The starting point for social scientists is concern with society, and hence with total populations, both the healthy and unhealthy, rather than with disordered segments of populations. They thereby gain breadth. Social scientists extend our vision also because they tend to begin with the independent variable a social factor and seek out its many effects, rather than begin with the dependent variable a health disorder and seek out what impinges on it. To start with health disorders is a narrowing view that is the norm in medicine, since its concerns arise there. Multifaceted society leads social scientists quickly and naturally to multivariate paradigms, rather than to the specificity of one-to-one relationships. Much of my own endeavour has been to bring the concepts and the methods of the social sciences into medicine and epidemiology. I shall here reverse myself and try and bring out for social scientists the notions and emphases of epidemiologists. One of these is the notion of the triad, agent-hostenvironment, and the need, in analysing the problems of health disorders and in constructing research designs, to segregate these three interdigitating elements. This triad is entrenched in epidemiology. Indeed, the triad appeared at last to have become an impediment to flexible multivariate analysis, to the extent that the subject needed to be freed from its tyranny. In counterpoint, the literature of 'stress' disorders shows that many investigators could have benefited from familiarity with the concept. There has been a lack of differentiation among the components of the triad: some studies concentrate on agents, namely stressful events; others on the environment, namely situations; while the characteristics of the host have been virtually ignored, and we know little of susceptibilities and resistance. In classifying agents, researchers conscious of the triad might less often have lumped stimuli indiscriminantly together, irrespective of force and direction. Stress disorders can be construed in the manner of Hans Selye as the intermediate result of some final common pathway, analogous to inflammation, which is a general pathological response to many agents and the antecedent of a variety of specific manifestations of disorder. Understanding of inflammation was much advanced when the agents were sorted one from another: chemicals, heat, ultraviolet and ionizing radiation, bacteria, viruses, protozoa and multicellular parasites all cause inflammation. In the field of stress, it is clear that specified events yield more convincing evidence than do undifferentiated ones: 'loss' emerges as the particular antecedent of depression (Paykel et al. 1969). In describing hosts, researchers conscious of the triad might have paid more attention to the varying susceptibility and immunity of individuals with different constitutions and experiences, including variability denoted by 'simple' indices like age, sex and other statuses and roles. No disease is uniform in all these respects.