Jimenez, and Claudio, 1998; Bullard, 1990, 1996). Another theme deals with the social gradient, exploring the health relevance of the social hierarchy (Singh and Siahpush, 2002; Steenland, Henley, and Thun, 2002; Pappas et al., 1993; Marmot et al., 1991; Marmot and Shipley, 1996). Other authors have examined populations defined by race, ethnicity, or gender, identifying their health needs and risks for preventable illness or disorder (Frumkin and Pransky, 1999; Centers for Disease Control [CDC], 1997; Polednak, 1989; Molina and Aguirre-Molina, 1994; Braithwaite and Taylor, 1992). Interrelationships among the social factors that influence health have also been the subject of research (Krieger et al., 1993; Kirkpatrick, 1994). Studies focused on work-related illness or disorder and injury within these populations remain relatively sparse; however systematic disparities in occupational health have been found to be related to race, class, and gender (Santiago and Muschkin, 1996; Burnett and Lalich, 1993; Robinson, 1984, 1989).
Older workers should not be regarded as a uniform population. Within the population of older workers, there are disparities related to social class, race, ethnicity, and gender, all of which have implications for how best to conduct research and develop policy for protecting older workers’ health. Although health research traditionally has used the individual as the unit of analysis, this methodological approach leaves societal-level factors un-examined (Schwartz, 1994; Needleman, 1997). The social context of exposure and health—the relevant history, cultural values, social networks, behavioral norms, economic and power relationships, and access to health-relevant resources—should be considered. For example, analyzing data on individuals’ social class as a static personal attribute is not the same as seeking insight into how the social system itself is stratified and the ways that individuals at different class levels interact. Examining individuals’ race and gender as personal attributes is different from studying the processes by which societal norms regarding race and gender can develop and change, differentially shaping work opportunities and health outcomes. To understand causes and possible solutions for health disparities within the aging workforce, it will be important to conduct research not only on detecting patterns among individual-level variables, but also on clarifying ecological, system-level variables having to do with social meanings, institutions, relationships, and interactions.
An examination of a social gradient for health is important for the older worker for at least five reasons. First, the nature of the work may play