Date of Degree
Doctor of Philosophy (Ph.D.)
Epidemiology and Biostatistics
Agency, adolescence, pregnancy, abortion, Zambia, India, United States
Purpose. The agency to decide one’s reproductive future is paramount, yet an estimated 75 percent of adolescent girls in the United States (U.S.) and 49 percent of those in low and middle-income countries experience mistimed or unwanted pregnancy. This dissertation characterized empirical representations of agency over the course of adolescence and investigated the relationship between agency and early and unwanted pregnancy and childbearing using data from three countries (India, Zambia and the U.S.).
Methods. Using latent class and latent transition analysis, multi-dimensional profiles of adolescent agency were assessed using data from: (1) a state representative cross-sectional cohort study of adolescents in Uttar Pradesh, India, (2) a cluster-randomized empowerment intervention of disadvantaged girls in Zambia and (3) a national longitudinal study of school-going youth in the U.S., including whether and when age and sex-related differences in agency profiles emerge. The influence of early life access to resources (e.g., educational attainment, household wealth, social support and access to quality health services) as well as time-varying exposures (e.g., physical or sexual violence, early marriage and early or unwanted pregnancy and childbearing) on baseline agency status as well as transitions in agency over time was investigated. To identify agency profiles most at risk of an early/unwanted reproductive health event and to provide insight into the directionality between agency and adolescent fertility outcomes, the observed proportion of early and unwanted reproductive health events was examined by agency profile membership at the time prior in the longitudinal studies.
Results. Findings supported the multidimensional and dynamic nature of agency, which was both age, sex and context dependent. Identified agency profiles were characterized by unique combinations of attributes associated with differential reproductive health risks dependent on setting. In general, protective assets were likely to augment the probability of membership to the highest agency profile. High agency profile members were also likely to experience a detraction in agency following life events including early marriage, physical or sexual violence as well as early/unwanted pregnancy and childbearing. However, among lower agency profiles, or those characterized by a mixed combination of agency attributes, exposure to these events were at times sources of agency. Together, findings illustrate a complex pattern of adolescent childbearing, where current agency status has implications for future reproductive health risk and is influenced by prior agency status.
Conclusion. Findings identified differential reproductive health risks associated with unique combinations of agency attributes, dependent on setting. Evidence of heterogeneity in the pathways by which agency influences early/unwanted adolescent pregnancy underscores the importance of program segmentation beyond age, sex and marital status in adolescence to include agency profile in order to identify and best engage those most-at risk. Findings further conceptual understanding of agency as a salient resource in adolescence and underscore the need for improved measurement among adolescent populations.
McCarthy, Katharine, "The role of agency in early and unwanted adolescent pregnancy: A multi country analysis" (2020). CUNY Academic Works.