Intimate partner violence (IPV) is a serious problem affecting many adolescents and young adults, especially in Uganda. Researchers have found IPV rates as high as 46% among clinic-based samples of adolescent girls and young women (AGYW) in Uganda aged 15 to 24 (Mayanja et al., 2020). Compared to adolescent girls in 16 low-and-middle-income countries, findings from a population-based study indicated that Ugandan adolescent girls had one of the highest rates of lifetime physical and sexual IPV, with 30% and 41% of adolescent girls aged 15 to 19 reporting sexual and physical IPV, respectively (Decker et al., 2015). The only country with a higher rate was Congo Democratic Republic (33% sexual IPV; 46% physical IPV). Consequently, a substantial proportion of AGYW in Uganda specifically may be at risk of IPV, considering Uganda has the world's youngest population, with more than half (69%) of its population under age 24 (Central Intelligence Agency, 2020).

Adolescent boys and young men living in Uganda also report experiences of IPV. A recent study using national data reported that 8% of Ugandan boys and men age 13 to 24 years experienced physical and/or sexual IPV (Cohen et al., 2020). This rate was significantly lower than that found among Ugandan AGYW in the study aged 13 to 24 (22%). However, there is research to suggest that married adolescent boys and young men living in Uganda report IPV at similar and sometimes higher rates than AGYW (UBOS & ICF, 2018). For instance, national data from the Uganda Demographic Health Survey (DHS) demonstrated that slightly more than half (52%) of ever-married Ugandan women age 20 to 24 and 42.5% of ever-married Ugandan men age 20 to 24 experienced lifetime physical, sexual, or emotional IPV (Uganda Bureau of Statistics (UBOS) & ICF, 2018). Married adolescent boys aged 15 to 19 living in Uganda reported slightly higher lifetime sexual IPV rates than married adolescent girls aged 15 to 19 (19.7% vs. 17.4%, respectively) and similar lifetime physical IPV rates (26.7% vs. 27.4%, respectively) (UBOS & ICF, 2018). However, adolescent girls 15 to 19 years old reported higher rates of emotional IPV than adolescent boys this age (27.9% vs. 23.1%).

Despite the clear evidence indicating young people living in Uganda experience IPV at high rates, less is known regarding the potential reasons for their IPV risks and whether these reasons differ by gender. Further, existing IPV research examining Ugandan adolescents and young adults does not provide detail regarding age-disaggregated associations. Gender-and-age-disaggregated IPV associations could help determine the appropriateness of IPV interventions and policies. Also, understanding the age of first occurrence of IPV may help identify when best to deliver an intervention targeting primary versus secondary IPV prevention.

The current study provides national estimates of the prevalence, characteristics, and correlates of physical and sexual IPV victimization among AGYW and adolescent boys and young men living in Uganda. Using data from the Uganda Violence Against Children Survey, the aims of the study are: (1) Determine the prevalence of experiencing lifetime IPV (physical and sexual) in adolescence and young adulthood. (2) Identify characteristics of the first reported episode of IPV among adolescents and young adults reporting a history of IPV. (3) Explore individual, relationship, community, and societal factors associated with reporting lifetime IPV among adolescents and young adults living in Uganda. (4) Investigate whether gender intersects with other significant factors increasing IPV risk among adolescents and young adults. Differences in IPV victimization by gender and developmental stage will be highlighted. This research is imperative as it may inform Uganda's policies and programs to prevent or curtail IPV among adolescents and young adults.

Adolescent and Young Adult Developmental Context

Adolescence and young adulthood are important stages to examine IPV experience. Adolescence, the 10 to 19 years old age group (WHO, 2021), is a unique developmental period in which the importance of peer relationships and desire for peer approval are at their peak, and romantic relationships are influential and formative (Furman & Shaffer, 2003). During this time, adolescents seek to develop intimacy, identity, and sexuality (Pempek et al., 2009). Age is a salient factor in IPV during adolescence, as research in the United States (U.S.) has shown that IPV rates increase from early to late adolescence and decrease across young adulthood (Hokoda et al., 2012). However, these trends are not consistent among all types of IPV, genders, and ethnicities (Foshee et al., 2009).

Young adulthood, or the developmental period from adolescence to around age 24 (WHO, 2021), is characterized by the desire to solidify and express identity and deepen intimate relationships (Arnett, 2000). This period is significantly impacted by cultural context, as young adults try out adult roles that vary by culture. There is often a shift in this period from prioritizing peers to prioritizing romantic relationships (Brown et al., 2020), and young adults in some cultures are more likely than adolescents to engage in romantic relationships and experiment with sexuality (Giordano et al., 2009). For instance, many U.S. young adults experience IPV for the first time during this stage (age 18–24), as more women (45.2%) and men (41.2%) have been shown to experience IPV for the first time in young adulthood rather than earlier in adolescence (Smith et al., 2018).

Adolescence and young adulthood are also when components of gender identity are formed, and beliefs about gender and dating influence dating, sex, and IPV behaviors. Adolescents might face pressure to conform to culturally specific gender norms (Priess et al., 2009). Adolescents also receive gendered messages from their primary socialization sources- parents, peers, and the media- particularly around dating and sexuality (Trinh & Ward, 2016). Adolescents rely on these messages as many do not have dating experience to draw from (Ward, 2003). Environments with gender norms that link masculinity to aggressive behavior and support IPV may intensify IPV risk, especially against AGYW (Solotaroff & Pande, 2014). Research across the globe, including sub-Saharan Africa, has linked adolescent boys' and young men's endorsement of traditional masculinity norms that support men's dominance over women and sexual violence behaviors (Tharp et al., 2013) to IPV perpetration (for reviews, see Dardis et al., 2015 and Malhi et al., 2020). Because of developmental changes in adolescence and IPV rates peaking for many groups around adolescence and young adulthood, it is a critical period to focus IPV research and interventions (Yount et al., 2017).

Theoretical Framework: Feminist Intersectionality Perspective and Ecological Framework

Our work is informed by the feminist intersectionality perspective, that IPV is gender-based and occurs due to patriarchal power and control and other forms of gender-based oppression. Beyond gender, IPV risk increases among AGYW with intersecting marginalized identities (Kelly, 2011). Components of gender identity related to adolescent development (described above) combined with cultural norms in Uganda that condone IPV and tolerate men's controlling behaviors increase risk of gender power imbalances among Ugandan AGYW, making it very difficult for AGYW to avoid or escape IPV (Mujuzi, 2014; Winchester, 2016). In Uganda, girls aged 15 to19 years were more likely (58%) than older women (45–49%) to agree with at least one reason for wife-beating (e.g., agree that a husband is justified in beating his wife if wife burns food); similar trends existed among adolescent boys (53%) relative to men (30–43%) (UBOS & ICF, 2018). Further, due to economic deprivation, Ugandan AGYW may feel compelled to remain in violent relationships to gain necessary financial or tangible needs (Bell, 2012; Winchester, 2016).

In accordance with the feminist perspective, researchers studying IPV in Uganda have found that AGYW 13 to 24 years old had a higher risk of IPV victimization than adolescent boys and young men 13 to 24 years old; adolescent boys and young men were twice as likely as AGYW to perpetrate IPV (Cohen et al., 2020). Further, findings from the Uganda DHS demonstrated that IPV risk increased by age, but only among women. There was no variation in IPV by age among men (UBOS & ICF, 2018). Overall, these findings suggest that IPV in Uganda is influenced by intersecting gender and age factors.

Additionally, adopting an ecological framework, the current study assumes that societal (e.g., cultural norms supporting violence; gender, social, and economic inequalities), community (e.g., poverty; high unemployment rates; high crime rates), relationship (e.g., low socioeconomic household status; peers that engage in violence; marital conflict), and individual (e.g., age; education; mental health; history of child maltreatment; substance use) factors interplay to influence IPV risk among young people in Uganda (Kelly, 2011; WHO, 2010). Supporting this framework, findings across studies suggest the following factors are associated with IPV among young people in sub-Saharan Africa: sociodemographic characteristics (i.e., level of schooling, marital status, age, urbanicity, number of children, and wealth status), alcohol use, perception that IPV is acceptable/attitudes around gender, partner's jealousy, borrowing money from outside the home, and being sexually active (Burns et al., 2018; Logie et al., 2019; Meinhart et al., 2020; Selin et al., 2019; Wubs et al., 2013; Wusu, 2015).

Method

Study Setting

It is not uncommon for Ugandan girls to marry or live with their romantic partners at an early age. For example, 20% of girls aged 15 to 19 reported being married/cohabiting versus 2% of boys in this age group (UBOS & ICF, 2018). The Ugandan economy mainly depends on subsistence farming as the primary source of livelihood (UBOS, 2018). In 2016, 41.3% of Ugandans lived below the extreme international poverty line of U.S. $1.90 a day The World Bank Group, 2021). Most Ugandans have either no formal education (19% of women and 13% of men age six or older) or only some primary education (44% of women and 54% of men) (UBOS & ICF, 2018).

Recognizing that many of its young people are vulnerable to violence, Uganda has enacted policies to protect children and vulnerable groups from violence, including the Domestic Violence Act of 2010 (Mujuzi, 2014). This Act makes it illegal to commit an IPV offense (Mujuzi, 2014). However, social and cultural norms that support gender-based violence make it difficult to enforce policies against IPV (Mujuzi, 2014). Further, judicial corruption negatively affects how laws are applied (Parikh, 2012).

Study Participants and Procedures

The Violence Against Children Survey (VACS) is a cross-sectional nationally representative household survey involving young people aged 13 to 24 years living in Uganda. Data were collected between September and November 2015 through face-to-face interviews with young people using electronic netbooks with CSPro software. Researchers introduced the survey to participants as an opportunity to learn about "young peoples' health, educational, and life experiences" (MGLSD, 2015, p. 94). The head of each participant's household provided household sociodemographic information.

A three-stage cluster sample survey design was used to collect VACS data. In the first sampling stage, 768 primary sampling units or enumeration areas (E.A.s) from the 2014 census were randomly selected. A split sample approach, with each E.A. consisting of either only females (n = 368 EAs) or males (n = 400 EAs), was used to protect participants' confidentiality and reduce the chances of interviewing a perpetrator and victim of violence living in the same community. In the second sampling stage, approximately 25 households were randomly selected from each E.A. using equal probability systematic sampling. In the third sampling stage, one eligible household member (age 13–24, fluent in one of seven survey languages, and capacity to understand survey questions) was randomly selected as a study participant. Overall, 5,804 young people (3,159 AGYW and 2,645 boys and young men) participated in the Uganda VACS. The response rate was high: 82.2% for girls/young women and 82.3% for boys/young men. VACS statisticians created weights to account for stratification and nonresponse and to ensure the sample was nationally representative. A detailed discussion regarding the Uganda VACS sampling procedures and methods is available elsewhere (MGLSD, 2015). The Arizona State University Institutional Review Board approved this study.

Analytic Sample

The current study only includes data from ever-partnered young people aged 13 to 24 (n = 3,404), as these were the only participants asked about IPV. Of these young people, 574 had missing data on variables included in the study. Rather than conducting listwise deletion, we gave the subpopulation of participants with missing data a weight set to zero. This subpopulation analytic technique is appropriate for handling missing data on complex sample survey data, such as VACS data, and does not lead to incorrect inferences (Bell et al., 2009). It allowed the VACS sampling weights to produce estimates that retain accurate variance estimations for the subpopulation of 2,830 young people with complete data included in the final analysis and accounted for the VACS sampling design (Bell et al., 2009; West et al., 2008).

Bivariate analyses comparing participants in the analytic sample (n = 2,830) to those given a weight of zero due to missing data (n = 574) showed that those with missing data were more likely to be female and married (p < 0.05). Therefore, it is important to note that young people aged 13 to 24 with missing VACS data were likely at higher risk for IPV than those with complete data. Nevertheless, VACS represents the best data available for exploring the IPV experiences of young people living in Uganda, as it is the first and currently the only national study of violence against children in Uganda (MGLSD, 2015).

Measures

The core VACS questionnaire is a standardized global tool utilized by VACS partners across several countries. The Uganda version of the questionnaire was adapted with the help of key stakeholders in Uganda familiar with violence against children, child protection, and the cultural context. The VACS questionnaire items were drawn from several well-respected survey tools, including the DHS, Youth Risk Behavior Survey (YRBS), and Add Health.

Independent Variables

We examined the association between lifetime IPV with ten correlates of IPV among young people living in sub-Saharan Africa identified in prior research (Burns et al., 2018; Logie et al., 2019; Meinhart et al., 2020; Selin et al., 2019; Wubs et al., 2013; Wusu, 2015).

Sociodemographic and background information

We used data regarding participants' education level, age, gender, marital status, region of residence, and household wealth. Household wealth was calculated based on ownership of nonproductive assets. Specifically, using procedures for calculating a wealth index for the full sample (n = 5,804; DHS, 2018), we conducted a factor analysis including six household characteristics (having a watch, bike, motorcycle or scooter, radio, television, mobile phone) as indicators. We then computed a score, with higher scores indicating higher household wealth. We used raw scores in regression models; however, we reported wealth position by quintiles for descriptive purposes.

Attitudinal acceptance of IPV

Attitudes accepting IPV was measured using the following items: Do you believe it is right for a man to hit or beat his wife if she: (1) goes out without telling him; (2) does not take care of the children; (3) argues with him; (4) refuses to have sex with him; (5) burns the food. Response options included yes (scored as 1) or no (scored as 0). Scores were summed such that higher scores reflect greater acceptance of IPV. Reliability of this scale in the current study was acceptable (α = 0.7).

Witnessing parental physical IPV

Witnessing IPV between parents was assessed using the following single item: "How many times did you see or hear your parent punched, kicked, or beaten up by your other parent, or their boyfriend or girlfriend?" Response options included never, once, a few times, or many times. We combined responses into a single dichotomized variable (witnessed parental physical IPV at least once/never witnessed parental physical IPV).

Mental health

Participants were asked six questions about their feelings and the frequency of these feelings during the past 30 days (e.g., During the past 30 days, how often did you feel so sad that nothing could cheer you up?). Response options ranged from 0 = None of the time to 4 = All the time. Scores were summed, with higher scores indicating poorer mental health. The internal consistency reliability of the mental health scale in this study was good (α = 0.8).

Alcohol use

A single item was used to measure alcohol use: "In the past 30 days, on how many days did you drink alcohol to the point that you became drunk?" Data were measured on a continuous scale, with participants who reported they never drank alcohol scored as 0.

Dependent Variables

Lifetime Physical IPV

The following lifetime physical IPV behaviors were measured into a single dichotomized variable (at least one of these behaviors happened/none of these behaviors happened): partner punched, kicked, whipped, or beat you with an object; partner strangled, suffocated, tried to drown you, or burned you intentionally; partner used or threatened you with a knife, gun or other weapon. The three items demonstrated acceptable internal consistency with a Cronbach alpha of 0.6. Data were also gathered regarding the frequency of these behaviors (once, few, or many times). Finally, we assessed the following characteristics of the first physical IPV episode: age, partner type, partner's age, injured (yes/no), and, if injured, type of injuries (e.g., cuts, broken bones).

Lifetime Sexual IPV

Sexual IPV was measured using four binary response items assessing whether participants ever experienced the following kinds of sexual violence (yes or no): (1) touched you in a sexual way without you wanting to/without permission, (2) tried to make you have sex against will but did not succeed, (3) pressured you to have sex, through harassment, threats or tricks and did succeed, and (4) physically forced you to have sex and did succeed (α = 0.6). Following each question, if a participant responded yes, they were asked about the relationship of the person who first (past) and last (recent) committed the sexually violent act against them. Participants who reported either a romantic partner, spouse, ex-romantic partner, or ex-spouse were classified as having experienced sexual IPV. For participants reporting that the first sexual violence experienced was committed by an intimate partner, we assessed the age of (a) participant during the first experience and (b) partner who committed the violent act. When a participant reported experiencing multiple forms of sexual IPV in the same time frame, we reported on characteristics of the most severe form experienced.

Data Analysis

Analyses were conducted using Stata 16.1 with weights allowing findings to be nationally representative of young people living in Uganda. Data were disaggregated by age, based on WHO's definition of adolescents (13–19 years) and young adults (20–24 years). First, we conducted bivariate analyses (f-test/corrected chi-square testFootnote 1 and t-test) to understand differences between adolescents' and young adults' characteristics and IPV experiences. Second, we used logistic regression models to explore IPV risk factors, including lifetime physical and/or sexual IPV (yes/no) as the outcome and independent variables as covariates. These analyses were conducted separately for adolescents and young adults. Finally, to examine potential gender differences, we created an interaction term between gender and independent variables found to be statistically significant (p < 0.05) in the second analysis step. Each interaction term was then entered separately in separate logistic regression models, including all covariates.

Results

Sample Characteristics

As shown in Table 1, although over half of the sample (62.5%) were in the three lowest wealth quintiles, adolescent households had higher wealth than young adult households (p < 0.05). Adolescents were also more likely than young adults to perceive IPV as acceptable. Conversely, more young adults reported symptoms of poor mental health than adolescents. Young adults also reported more days getting drunk and were more likely to be married/cohabiting.

Table 1 Sample characteristics of ever-partnered adolescents (aged 13–19) and young adults (aged 20–24) living in Uganda
Full size table

Lifetime IPV Prevalence and Characteristics

Overall, 19.5% of participants reported experiencing at least one type of physical IPV. AGYW were more likely than adolescent boys and young men to report physical IPV: 18.1% and 6.1% of adolescent girls and boys, respectively, reported physical IPV (F (1, 752) = 22.2, p < 0.001); 34.3% and 11.4% of young women and men, respectively, reported physical IPV (F (1, 752) = 53.4, p < 0.001). Young adults (24.7%) were more likely than adolescents (12.3%) to report lifetime physical IPV. Most young people who experienced physical IPV reported experiencing physical IPV once (47.1%) or few times (34.4%); nearly one-fifth (18.6%) reported experiencing physical IPV many times.

The prevalence of specific types of lifetime physical IPV is shown in Fig. 1. Except for being threatened with a knife, gun, or other weapon, young adults were more likely than adolescents to experience all types of physical IPV. Gender differences were also present. AGYW were more likely than adolescent boys and young men to report all physical IPV types.

Fig. 1
figure 1

Proportion of young people reporting physical IPV in Uganda by physical IPV type. Estimates based on small numbers should be interpreted with caution

Full size image

Overall, 15.6% of young people reported an experience of sexual violence. There was no significant difference comparing adolescents' (16.9%) and young adults' (14.6%) lifetime sexual IPV rates. However, when examining specific sexual IPV types (Fig. 2), adolescents were more likely than young adults to report unwanted touching (F (1, 752) = 4.2, p = 0.04). Additionally, gender differences were present. AGYW reported significantly higher rates of sexual IPV than adolescent boys and young men, overall and across most types.

Fig. 2
figure 2

Proportion of young people reporting sexual IPV in Uganda by sexual IPV type. Estimates based on small numbers should be interpreted with caution

Full size image

There were also notable gender and age group differences in characteristics of the first episode of physical (Table 2) and sexual IPV (Table 3). For example, adolescents were more likely than young adults to report that the partner who perpetrated the first physical IPV episode was about the same age and a boy/girlfriend. Although they reported similar rates of injury during the first physical IPV episode, young adults who suffered an injury were more likely to report that the injury led to being permanently injured or disfigured than adolescents who suffered an injury. Young adults were more likely to report an injury during the first IPV episode if they were women versus men.

Table 2 Characteristics of first reported physical IPV episode among ever partnered adolescents (aged 13–19) and young adults (aged 20–24) living in Uganda who reported physical IPV
Full size table
Table 3 Characteristics of first reported sexual IPV episode among ever partnered adolescents (aged 13–19) and young adults (aged 20–24) living in Uganda who reported sexual IPV
Full size table

Regarding the first sexual violence experience, young adults reported significantly older age than adolescents across all sexual IPV types. However, there was no significant difference in perpetrator's age across sexual IPV types comparing young adults to adolescents. Conversely, gender differences were present (Table 3). For instance, AGYW were more likely than adolescent boys and young men to report that someone older perpetrated the first IPV episode (physical and most sexual IPV types). Typically, AGYW reported that a spouse rather than a boy/girlfriend perpetrated the first physical IPV episode.

Correlates of IPV

The following sections describe findings related to regression models for (1) adolescents and (2) young adults (Table 4). We also discuss moderation results (not shown in table).

Table 4 Multivariable logistic regression models predicting lifetime physical and/or sexual IPV among adolescents (aged 13–19) and young adults (aged 20–24) living in Uganda
Full size table

Adolescents

Significant correlates of IPV among adolescents (n = 1,182) were older age (aOR = 1.3, 95% CI: 1.1–1.5), being female (aOR = 3.9, 95% CI: 2.2–6.8); witnessing parental physical IPV (aOR = 1.8, 95% CI: 1.1–3.0); being widowed, divorced, or separated (aOR = 4.6, 95% CI: 1.6–13.5); and poorer mental health (aOR = 1.1, 95% CI: 1.04–1.1). Age was the only variable significantly moderated by gender (aOR = 1.4, 95% CI:1.1–1.9). Specifically, girls’ IPV risk increased with age; however, boys’ IPV risk remained nearly the same from age 13 to 19 (Fig. 3).

Fig. 3
figure 3

Interaction effect of age by gender on IPV exposure among adolescents (aged 13–19) living in Uganda

Full size image

Young Adults

Similarly, among young adults (n = 1,648), being female (aOR = 3.1, 95% CI: 2.1–4.4), witnessing parental physical IPV (aOR = 1.8, 95% CI: 1.2–2.6), being widowed, divorced, or separated (aOR = 5.2, 95% CI: 2.5–10.8), and poorer mental health (aOR = 1.1, 95% CI: 1.1–1.2) were significantly associated with IPV. Different from adolescents, there was no statistically significant association between age and IPV among young adults. Also, our analysis testing gender as a moderator revealed no statistically significant findings.

Discussion

This study provides further evidence from national data that IPV is prevalent among young people living in Uganda, particularly among AGYW. Additionally, it is one of the first studies to examine risk factors associated with Ugandan boys and young men's IPV experiences. Results highlight the need for IPV primary and secondary prevention programs across all Ugandan regions, beginning in primary school, as adolescents in our study were as young as 13. However, recognizing that many Ugandans may not receive or complete primary school education (UBOS & ICF, 2018), other opportunities for IPV prevention should be considered, such as intervening in the home or using public media campaigns.

Overall, young adults reported higher rates of lifetime physical IPV than adolescents. However, the two age groups did not differ in lifetime sexual IPV rates. Age only influenced IPV risk for adolescents. It was not a factor contributing to IPV risk among young adults. Outside of age, adolescents and young adults shared similar risk factors for IPV. However, they reported different characteristics of their first IPV experience. These characteristics also varied by gender.

Supporting findings from past studies conducted in Uganda (e.g., Gubi et al., 2020; Logie et al., 2019) and the ecological framework, we found an association between IPV (physical and/or sexual) and witnessing parental IPV and IPV and poor mental health among both adolescents and young adults. Also, corroborating our results related to characteristics of the first IPV incident and the feminist perspective, we found that AGYW were at greater risk of IPV than adolescent boys and young men. Marital status was also associated with IPV among both age groups.

Our findings are consistent with Uganda DHS trends (UBOS & ICF, 2018) and support the feminist intersectionality perspective, suggesting age and gender intersect to increase IPV risk. IPV risk increased among adolescent girls as age increased. Yet, IPV risk remained nearly the same for boys throughout adolescence. Findings also highlight developmental differences between boys' and girls' IPV risk. Adolescent boys and young men living in Uganda are twice as likely as AGYW to perpetrate IPV (Cohen et al., 2020). Developmentally, adolescents are at a stage where they are navigating social norms and learning what it means to be a "man" versus "woman." Perceptions of the traditional meaning of "man" in Uganda sometimes includes aggressive behaviors (Bonnevie & Wagman, 2019). Further, gender norms reduce girls' power in relationships by reinforcing passivity and the acceptability of male aggression. In these ways, gender norms contribute to adolescent boys' IPV perpetration risk and adolescent girls' IPV victimization risk.

While Uganda has taken a human rights-based approach (Androff, 2018) to gender equality, friction may exist between such efforts and young people's intimate relationships. For example, relative to other age brackets, married/cohabiting Ugandan women aged 15 to 19 (17.5%) and 20 to 24 (12.1%) who received cash earnings for employment reported higher rates of allowing their husband/partner to decide how their cash earnings are used (UBOS & ICF, 2018). Therefore, interventions need to move beyond increasing economic opportunities for women and target gender norms as well. Friction may also exist between young people's human rights and the application of laws. Multiple laws in place to protect Ugandan children against violence only protect children under age 12 (e.g., the penal code and Part V of the Children Act), ignoring victims between age 12 and 17 (The Republic of Uganda, 2016; Uganda, 2007). Societal and structural change is required to reduce IPV risk among young people. Otherwise, the positive effects associated with current IPV programs might be reduced (Levey et al., 2019).

We encourage IPV programs for young people in Uganda that promote healthy gender norms, including norms rejecting IPV acceptability. Programs promoting gender equality may improve AGYW's agency and coping response to IPV. These programs should incorporate gender-transformative approaches, which have proven effective in reducing IPV among young people across the globe (for review, see Levey et al., 2020). Such programs are especially imperative in Uganda, considering a recent systematic review revealed zero high-quality gender-transformative programs focused on young people in Uganda (Levy et al., 2020). Moreover, current gender-transformative programs used with young populations often restrict definitions of gender, ignoring the role of masculinity in IPV and heavily focusing on women’s empowerment (Levy et al., 2020). To increase chances of success and shift gender norms, these programs must engage boys and young men (Bonnevie & Wagman, 2019). As highlighted by our study findings, unique subgroups of AGYW and adolescent boys and young men might be at heightened risk of experiencing IPV. Therefore, IPV interventions targeting young people in Uganda should also consider gender in the context of intersectionality.

Despite gender inequalities, AGYW in Uganda seem to be becoming more empowered to leave or end unhealthy marriages. Ugandan women are more likely than men to report being divorced or separated (11% versus 5%) (UBOS & ICF, 2018). However, we found that being divorced, separated, or widowed was positively associated with IPV among adolescents and young adults. Other researchers have also found a positive association between IPV and divorce/separation among Ugandan women (Wagman et al., 2016). Therefore, separation or divorce might be a dangerous time for Ugandan women, and women need formal support if they decide to leave a relationship. There are limited formal resources available in Uganda for women experiencing IPV (Lane, 2015). As previously mentioned, early marriage among girls is common in Uganda. The legal framework for marriage in Uganda is complex, as many marriage laws vary depending on custom and religion. However, the official legal age of marriage is 21 or 18 with parental consent (Uganda Legal Information Institute, 2021). Nearly one-quarter (24.3%) of adolescents in our study reported being married or cohabiting, and most were girls (87.3%), indicating that many adolescent girls might be in a situation of emotional and/or economic dependence to their male partners.

Interventions, formal supports, and prevention programs for IPV in Uganda may also help prevent child maltreatment and child exposure to IPV. Despite having laws prohibiting child maltreatment, Uganda's policies do not speak specifically to child exposure to IPV. Policies in Uganda implementing and enforcing consequences for children witnessing IPV may help prevent IPV and other forms of child maltreatment (Devries, 2017). Additionally, interventions like SASA!, designed and tested in Uganda, could help reduce child exposure to IPV (Kyegombe et al., 2015). Regardless if poor mental health is a predictor or consequence of IPV (Devries et al., 2013; Rodriguez et al., 2020), increasing the number of available IPV resources in Uganda, including IPV shelters, may improve mental health issues associated with IPV. In addition to providing a safe haven for those wanting to escape IPV, shelters can provide counseling for AGYW experiencing IPV.

Strengths and Limitations

The current study has several strengths, including the use of a large nationally representative sample. Also, different from prior IPV research focused on young people in sub-Saharan Africa, we used age-and-gender-disaggregated methods to assess IPV risk. Nonetheless, there are limitations. First, we used cross-sectional data; thus, we cannot infer the directional relationship between IPV and other variables. Second, VACS only measured sexual IPV based on the first and last experience. If a participant experienced more than two incidents and the first and most recent incidents were not with an intimate partner, this participant was not identified as experiencing sexual IPV. Thus, sexual IPV rates are likely greater than those reported. Relatedly, IPV characteristics were based on the first incident. IPV characteristics outside of the first are not represented. Third, our IPV definition only includes physical and sexual IPV, excluding other forms of IPV. Fourth, we relied on self-reported data. There could be response bias producing under-reporting of IPV, and we may have excluded those with the most severe IPV experiences. Thus, results may not generalize to the most marginalized young people. Fifth, although measures were drawn from existing validated instruments and reliability levels were acceptable, they were not psychometrically tested. Consequently, estimates may be less precise than ideal.

Given these study limitations, we recommend future IPV research with young people in Uganda using longitudinal data. This research should include psychometrically tested measures and assess multiple IPV forms and characteristics (not only that related to a single incident). Additionally, it should consist of IPV data collected across various informants or sources.

Conclusions

Overall, one in five and one in six young people aged 13 to 24 years in Uganda reported lifetime physical and sexual IPV, respectively. Further, among young people reporting physical IPV, more than half experienced greater than one IPV incident. Although AGYW are at greater risk of IPV than adolescent boys and young men, adolescent boys and young men are also vulnerable to IPV. IPV risk increases for adolescent girls as they get older. Primary and secondary IPV prevention programs are necessary to combat IPV among young people in Uganda, and primary IPV prevention should begin as early as age 13. If not, IPV may become more severe, as young adults report higher rates of severe and permanent IPV consequences than adolescents. This study highlights risk factors that professionals in Uganda working with young people can use as indicators for the need to assess for IPV and possibly intervene.