Background Both women's and men's occupational health problems merit scientific attention. Researchers need to consider the effect of gender on how occupational health issues are experienced, expressed, defined, and addressed. More serious consideration of gender-related factors will help identify risk factors for both women and men. Methods The authors, who come from a number of disciplines (ergonomics, epidemiology, public health, social medicine, community psychology, economics, sociology) pooled their critiques in order to arrive at the most common and significant problems faced by occupational health researchers who wish to consider gender appropriately. Results This paper describes some ways that gender can be and has been handled in studies of occupational health, as well as some of the consequences. The paper also suggests specific research practices that avoid errors. Obstacles to gender-sensitive practices are considered. Conclusions Although gender-sensitive practices may be difficult to operationalize in some cases, they enrich the scientific quality of research and should lead to better data and ultimately to well-targeted prevention programs. Am. J. Ind. Med. 43:618-629, 2003. KEY WORDS: gender; sex; women; men; research methodology; epidemiology; ergonomics; confounding; effect modification; gender-based analysis
INTRODUCTIONBoth women's and men's occupational health merit scientific attention. In the United States, women constitute 46% of the paid workforce [United States Department of Labor, 2002], and have one third of compensated occupational health and safety problems, resulting in 81% of claims on a per hour basis [McDiarmid and Gucer, 2001]. These injuries entail direct and indirect costs to workers and employers, as well as human suffering [deCarteret, 1994]. Therefore, appropriately including sex and gender is increasingly relevant for occupational health research. Although researchers are interested in developing studies involving these variables, they may not know exactly how to do this. This article supplies some suggestions. Many of the arguments presented here will apply to other sources of socially defined diversity such as age, race/ ethnicity, and social class [Krieger et al., 1993;Kilbom et al., 1997;Wegman, 1999; Chaturvedi, 2001]. Each of these factors has its own interactions with the work environment and health effects, but their discussion is beyond the scope of this paper.We have identified three types of problems in the way occupational health research has dealt with sex and gender. First, hazards in women's work have been underestimated [Rosenstock and Lee, 2000;Bäckman and Edling, 2001;London et al., 2002;McDiarmid and Gucer, 2001]. Women have been less often studied by occupational health scientists [Zahm et al., 1994;Messing, 1998a;Niedhammer et al., 2000]. Under-reporting and under-compensation, recognized problems in occupational health [Biddle et al., 1998;Davis et al., 2001;Harber et al., 2001], may be more of a problem for women [Lippel and Demers, 1996;Gluck and O...
