Esplendor e miséria da Epidemiologia para avaliação da promoção da saúde
Louise Potvin, Patrick Chabot

Splendor and misery of Epidemiology for the evaluation of health promotion
Louise Potvin, Patrick Chabot

During the past 30 years, the field of public health has been under enormous pressure to move toward a more "social" approach to health. This is true of the two fundamental areas of the field: researching the “causes” of disease and ill health and intervening to improve health. In terms of research, social epidemiology has broadened the traditional domain of classic epidemiology to include social determinants (Berkman & Kawachi, 2000a) in studies looking at what causes unhealthy societies (Kawachi, 2002).
Given the realm of public health intervention, the Ottawa Charter (WHO 1986) and health promotion were created with the specific goal of changing the way health professionals and decision makers think about health and “to transform the complex knowledge of social epidemiology into practice and at the same time be able to document an effect” (Kickbush, 1994, p. 13). Despite the growing support from research agencies and health decision makers , both social epidemiology and health promotion still struggle to put into practice their social and population perspectives on health. Both have yet to achieve their transformation from classical epidemiology for one, and disease prevention for the other, both being based on individualistic models of health and public health intervention. Social epidemiology has yet to demonstrate that unpacking the social determinants of health leads to a better understanding of health and health promotion. It also faces the need to demonstrate whether and how it improves health.
Underlying this paper is the proposition that the challenges facing both social epidemiology and health promotion are closely linked. Both areas are experiencing difficulties in developing a satisfying conception of the social aspects of health. Although social epidemiology proposes innovative conceptualizations of health and disease (McKinlay & Marceau, 1999; 2000), causality (Krieger, 1994) and social categories as fundamental causes of disease (Link & Phelan, 1995), most studies make use of these social categories as just another layer of risk factors in predictive models (Macintyre, Ellaway, Cummins, 2002). There is little discussion on whether these categories are of the same nature as the risk factors that are usually produced by classical epidemiological studies (Potvin & Frohlich, 1998).
Similarly, in the realm of public health intervention, from disease prevention to health education, and to health promotion, approaches to improving population health have evolved as well as our conception of health and disease (Green & Kreuter, 1999). Although some “avant garde” practices in health promotion are leading the way into radically new conceptions of health and public health interventions, these practices still lack proper tools to reflect on their process (McQueen & Anderson, 2001) and produce the much awaited positive results that will legitimate public spending (IUHPE, 2000).
In this paper, we propose that a careful examination of the barriers encountered by health promotion to complete its transformation away from disease prevention also provides insights that will help social epidemiology achieve its own transformation away from classical epidemiology. In so doing we identify two epistemological blind spots that are common to health promotion and to social epidemiology. These two blind spots are reflexivity and historicity, two notions that contemporary social theory has developed extensively to further our understanding of the complex relationship between human practices and the social structure. The former pertains to the absence of an absolute determinism between the social structure and human practices given the human capacity to reflect on its own experience with abstract categories, thus creating agency and capacity to transform the structure. The latter refers to the conception that at any time, the state of an object (program, health status or other) cannot be isolated from the contexts that give it meaning: its previous states and its transformation.
In addition to their relevance for an appropriate evaluation of health promotion, these two notions could help debug some of the issues in the study of health and place (Macintyre et al., 2002). Our hope is that by achieving its own transformation from classical epidemiology, social epidemiology will contribute to freeing the evaluation of health promotion from the models that were designed to evaluate disease prevention and health education interventions.