摘要：We reviewed the recent assertion that population strategies of prevention may inadvertently widen social inequalities in health. We used folate intake as a case example to examine what is known about the impact on inequalities of 2 population strategies: one agentic (public information campaign) and the other structural (mandatory fortification policy). We found some support for our hypothesis that the mandatory fortification policy was less likely than were the information campaigns to lead to worsening inequalities in health by socioeconomic status or race/ethnicity; however, conclusions were complicated by different outcome variables and different economic and political regimes in which interventions took place. Rose's population strategy of prevention 1 , 2 describes prevention activities that target a whole population regardless of variation in individual risk status. In contrast with a high-risk prevention strategy, which targets individuals identified (e.g., through screening) as having elevated risk for some adverse health outcome, the population strategy targets whole populations (areas, groups, jurisdictions, institutions, etc.) and acts on underlying causes of major health problems. Notwithstanding critiques emanating from recent improvements in risk identification, 3 , 4 the population strategy of prevention theoretically has significant leverage to impact prevalence of health outcomes by shifting the population's risk distribution in a direction of reduced risk. Recently, it has been suggested that the population strategy of prevention may inadvertently widen social inequalities in health. 5 Drawing on examples including public information campaigns regarding smoking, Frohlich and Potvin 5 made the argument—rooted in the concept of fundamental causes 6 —that persons who are socioeconomically disadvantaged are generally less able to take up or otherwise benefit from a population-level intervention than are socioeconomically advantaged persons, and, therefore, an increasing gap in health status between socioeconomic groups, driven by the intervention, may ensue. We have asserted that the population strategy of prevention will not necessarily widen social inequalities in health, and that the likelihood of it doing so will depend on whether the prevention strategy is more structural (targets conditions in which behaviors occur) or agentic (targets behavior change among individuals) in nature. 7 An agentic intervention, such as a public information campaign, may well widen inequalities because the effectiveness of such an intervention rests on individuals’ ability to act on information provided, potentially in the face of social or economic barriers. A structural intervention, by contrast, “aims to remove the underlying impediments to healthier behavior, or to control the adverse pressures,” 2 (p100) rather than targeting a behavior itself. As such, the issue of uptake of the intervention by individuals is largely obviated with structural interventions, in contrast with agentic interventions in which the individual must attend to, and act on, for example, information provided. A structural intervention may, therefore, actually narrow social inequalities in health through selective advantage to those who are socioeconomically disadvantaged. Some support appears to exist for the contention that interventions of a structural nature may narrow—or at least not widen—social inequalities in health. 7 With smoking, for example, for which large and growing social inequalities are evident, 8 available evidence suggests that the impact of structural interventions such as strong clean indoor air laws on smoking outcomes does not vary by socioeconomic status, suggesting that this approach may not widen inequalities. 9 , 10 There is some evidence that fluoridation of drinking water—another example of a structural intervention—may be associated with a reduction in oral health disparities in Britain, 11 – 14 Australia, 15 and possibly New Zealand, 16 although findings varied according to age and outcome variable. In both the United States and Canada, fluoridation of drinking water has been recommended as a practical means of reducing oral health disparities. 17 , 18 Different health issues, however, may vary in their degree of structural and agentic qualities, which in turn would have implications for the potential impact of structural versus agentic interventions on inequalities in that health issue. Our objective was to examine what is known about the impact of an agentic versus a structural population-level intervention on social inequalities for a particular health issue: outcomes related to folate intake among women of childbearing age. Folate is a B vitamin that facilitates nucleic acid synthesis and is necessary for normal cell replication. 19 Low folate status has been shown to be associated with health risks including high homocysteine, an amino acid shown to increase risk of ischemic heart disease, stroke, dementia, and birth abnormalities. The evidence demonstrating the protective effect of periconceptional folate intake in the prevention of neural tube defects (NTDs) is overwhelmingly strong, with spina bifida and anencephaly being the most commonly reported NTDs overall. 20 We selected outcomes related to folate intake because interventions of both an agentic (public information campaign) and a structural (mandatory fortification policy) nature to improve folate intake exist, thus lending itself to our objective. A secondary objective was to consider the interventions’ impact on inequalities by using both an absolute and a relative scale, because, as reported elsewhere, 21 the metric used can influence the conclusions drawn about the impact (equitable or not) of the intervention.
