The relationships between poor mental health, child abuse, women's experiences and male perpetration of intimate partner violence in Gauteng province of South Africa

Abstract

Aims and objectives To address, the aforementioned knowledge gaps, this thesis aimed to investigate the relationships between childhood trauma, mental ill-health, women’s experiences and male perpetration of intimate partner violence in Gauteng province of South Africa in 2010. Specific objectives were a) to conduct a global systematic review and meta-analyses to estimate the associations between alcohol abuse, mental ill health symptoms and perpetration of IPV as an outcome; b) to investigate the relationships and pathways between childhood trauma and men’s IPV perpetration including the mediating effects of mental ill health; c) to investigate the inter-relationships of childhood trauma, mental ill health and women’s experiences of IPV and lastly d) investigate the factors associated with resilience that was measured as absence of psychopathology among women survivors of IPV. Existing evidence shows that South Africa has high prevalence of different forms of violent and traumatic exposures which include violence against children (VAC), violence against women (VAW) and interpersonal violence or violent crime which are risk factors for a high burden of mental ill health in the population. VAC and VAW co-occur or intersect by having shared common risk factors and mental ill-health consequences including depression, post-traumatic stress disorder (PTSD), and binge drinking. Yet, many VAC or VAW victims do not report mental ill health symptoms or recover from them more expediently than others, showing different responses to traumatic exposures and resilience. Global reviews of literature suggest that mental ill health may increase women’s risk for repeated or on-going intimate partner violence (IPV) experiences. Research in other settings and especially among military and clinical samples has indicated that mental ill health symptoms increase risk of men’s perpetration of violence against intimate partners. However, pathways between childhood trauma exposures, mental ill health and men’s IPV perpetration or women’s IPV experiences have been little studied from population-based and even less in longitudinal studies conducted in Africa or in low and middle-income country settings (LMICs) where the prevalence of IPV is higher than that documented in the high-income countries. Methods Systematic review, meta-analytic techniques and quantitative research methods were employed in the study. For the systematic review and meta-analysis, studies that were published in peer reviewed journal articles before April 2018 and that reported associations between alcohol abuse, depression or PTSD and IPV were searched for on the PubMed and Science Direct. Only studies measuring IPV in the past year were included in the meta-analysis. Study data were extracted and odds ratios (OR) were pooled using meta-analysis commands in STATA 15. The source of heterogeneity of included studies was further assessed through tests for publication bias. Secondary analysis was conducted on data collected in household survey employing a multi-stage random sampling approach to select 511 women and 416 men from Gauteng, South Africa in 2010. IPV was measured using the WHO Multi-country Study on Women's Health and Domestic Violence Questionnaire. Childhood trauma was measured using a short form of the Childhood Trauma Questionnaire. Depression was measured using the Centre for Epidemiologic Studies Depression Scale (CESD). PTSD symptoms were measured using the Harvard Trauma Questionnaire. Resilience was measured as scoring below the threshold for the Centre for Epidemiological Studies Depression Scale and the Harvard Trauma Questionnaire. Binge drinking was measured using the Alcohol Use Disorders Identification Test (AUDIT) scale. All statistical analyses were conducted in Stata taking into account the survey design. Regression modelling was used to test the associations between variables whilst adjusting for potentially confounding variables identified from literature. The investigation of associations and path analysis of the inter-relationships of variables was conducted using structural equation modelling (SEM) with full information maximum likelihood estimation accounting for missing data. Results In meta-analytic work, 15 studies showed a positive association between alcohol abuse and past year IPV perpetration (OR 1.77 [ 95% CI 1.37-2.28] I2=86.3; p=0.000). Nine studies showed no pooled association between current depressive symptoms and past year IPV perpetration. Four studies showed a positive marginal association between current PTSD symptoms and past year IPV perpetration (OR 1.02 [95%CI 0.96-1.54] I2=83.5%; p=0.000). From the secondary analysis of survey data, 62% reported lifetime IPV perpetration. In contrast, 50% of women reported lifetime experience of IPV The prevalence of men’s IPV perpetration (19.7%) and women’s IPV victimisation (18%) in the past year were comparable. More men participating in the study reported experiencing childhood physical abuse (men 88% vs women 72%) and emotional abuse (men 63% vs women 54%) compared to women. Women reported more childhood neglect experience (men 55% vs women 59%), and child sexual abuse experience (men 20% vs women 28%) compared to men. The trauma event reported more frequently by men was being robbed/carjacked at gunpoint or knifepoint or being kidnapped (49.3%). The most common trauma event reported by women was witnessing abuse of their mother (35.7%) and this was higher than the proportion of men reporting this (25%). A higher proportion of men witnessed murders of someone close (men 26% vs 15.8% women), witnessed murders of strangers (men 34.4% vs 9.7% women,) being robbed or carjacked at gunpoint or knifepoint or being kidnapped (men 49.3% vs women 19.2%), and being close to death (men 28.7% vs women 13.3%). Overall, men reported a higher prevalence of mental ill health symptoms. Thirty six percent of men compared to 14% of women reported binge drinking in the past year. Twenty four percent of men compared to 11.6% of women in our study reported PTSD symptoms. Twenty four percent of men and 23% of women reported depressive symptoms. There was also evidence of co-occurring mental ill health among women and men, with co-occurring symptoms being higher among men than women. Nine percent of women and 14.2% men reported PTSD and depressive symptoms. Five percent of women and 10% of men reported both binge drinking and depressive symptoms. Three percent of women and 10.6% of men reported both binge drinking and PTSD symptoms. Men’s childhood trauma had direct effects on perpetration of IPV and other indirect effects through pathways mediated by personal gender attitudes, other life trauma and mental ill health symptoms. Men’s childhood trauma had direct effects on PTSD symptoms (β 18; p=0.003) and depressive symptoms (β 4.6, p=0.01). Both PTSD and depressive symptoms had direct effects on binge drinking. PTSD had direct effects on IPV perpetration, thus being a mediator of the relationship of childhood trauma or other life trauma and lifetime IPV perpetration. Women’s childhood trauma score was positively associated with PTSD (β 0.45; 95 CI 0.33-0.57; p<0.0001) and depression scores (β 0.39; 95 CI 0.26-0.51; p<0.0001). PTSD and depressive scores fully mediated the relationship of childhood trauma and IPV victimisation. Childhood trauma also had indirect effects on women’s IPV experience through a pathway mediated by women’s power in their current relationship. Women’s IPV victimisation in the past year had direct effects on PTSD (β 0.24; p <0.0001), depressive symptoms (β 0.24; p <0.0001) and binge drinking (β 0.19; p <0.0001). Forty two percent of women who reported lifetime IPV experiences scored below the threshold for PTSD or depressive symptoms at the time of the survey and so were categorized as resilient. Increased resilience was associated with social support indicators i.e women who perceived that their communities were supportive, and they would easily find money in an emergency were more likely to be resilient. Women who binge drank, experienced severe IPV in past 12 months, received negative reactions to disclosure and utilized medical or psycho-social services were less likely to be resilient. Conclusions Our findings show women and men’s high reporting of cumulative and multiple trauma exposures in the life course which include childhood trauma. PTSD symptoms which are the effects of violent exposures, were co-occurring with depression and alcohol and were a central risk factor for men’s perpetration of IPV or women’s experience of IPV. The findings show the importance of mental health response and intervention to the VAW field. Therapies targeting women’s mental disorders and strengthening women’s resilience should be prioritised as an essential component of packages of interventions to improve women’s health and wellbeing, prevent IPV, along with other gender transformative, parenting, and alcohol reduction programmes. Pharmacotherapeutic and psychotherapeutic interventions aimed to treat multiple mental ill health symptoms including PTSD, depression and alcohol abuse hold promise for reducing male-perpetrated IPV in these settings. IPV prevention and violence reduction programmes are critical to reduce the burden of mental ill health in the population. This may entail development or adaptation of multi-component prevention programmes that combine transdiagnostic mental health and gender transformative components in their design. Positive social support from family and perceptions of a supportive community were shown to protect women from the potential mental ill health effects of IPV, therefore women’s resilience could be fostered through supportive responses to IPV victims at all the levels of social capital. The VAW field would benefit immensely from further research investigating women’s psychological resilience the factors associated with it and using the evidence to develop models to foster victim resilience and reduce revictimisation.