dcyphr | Sexual violence among adolescent girls and young women in Malawi: a cluster-randomized controlled implementation trial of empowerment self-defense training


The authors investigated the efficacy of empowerment self-defense (ESD) in reducing the prevalence and risk of sexual violence (primary outcomes) among adolescent girls in Malawi. ESD’s effect on confidence, knowledge on self-defense, and disclosure about violence was also investigated. The results supported the hypothesis that ESD would reduce the primary outcomes and improve secondary outcomes. ESD is effective in reducing sexual violence; thus, implementation into education systems is advocated. 


Sexual violence, or gender-based violence (GBV), is a recognized global health and human rights crisis that occurs at alarming rates in low- and middle-income countries (LMICs) in particular. GBV is associated with many co-occurring mental, physical, and occupational problems (i.e., unintended pregnancy) that can continue to have lasting detrimental effects. Adolescents in Malawi are particularly at risk for GBV and need interventions as there are sharp gender-based power disparities and approximately 1 in 5 girls will have experienced sexual violence before the age of 18.

 ESD cultivates power and a strong voice through an interactive training experience. It prepares participants for mental, physical, and verbal self-defense through safety promotion, learning safety skills, and having the confidence to implement them instilled in them. ESD pushes back against a culture that perpetuates gender disparities and takes power away from the abuser. 


151 schools with 7,832 participants were enrolled in the study, with student to ESD instructor ratio being 20:1. Ultimately, 72 experimental schools (2195 participants) and 69 control schools (2083 participants) were followed up for analysis. 

IMPower intervention activities 

IMPower includes weekly (6 in total) 2 hours sessions of empowerment self-defense training. It emphasizes boundary testing/setting (i.e., warning about harmful behaviors), negotiation and diffusion tactics, and verbal assertiveness, with physical violence being the last resort. 

Control condition activities 

Controls were instructed in the standard Malawi LifeSkills program. This 2 hour program consisted of teaching about adolescent health topics (e.g., puberty, hygiene, STIs, and menstruation). 

Supporting sexual violence survivors: The SASA Program 

IMPower participants who disclosed sexual violence were referred to the Sexual Assault Survivors Anonymous program, a weekly support group with the goal of healing. 


Individual baseline characteristics 

The average age for primary and secondary school students was 15.28 and 19.55, respectively. 16% of primary school girls and 26% of secondary school girls had reported ever having been forced to have sex. Forced sex victimization in the past year was reported at 10% and 18% for primary and secondary school girls respectively. Of girls who reported forced sex, 49% of girls in primary schools reported only one incident, whereas 63% of girls in secondary school reported multiple incidents. 

Primary outcomes

Baseline rates of past-year and overall incident rates of sexual violence were around the same for the control and intervention groups (~14% for past-year and ~19.5% for overall). ESD resulted in significant reductions in the rates of sexual violence (9.2% at follow up and trending down), whereas the control group led to steady incident rates that trended upwards. 

Secondary Outcomes (Intervention v. Control)

Confidence in self-defense:

  • Baseline: 41.2% vs. 26.2%

  • Increased by: 1.33 v. 1.26

Knowledge in self-defense:

  • Baseline: 15% v. 7%

  • Increased by: 3.33 (threefold) v. 1.27

Sexual violence disclosure:

  • Baseline: 73% v. 68.2% 

  • Increased to: 83.7% v. 74.8% 

The pattern of primary and secondary outcomes was consistent among primary and secondary school girls. 

Skills mix 

Within the ESD group, 43% had reported using the learned skills to stop forced sex since completing the training, with 52% using them more than once. 49%, 13%, and 38% of those who used the skills employed verbal only, physical only, or both respectively. 


Results support the implementation of a school-based empowerment self-defense program for reducing the risk and prevalence of sexual violence among primary and secondary school girls in Malawi. There was no significant effect from the intervention on self-defense confidence, so the minor increase may be due, in part, to natural maturation. The authors’ measures may have also been insufficiently sensitive to capturing ESD intervention effects. 

Most of the perpetrators of sexual violence for this population were known to victims, with boyfriends being the most common. More research into how ESD intervention may affect different groups of perpetrators is warranted.