Abstract
School enrollment, mobility, and occupation are each important factors to consider when examining HIV (human immunodeficiency virus) infection risk among youth in sub-Saharan Africa. Through an analysis of narrative life histories from 30 HIV-positive and 30 HIV-negative youth (aged 15-24 years), matched on gender, age, and village and purposively selected and interviewed from the Rakai Community Cohort Study, this article shows the complex connection between leaving school, mobility, and occupation with implications for HIV risk. We identified a pattern of risk factors that was present in many more HIV-positive than HIV-negative youth life stories. These HIV-positive youth shared a similar pathway during their transition to adulthood: After leaving school, they moved in search of occupations; they then engaged in risky occupations before eventually returning to their home village. Linking the lines of inquiry on school enrollment, mobility, and risky occupations, our findings have important implications for adolescent health research, practice, and policy in Uganda and across sub-Saharan Africa and the developing world.
Keywords
Across the globe, as youth transition to adulthood, leaving school and finding an occupation are two major milestones (Arnett, 2000; Patton et al., 2016). Despite similarities, adolescents in distinct sociocultural contexts face unique rewards and challenges in achieving these milestones, with specific consequences for their health (Golden & Earp, 2012; Juárez, LeGrand, Lloyd, & Singh, 2008; Raneri & Wiemann, 2007; Salazar et al., 2010). While education has been linked to improved health outcomes (Patton et al., 2016), mobility, which is often required to find an occupation (Barratt, Mbonye, & Seeley, 2012; Farrugia, 2016), has long been connected to the transmission of infectious diseases (Gushulak & MacPherson, 2000; Moreno et al., 2017). In the context of the HIV (human immunodeficiency virus) epidemic in sub-Saharan Africa, the transition to adulthood involves increased risk for infection as youth move in search of work (Olawore et al., 2018; Schuyler et al., 2015).
Region Description and Demographics
The Rakai District in southern Uganda is largely rural, with dispersed trading centers. The district borders Lake Victoria on the East and Tanzania in the south with a population of approximately 516,000 (Uganda Bureau of Statistics, 2016, 2017). The majority of the population (57%) is between 0 and 17 years old, 19% are 18 to 30 years old, and 24.2% are 31 years and older. HIV prevalence from 2011-2013 ranged from 10% to 25%, depending on the community (Chang et al., 2016), and HIV incidence rates ranged from 1.2 to 1.1 cases per 100 person-years (Grabowski et al., 2017). By age-group, incident rates for 15- to 19-year-olds and 20- to 24-year-olds were 0.5 and 1.2, respectively (Grabowski et al., 2017). In light of these high rates of HIV, researchers have focused on factors that may contribute to HIV risk and protection.
Educational Enrollment
In Rakai, Uganda, the benefits of school enrollment appear to extend beyond intellectual and economic domains: For youth in this region, school enrollment is protective against HIV and, conversely, leaving school is associated with increasing HIV risk (Santelli, Edelstein, et al., 2015; Santelli, Mathur, et al., 2015). Drawing on the Rakai Community Cohort Study (RCCS), which contains data from 7,500 youth (15-24 years old) over 20 years, Santelli, Edelstein, et al. (2015) and Santelli, Mathur, et al. (2015) found that school enrollment was a strong protective factor against HIV infection. Specifically, youth who left school, as compared to enrolled peers, had higher HIV incidence, increased prevalence of sexual experience, increased alcohol use, and decreased condom use. For young men, school leaving was also associated with increases in multiple partners and sexual concurrency.
Mobility
Since the early phases of the epidemic to the current moment, migration has been understood as a critical element in the spread of HIV throughout sub-Saharan Africa (Barongo et al., 1992; Mann et al., 1986). More recently, researchers have recognized both migration and short-term travel as potential pathways to increased risk for HIV infection and transmission (Cassels, Manhart, Jenness, & Morris, 2013; Schuyler et al., 2015). Consistent with this literature, we use the term mobility in this article to encompass both long-term migration and short-term travel. For youth in Rakai, mobility has been linked with HIV risk behaviors: Mobile youth reported increased rates of alcohol use, sexual experience, sexual concurrency, and inconsistent condom use (Schuyler et al., 2015). Furthermore, recent migrants to the region had higher HIV incidence rates than long-term residents (Olawore et al., 2018).
Risky Occupations
The search for work is one of the main drivers of youth mobility in Uganda (Nyanzi, Nyanzi, Wolff, & Whitworth, 2005; Olawore et al., 2018; Schuyler et al., 2015) and across Eastern Africa (Boerma et al., 2002; Camlin, Kwena, Dworkin, Cohen, & Bukusi, 2014). A long line of previous literature from the region has shown that people involved in certain professions are at higher risk for HIV infection (Kagaayi et al., 2014; Ntozi & Kirunga, 1997; Serwadda et al., 1992). Occupations such as motorcycle taxi drivers (locally called boda-boda cyclists), truck drivers, traders, police and security officers, and those in hospitality service industries (bar, restaurant, and hotel workers) place people at greater risk for HIV infection. In contrast, low-risk occupations include jobs such as engineering and mechanics, subsistence farming, and government employees like health workers or educators. For the purposes of this article, we refer to these two groups as risky occupations and low-risk occupations.
Life History Theoretical Framework
A recent special issue of Health Education & Behavior brought critical attention to the importance of qualitative inquiry for understanding and remediating the complex processes that may contribute to increased health equity (Griffith, Shelton, & Kegler, 2017). It is well documented that complex structural factors play a critical role in the HIV epidemic (e.g., Blankenship, Friedman, Dworkin, & Mantell, 2006; Dworkin & Ehrhardt, 2007; Underwood, Skinner, Osman, & Schwandt, 2011) and that the epidemic is unequally distributed, with people in sub-Saharan Africa—and specifically young women—bearing the brunt of the burden (Dellar, Dlamini, & Karim, 2015; Pettifor et al., 2005; Santelli, Edelstein, et al., 2015).
It is through narratives that people interpret and share their multilayered and diverse lived experiences (Bruner, 1991; Daiute, 2011; Wertsch, 1997). Thus, through systematic analysis of life history narratives, researchers can gain insights into the multilayered features of participants’ lives, which may relate to the complex array of HIV-protective and risky behaviors. In other African contexts, the life history approach has been used to examine HIV risk and protective factors for older adults (Mojola, Williams, Angotti, & Gómez-Olivé, 2015) and for orphaned Kenyan schoolgirls (Mojola, 2011). In the current study, the analysis of life history narratives coupled with the ethnographic case control (ECC) design shows the developmental relationship between leaving school, mobility, risky occupations, and HIV for youth in sub-Saharan Africa.
Method
Ethical Considerations
Approvals for the current analysis were obtained from Research and Ethics Committee of the Uganda Virus Research Institute, the Uganda National Council for Science and Technology, and the Institutional Review Boards at Columbia University and Johns Hopkins University, and Western Institutional Review Board in the United States.
Data Sources
The demographic data used in the current analyses were drawn from the RCCS. The RCCS, which began in 1994 and continues to date, records demographics and sexual and reproductive health responses from an open cohort of residents 15 to 49 years of age from 50 communities in the Rakai district of southwestern Uganda. The RCCS research design and procedures have been described in detail elsewhere (Grabowski et al., 2017; Wawer et al., 1999). In brief, during survey rounds, which occur every 12 to 18 months, people aged 15 to 49 years in the RCCS communities consent, and minors (ages 15-17 years) assent, along with parent or guardian consent for participation. Participants (n ~ 14,000 per year) then complete a private face-to-face interview conducted by same-sex interviewers fluent in the local language and are asked to provide biological specimens for HIV and STI (sexually transmitted infection) testing. Among all residents in the study communities, RCCS achieves over 85% coverage, with 99% of consenting participants responding to the complete questionnaire and over 90% agreeing to specimen collection. In addition, all members of the study communities receive voluntary HIV prevention education, counseling, and testing. We conducted an analysis of qualitative interviews from participants (N = 60) who were purposively sampled from the RCCS.
Sampling
The sampling approach and the interview procedure have been described in more detail elsewhere (Higgins et al., 2014). In brief, the participants were selected using an ECC approach for biomedically confirmed HIV-positive (cases) and HIV-negative (controls) young men and women aged 15 to 24 years. HIV-positive participants who had seroconverted within the past year (cases) were identified and then matched with HIV-negative participants (controls) based on gender, marital status, and community location (see Table 1 for detailed breakdown of participants by their demographic characteristics). This approach allowed for a comparison of HIV-positive and HIV-negative narratives, while controlling for other potentially influential variables such as gender, marital status, and location. A team of trained qualitative researchers, blind to participants’ HIV status, conducted life history interviews in Luganda, the most widely used language in the district; transcribed the interviews; and then translated the transcripts into English.
Rakai Community Cohort Youth Life History Respondent Sampling Frame (N = 60).
Narrative Analysis
Life history narratives have long been used in social science research (Bertaux, 1981) and are especially appropriate when working with vulnerable populations such as youth (McLean, Wood, & Breen, 2013) and people living with HIV (Sprague, Scanlon, & Pantalone, 2017; Sprague, Scanlon, Radhakrishnan, & Pantalone, 2016). It is through narratives that people make sense of, and construct, their worlds (Bruner, 1986). Therefore, through systematic analysis of these narrative constructions, researchers can gain insights into participants’ lives and, specific to the current study, the complex array of HIV protective and risky behaviors. In the current analysis, we focused on the plot elements of youth narratives to parse the complex dynamics and relations between leaving school, mobility, risky occupations, and HIV.
A plot analysis is one technique for systematically studying life stories and identifying patterns across multiple participants (Daiute, 2014). Interviews are one type of narrative organized by the process of asking and answering questions (Mishler, 1991) and therefore are well suited for a plot analysis (Daiute, 2010, 2014; Messina, 2014; Murray, 2017). Narratives often follow a general arc and contain specific components that together comprise the plot: They begin with an initiating action, followed by a complicating action or actions, followed by a climax, and they conclude with subsequent resolution strategies (Daiute, 2010, 2014).
To answer the research question of how youth made the transition from school to their lives beyond, we selected a specific question of the semistructured life history interview for analysis: “What were your goals regarding school?” The question about goals for schooling was selected for analysis because it prompted participants to tell stories about their educational journeys. We envisioned that an analysis of these stories would illuminate the relationship between school enrollment, HIV risks, and youths’ transition to adulthood. We present one example of a fully coded narrative to provide a sense of the overall coding process. This will be followed by examples from different youth that demonstrate each of the plot elements and finally a table with the seven different resolution strategies gathered across the 60 narratives. Our approach ultimately yielded insights into the relationship between school enrollment, mobility, risky occupations, and HIV status.
Crispus’s Example
Here we present the following response to the question “What were your goals regarding school?” from a married, 23-year-old, HIV-positive young man, whom we have given the pseudonym Crispus: At first I thought I would be able to complete school and become a qualified teacher. But unfortunately my parents died when I was still young. I cannot remember the year my father died. My mother died in 1997. By the time my mother died I was in primary three . . . My maternal aunt wanted me to continue schooling but I had no money. Our father had left us a house where we could collect money but we were quite many and the money was so small to cater for us all. I had no one to go to for education money. By the time school leaving examinations results were brought: I had left my maternal aunt’s home in Bakka and I had gone to stay with my elder brother. This one introduced me to the coffee business. I started trading in coffee and gave up with studies. I performed so well in the coffee business and up until now I am still in the coffee business. After five years I decided to come back to our family house which our father had built for us . . . From that time I started life as a coffee trader and that is how I earn money. From that time I admired getting a wife because my business could not progress properly when I am alone. I looked for a woman and I got one.
Initiating action
An initiating action is what starts a story. As noted above, the young man’s educational story began in response to the question “What were your goals regarding schooling?” Crispus initiated the action by explaining that he “thought [he] would be able to complete school and become a qualified teacher.”
Complicating actions
After initiating their stories, youth then described obstacles they faced, or complicating actions. A narrator may include numerous complicating actions in one narrative. Crispus, like many participants, described the death of his parents as one complicating action that then led to his struggle to pay for his education, which was coded as another complicating action.
Climax
A climax is a critical element of any narrative. Climaxes often involve some type of shift or change, which are frequently the high point of a story, though in some narratives the climax is actually a frustrating moment or low point. Each narrative has one climax. In the current corpus of educational journey narratives, we identified five distinct types, including aspirations, frustrations, physical struggles such as illness, interpersonal struggles such as negotiating with an abusive parent or partner, and interpersonal successes such as when a friend or family member helped a narrator. Crispus’s narrative reaches a climax, which was coded as an interpersonal success, when he leaves his “maternal aunt’s home in Bakka” and goes to stay with his elder brother who then introduces him to the coffee business. This narrative move creates the opportunity for Crispus to resolve the narrative, and he does so by entering into the coffee-trading business.
Resolution strategies
Resolution strategies are how a narrator wraps up their story. How did they solve, or resolve, the complicating actions? Crispus like many other youth used multiple resolution strategies. The resolution strategies used by Crispus included finding an occupation, specifically trading, which was among professions we categorized as a risky occupation; returning to his family home; and getting married. In Crispus’s case, as we will show with many other HIV-positive youth, mobility was related to the risky occupation as well as returning home, and we suggest that this is a potential pathway that increased risk of HIV infection.
Interrater Reliability
The lead author randomly selected 20% of the narratives and established interrater reliability with the second author, with an average Cohen’s k across codes = 0.80, p < .0001 (Landis & Kock, 1977). The lead author coded the remaining interviews. The results below focus on the plot analysis resolution strategies for HIV-positive and HIV-negative participants.
Results
Our analysis focused on the transition from schooling to life beyond school for youth in Rakai and how this transition relates to potential risk for HIV infection. Each element of the plot analysis provides necessary context for understanding this youth transition. We will devote most attention to the resolution strategies as these appear to be critical for understanding the relationship between leaving school and risk for HIV infection.
Initiating Actions
As might be expected when asked about their goals relating to school, the majority of youth (33) focused their initiating actions on their goals for schooling while the remaining 27 youth focused on careers. The initiating actions did not differ meaningfully by HIV status. The most common career goals noted in order of times mentioned were to become a teacher, health worker, or mechanic, all three of which are categorized as low-risk occupations. Unfortunately, for the vast majority of interviewed youth, complicating actions led them to reevaluate their education and career goals. One excerpt from an HIV-positive married mother with one child, with the pseudonym Agnes, who was 21 years at the time of the interview, succinctly illustrates this pattern:
You told me that when still a young person you wanted to become a health worker, how has this goal regarding work changed over time?
I did not achieve what I wanted, my goal failed and I decided to work in a saloon.
Here the interviewer reiterates what Agnes had described as her career goal earlier in the interview. While Agnes had hoped to enter the low-risk profession of a health worker, she was not able to achieve her goal and instead turned to working in a saloon, a risky profession.
Complicating Actions
For Agnes, and many other participants, the most frequent complicating action was lack of money for school, with 51 out of 60 participants noting this as a challenge. The complicating actions did not differ by gender, marital status, or HIV status. The second most frequent complicating action, which often co-occurred with financial challenges, was the death of one or both parents, with 22 out of 60 participants noting this complicating action. Only five participants noted academic challenges as a complicating action. The complicating actions did not differ meaningfully by HIV status.
Climaxes
The majority of climaxes focused on youths’ frustrations, 20, while the second most common climax, 18, communicated their aspirations. Ten climaxes focused on interpersonal struggles, six climaxes focused on interpersonal successes, and six focused on physical struggles. The climaxes did not differ meaningfully by HIV status.
Resolution Strategies
Overall, we coded for seven types of resolution strategies including low-risk occupations, risky occupations, ending schooling, moving, returning home, continuing education, and getting married. Table 2 presents the resolution codes with excerpted examples.
Rakai Youth Resolution Codes and Example Quotations.
While HIV-positive and HIV-negative youth were equally likely to resolve their narratives by securing occupations, double the number of HIV-positive compared with HIV-negative youth engaged in risky occupations as shown in Table 3. Furthermore, more than twice as many HIV-positive youth, as compared with HIV-negative, used moving as a resolution strategy. As noted previously, youth often used more than one resolution strategy.
Rakai Youth Resolution Codes Counts.
HIV-Positive Youth and a Pattern of Co-Occurring Resolution Strategies
Many HIV-positive youth used a similar pattern of resolution strategies: They left school, moved in search of work, entered risky occupations, and eventually returned to their home village. Of the nine HIV-positive youth engaged in risky occupations, six also used moving as a resolution. In contrast, only two HIV-negative youth resolved their narratives with risky occupations and moving. Furthermore, of these six HIV-positive youth, five concluded their narratives by returning to their home village.
Our analysis sheds light on a specific developmental trajectory that was much more common for HIV-positive as compared to HIV-negative youth. We theorize that this trajectory exposes these youth to multiple points of increased HIV risk and potential infection as they transition to adulthood. They become increasingly vulnerable as they move to new locations and again as they work in risky jobs. The following two excerpts from HIV-positive narratives show how after youth leave school, moving, risky occupations, and returning home are deeply interrelated. A previously married HIV-positive young man, who we have given the pseudonym Marcus, explained, I started cultivation and selling of “matooke.” Later on I started operating a boda boda. I used to do them from here in Kasasa. I also went to Kampala . . . I got some problems. I was not earning like I had expected to earn so I decided to come back to the village.
This pattern of moving, engaging in risky occupations, and then returning to a home village was also present in the narrative of a young man who had never been married, who we have given the pseudonym Gerald: I did not stay long here in the village. I went to the city. I stayed in Kampala working for about a month and then I came back to the village. Then I went back to Kampala the second time but after seeing that things were not working out well I came back to the village. I first went to Kampala four years ago. I was selling fried groundnuts in Kampala. It was in the same place and it was within the same year but I saw that things were not working out well, I decided to come back to the village.
The following narrative from a married, HIV-positive young woman, with the pseudonym Mariah, shows a similar pattern of mobility, risky occupation, and returning home: When I left school, I stayed at home then I got a job and worked there for some time. I was working in Mutukula [at the Tanzanian boarder] in a restaurant but I later returned home. I was there for three months . . . I was not being paid. So I left . . . I stayed at home for a short time, then got married.
Across these narratives, youth are communicating the challenges of transitioning from school to the workforce, and those who are HIV-positive frequently take a similar route of moving to find risky occupations and eventually returning home.
Discussion and Significance
Our analysis of youth narratives showed that after leaving school, youth who eventually become HIV-positive moved more frequently than youth who remain HIV-negative, and this movement typically co-occurred with engagement in risky occupations. Youth who moved and engaged in risky occupations were also more likely to eventually return to their home village. Previous research in the region has shown that for youth, school enrollment is the strongest protective factor against HIV infection (Santelli, Edelstein, et al., 2015; Santelli, Mathur, et al., 2015). A second line of inquiry has indicated that mobility is associated with HIV risk behaviors and HIV incidence (Olawore et al., 2018; Schuyler et al., 2015). A third line of inquiry has demonstrated that there are certain jobs that can put people at increased risk for HIV infection (Kagaayi et al., 2014; Ntozi & Kirunga, 1997; Serwadda et al., 1992). The ECC design combined with our qualitative analysis demonstrates how these HIV risk factors are related along the path to adulthood for youth in Uganda.
Across narratives, youth expressed an overwhelming desire to complete secondary school and obtain low-risk jobs such as being a teacher, health worker, or mechanic. Youth were acutely aware that education was a critical means for social and economic mobility, or as one participant put it, education is “the light.” However, few youth were able to achieve their educational goals and enter their careers of choice. This finding is consistent with literature on education and career aspirations across East Africa (Human Rights Watch, 2017; Mojola, 2014).
Youth in the region face increased risk for HIV due to structural factors, including an economic landscape that leaves them with few options for work. Consequently, youth are forced to move in search of job opportunities, and it is in this search that we suspect they are most vulnerable, and all too frequently become infected with HIV. The complicating actions detailed by participating youth corroborate previous findings from Uganda and indicate that despite establishing free primary education in 1997—known as Universal Primary Education (UPE)—poverty is the main driver of school attrition (Mpyangu, Ochen, Onyango, & Lubaale, 2014). Notwithstanding UPE and similar policies in countries across East Africa, including Kenya, Tanzania, and Rwanda, many youth still struggle with the hidden costs of education (Deininger, 2003; Grogan, 2009; Williams, Abbott, & Mupenzi, 2015). Unable to afford these hidden costs that include purchasing basic hygiene necessities, school supplies, and uniforms, the majority of youth in our study left school in search of occupations. After leaving school, youth faced challenging decisions in their transition to adulthood.
A key element of autonomy is earning an income: For youth across the globe, transitioning to adulthood means finding a way to support oneself financially. For Ugandan youth, this frequently involves moving in search of an occupation. Our results show that after moving, youth often found risky jobs that were generally undesirable in the long term. Thus, youth frequently worked these risky jobs for a short period of time before moving again, only to eventually return to their home villages. It is in this transitional period, while searching for and ultimately engaging in risky occupations, and again in their transition back to their home village that we theorize youth are particularly vulnerable to HIV infection.
While we compared the narratives from youth who had become HIV-positive in the past year with HIV-negative youth, we were not able to determine if HIV-positive participants seroconverted while engaged in risky jobs or when they returned to their home village for two reasons: First, HIV testing in the RCCS is conducted only every 12 to 18 months. Second, given the nature of life history narratives and the details provided by youth, it is not possible to determine the exact dates for each of the plot elements.
One strength of this article was the ECC design, which allowed for a controlled comparison of narratives from youth who shared similar sociodemographic characteristics but differed in terms of their HIV status. A second strength was our analytic approach, adapted from developmental and life course research (Daiute, 2010; 2014; Messina, 2014; Murray, 2017), which allowed us to link previously disparate lines of study on adolescent life transitions and HIV. As suggested in a previous special issue of this article, qualitative approaches “offer insight into the complex, layered, and multifaceted social, economic, and political factors that shape health disparities” (Griffith et al., 2017, p. 673). Using the ECC and a fine-grained qualitative analysis, we show that for youth in Uganda, leaving school, mobility, and risky occupations are part of a transition to adulthood that often involves increased risk for HIV infection.
Implications
Considering the mechanisms and pathways to HIV infection in Uganda, there is a continued need to understand how educational enrollment and subsequent mobility lead to HIV infection. Despite the UPE and similar policies in countries across East Africa including Kenya, Tanzania, and Rwanda, many youth still struggle with hidden costs of education (Deininger, 2003; Grogan, 2009; Williams et al., 2015). It is in the transition to adulthood, when youth leave school and search for occupations, that they are most vulnerable to HIV infection. Our current analysis suggests that this vulnerability occurs as youth move to new locations and enter risky jobs.
To mitigate the risk for HIV infection, future research and practice-based initiatives must focus on understanding these risks during transitional moments and designing health, education, and economic interventions to support youth in their transition to adulthood. One successful approach for orphaned youth in Uganda involved financial literacy training, mentorship, and cash transfers, which resulted in lower dropout rates and higher HIV knowledge scores (Ssewamala et al., 2018). While these findings were promising, enrolling mobile youth may require integrating new approaches, such as cell phone–based interventions. Though data were limited for sub-Saharan Africa, a recent systematic review of digital interventions indicated that they were generally feasible and improved both prevention and care of HIV/STIs, especially for hard-to-reach populations (Daher et al., 2017). Youth in sub-Saharan Africa bear the brunt of the global HIV epidemic; given the structural factors that force youth to move in search of work, tailored digital interventions may present a tool for targeting mobile youth. However, designing interventions for highly mobile youth will necessarily present unique challenges.
Footnotes
Acknowledgements
We would like to thank the participants who shared their stories and the Rakai Health Sciences researchers who conducted the interviews. We like to thank Dr. Susie Hoffman and the members of the HIV Center Manuscript workshop for their contributions to the development of this manuscript. We would like to thank Dr. Colette Daiute for her comments on an earlier version of the manuscript and Esther Spindler for the copyedits she provided.
Authors’ Note
The content is solely the responsibility of the authors and does not necessarily represent the official views of National Institute of Mental Health or the National Institutes of Health or Columbia University.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was supported by Award No. R01HD091003 (Principal Investigator, John Santelli, MD, MPH) from the National Institute of Child Health and Human Development. In addition, Philip Kreniske’s contribution was supported by Award No. T32 MH019139 (Principal Investigator, Theodorus Sandfort, PhD) from the National Institute of Mental Health and a grant from the National Institute of Mental Health to the HIV Center for Clinical and Behavioral Studies at NY State Psychiatric Institute and Columbia University (P30-MH43520; Principal Investigator, Robert Remien, PhD).
