UC Berkeley

Abstract

Systems of Influence in Global Health:Examining the Impact of Familial, Community, and Societal Structures on Health Outcomes

by

Brigid Ellen Cakouros

Doctor of Public Health

University of California, Berkeley

Professor Lia Fernald, Chair

This dissertation is a collection of three essays exploring topics in global health. Each essay aims to examine the associations between an individual, the systems in which they live, and potential ways to intervene on the systems in order to achieve better health and cognitive outcomes worldwide.

The first study uses qualitative data from Zimbabwe to examine the experiences of women with partners of unknown human immunodeficiency virus (HIV) status accessing both HIV testing and preexposure prophylaxis (PrEP) in Zimbabwe. This analysis aimed to identify themes of structural vulnerability and social drivers within this population in order to understand the best ways in which to conduct outreach for this community. Results showed that the majority of women wanted to be tested for HIV and use PrEP as a result of their partner’s non-monogamous lifestyle and their lack of ability to negotiate condom use. In order to increase access to PrEP, these women pointed to the need to carry out sensitization efforts within the community. Specifically, many women spoke about the need for PrEP to be discussed more frequently by clinicians at any point of contact in a healthcare setting. In addition, community health workers were considered a reliable support system within many communities. Findings suggest that community health workers could function as a key group for delivering information for accessing PrEP while also further desensitizing the community to HIV prevention and care.The second study included in this dissertation was a descriptive analysis of the Young Lives Study, using longitudinal datasets from four countries (Ethiopia, India, Peru and Vietnam) in order to explore patterns of family structure trajectories for children between ages of approximately 1 and 15 years. We categorized the children into seven household structures and documented the proportion of children who experienced household stability, those who had transitions, and the frequency of transitions. In all four countries, fewer than half of the children lived in nuclear households throughout childhood. In most countries, co-residence with a grandparent was typical in early childhood, with transition into family types without grandparents as time passed. Ethiopia had larger proportion of children who have lived in single-mother households and in households without their biological mother. These results suggest that research on childhood in lower- and middle-income countries ought to consider the dynamics of household structure as a variable that could have an association with child outcomes.

The third study, building on the second study, is a secondary analysis of the Young Lives Study longitudinal dataset from Ethiopia in order to explore associations between family structure trajectories and cognitive outcomes (vocabulary, math, and reading scores) at age 15 years. We categorized children as either having both parents present or not always having both parents present in all five rounds of data collection. We also explored the occasional presence or absence of grandparents. Our results show that urban children in family structures with an absent parent had statistically significant lower average test scores at age 15 when compared to results from children who always had both parents present. However, the addition of wealth index measures explained this association, reinforcing the link between socioeconomic status and cognitive measures. When a parent was absent, children with grandparents’ present had higher cognitive outcomes than children who did not have grandparent’s present. These findings suggest that educational attainment in Ethiopia could have associations with a child’s family structure beyond just the parental structure.