Date Available


Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation





First Advisor

Dr. Kristin B. Ashford


Globally, in low and middle-income countries 4 out of every 10 pregnancies is reported to be unintended. Having an unintended pregnancy increases the risk of maternal and infant morbidity and mortality, preterm birth, low birth weight, and decreases rates of breast-feeding. The United States (U.S.) consistently has some of the highest rates of preterm birth, infant and maternal mortality of all high-income countries and 45% of all pregnancies in the U.S. are reported to be unintended. The etiology of these outcomes and their relationship to pregnancy intention are complex and multifactorial, but we know this disproportionately effects women living in poverty both in the U.S. and globally.

When couples have the knowledge, access, and power to decide when and whether to become pregnant they are more likely to seek preconception care, thus increasing the likelihood of planned pregnancies leading to improved maternal and child health outcomes. Primary prevention strategies to improve maternal/child health outcomes in the U.S. include sexual and reproductive health considerations such as increasing access to birth control. Globally, strategies include expanding access, as well as focusing on the empowerment of women and improving gender social norms. Focusing on community level norms and individual empowerment can lead to greater reproductive autonomy, which in turn leads to an increase in the uptake of birth control and family planning. This broader consideration of multiple levels of power or autonomy is often lacking in approaches taken in the U.S. More information is needed about the social context and determinants of pregnancy intention in our communities, particularly of women living in poverty.

The purposes of this dissertation were to 1) to describe reproductive autonomy and family planning challenges in a population of marginalized Ecuadorian women; 2) develop a conceptual framework of reproductive autonomy from the global literature; 3) to validate a shortened form of an interpersonal violence scale used in a study of low-income pregnant women in Kentucky; and finally 4) to investigate the association between pregnancy intention and individual, interpersonal and community factors of impoverished women living in Kentucky.

The qualitative study of women in Ecuador identified barriers and facilitators to family planning in a low-resource community. The major themes that emerged were that women’s autonomy was limited by men, shame was ‘keeping women quiet’, systems failed women, and as women aged they were able to build resilience in spite of these challenges. Many reported reproductive coercion, gender-based violence, and regret. Those who could leave unsupportive partners and found social support were more effective at planning their pregnancies. Evidence supports these themes are relatively common in the global literature, particularly of women living in poverty. The comprehensive review of these findings was used to develop a conceptual framework of reproductive autonomy. The Socio-Ecological Model was used to organize the data based on individual, interpersonal or community level determinants of pregnancy intention and reproductive autonomy. This new conceptual model, called the Power and Reproductive Autonomy (PARA) model, was used as a guide to analyze multiple levels of data in a secondary analysis of pregnant women living in poverty in Kentucky. Prior to this secondary analysis study, a measure used in the parent study needed to be validated. A short form of the Women’s Experience with Battering (WEB) scale was found to be psychometrically valid to measure of the impact of intimate partner violence for this population. Findings from the secondary analysis included high rates of unintended pregnancy (66%), and women with unintended pregnancy were more likely to report exposure to interpersonal violence, poor social support, and anxiety at the bivariate level. At the community (county) level those with an unintended pregnancy were more likely to live in counties with fewer social associations, and in rural communities. None of the access, gender equity, income inequality, or violence variables were correlated to pregnancy intention. In the final multilevel model, controlling for demographic variables, only being unmarried and answering the question in English were significant predictors of unintended pregnancy. The rate of social associations in a county was marginally significant with pregnancy intention, in that the presence of social associations appeared to decrease the likelihood of unintended pregnancy.

Operationalizing the PARA framework to examine predictors of unintended pregnancy in Kentucky proved to not yield expected results; county level variables related to access, gender equity, and violence were not found to be significantly correlated. Women answering the question in Spanish had significantly higher rates of planned pregnancy, which is a new finding. Having opportunities for social engagement also seemed to be a protective factor in preventing unintended pregnancies. Limitations of cross-sectional data also make it a challenge to capture cumulative life stressors which could contribute to poor reproductive autonomy. Future studies may yield a greater understanding of the social context of pregnancy intention if more interpersonal data related specifically to reproductive autonomy are in the model, such as reproductive coercion, relationship power, communication, and contraceptive decision making. Additionally, further examination of structures or systems that provide economic opportunities in the community is a promising area of reproductive autonomy and pregnancy intention research.

Digital Object Identifier (DOI)