Individual and Parental Risk Factors for Sexual Exploitation Among High-Risk Youth in Uganda

Abstract

This study examined risk factors to determine associations with commercial sexual exploitation of children and youth (CSEC) in a convenience sample of adolescents living in the slums in Kampala, Uganda. Individual-level factors included demographic, adverse experiences (ever living on the streets; victim of dating violence, parental abuse, or rape), and behavioral risk (social media, alcohol use, age at first intercourse). Parental-risk factors included parent alcohol use and approval attitudes toward youth sex. Analyses included those who self-reported sexually active adolescents (n = 593) of whom 39% reported CSEC history. CSEC was significantly associated with being female (odds ratio [OR] = 6.85, 95% confidence interval (CI) = [4.22, 11.12]), living on the streets (OR = 2.68; 95% CI = [1.65, 4.36]), using social media (OR = 1.48; 95% CI = [0.94, 2.35]), being a victim of physical dating violence (OR = 1.74; 95% CI = [1.08, 2.80]), and ever being raped (OR = 4.03; 95% CI = [2.51, 6.47]). Further analyses suggested differential risk associates among females and males. This study contributes to our knowledge of risk factors for CSEC among adolescents living in high-risk circumstances in low-resource countries and suggests that preventive efforts should prioritize adolescents with a history of living on the streets who engage in social media, use alcohol, and have a history of trauma.

Keywords

prostitution/sex work violence exposure anything related to child abuse child abuse

Introduction

Commercial sexual exploitation of children and youth (CSEC) is a growing human rights problem throughout the world. CSEC is defined as when a child, younger than 18 years of age, is recruited, enticed, or harbored into a lifestyle whereby commercial sex acts are induced by force, fraud, or coercion, in exchange for anything of value (Trafficking Victims Protection Reauthorization Act, 2013). Studies indicate that victims of CSEC are at higher risk for substance use (Cole, Sprang, Lee, & Cohen, 2016; Miller et al., 2011), suicide (Clarke, Clarke, Roe-Sepowitz, & Fey, 2012), posttraumatic stress disorder (Lederer & Wetzel, 2014), and a broad array of health problems, including HIV (Swahn, Culbreth, Salazar, Kasirye, & Seeley, 2016).

Determining accurate global prevalence estimates of CSEC is inherently challenging due to the covert nature of these acts, as well as the lack of uniformity in classification and national reporting requirements (Institute of Medicine and National Research Council, 2013). In the United States, commercial CSEC is considered a form of child sexual abuse by state statutes (Murray, Nguyen, & Cohen, 2014). Conversely, in low-resource countries, where economic and legislative factors affect government and institutional frameworks’ ability to sufficiently protect children, acts of exploitation may be classified as transactional sex (United Nations Office on Drugs and Crime [UNODC], 2016). Notably, prevalence rates have been found to be higher in low-resource countries than in high-resource countries, likely due to factors that elevate risk including, but not limited to, poverty, political unrest, migration, domestic violence, and poor education (“The Advocates for Human Rights,” 2017), as well as within cultures with more entrenched sexism, racism, and classism (Farley, 2006). The elimination of CSEC has been delineated by the United Nations Sustainable Development Goals as an imperative issue of focus to promote just, peaceful, and inclusive societies (http://www.un.org/sustainabledevelopment/peace-justice/).

Inconsistent CSEC definitions and screening processes in research has also limited the understanding of the magnitude of CSEC worldwide. Best practices for screening and assessment questions pertaining to the research of sensitive topics indicate the importance of using a behaviorally specific approach, an approach that clearly describes experiences without labeling them as abuse (e.g., Fricker, Smith, Davis, & Hanson, 2003; Tourangeau, 2000; Tourangeau & Smith, 1996). For example, Fricker and colleagues (2003) found that when examining history of child sexual abuse, respondents endorsed behaviorally specific questions significantly more (32%) than labeled questions (9%), an increase of nearly 28%. Clearly, research that focuses on the incidence of CSEC and identification of risk factors, especially in low-resource countries, is warranted to better understand how to protect adolescents from this type of traumatic victimization. The purpose of this study is to build upon the existing literature by examining the prevalence of CSEC among adolescents living in the slums of Kampala, Uganda, as well as to identify individual and parental risk factors for this type of victimization. The current state of the field pertaining to CSEC risk factors from a global perspective is summarized in the next section.

Risk Factors for CSEC

A burgeoning body of literature has focused on factors that contribute to CSEC in low- and high-resource countries (e.g., Adjei & Saewyc, 2017; Cole et al., 2016; Fielding-Miller, Dunkle, Cooper, Windle, & Hadley, 2016; Hossain, Zimmerman, Abas, Light, & Watts, 2010; Ranganathan et al., 2017; Reid & Piquero, 2014; Svedin & Priebe, 2007; Swahn et al., 2016; Williams, Binagwaho, & Betancourt, 2012). These studies suggest that CSEC victimization is dependent on a convergence of socioecological factors, including individual- and family-level factors, as further delineated below.

Individual Demographic Factors and CSEC

With regard to individual demographic risk factors, women and girls are overrepresented as CSEC victims (UNODC, 2016). However, emerging research suggest a more complicated picture, with greatly increased incidence rates reported among U.S. minority males with a history of delinquency and adverse child experiences (Reid & Piquero, 2014). Furthermore, research from a school sample in rural western Canada documented comparable prevalence rates of CSEC among males and females (Homma, Nicholson, & Saewyc, 2012). Educational attainment has also emerged as an important predictor of CSEC victimization in the United States and globally, though findings have been mixed. In the United States, low educational attainment increased CSEC risk (Reid & Piquero, 2014). In contrast, research in South Africa indicates that women with no education were less likely to experience sexual violence than those with higher levels of education, as higher education and emerging women’s rights has been a threat to patriarchal mores and increased violence exposure risk (Bhana, de Lange, & Mitchell, 2009; Jewkes & Abrahams, 2002). Increased risk of SEC among the more educated in South Africa may also be attributed to the relatively high rates of sexual abuse perpetrated by teachers, with one study reporting that 33% of children identified teachers as the perpetrators of their abuse (Jewkes, Levin, Mbananga, & Bradshaw, 2002).

Individual Adverse Experiences and CSEC

Adverse experiences, such as living on the streets and a history of violence or abuse, can greatly increase the risk of CSEC victimization among youth. Studies documenting the associations between CSEC and a history of running away/street living have indicated striking increases in risk for CSEC (Bigelsen, 2013; Greene, Ennett, & Ringwalt, 1999), likely due to the inability of youth living on the streets to meet their hierarchy of needs (i.e., food, shelter). Furthermore, several studies have found that children with a history of childhood sexual abuse or rape are at an increased risk of CSEC (Adjei & Saewyc, 2017; Lalor & McElvaney, 2010; Stoltz et al., 2007; Svedin & Priebe, 2007; Swahn et al., 2016; Tjaden & Thoennes, 2000; Walker, 2013). For example, Walker (2013) found that children who had a history of childhood sexual abuse are 28 times more likely to be arrested for transactional sex at some point in their lives than children with no history of childhood sexual abuse. Physical abuse also increases CSEC risk. Varma, Gillespie, McCracken, and Greenbaum (2015) found that 44% of CSEC youth had a history of physical abuse. Dating violence as an independent predictor of CSEC has been less explored in the literature and more is needed to understand how this specific type of violence exposure might increase risk. The most relevant published work to date is a qualitative study examining risk trajectory in low-income African American girls. This study demonstrated that girls who resided in communities that place them at increased risk of sexual exploitation also experienced chronic exposure to physical aggression in peer relationships (Kruger et al., 2013). Thus, it is imperative to assess the intersections between exposure to multiple forms of violence, or cumulative trauma, and sexual exploitation.

Adolescent Risk Behaviors and CSEC

Age of first sexual intercourse and alcohol use have also been identified as behaviors predictive of CSEC. For instance, Reid and Piquero (2014) found that higher age of first sexual encounter reduced the likelihood of initiation of CSEC. Similarly, in a study involving high school youth in Sweden, Svedin and Priebe (2007) found that the mean age of first sexual intercourse was lower (14.4 years) in youth who reported transactional sex compared with those who reported no transactional sex history (15.6 years). Moreover, results from this same study indicated higher levels of alcohol consumption at an earlier age for youth who had been exploited. Swahn and colleagues (2016) also documented a statistically significant association between alcohol use and engaging in sex work among youth in Uganda (Swahn et al., 2016). Such behavioral patterns can serve as gateways to sexual exploitation that place children on a continuum of risk (Estes & Weiner, 2005).

A less explored, yet increasingly important youth behavior that may serve as a risk factor for CSEC, is youth engagement with social media. Global estimates of social media use worldwide suggest that over 2 billion people use social media (https://www.statista.com/statistics/278,414/number-of-worldwide-social-network-users/), and recent Pew Research Center estimates suggest that 81% of youth ages 12 to 18 years use social media (http://www.pewinternet.org/2016/11/11/social-media-update-2016/). Technology and social media use can increase youth risk and exposure to sexual messaging, contact with strangers, and pornography (Livingstone & Smith, 2014). Online platforms are increasingly being used to sell, recruit, and market CSEC victims, leading to opportunities for exploitation of a greater number of victims (Latonero, 2011; Wells, Mitchell, & Ji, 2012). Importantly, one recent report from the United States found that 80% of CSEC investigations included technology (Mitchell & Boyd, 2014). In 2012, the UNODC estimated that ~76% of transactions for commercial sex exploitation of underage girls were initiated on the Internet. In Uganda, there was an increasing trend of young girls taking nude self-photos in exchange for money and using these images to recruit clientele via the Internet (Uganda Youth Development Link [UYDEL], 2011). Thus, greater understanding of how youth social media behavior affects CSEC risk is important to explore.

Parental Risk and CSEC

Unstable family factors, such as parental alcohol and substance abuse and parental attitudes toward sex, heighten risks for CSEC. Parental substance and alcohol abuse may increase youths’ susceptibility to CSEC, rendering parents incapable of adequate guardianship, including limited ability to effectively monitor youth (Reid, 2011; Smith, Vardman, & Snow, 2009; Williamson & Prior, 2009). Adjei and Saewyc (2017) examined sexual exploitation of adolescents in sub-Saharan African and found that youth with low levels of parental monitoring had a higher prevalence of CSEC compared with youth with higher levels of parental monitoring. Furthermore, parents with substance abuse dependence may exploit their children as a way to support their addiction. Parent attitudes toward sex may also play an important role in CSEC risk, as this has been an important determinant for youth sexual risk factors including pregnancy, sexually transmitted infection risk, and promiscuity (Sorbring, Hallberg, Bohlin, & Skoog, 2015; Tobey, Hillman, Anagurthi, & Somers, 2011; Whitaker & Miller, 2000). Clearly, further work is necessary to understand how parental factors are associated with youth risk for CSEC victimization.

Purpose of Current Study

CSEC is pervasive in both high- and low-income countries. While the literature on risk factors for CSEC is expanding, especially in high-resource countries where there is stronger infrastructure for data collection, less research has been conducted in low-resource countries that can directly identify unique risk factors and better inform prevention strategies for youth residing in these settings. The current study builds upon the work by Swahn and colleagues (2016), which is one of the few population-based studies involving CSEC youth in the developing country of Uganda, and expands on the prior work in several important ways. First, a behaviorally specific assessment question for CSEC is used to determine the prevalence of CSEC among Ugandan youth. The main outcome of interest in the current study is the youth reporting a history of sexual exploitation more broadly, compared with the previous study of examining youth specifically reporting commercial sex work engagement (Swahn et al., 2016). Second, risk factors at two levels of the adolescent social ecology were examined, including risk at the individual and parental level. Third, risk is examined specifically for females and males separately, with the goal of expanding on the sparse information, particularly with respect to male CSEC risk.

Several hypotheses were posited:

Hypothesis 1: Among youth living in the slums, females, and those youth with less education (based on the work of Swahn et al., 2016), would report higher rates of SEC.

Hypothesis 2: Youth with prior adverse experiences, including those who have ever lived on the streets,or who have a, history of rape, history of parental abuse or dating violence, would be at a greater risk for reporting SEC.

Hypothesis 3: Youth reporting engagement in high-risk behaviors, including earlier age of first sexual experience and alcohol use, as well as higher levels of social media use, would be at a greater risk for reporting SEC.

Hypothesis 4: Youth with parents who report a casual attitude toward youth sex and report high levels of alcohol use would report higher levels of SEC. As an exploratory hypothesis, the risk factors among males and females who report SEC will be examined.

Method

Setting, Sample, and Data Collection

This article is based on data from the “Kampala Youth Survey,” a cross-sectional study conducted in 2014 among youth living in the slums of Kampala. The main purpose of this survey was to assess high-risk behaviors among youth, including sexual risk behaviors, alcohol use, and violence victimization and perpetration. The survey also collected self-reported health outcomes, such as HIV, sexually transmitted infections, and mental health outcomes. Survey measures were based on previously validated instruments, including questions obtained from the Global School-based Student Health Survey (GSHS), Kampala Youth Survey 201, Monitoring Alcohol Marketing Practices in Africa (MAMPA) 2012 Questionnaire, Alcohol Use Disorders Identification Test (AUDIT) Questionnaire, “Cut-Down, Annoyed, Guilty, and Eye-Opener” (CAGE) Questionnaire, Imppacs, AIDS Indicator Survey, and the Demographic Health Survey.

The sample consisted of a convenience sample of youth living in an urban area who were participating in UYDEL drop-in centers. UYDEL centers provided vocational training, mental health counseling, HIV/STI testing, and other support services for youth. The youth were recruited from one of the six drop-in centers. The only inclusion criteria included youth who were 12 to 18 years of age at the time of the survey. The youth who agreed to participate received a small snack for incentive.

During the data collection period, 1,628 youth were asked to participate in the study. Among the 1,628, 131 declined. The total participation rate was 92% (1,497 surveys collected). Approximately, 320 surveys were erased due to technical problems with the offline server connection, yielding a total number of 1,134 surveys of youth (44% boys, 56% girls).

Social workers and peer educators were trained on the survey methodology and conducted the surveys using in-person interviews due to anticipated literacy issues with recruited participants. The social workers and peer educators were employed by UYDEL and have experience working with youth in UYDEL and surrounding communities. All survey questions were read in Luganda (local Uganda language) if necessary. Participants were fully informed of the study and read the consent forms to participate in the study. For youth who could not read, the consent forms were read to them. All youth participants gave verbal consent to participate in the survey. Youth who “cater to their own livelihood” are considered emancipated under Ugandan law and could give consent to participate without parental consent. Institutional review board (IRB) approvals were obtained from both Georgia State University and the Uganda National Council for Science and Technology to perform this study in Kampala, Uganda.

Measures

Sexual exploitation was the main outcome of interest and was assessed using the question, “Have you ever gotten money, food, or other things for having (in exchange for) sexual intercourse with someone?” Youth could answer yes or no. Demographic characteristics were also assessed, including gender and education. Predictors for SEC were ascertained from the literature and include individual adverse experiences or prior trauma including ever living on the streets, parental abuse of youth, ever being raped, and being a victim of dating violence. In addition, individual risk factors were examined including youth social media use, alcohol use, and age at first sexual intercourse. Two factors at the parental level were also examined, parental alcohol use and parental attitudes toward youth engagement in sex.

Ever living on the streets was assessed using, “Have you ever lived on the streets with no other place to go?” Parental physical abuse of youth was measured using, “Did your parents ever beat you so hard you had bruises or marks?” Ever being raped was measured using, “Has someone ever raped you or forced you to have sex with him or her?” Being a victim of dating violence was measured using, “In the past year, did your boyfriend/girlfriend hit, slap, or hurt you?” Response options for the previous four questions were dichotomous (yes or no)

Participants were asked about their social media use using, “Do you use any type of Social Media (e.g., Twitter, Facebook)?” Consistent with other published research (Swahn et al., 2016), current alcohol use was measured by the question, “Have you had a drink of alcohol in the past year?” All participants could respond yes or no to social media use and alcohol use. Age at first intercourse was measured using, “How old were you when you had sexual intercourse for the first time?” Participants could answer <12, 13 to 14, 15 to 16, or 17 to 18 years.

Parental alcohol use was measured using, “Did your parents/caretakers drink a lot of alcohol when you were growing up?” All participants could answer yes or no. Parental attitudes toward sex were measured using, “My parents would be upset if they found out that I am having sex.” Youth could answer, agree, neither agree nor disagree, or disagree.

Data Analysis

The analyses were restricted to only sexually active youth (n = 593) to enable comparability of the outcome and reference group. Descriptive statistics among youth who reported a history of sexual exploitation were computed for sociodemographic variables and predictor variables. Bivariate and multivariable logistic regression models were computed to determine statistically significant predictors for reporting sexual exploitation among youth. All analyses were conducted in SAS 9.4 (Cary, NC). Males and females were stratified to compute the prevalence of sexual exploitation, sociodemographic variables, and predictor variables. Chi-square and Fisher Exact Tests were conducted to assess the differences between youth who report a history of sexual exploitation compared with youth who do not report this history among males and females separately.

Results

Among sexually active adolescent participants (n = 593, average age 17 years), 39% of youth (n = 231) reported a history of sexual exploitation (Table 1). The majority of adolescents who reported a SEC history were female (82%). Youth who reported SEC also reported a higher likelihood of ever living on the streets compared with youth reporting no SEC history (40% vs. 23%, respectively). Youth with a history of SEC also reported a higher prevalence of alcohol use (59%), parental alcohol use (59%), and parental physical abuse (41%) compared with youth who did not engage in them. In addition, the prevalence of ever being raped was nearly 4 times higher for youth with history of SEC as compared with those who did not.

Table 1.

Descriptive Statistics Among Adolescents Who Report Sexual Exploitation in the Slums of Kampala, Uganda (n = 593).

	Commercial Sexual ExploitationYes (n = 231) 39.0%	Commercial Sexual ExploitationNo (n = 362) 61.0%
Individual demographic factors
Gender
Male	41 (17.8%)	203 (56.1%)
Female	190 (82.3%)	159 (43.9%)
Level of education
Less than primary	81 (35.7%)	117 (32.6%)
Completed primary	42 (18.5%)	78 (21.7%)
Some Secondary or higher	104 (45.8%)	164 (45.7%)
Individual adverse experiences
Ever lived on the streets
Yes	92 (39.8%)	84 (23.2%)
No	139 (60.2%)	278 (76.8%)
Parental abuse of youth
Yes	94 (40.7%)	139 (38.5%)
No	137 (59.3%)	222 (61.5%)
Ever being raped
Yes	109 (47.2%)	45 (12.4%)
No	122 (52.8%)	317 (87.6%)
Victim of dating physical violence
Yes	76 (32.9%)	69 (19.1%)
No	155 (67.1%)	293 (80.9%)
Adolescent risk behaviors
Social media use
Yes	75 (32.9%)	99 (27.5%)
No	153 (67.1%)	261 (72.5%)
Alcohol use
Yes	136 (58.9%)	160 (44.2%)
No	95 (41.1%)	202 (55.8%)
Age at first sexual intercourse
<12	33 (14.3%)	38 (10.5%)
13-14	77 (33.3%)	99 (27.4%)
15-16	89 (38.5%)	159 (43.9%)
17-18	32 (13.9%)	66 (18.2%)
Parental risk factors
Parental alcohol use
Yes	137 (59.3%)	174 (48.2%)
No	94 (40.7%)	187 (51.8%)
Parents upset about youth having sex
Agree	160 (69.3%)	269 (74.3%)
Neither agree/disagree	46 (19.9%)	56 (15.5%)
Disagree	25 (10.8%)	37 (10.2%)

Table 2 presents the results of bivariate and multivariable associations between predictor variables and SEC. In the multivariable model, SEC was significantly associated with being female (odds ratio [OR] = 6.85; 95% confidence interval (CI) = [4.22, 11.12]), living on the streets (OR = 2.68; 95% CI = [1.65, 4.36]), ever being raped (OR = 4.03; 95% CI = [2.51, 6.47]), being a victim of physical dating violence (OR = 1.74; 95% CI = [1.08, 2.80]), and using social media (OR = 1.48; 95% CI = [0.94, 2.35]).

Table 2.

Bivariate and Multivariable Associations of Adolescents Who Report Sexual Exploitation in the Slums of Kampala, Uganda (S = 593).

	Unadjusted Odds Ratios(95% CI)	Adjusted Odds Ratios(95% CI)
Individual demographic factors
Gender
Male	1.00	1.00
Female	5.92 [3.98, 8.79]	6.85 [4.22, 11.12]
Level of education
Less than primary	1.00	1.00
Completed primary	0.78 [0.49, 1.25]	0.71 [0.40, 1.26]
Secondary or higher	0.92 [0.63, 1.33]	1.13 [0.71, 1.81]
Individual adverse experiences
Ever lived on the streets
Yes	2.18 [1.52, 3.14]	2.68 [1.65, 4.36]
No	1.00	1.00
Parental abuse of youth
Yes	1.10 [0.78, 1.54]	0.70 [0.46, 1.08]
No	1.00	1.00
Ever being raped
Yes	6.29 [4.20, 9.44]	4.03 [2.51, 6.47]
No	1.00	1.00
Victim of dating physical violence
Yes	2.08 [1.42, 3.04]	1.74 [1.08, 2.80]
No	1.00	1.00
Adolescent risk behavior
Social media use
Yes	1.25 [0.87, 1.80]	1.48 [0.94, 2.35]
No	1.00	1.00
Alcohol use
Yes	1.81 [1.29, 2.54]	1.26 [0.81, 1.97]
No	1.00	1.00
Age at first sexual intercourse
<12	1.00	1.00
13-14	0.90 [0.52, 1.56]	0.84 [0.41, 1.71]
15-16	0.65 [0.38, 1.10]	0.94 [0.47, 1.86]
17-18	0.56 [0.30, 1.05]	0.79 [0.35, 1.75]
Parental risk factors
Parental alcohol use
Yes	1.57 [1.12, 2.19]	1.52 [0.98, 2.37]
No	1.00	1.00
Parents upset about youth having sex
Agree	1.00	1.00
Neither agree/disagree	1.38 [0.89, 2.14]	1.33 [0.77, 2.32]
Disagree	1.14 [0.66, 1.96]	0.82 [0.41, 1.65]

Note. Model fit statistics: Akaike information criterion (AIC) = 617.37, Global Wald Test: $χ^{2}$ = 122.90, df = 15, p < .0001. CI = confidence interval. Significance level set at p < 0.05.

Tables 3 and 4 present chi-square statistics for differences of females and males reporting a history of SEC compared with those who did not. For females, statistically significant associations with SEC were found for ever living on the streets ( $χ^{2}$ = 27.25, df = 1, p < .0001), ever being raped ( $χ^{2}$ = 27.04, df = 1, p < .0001), being a victim of dating physical violence ( $χ^{2}$ = 10.21, df = 1, p = .001), social media use ( $χ^{2}$ = 7.85, df = 1, p = .005), alcohol use ( $χ^{2}$ = 22.92, df = 1, p < .0001), parental alcohol use ( $χ^{2}$ = 10.65, df = 1, p = .001), and parental attitudes toward sexual engagement ( $χ^{2}$ = 7.47, df = 2, p = .02). For males, statistically significant associations with SEC were found for previously living on the streets ( $χ^{2}$ = 8.58, df = 1, p = .003) and ever being raped (<.0001).

Table 3.

Descriptive Statistics and Chi-Square Results for Sexual Exploitation Among Sexually Active Adolescent Females in the Slums of Kampala, Uganda (n = 349).

	Females	Chi-Square, df, p Value
	Transactional SexYes (n = 190) 54.4%	Chi-Square, df, p Value
Individual demographic factors
Level of education
Less than primary	69 (36.9%)	1.32, (2), p = .52
Completed primary	36 (19.3%)
Secondary or higher	82 (43.9%)
Individual adverse experiences
Ever lived on the streets
Yes	69 (36.3%)	27.25, (1), p < .0001
No	121 (63.7%)
Parental abuse of youth
Yes	76 (40.0%)	0.07, (1), p = .79
No	114 (60.0%)
Ever being raped
Yes	93 (49.0%)	27.04, (1), p < .0001
No	97 (51.0%)
Victim of dating physical violence
Yes	62 (32.6%)	10.21, (1), p = .001
No	128 (67.4%)
Adolescent risk behaviors
Social media use
Yes	59 (31.6%)	7.85, (1), p = .005
No	128 (68.5%)
Alcohol use
Yes	117 (61.6%)	22.92, (1), p < .0001
No	73 (38.4%)
Age at first sexual intercourse
<12	26 (13.7%)	5.41, (3), p = .14
13-14	65 (34.2%)
15-16	72 (37.9%)
17-18	27 (14.2%)
Parental risk factors
Parental alcohol use
Yes	111 (58.4%)	10.65, (1), p = .001
No	79 (41.6%)
Parents upset about youth having sex
Agree	133 (70.0%)	7.47, (2), p = .02
Neither agree/disagree	36 (19.0%)
Disagree	21 (11.0%)

Note. Significance level set at p < 0.05.

Table 4.

Descriptive Statistics and Chi-Square Results for Sexual Exploitation Among Sexually Active Adolescent Males in the Slums of Kampala, Uganda (n = 244).

	Males	Chi-Square, df, p Value
	Transactional SexYes (n = 41) 16.8%	Chi-Square, df, p Value
Individual demographic factors
Level of education
Less than primary	12 (30.0%)	0.83, (2), p = .66
Completed primary	6 (15.0%)
Secondary or higher	22 (55.0%)
Individual adverse experiences
Ever lived on the streets
Yes	23 (56.1%)	8.58, (1), p = .003
No	18 (43.9%)
Parental abuse of youth
Yes	18 (43.9%)	0.43, (1), p = .51
No	23 (56.1%)
Ever being raped^a
Yes	16 (39.0%)	<.0001
No	25 (61.0%)
Victim of dating physical violence
Yes	14 (34.2%)	3.80, (1), p = .05
No	27 (65.9%)
Adolescent risk behaviors
Social media use
Yes	16 (39.0%)	0.28, (1), p = .59
No	25 (61.0%)
Alcohol use
Yes	19 (46.3%)	0.26, (1), p = .61
No	22 (53.7%)
Age at first sexual intercourse
<12	7 (17.1%)	1.38, (3), p = .71
13-14	12 (29.3%)
15-16	17 (41.5%)
17-18	5 (12.2%)
Parental risk factors
Parental alcohol use
Yes	26 (63.4%)	1.23, (1), p = .27
No	15 (36.6%)
Parents upset about youth having sex
Agree	27 (65.9%)	0.58, (2), p = .75
Neither agree/disagree	10 (24.4%)
Disagree	4 (9.8%)

Fisher’s Exact Test performed due to expected cell counts < 5.

Note. Significance level set at p < 0.05.

Discussion

The SEC is a well-established public health problem (Felner & DuBois, 2017). The goal of this study was to explore rates of sexual exploitation among a sample of youth living in the slums of Kampala, Uganda, and to identify factors at the individual and parental level, based on prior research, that could influence SEC victimization. This study builds upon the initial work completed by Swahn and colleagues (2016), which found that 13.8% of this youth sample reported engagement in transactional sex work and that being female, having less education (never been to school), and a history of being orphaned, raped, and using alcohol led to greater risk of youth engaging in transactional sex.

The current study took a more behaviorally specific approach to the assessment of SEC, given the existing literature supports this assessment format for questions pertaining to sensitive topics (e.g., Fricker et al., 2003; Tourangeau, 2000; Tourangeau & Smith, 1996). More specifically, Swahn et al. (2016) reported that 13% of the study participants responded affirmatively to “Are you currently engaged in commercial sex work?” In contrast, 39% of participants in the current study indicated that they had “gotten money, food, or other things in exchange for having sexual intercourse with someone,” an increase of 33% in affirmative responses, respectively. The question used for this study is framed in a format commensurate with other recently published work examining prevalence of SEC (e.g., Fogel, Martin, Nelson, Thomas, & Porta, 2016) as well as supported by best practices in assessment of trauma endorsed by the National Child Traumatic Stress Network (http://www.nctsn.org/resources/online-research/reading-lists/core-data-set). Thus, a primary contribution of this study is the validation of the use of behaviorally specific questions in screening of SEC among youth living in low-resource countries.

A second major contribution of this article is the further illumination of individual- and parental-level risk factors associated with SEC in a low-resource country. Findings from this study suggested that individual-level risk factors were more relevant to SEC than parental factors for high-risk youth in Uganda. Specifically, being female, as well as a history of adverse experiences, such as ever living on the streets, being raped, and physical dating violence, significantly increased SEC risk. The increased risk among females and youth who have experienced housing instability and prior trauma is commensurate with prior research (e.g., Basson, Rosenblatt, & Haley, 2012; Estes & Weiner, 2001; Hossain et al., 2010; UNODC, 2016). However, this was the first known study to examine dating violence history as a risk factor, with only one other study documenting the relationship between SEC risk and a history of peer-to-peer violence (Kruger et al., 2013). Further work is warranted to better clarify whether the experiences of dating violence occurred as part of the exploitation acts or whether violence in youth romantic relationships serves as a precursor for youth involvement in SEC.

Importantly, of the adolescent risk behaviors examined, only social media use emerged as a significant predictor in the multivariable model. This is unexpected, given that prior research has consistently documented that the risk behaviors of early sexual intercourse and alcohol use are important contributors to SEC risk (Estes & Weiner, 2001; Reid & Piquero, 2014; Svedin & Priebe, 2007). The association between social media use and SEC has not been comprehensively explored in prior research, although emerging research in the United States found that an average of 9% of youth ages 10 to 17 have been sexually solicited online (Jones, Mitchell, & Finkelhor, 2012). Social media likely does not function in isolation of other risk factors. For instance, one study found that lack of parental involvement is associated with youth vulnerability to being groomed online for sexual abuse victimization (Whittle, Hamilton-Giachritsis, Beech, & Collings, 2013). Furthermore, engagement in social media in and of itself, is likely not as important as youth actual behavior while using social media. It is highly likely that youth who have higher risk engagement, in general, take more risks online, which consequently increases risk of exposure to inappropriate content and predators. Thus, further work is needed to explore how youth sensation seeking and other personality factors, as well as parental supervision and monitoring, interact with youth social media behavior to predict SEC risk. This work is imperative given that youth worldwide have increased technological access and use of social media platforms, which can easily increase youth exposure to victimization.

In regard to findings for the parental factors, neither parental attitudes toward youth sex nor parent substance use emerged as predictors of SEC in the full sample. This could be an indication of the lessening importance of parental influence during the adolescent developmental period due to adolescents’ growing need for autonomy and independence (Hill et al., 2004), or it could be that the types of parental risk examined in this study are not as important to youth SEC risk as parental monitoring (Adjei & Saewyc, 2017). While these factors did not emerge in the full sample multivariable model, it is important to continue studying various factors relevant to youth risk from a socioecological perspective, so that preventive interventions can be comprehensive and target individual and relational youth risk.

In terms of the exploratory hypotheses examined, differential risk factors distinguished male and female youth with and without an SEC history. Prior research has noted that despite growing evidence that both male and female youth are susceptible to SEC, information regarding risks for males is distinctly missing from scholarly research (Reid & Piquero, 2014). For males and females who endorsed SEC as compared with those youth who did not, a history of ever living on streets and ever being raped emerged as highly prevalent risk factors both, which are well-supported by prior research (e.g., Greene et al., 1999; Reid & Piquero, 2014). However, for females, several additional risk factors emerged as variables that distinguished between youth with and without an SEC history, including a history of dating violence, as well as social media and alcohol use. Furthermore, for females with an SEC history, the two parental variables, parent alcohol use and parent negative attitudes toward child having sex, were both associated with victimization. As expected, increased parent alcohol use was associated with higher SEC risk, but the reverse was true for parental attitudes toward sex (parents with negative attitudes had youth with greater SEC risk). These findings suggest that distinct pathways result in increased SEC risk for males, but that the pathways for female risk are much more complicated and diverse. Importantly, the parental factors emerged for females, perhaps suggesting that the relational risk may make females more vulnerable than males. Future research would benefit from a larger sample of males to ensure there is appropriate power to detect potential risk predictors.

Limitations of Current Study

This study used a cross-sectional design, and consequently, causal relationship between variables cannot be ascertained. All data are self-reported. Due to the low literacy rates of youth participants, interviewers read aloud the survey questionnaire, which may lead to biases or underreporting by youth. Orphan status was not assessed as a predictor in this study, and because approximately 1 in 4 of youth living in slums in Uganda are orphaned (Swahn et al., 2016), the parental risk factors may not have been as relevant for these youth depending on what age they were orphaned. The parental variable related to parent attitudes toward sex was not explored from a developmental perspective, which may be important to examine in future studies, given that there may be differential parental attitudes about sex for a child of 12 versus 17 years. In addition, the question used to assess adolescent alcohol use measured the prevalence, but not extent of alcohol use. Further work is needed to better understand how variability in alcohol use is associated with SEC. Finally, the study variables were limited by the questions asked in the survey. According to one recent study, the greatest predictor of SEC for youth ages 12 to 25 years is to have friends and families involved in prostitution (Klatt, Cavner, & Egan, 2014), a variable not assessed in this study.

Implications and Future Research Directions

These findings have important public health implications, especially for youth living in the slums of Kampala. The rate of reported sexual exploitation was high, with more than one third of youth living in the slums reporting a victimization history. In terms of prevention, as reported in Swahn and colleagues (2016), Uganda created a Prevention of Trafficking Person Act in 2009, followed by a child help hotline in 2014, and an update of the 2009 law in 2016 (Child Amendment Act) to better identify youth and traffickers/perpetrators responsible for trafficking children for sex or labor. However, investigations, prosecutions, and convictions to date have been very limited. Thus, it is important to consider how the current findings could add to the existing resources. In regard to screening for SEC, it is important to include behaviorally specific questions. In addition, it is imperative to consistently screen youth who have any history of living on the streets and who have a history of rape, the two risk factors that emerged as important in both the multivariate analyses of the full sample and in the comparative analyses among male and female SEC victims. Screening of youth social media use as a potential risk factor is also warranted, and efforts to eliminate the risk incurred through social media may be valuable.

For next steps in research, a deeper understanding of the complex living and social histories of adolescents living in low-resource countries, and how their social ecology influences SEC risk, is warranted. Some minors may “willingly” engage in transactional sex as a means of survival, and this behavior is directly influenced by prior trauma, poverty, and use of social media. An especially imperative focus in future research, given its worldwide proliferation, is a deeper understanding of the role of social media in SEC. Delineation of youth social media use and engagement in low-resource countries can inform future strategic prevention and policy efforts.

Footnotes

Authors’ Note

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research reported in this article was supported by the National Institute on Alcohol Abuse and Alcoholism of the National Institutes of Health under Award Number R21AA22065 (to Dr. Swahn).

Author Biographies

Shannon Self-Brown, PhD, is a Professor and the doctoral program director in the School of Public Health at Georgia State University. Her research interests include examining risk and protective factors for youth exposed to community and family violence, child maltreatment, and disaster, as well as the evaluation of dissemination/implementation efforts for child maltreatment prevention programs. She has over 60 publications focused on these topics.

Rachel Culbreth is a PhD student and Second Century Initiative fellow in the School of Public Health. Her research interests include examining health disparities among adolescents pertaining to mental health, HIV/STIs, and interpersonal violence.

Rebecca Wilson, MPH, is a fourth-year PhD student in the School of Public Health at Georgia State University. Her research focuses on child maltreatment prevention among children and adolescents.

Lisa Armistead, PhD, is a Professor of Psychology with a family-focused scholarship has two prongs: (a) descriptive work on family functioning and (b) intervention development and efficacy testing. The descriptive work targets caregiver and child functioning in the context of various stressors (e.g., maternal HIV infection) and includes the identification of mechanisms involved in the stressor–family functioning relationship. This work is then leveraged to build and empirically evaluate interventions for HIV prevention and maternal disclosure of HIV status.

Rogers Kasirye, MA, is an experienced NGO leader and expert consultant on youth in sub-Saharan Africa and he has worked extensively to research, implement programs, and inform policy on pervasive health risks that face the youth population targeted in this proposal. He has acted as lead NGO expert for UNODC on drug abuse and trafficking in sub-Saharan Africa; as a national trainer for a UNFPA project on reproductive health, drug abuse, and at-risk youth; and as a contributor to WHO’s publication on alcohol use among genders in Uganda.

Monica H. Swahn, PhD, is a Distinguished Professor in the School of Public Health at Georgia State University. She conducts research on the risk factors and structural drivers of alcohol use and other health-risk behaviors such as high-risk sex and violence among youth and young adults in the United States and in sub-Saharan Africa. She is currently assigned to Makerere University in Kampala, Uganda as a Fulbright Scholar to strengthen research capacity and to build a research center focused on alcohol and substance research and treatment.

References

Adjei

J. K.

Saewyc

E. M.

(2017). Boys are not exempt: Sexual Exploitation of adolescents in sub-Saharan Africa. Child Abuse & Neglect, 65, 14-23.

The advocates for Human Rights: Stop violence against women. (2017). Retrieved from http://hrlibrary.umn.edu/svaw/trafficking/explore/3factors.htm

Basson

Rosenblatt

Haley

(2012). Research to action: Sexually exploited minors needs and strengths. Oakland, CA: West Coast Children’s Clinic.

Bhana

de Lange

Mitchell

(2009). Male teachers talk about gender violence: Zulu men demand respect. Educational Review, 61, 49-62.

Bigelsen

(2013). Homelessness, survival sex and human trafficking: As experienced by the youth of Covenant House New York. New York: Covenant House New York. Retrieved from https://humantraffickinghotline.org/sites/default/files/Homelessness%2C%20Survival%20Sex%2C%20and%20Human%20Trafficking%20-%20Covenant%20House%20NY.pdf

Clarke

R. J.

Clarke

E. A.

Roe-Sepowitz

Fey

(2012). Age at entry into prostitution: Relationship to drug use, race, suicide, education level, childhood abuse, and family experiences. Journal of Human Behavior in the Social Environment, 22, 270-289.

Cole

Sprang

Lee

Cohen

(2016). The trauma of commercial sexual exploitation of youth: A comparison of CSE victims to sexual abuse victims in a clinical sample. Journal of Interpersonal Violence, 31, 122-146.

Estes

R. J.

Weiner

N. A.

(2001). Commercial sexual exploitation of children in the United States, Canada, and Mexico. School of Social Work, University of Pennsylvania. Retrieved from http://www.gems-girls.org/Estes%20Wiener%202001.pdf

Estes

R. J.

Weiner

N. A.

(2005). The commercial sexual exploitation of children in the United States. In Cooper

S. W.

Estes

R. J.

Giardino

A. P.

Kellogg

N. D.

Vieth

V. I.

(Eds.), Medical, legal & social science aspects of child sexual exploitation: A comprehensive review of child pornography, child prostitution, and internet crimes against children (pp. 95-128). St. Louis, MO: GW Medical.

10.

Farley

(2006). Prostitution, trafficking, and cultural amnesia: What we must not know in order to keep the business of sexual exploitation running smoothly. Yale Journal of Law and Feminism, 18, 109-143.

11.

Felner

J. K.

DuBois

D. L.

(2017). Addressing the commercial sexual exploitation of children and youth: A systematic review of program and policy evaluations. Journal of Child and Adolescent Trauma, 10, 187-201.

12.

Fielding-Miller

Dunkle

K. L.

Cooper

H. L.

Windle

Hadley

(2016). Cultural consensus modeling to measure transactional sex in Swaziland: Scale building and validation. Social Science & Medicine, 148, 25-33.

13.

Fogel

K. F.

Martin

Nelson

Thomas

Porta

(2016). “We’re automatically sex in men’s eyes, we’re nothing but sex . . .”: Homeless young adult perceptions of sexual exploitation. Journal of Child and Adolescent Trauma, 9, 1-10.

14.

Fricker

A. E.

Smith

D. W.

Davis

J. L.

Hanson

R. F.

(2003). Effects of context and question type on endorsement of childhood sexual abuse. Journal of Traumatic Stress, 163, 265-268.

15.

Greene

J. M.

Ennett

S. T.

Ringwalt

C. L.

(1999). Prevalence and correlates of survival sex among runaways and homeless youth. American Journal of Public Health, 89, 1406-1409.

16.

Hill

N. E.

Castellino

D. R.

Lansford

J. E.

Nowlin

Dodge

K. A.

Bates

J. E.

Pettit

G. S.

(2004). Parent academic involvement as related to school behavior, achievement, and aspirations: Demographic variations across adolescence. Child Development, 75, 1491-1509.

17.

Homma

Nicholson

Saewyc

E. M.

(2012). Profile of high school students in rural Canada who exchange sex for substances. The Canadian Journal of Human Sexuality, 21, 29-40.

18.

Hossain

Zimmerman

Abas

Light

Watts

(2010). The relationship of trauma to mental disorders among trafficked and sexually exploited girls and women. American Journal of Public Health, 100, 2442-2449.

19.

Institute of Medicine and National Research Council. (2013). Confronting commercial sexual exploitation and sex trafficking of minors in the United States. Washington, DC: The National Academies Press.

20.

Jewkes

Abrahams

(2002). The epidemiology of rape and sexual coercion in South Africa: An overview. Social Science & Medicine, 55, 1231-1244.

21.

Jewkes

Levin

Mbananga

Bradshaw

(2002). Rape of girls in South Africa. The Lancet, 359, 319-320.

22.

Jones

L. M.

Mitchell

K. J.

Finkelhor

(2012). Trends in youth internet victimization: Findings from three youth internet safety surveys 2000-2010. Journal of Adolescent Health, 50, 179-186.

23.

Klatt

Cavner

Egan

(2014). Rationalising predictors of child sexual exploitation and sex-trading. Child Abuse & Neglect, 38, 252-260.

24.

Kruger

Harper

Harris

Sanders

Levin

Myers

(2013). Sexualized and dangerous relationships: Listening to voices of low-income Afrian American grils placed at risk for sexual exploitation. Western Journal of Emergency Medicine, 14, 370-376.

25.

Lalor

McElvaney

(2010). Child sexual abuse, links to later sexual exploitation/high-risk sexual behavior, and prevention/treatment programs. Trauma, Violence, & Abuse, 11, 159-177.

26.

Latonero

(2011). Human trafficking online: The role of social networking sites and online classified. Center on Communication Leadership and Policy, University of Southern California School for Communication and Journalism. Retrieved from https://technologyandtrafficking.usc.edu/files/2011/09/HumanTrafficking_FINAL.pdf

27.

Lederer

Wetzel

(2014). The health consequences of sex trafficking and their implications for identifying victims in healthcare facilities. Annals of Health Law, 23, 61-91.

28.

Livingstone

Smith

(2014). Annual research review: Harms experienced by child users of online and mobile technologies: The nature, prevalence, and management of sexual and aggressive risks in the digital age. Journal of Child Psychology and Psychiatry, 55, 635-654.

29.

Miller

C. L.

Fielden

S. J.

Tyndall

M. W.

Zhang

Gibson

Shannon

(2011). Individual and structural vulnerability among female youth who exchange sex for survival. Journal of Adolescent Health, 49, 36-41.

30.

Mitchell

K. J.

Boyd

(2014). Understanding the role of technology in the commercial sexual exploitation of children: The perspective of law enforcement. Durham: Crimes against Children Research Center, University of New Hampshire.

31.

Murray

L. K.

Nguyen

Cohen

J. A.

(2014). Child sexual abuse. Child and Adolescent Psychiatric Clinics, 23 (2), 321-337.

32.

Ranganathan

MacPhail

Pettifor

Kahn

Khoza

Twine

. . . Heise

(2017). Young women’s perceptions of transactional sex and sexual agency: A qualitative study in the context of rural South Africa. BMC Public Health, 17, Article 666.

33.

Reid

J. A.

(2011). An exploratory model of girls’ vulnerability to commercial sexual exploitation in prostitution. Child Maltreatment, 16, 146-157.

34.

Reid

J. A.

Piquero

A. R.

(2014). Age-graded risks for commercial sexual exploitation of male and female youth. Journal of Interpersonal Violence, 29, 1747-1777.

35.

Smith

L. A.

Vardman

S. H.

Snow

M. A.

(2009). The national report on domestic minor trafficking: America’s prostituted children. Shared Hope International. Retrieved from https://sharedhope.org/wp-content/uploads/2012/09/SHI_National_Report_on_DMST_2009.pdf

36.

Sorbring

Hallberg

Bohlin

Skoog

(2015). Parental attitudes and young people’s online sexual activities. Sex Education, 15, 129-143.

37.

Stoltz

J. M.

Shannon

Kerr

Zhang

Montaner

J. S.

Wood

(2007). Associations between childhood maltreatment and sex work in a cohort of drug using youth. Social Science & Medicine, 65, 1214-1221.

38.

Svedin

C. G.

Priebe

(2007). Selling sex in a population-based study of high school seniors in Sweden: Demographic and psychosocial correlates. Archives of Sexual Behavior, 36, 21-32.

39.

Swahn

M. H.

Culbreth

Salazar

L. F.

Kasirye

Seeley

(2016). Prevalence of HIV and associated risks of sex workers among youth in the slums of Kampala. AIDS Research and Treatment, 2016, Article 5360180.

40.

Tjaden

Thoennes

(2000). Full report of the prevalence, incidence and consequences of violence against women: Findings from the National Violence Against Women Survey. Washington, DC: National Institute of Justice, Office of Justice Programs, United States Department of Justice and Centers for Disease Control and Prevention.

41.

Tobey

Hillman

S. B.

Anagurthi

C. C.

Somers

(2011). Demographic differences in adolescents’ sexual attitudes and behaviors, parent communication about sex, and school sex education. Electronic Journal of Human Sexuality, 14(3). Retrieved from http://www.ejhs.org/volume14/Sexual%20attitudes.htm

42.

Tourangeau

(2000). Remembering what happened: Memory errors and survey reports. In Stone

A. A.

(Ed.), Science of self-report: Implications for research and practice (pp. 29-47). Mahwah, NJ: Lawrence Erlbaum.

43.

Tourangeau

Smith

T. W.

(1996). Asking sensitive questions: The impact of data collection mode, question format, and question context. Public Opinion Quarterly, 60, 275-304.

44.

Trafficking Victims Protection Reauthorization Act. (2013). Retrieved October 27, 2017 from https://www.congress.gov/bill/113th-congress/house-bill/898

45.

Uganda Youth Development Link. (2011). Commercial sexual exploitation of children in Uganda: A critical review of efforts to address CSEC 2005-2011.

46.

United Nations Office on Drugs and Crime. (2012). Global report on trafficking in persons. Vienna, Austria: Author. Retrieved from https://www.unodc.org/documents/data-andanalysis/glotip/Trafficking_in_Persons_2012_web.pdf

47.

United Nations Office on Drugs and Crime. (2016). World Drug Report (United Nations publication, Sales No. E.16.XI.7). Retrieved from https://www.unodc.org/doc/wdr2016/WORLD_DRUG_REPORT_2016_web.pdf

48.

Varma

Gillespie

McCracken

Greenbaum

(2015). Characteristics of child commercial sexual exploitation and sex trafficking victims presenting for medical care in the United States. Child Abuse & Neglect, 44, 98-105.

49.

Walker

(2013). California child welfare council, ending the commercial sexual exploitation of children: A call for multi-system collaboration in California. Retrieved from http://www.chhs.ca.gov/Child%20Welfare/Ending%20CCSEC%20-%20A%20Call%20for%20Multi-System%20Collaboration%20in%20CA%20-%20February%202013.pdf

50.

Wells

Mitchell

K. J.

(2012). Exploring the role of the Internet in juvenile prostitution cases coming to the attention of law enforcement. Journal of Child Sexual Abuse, 21, 327-342.

51.

Whitaker

D. J.

Miller

K. S.

(2000). Parent-adolescent discussions about sex and condoms: Impact on peer influences of sexual risk behaviour. Journal of Adolescent Research, 15, 251-273.

52.

Whittle

Hamilton-Giachritsis

Beech

Collings

(2013). A review of young people’s vulnerabilities to online grooming. Aggression and Violent Behavior, 18, 135-146.

53.

Williams, T., Binagwaho, A., & Betancourt, T. S. (2012). Transactional sex as a form of child sexual exploitation and abuse in Rwanda: Implications for child security and protection. Child Abuse & Neglect, 36, 354-361.

54.

Williamson

Prior

(2009). Domestic minor trafficking: A network of underground players in the midwest. Journal of Child and Adolescent Trauma, 2, 46-61.