摘要：Objectives. We examined relationships between US women’s exposure to midlife work–family demands and subsequent mortality risk. Methods. We used data from women born 1935 to 1956 in the Health and Retirement Study to calculate employment, marital, and parenthood statuses for each age between 16 and 50 years. We used sequence analysis to identify 7 prototypical work–family trajectories. We calculated age-standardized mortality rates and hazard ratios (HRs) for mortality associated with work–family sequences, with adjustment for covariates and potentially explanatory later-life factors. Results. Married women staying home with children briefly before reentering the workforce had the lowest mortality rates. In comparison, after adjustment for age, race/ethnicity, and education, HRs for mortality were 2.14 (95% confidence interval [CI] = 1.58, 2.90) among single nonworking mothers, 1.48 (95% CI = 1.06, 1.98) among single working mothers, and 1.36 (95% CI = 1.02, 1.80) among married nonworking mothers. Adjustment for later-life behavioral and economic factors partially attenuated risks. Conclusions. Sequence analysis is a promising exposure assessment tool for life course research. This method permitted identification of certain lifetime work–family profiles associated with mortality risk before age 75 years. Over the past half-century, US women have fundamentally shifted the ways in which they combine the roles of employee, spouse, and parent. In 1935, 35% of women aged 25 to 54 years worked for pay; by 1970, participation climbed to 50%, and, by 1990, work rates had risen to the current level of 75%. 1 Marriage patterns also changed: women delayed age at first marriage, chose cohabitation over marriage, and divorced at higher rates than did their predecessors. 2–4 Finally, fertility remained at approximately 2 children per woman, although age at first birth rose. 5 Whereas for a woman born in 1930 a normative work–family trajectory was to marry in her early 20s and exit the labor force at the birth of her first child, patterns for subsequent birth cohorts have diversified. 6,7 As the United States has few public social policies that support working parents (such as paid maternity leave or subsidized child care), certain work–family situations may generate competing demands, creating burdens of time, finances, and other resources. 8 Work–family strain, in turn, may predict future disadvantage and disease, including cardiovascular disease, psychiatric disorders, and health-damaging behaviors, 9–13 increasing mortality risk as a result. Consequences of work–family conflict may be more pronounced among single mothers without offsetting spousal, social, financial, or emotional support, 14 putting them at higher risk of smoking, 15 cardiovascular disease, 16 and mortality. 17 The same objective demand profile may have more profound health effects in lower–socioeconomic status (SES) versus higher-SES women. 11 Research has identified long-term health impacts of work–family strain and has attempted to explain the phenomenon. The role accumulation hypothesis posits that marriage, children, and work promote health, individually and together. 18 The multiple role hypothesis suggests that, although these roles are salubrious individually, certain combinations of roles can impose competing health-damaging demands. 19 A major weakness in studies testing both theories is that work–family demands are typically assessed at a single point in midlife. 16,18 However, work–family circumstances often change many times during early life and midlife; one-time assessments of demands could obscure underlying patterns. Life course frameworks suggest that both exposure dose and timing can have an impact on later disease risk. 20,21 Thus, analysis of ages during which certain work–family demands might be especially toxic—and analyzing for whom toxicity is most profound—could enhance understanding of relationships between work–family demands and health. We used the role accumulation and multiple roles frameworks to test whether patterns and timing of work–family combinations during early life and midlife (16 to 50 years) predicted subsequent mortality risk among US women. We hypothesized that high demands from work and home, with little offsetting support—for example, single working mothers—would be at greater risk for mortality between ages 55 to 75 years when compared with women in lower-strain work–family circumstances. We also hypothesized that hazards in higher-risk groups would be partially explained by sociodemographic characteristics, health behaviors, and economic factors at older ages.
