Social structure, the positions people hold within it, and the nature of the social relations that result from these positions are important antecedent factors in determining resources that affect health (Link and Phelan, 1995; Kennedy et al., 1996; Kiwachi and Kennedy, 1997). Gender, socioeconomic position, and racial segregation contribute to the ideological and cultural context in which social relationships occur (Williams, 1990; Benderly, 1997). These, in turn, affect people 's everyday lives. Thus, it is not difficult to see health as “the product of social relationships between…groups, with these relationships expressed through people's everyday living and working conditions, including daily interactions with others” (Krieger, 1994). Social and economic relations are causal, explicitly shaping the production and distribution of individual and population health and disease at different points across the life span (Evans, 1995; Moss, 2000a).

The resources that are differentially associated with social and economic status include knowledge, income, wealth and assets, power, prestige, social networks, and psychological well-being. Availability and deprivation of these resources over time structure and differentiate the life course of individual men and women (Table 1) (Link and Phelan, 1995; Moss, 2000a). The cumulative effect of deprivation of social and material support over the life course is associated with stressful living and working conditions. Trajectories of stress and deprivation across a number of dimensions, including social class and ethnicity, may explain a higher prevalence of poor birth outcomes in groups such as African Americans (Geronimus, 1992). By contrast, resources are hypothesized to influence the ability of people to avoid health risks such as stress and to minimize their consequences if they occur (Rowley, 1998). In this framework, stress and health behaviors are mediating factors that occur in the context of an individual's social and economic position, the socioeconomic characteristics of a

FIGURE 2. Onset of prenatal care use: United States, 1984–1996. SOURCE: National Center for Health Statistics, prepared by March of Dimes, Perinatal Data Center, 1997.