Intimate partner violence (IPV) is the intentional use of physical force or power, threatened or actual, against an intimate partner (e.g., spouse, cohabitating partner, date), that either results in or has a high likelihood of resulting in injury, death, psychological harm, maldevelopment, or deprivation. This definition encompasses physical, sexual, and emotional or psychological abuse [1,2]. Worldwide, 30% of ever-partnered women have experienced physical or sexual violence, or both, perpetrated by an intimate partner [3]. While both men and women may perpetrate IPV [4], globally the greatest burden and consequences of IPV are experienced by women [3]. In Malawi, the focus of this research, 40% of ever-married women report having experienced physical and/or sexual IPV in their lifetimes [5]. IPV can have short- and long-term physical and mental health effects on women and their children, including, but not limited to, low birth weight and preterm birth [6,7], depression [8,9], and increased risk for sexually transmitted infections, including HIV [8,10–12]. In general, research in lower income countries has focused primarily on women’s exposure to IPV, not men’s perpetration [13]. Similarly, the relationship between men’s exposure to violence in childhood, including experiencing and/or witnessing violence, and their perpetration of IPV against women is understudied in these contexts.

Violence against children is a significant problem worldwide, one that has devastating consequences for children’s life-long health and well-being [14]. Children may be exposed to physical, emotional, or sexual abuse perpetrated by their parents or adult caregivers, adults in the community, or peers. Children may also witness violence between adults or peers in their communities. Worldwide, an estimated 37% of children experience emotional abuse, 23% of children experience physical abuse, and 18% of girls and 8% of boys experience sexual abuse [15]. In a nationally representative survey, 56.1% of females and 67.9% of males aged 13–17 in Malawi reported experiencing some form of emotional, physical, and/or sexual abuse in the past 12 months [16]. Authors of a population-based study of six countries in Asia and the Pacific reported that men’s childhood experiences of violence, including experiencing emotional, physical, or sexual violence, or witnessing abuse of his mother were significantly associated with physical, sexual, emotional, and/or economic IPV perpetration in at least one or more countries [17]. In another multicountry study, which included data from two African countries, witnessing parental violence was a risk factor for men’s perpetration of physical IPV [18]. Additionally, research studies from Sub-Saharan Africa demonstrate associations between both experiencing violence in childhood [19,20] and witnessing parental IPV [21] with men’s perpetration of IPV. No population-based studies, to our knowledge, have explored the relationship between childhood exposures to violence and men’s perpetration of IPV against women in Malawi.

There is a growing body of literature describing the negative impact of adverse childhood experiences on a range of health outcomes in adulthood, including chronic disease [22], infectious diseases including HIV [23], reproductive health [24], mental health [25], suicide attempts [22], and substance abuse [22]. Additionally, evidence suggests a dose-response relationship. As the number of categories of adverse childhood experiences increases, so does the risk for a range of poor health outcomes [22]. There is evidence from the United States and England to suggest this graded relationship exists between exposures to multiple adverse childhood experiences, including experiences of violence, and perpetration of violence, including IPV [26,27]. Researchers have theorized that multiple exposures to early-life stressors, including experiences of violence and witnessing violence, may lead to changes in the structure and physiology of the brain that may play a role in sexual and aggressive behavior, including the perpetration of IPV [28]. To our knowledge, this graded relationship has not been explored in population-based studies in Africa. Thus, the objectives of this study were to estimate the prevalence of young men’s perpetration of physical and sexual IPV in Malawi and to examine the association between exposures to violence in childhood, including exposure to multiple forms of violence, with young men’s perpetration of IPV.

The Malawi Violence Against Children and Young Women Survey (VACS Malawi) is a collaboration between the Malawi Ministry of Gender, Children, Disability and Social Welfare, United Nation’s Children’s Fund, The Center for Social Research of the University of Malawi, and the U.S. Centers for Disease Control and Prevention. VACS Malawi is a nationally representative, cross-sectional household survey designed to produce national-level estimates [16]. The survey, which used a four-stage cluster sample survey design, was conducted September–October, 2013. The survey followed a split sample approach; the survey for females was conducted in different enumeration areas than for the males. This approach served to protect the confidentiality of respondents and reduce the chance that a male perpetrator of a sexual assault and the female who was the victim of that assault (or vice versa) would both be interviewed. Males and females aged 13e24 years old living in selected households were eligible for participation in the study. For those aged 18 years and older, lifetime prevalence estimates of childhood violence were based on their reported experiences before age 18. A total of 2162 individuals aged 13–24 years old participated in the Malawi VACS; 1029 females and 1133 males completed the individual questionnaire, yielding a combined household and individual response rate of 84.4% for females and 83.4% for males. The current analysis focuses on the responses from 450 ever-partnered 18- to 24-year-old men whose responses were weighted to represent all ever-partnered 18- to 24-year-old men in Malawi.

Ethical protections were prioritized in the planning, design, and implementation of the study. The study adhered to World Health Organization guidelines on ethics and safety for research on violence against women and received Institutional Review Board approval from the Centers for Disease Control and Prevention’s Institutional Review Board and the Malawian National Commission for Science and Technology Ethical Review Board [16]. To help facilitate trust with respondents, care was taken to select interviewers who spoke the local languages, had experience with confidential data and health issues, and looked physically young. Training for the interviewers was specific to violence research. For respondents under 18 years of age, the team first obtained the permission of the primary caregiver to speak with the eligible respondents. At this stage, the survey was described as an opportunity to learn about “young people’s health, educational, and life experiences” and “community violence” was included in a list of broad topics included in the interview, and no specific mention was made of violence occurring in the home. These steps were taken to inform parents and care-givers about the content of the study without risking possible negative consequences against children for their participation. Once a respondent was selected, a gender-matched trained interviewer initially introduced the survey, and verbal assent or consent to provide more information about the study was obtained from respondents. Once the interviewer and respondent ensured privacy, the trained interviewer read the contents of a verbal assent or consent form, which informed the respondents about the voluntary nature of the study, specific topics covered in the survey, and reinforced that the information they shared was confidential.

Items from the Malawi VACS measuring violence perpetration and exposures to violence are described in Figure 1. These items, which have been used to measure violence against children in similar settings [29,30], were selected and adapted from a range of well-respected survey tools, and were pilot tested to ensure appropriateness for the Malawian context [16]. The outcome variables for this analysis were lifetime perpetration (yes/no) of physical or sexual IPV against a current or previous girlfriend, romantic partner, or wife. Exposure to violence in childhood was considered based on whether the respondent reported, before age 18 years, witnessing IPV, witnessing an attack in the community, or experiencing sexual abuse, physical violence, or emotional violence. Consistent with literature on adverse childhood experiences in developing country settings [31], we created a summative scale for cumulative exposures to violence, ranging from no exposures to violence, to exposure to four or more forms of violence during childhood. Demographic controls included age in years, if the respondent ever married, if he completed or attended secondary school or higher, and whether he ever begged in the street [19,31]. Potential confounders captured other adverse childhood experiences such as whether the respondent was orphaned (one or both parents died) before age 18 years, and the presence of social support from friends and family (e.g., if he felt very close to his biological mother and father, or if he reported talking with friends often about important things) [19].

We first estimated the weighted prevalence for perpetration of physical and/or sexual IPV and experiences of violence in childhood among young adult men. We then examined the associations between childhood experiences of violence and perpetration of IPV using logistic regression. We found very little overlap between young men’s reported perpetration of different types of IPV, with only 3% of young men reporting perpetration of both physical and sexual IPV. For this reason, we analyzed physical and sexual IPV perpetration as separate outcomes in logistic regression models. We explored whether potential confounders influenced the association between exposure to violence in childhood and perpetration of IPV and included variables significantly associated with both the exposure and the outcome in the adjusted models. In preliminary analyses exploring the bivariate relationships between exposure to multiple forms of violence in childhood and men’s perpetration of physical IPV, we found high estimated standard errors for the estimates, suggesting they may be unstable. Therefore, we present here the adjusted measures for exposures to multiple forms of violence in childhood and men’s perpetration of sexual IPV. Analyses were run in SAS 9.3 [32], and all analyses accounted for the complex survey design.

On average, the ever-partnered 18- to 24-year-old men in our sample were 20 years old, a third were married, 42% had completed or attended secondary school or higher, and 7% percent reported begging for alms in the street at some point in their life (Table 1). Forty-eight percent of young men talked with their friends about important things, 81% felt very close to their biological mother, 62% felt very close to their biological father, and 28% were orphaned before age 18 years. Young men in Malawi reported high levels of both experiencing and witnessing violence in childhood. Lifetime prevalence for experiencing violence in childhood ranged from 15% for sexual abuse to 28% for emotional violence and 65% for physical violence. As a child, 39% of young men witnessed an attack in their community and 32% witnessed IPV (Table 1). Only 18% of young men reported no exposure to violence in childhood, while the majority of young men experienced multiple forms of violence in childhood, with 10% exposed to four or more different forms of violence (Table 1). More than a quarter of young men (29%) reported perpetrating physical and/or sexual IPV in their lifetime. Lifetime prevalence for perpetration of sexual IPV (24%) was higher than for perpetration of physical IPV (9%) and very few men (3%) reported perpetrating both physical and sexual IPV (Table 1).

In Malawi, nearly all young men—over 8 of 10—reported exposure to one or more forms of violence as a child, with experiencing any physical violence as a child reported most frequently (65%). Likewise, perpetration of IPV, particularly sexual IPV, is common among young men. Twenty-four percent of young men reported perpetrating sexual IPV, a lifetime prevalence that is particularly concerning considering both the young age of the respondents (18–24 years) and the high prevalence of HIV in Malawi, at 11% [5].

Fewer young men reported perpetrating physical IPV compared to sexual IPV. The lower prevalence of lifetime physical IPV may reflect the young age of our sample and is consistent with findings from a multicountry study of men’s perpetration of IPV in Asia and the Pacific [15]. In five of six countries, the prevalence for perpetration of physical IPV only for men aged 18–24 years ranged from4.8% to 10.9%, while the prevalence for perpetration of sexual IPV only ranged from 15.6% to 23.9% [15]. We also found unique patterns of association between exposures to different forms of violence in childhood and young men’s perpetration of physical IPV compared to sexual IPV. For example, young men’s perpetration of physical IPV was associated with childhood physical violence, but not with other exposures to violence in childhood. In contrast, young men’s perpetration of sexual IPV was associated with childhood sexual abuse and emotional violence. While not all men exposed to violence in childhood reported perpetration of IPV later in life, our findings demonstrate a dose-response relationship between young men’s exposure to multiple forms of violence as a child and perpetration of sexual IPV. This finding is consistent with research from the U.S. and England [20,21] and is, to our knowledge, the first research to demonstrate a graded relationship between the number of exposures to different forms of violence in childhood and men’s perpetration of sexual IPV in Malawi. Small numbers for perpetration of physical IPV did not allow us to examine the graded relationship between exposures to multiple forms of violence in childhood and young men’s perpetration of physical IPV; this may represent a focus for future research.

This study has some limitations. Our analysis included a measure of men’s lifetime perpetration of IPV, and it is possible that in some cases, the outcome—young men’s perpetration of IPV, may have occurred before exposures to violence in childhood. Future research may address this limitation by including, for men aged 19 years and older, a measure of young men’s past year perpetration of physical and/or sexual IPV. Reporting of childhood experiences of violence was retrospective, and there exists a potential for recall bias and a possible underestimation of the occurrence of violence [28]. In addition, the items measuring IPV, while capturing lifetime perpetration, did not capture the severity, frequency, or duration of this violence, nor were data available to assess perpetration of emotional IPV. An empirical exploration of the pathways by which exposure to violence in childhood may impact perpetration of IPV is beyond the scope of this analysis but may be addressed in future research. Finally, for this subsample of ever-partnered young men, in some instances, the number of responses for each category was small, impacting both the stability of some estimates and power to detect differences. Despite these limitations, our results are consistent with findings from Kenya and Tanzania measuring young men’s exposure to violence in childhood [29,30] and Asia and the Pacific measuring young men’s perpetration of physical and sexual IPV [17], lending support to the our findings. Future research may benefit from exploring these relationships in a larger sample of men, including older men.

These findings highlight opportunities for future research. This analysis may be replicated using VACS data from other individual countries, or in a multicountry analysis. Future research may further explore the associations between exposures to violence in childhood and young men’s perpetration of IPV with a particular focus on what factors, if any, may mitigate the effect of exposure to violence in childhood on men’s perpetration of IPV.

Programs targeting the prevention of sexual IPV against women in Malawi can benefit from taking into account the high levels of young men’s exposure to violence in childhood. For young men already exposed to violence in childhood, programs and policies can focus on interrupting the intergenerational transmission of violence. In addition, prevention of violence against boys may be seen as primary prevention of IPV against women and prioritized accordingly.

We know that violence is preventable, and there is growing evidence regarding violence prevention programs that are scientifically grounded and politically feasible. The INSPIRE technical package [33], for example, provides examples of evidence-informed actions to prevent violence against children globally, and may be used to help countries, including Malawi, focus their priorities on programs and policies with the greatest potential to reduce violence against children.