Giving Birth and Returning to Work: The Impact of Work–Family Conﬂict on Women’s Health After Childbirth MIRA M. GRICE, PHD, DENISE FEDA, MS, PATRICIA MCGOVERN, PHD, BRUCE H. ALEXANDER, PHD, DAVID MCCAFFREY, BA, AND LAURIE UKESTAD, MS PURPOSE: Since 1970, women of childbearing age have increasingly participated in the workforce. How- ever, literature on work–family conﬂict has not speciﬁcally addressed the health of postpartum women. This study examined the relationship between work–family conﬂict and mental and physical health of employed mothers 11 weeks after childbirth. METHODS: Employed women, 18 years and older, were recruited while in the hospital for childbirth (N Z 817; 71% response rate). Mental and physical health at 11 weeks postpartum was measured using SF-12 version 2. General linear models estimated the associations between the independent variables and health. A priori causal models and directed acyclic graphs guided selection of confounding variables. RESULTS: Analyses revealed that high levels of work interference with family were associated with sig- niﬁcantly lower mental health scores. Medium and high levels of family interference with work revealed a dose-response relationship resulting in signiﬁcantly worse mental health scores. Coworker support was strongly and positively associated with better physical health. CONCLUSIONS: Work–family conﬂict was negatively associated with mental health but not signiﬁ- cantly associated with physical health. Availability of social support may relieve the burden women can experience when balancing work roles and family obligations. Ann Epidemiol 2007;17:791–798. Ó 2007 Elsevier Inc. All rights reserved. KEY WORDS: Job Satisfaction, Maternal Welfare, Minnesota, Postpartum Period, Women’s Health, Work. INTRODUCTION Over the past several decades labor force demographics have changed signiﬁcantly as women entered the workforce. In 1970 43% of women 16 years or older were in the labor force compared with 60% in 2002 (1). In 2003 an estimated 53.7% of women with infants participated in the labor force (2). The increasing number of employed women of child- bearing age highlights the emerging necessity to understand the health-related consequences that result from merging employment with family life, especially for postpartum women. In previous studies, work–family conﬂict has been deﬁned as conﬂict resulting from balancing both work and family roles (3–10). This deﬁnition may be best understood as a speciﬁc type of interrole conﬂict, in which participation in one role is incompatible with participation in another role (11). Greenhaus and Beutell (12) established that in- compatibility between the work and family domains may originate from time-, strain-, or behavior-based conﬂict. Time-based conﬂict occurs when role pressures stemming from involvement in either domain compete for the individ- ual’s time. For example, the number of hours worked in a job or an inﬂexible work schedule may compete for a mother’s time with her young children at home. Low levels of social support, whether at work (e.g., from supervisors and coworkers) or at home (e.g., from spouse) illustrate types of strain-based conﬂict, in which strain emanating from one role affects the individual’s performance in another role. Behavior-based conﬂict arises from exhibiting behavior that may be appropriate in one role but unacceptable in another role. For example, unemotional behavior may be acceptable at work but not acceptable in parenting (12). Measures developed by Frone and colleagues (13) were used to assess the two directions of work–family conﬂict; work interference with family (WIF) and family interfer- ence with work (FIW). Previous research has suggested that balancing work and family roles may lead to adverse outcomes for the family and the employer (5, 14, 15). Negative outcomes that have been associated with work– family conﬂict are job dissatisfaction, marital dissatisfaction, From the School of Public Health, University of Minnesota, Minneapolis. Address correspondence to: Mira M. Grice, PhD, University of Minne- sota, School of Public Health, Mayo MMC 807, 420 Delaware Street SE, Minneapolis, MN 55455. Tel.: (612) 626-4824; fax: (612) 626-4837. E-mail: gric0001@umn.edu. This research was supported by grant 5 R18 OH003605-05 from the National Institute for Occupational Safety and Health (NIOSH). Its contents are solely the responsibility of the authors and do not necessarily represent the ofﬁcial views of NIOSH. Gratitude is extended to the doctoral training program in occupational health services research and policy made possible through the Midwest Center for Occupational Health and Safety and Edu- cational Research Center supported, in part, by NIOSH (T24/CCT 510422-04-01). Received November 10, 2006; accepted May 13, 2007. Ó 2007 Elsevier Inc. All rights reserved. 1047-2797/07/$–see front matter 360 Park Avenue South, New York, NY 10010 doi:10.1016/j.annepidem.2007.05.002