Interrupting the Intergenerational Transmission of Violence Surveys, 2022-2024

The INTERRUPT_VIOLENCE study builds on a longitudinal cohort established in 2010/11 to investigate how violence, adversity and mental health difficulties are experienced and transmitted across generations in a disadvantaged health district in South Africa. The original cohort comprised 1,664 adolescents aged 10–17 years, recruited through door-to-door sampling from randomly selected census enumeration areas using a multi-stage sampling approach. Participants were followed up in 2011/12, with 97.9% retention, and were subsequently traced for a third wave of data collection in 2022–2024, when they were aged 20–33 years (INTERRUPT_VIOLENCE). Of the original cohort, 1,304 young adults (Generation 2; G2) participated in the third wave, representing 78.4% retention. The third wave expanded the study to capture experiences and use of violence across three generations. Former primary caregivers of the young adults (Generation 1; G1) and the oldest child aged six years or older whom the G2 participant reported caring for in a parental role (Generation 3; G3) were invited to participate. This resulted in interviews with 602 G1 caregivers and 280 G3 children. Data collection examined experiences and perpetration of violence within and across generations, alongside mental health, parenting and gender norms, household circumstances, economic and food security, and potential protective factors that may interrupt the intergenerational continuity of violence. Young adults were first contacted and invited to consent to participation; following their interview, permission was sought to approach their former caregiver and child, where eligible, and to link information across generations. Where young adults had died, their caregivers could still participate. Children provided assent in addition to parental consent. Information and consent materials were developed for communities with low literacy and adapted for different age groups, and were available in participants’ preferred language (SiSwati, XiTsonga or English). Interviews were conducted in local languages by trained local interviewers with specialist preparation in researching violence with children, adults and across generations. The sensitive nature of the study required extensive safeguarding and participant-support procedures. Two full-time social workers were employed to ensure that urgent and mandated referrals were acted upon and that participants could access follow-on support. All participants received information about local self-referral organisations. In total, 1,358 referrals were received and triaged, of which 765 were attended to by study social workers; approximately one in five participants receiving social-work support had multiple visits. Participants were linked to services addressing food insecurity, mental health, child protection, physical health and lack of birth identification. Findings from the resulting data demonstrate that intergenerational violence is complex but not deterministic. Network analyses identified highly interconnected patterns of violence exposure, perpetration and mental health difficulties, with gendered parenting beliefs and endorsement of corporal punishment emerging as central mechanisms connecting violence and poor mental health across generations. Path analyses further identified caregiver death, HIV-affected household contexts, early pregnancy, suicidality, anxiety, PTSD, substance use, financial stress, inequitable gender beliefs and parenting attitudes as risks for intergenerational continuity, with some pathways differing between women and men. Latent Transition Analyses showed that polyvictimization during adolescence was associated with increased likelihood of being in a high intimate partner and sexual violence class in young adulthood. The third generation data highlighted the continued burden of violence in childhood: 56% of children reported emotional abuse and 55% physical abuse by a caregiver, while 53% experienced repeated peer violence and 21% sibling violence. Community violence was also common (27%), and peer, community and domestic violence were consistently associated with poorer mental health. Importantly, the study also identified multiple potential “interrupters” of intergenerational violence. Educational continuity, food sufficiency, gender-equitable norms, egalitarian parenting beliefs, household economic security, HIV-free household status, school safety, and child interpersonal and environmental safety were associated with reduced violence continuity. No single protective factor weakened all pathways, but combinations of protective conditions were associated with substantially lower probabilities of intergenerational violence, with some combinations reducing risk by up to 50 percentage points. Overall, INTERRUPT_VIOLENCE provides a rare three-generation dataset linking longitudinal experiences from adolescence with young adulthood, parental data and in the next generation. The data demonstrate the importance of examining violence within its wider social, relational, economic and developmental context. While violence can persist across generations, the findings indicate that this continuity is not inevitable. Layered protective conditions across material, relational, normative and environmental domains may help interrupt cycles of violence, highlighting the potential value of integrated prevention strategies that address household stress, gender norms, caregiving practices, economic insecurity and child safety simultaneously.