Multiple disadvantage, substance use disorders, and attrition of low-income women in health policy research

Background: Policy makers have been urged to learn more about the needs of poor women for substance use treatment and other services. Poor women, and those with problems related to substance use, may be harder to recruit and retain in longitudinal research.

Purpose: Both welfare receipt and substance use disorders stigmatize and marginalize women. This study examines if these factors act with other social disadvantages to lead to differential attrition in longitudinal research.

Methods: The effects on attrition of multiple disadvantage and substance use are analyzed in a cohort of 498 female recipients of Temporary Assistance for Needy Families (TANF) and General Assistance (GA) in a California county. Multiple disadvantage is operationalized as four or more of nine socially marginalizing health determinants: very low income, long-term unemployment, low educational attainment, unstable housing, living alone, social isolation, poor neighborhood environment, disability, and minority race/ethnicity.

Results: Problem drinking, never drug use, and heavy drug use were associated with multiple disadvantage. In multivariate analysis, problem drinking (OR=2.16, CI=1.09, 4.32), drug dependence (OR=2.76, CI=1.25, 6.14), and multiple disadvantage (OR=2.34, CI=1.03, 5.28) were associated with attrition 12 months post-baseline. Multiple disadvantage was also associated with poor health.

Conclusion: Despite extended follow-up efforts there was differential attrition among women who experienced multiple disadvantages and those who reported alcohol and drug problems. Multiple disadvantage and substance use disorders may be factors in low response rates in studies of low-income women, and consequently, difficulty in developing accurate, research-driven health policies and services for poor and marginalized women.