Background Major Depressive Disorder is a leading cause of disease burden for women of childbearing age. The prevalence of perinatal depression is significantly higher in women from low and middle income countries than from high income countries. Antenatal depression is associated with low birth weight and preterm delivery. There is also evidence that depression during pregnancy can alter the development of a woman's fetus and her child, with an increased risk for child emotional and behavioral problems. Detection of perinatal depression in primary health care (PHC) is very low. This research gathered evidence on the current practice of detection of pregnant women with moderate-severe depression in the Sodo district and explored potential solutions to the identified challenges. Methods A qualitative study was conducted in four health centres and a primary hospital, located in the Sodo district, Ethiopia. In-depth interviews were carried out with pregnant women diagnosed to have depression. Focus group discussions (FGDs) were conducted with antenatal care (ANC) PHC workers. A framework approach to qualitative data analysis was used. OpenCode computer software was used for data handling during the analysis. Results Nine pregnant women were interviewed in-depth and twelve ANC PHC workers participated in two FGDs. Identified themes were categorized under client, provider and system level barriers, and illness related influences. Women's reluctance to disclose symptoms and seek help and not thinking what they had was an illness and/or not thinking they would get help from ANC clinic were client-side barriers to identifying depression. ANC PHC workers not asking about depressive symptoms, their reluctance to disclose a diagnosis of depression to the affected women and not thinking women with mental health problems are in their clinics were the main provider level barriers. The setting not being conducive for depression assessment, inadequate mental health training and ANC assessment format not having a place for recording mental health assessment were system level barriers raised. Difficulty in differentiating clinical depression from normal emotional reaction emerged as the main illness related influence. ANC PHC workers proposed community awareness-raising, improving communication between women and PHC workers and making mental health assessment part of the routine ANC service, among others, as possible solutions to improve detection of antenatal depression in ANC clinics. Conclusion Multiple client, provider and system level barriers to detection of depression in pregnant women in ANC clinics were identified in the current study. The identified barriers call for interventions at different levels; awareness raising at a community level, training on communicating about emotional concerns and mental health care at a provider level and enabling the health care system to support integration of mental health care into maternal health services at a higher level. Follow-up studies are required to better understand the experiences of women and providers, to quantify the magnitude of the problem and to develop and evaluate contextually appropriate interventions to improve detection of depression in pregnant women.

Background: Methamphetamine dependence is associated with impairment in executive function, as well as brain functional and structural alterations, findings on the relationship between executive function impairment and brain alterations seem inconsistent. Methamphetamine dependence may respond to contingency management, yet it is unclear if the treatment response is predicted by these neuropsychological, and brain functional and structural changes, and whether treatment alters neuropsychological impairment. I first conducted a systematic review to rigorously assess available findings on the relationship between executive function impairment and brain functional changes. I then explored data from a study of contingency management in methamphetamine dependence with the aims of determining 1) whether treatment response was predicted by executive function impairment and brain functional and structural alterations, and 2) whether treatment led to changes in executive function and brain functional and structural impairment in treatment responders and non-responders. Methods: The systematic review involved a rigorous search and assessment of articles on the association of stimulant use and resting state functional connectivity. In the empirical study, 33 subjects underwent executive function testing, resting state-fMRI, and structural neuroimaging prior to contingency management treatment. Executive function was assessed with the trail making task, the Stroop-word task, and the Connors continuous performance task. Seed-based analysis was used for functional MRI, with a focus on brain regions associated with executive function, and brain structural alterations were assessed using measures of cortical thickness and surface area. In the statistical analysis, first associations of baseline executive function, rs-fMRI, and brain structural alterations with treatment outcome were assessed using linear regression, and second, comparison of executive function, rsfMRI, and brain structural parameters at baseline versus at treatment end in treatment responders and non-responders was undertaken using linear regression, Cohen's d and a change score. Results: The systematic review noted specific associations between executive function impairment and resting state-fMRI. While in the study, treatment responders had improved executive function at baseline as assessed by two measures (faster completion times on the trail making, and greater accuracy on the Connors continuous performance task), but worse executive function on a third measure (lower accuracy on the Stroop word task) when compared with non-responders. No statistically significant differences between groups was found with regards to rsFC, however greater cortical thickness was found in responders brain regions associated with executive function, in comparison to non-responders. Analysis of pre vs post treatment findings showed that in treatment responders there was better executive function after treatment, in comparison to non-responders (as assessed by greater accuracy on the Connors continuous performance task). Furthermore, in treatment responders there was greater increase in cortical volume in regions associated with executive function, than in non-responders. Conclusion: These findings support the hypothesis that better executive function at baseline (task switching and selective attention) is associated with better outcomes in a contingency management trial of 8-weeks. There is also evidence of improved executive function post trial (in selective attention and cortical thickness findings support improved executive function) implying that abstinence as a consequence of a contingency management trial of 8-weeks may improve executive function, a larger sample size would be needed to determine if improvements extend to other regions of executive function

Background HIV-associated neurocognitive disorder (HAND) is the consequence of the effects of HIV-1 within the central nervous system (CNS). HIV-associated neurocognitive impairment differs in severity with milder forms presenting in 50% of people living with HIV (PLWH), regardless of treatment status. Chronic immune dysregulation has been associated with HAND; in particular, it has been noted that inflammation persists despite the successful treatment with antiretroviral therapy (ART). However, the nature to which (neuro)inflammation influences cognitive performance and brain integrity remain unclear. Further, it is not clear how sequence variation in neurotoxic viral proteins, including Tat, affects inflammation in PLWH. This study aimed to 1) perform a systematic review of the existing literature to identify changes in peripheral immune markers that are associated with HAND in ART-experienced PLWH, 2) determine the association of blood peripheral immune markers with domain-based neurocognitive performance and structural brain changes in South African PLWH, and 3) lastly, to evaluate the possible influence of Tat sequence variation on a dysregulated immune profile in HIV-1C infection (i.e. Tat-C). Methods A systematic review of the published literature was performed to identify the most common markers associated with HAND in the ART-era. A panel of markers was measured in a treatment naïve South African cohort by enzyme-linked immunosorbent assays (ELISA). Cognitive performance was established using a battery of tests sensitive to HIV-associated neurocognitive impairment, with domain based scores utilized in analysis. Thickness and surface area of all cortical regions were derived using automated parcellation of T1-weighted images acquired at 3T. Markers were correlated with neurocognitive performance and cortical thickness and surface area. Further, a prospective review of the literature was performed to determine the association between Tat sequence variation and underlying mechanisms (and inflammation) of HAND. The HIV-1 was genotyped and the influence of Tat sequence variation on immune marker levels was evaluated in a subset of South African participants. Results A systematic review of the existing literature suggested that peripheral immune markers of monocyte activation (sCD14 and sCD163) and inflammation (IL-18 and IP10) were associated with HAND in the majority of studies. Evaluation of blood immune markers in a treatment naïve South African cohort showed that thymidine phosphorylase (TYMP) and neutrophil gelatinase-associated lipocalin (NGAL) levels were significantly higher, while matrix metalloproteinase (MMP)9 levels were significantly lower in PLWH. The results further showed that in PLWH, worse psychomotor processing speed was associated with higher TYMP and NGAL levels and worse motor function was associated with higher NGAL levels. Further, in imaging analysis, it was reported that higher NGAL levels were associated with the reduced thickness of the bilateral orbitofrontal cortex. The association of NGAL withworse motor function was mediated by the cortical thickness of the bilateral orbitofrontal cortex. The associations between higher NGAL and TYMP levels with cortical thickness were largely found in the regions of the frontal cortex. A review of the literature suggests that key protein signatures (C31S and R57S) present in the Tat protein from HIV-1 subtype C (Tat-C) infection may contribute to the lowered inflammation. Supporting this hypothesis, the results from this thesis showed that HIV-1C participants with the R57S mutation had lower peripheral TYMP levels. Conclusions Current literature supports the premise that chronic inflammation may be an important contributor to the development of the milder forms of HAND. For patients on ART, other strategies are required to address the ongoing peripheral inflammation, in addition to simply suppressing the viral load. In a South African context, TYMP and NGAL may be promising markers for their involvement in HAND. Patients were largely treatment-naïve; therefore, these markers may represent HIV related effects without the potential confounding effects of ART. Therefore, these findings may represent long-standing effects which might persist in treatment experienced participants. In HIV-1C infection, the level of certain inflammatory markers may be influenced by the R57S Tat protein signature. To our best knowledge, this is the first thesis to report the association of these markers with HAND. These immune markers need to be investigated for their potential role in the underlying mechanisms of HAND.

Includes bibliographical references. / To repeat the survey performed in 1993 in order to determine whether or not the last 15 years have resulted in a greater number of psychiatrists being more accessible to the majority of the South African people. In addition, attention will be paid to structure of each psychiatrist's practice to determine whether or not there has been any shift towards the proposed changes in the way of working as outlined by Flisher et al in 1997.

Introduction: Over the last few decades, prenatal alcohol exposure (PAE) has been a major public health problem both globally and in low- to-middle-income countries (LMICs) such as South Africa. Pregnant women and new mothers are particularly vulnerable; and PAE may be associated with adverse child neurodevelopmental outcomes. However, few studies have explored the association of PAE, including risk factors, and subsequent neurodevelopmental trajectories over multiple timepoints in the early years. Given the high burden of PAE and associated risk factors, and the relative paucity of empirical data, further work in South African populations is warranted. This thesis aimed to investigate the association between PAE and early neurodevelopmental outcomes in the Drakenstein Child Health Study (DCHS), a South African birth cohort. The specific objectives included: 1. a systematic review on the available longitudinal studies exploring the impact of PAE on language, speech and communication development (Chapter 3 Manuscript 1); 2. an exploration of the association between PAE and motor, language and cognitive outcomes in infancy (Chapter 4- Manuscript 2); 3. an investigation of the association between PAE, including interactions of tobacco smoking exposure, and the neurodevelopmental trajectories (motor, language and cognitive outcomes) of children across the first 4 years of life (Chapter 5 Manuscript 3); 4. a comparison of the conversational turn-taking between mothers and their alcohol exposed children compared to those between mothers and their unexposed children (Chapter 6 Manuscript 4). Methods: This thesis included four publications, three of which present data from the DCHS. Pregnant women were enrolled from two public primary healthcare clinics, Mbekweni (a predominantly black African population) and TC Newman (a mixed-ancestry population), and more than 1000 mother-child dyads were followed longitudinally from birth through the first 5 years of life. For this study, both antenatal and postnatal maternal measures were used to assess moderate-to-severe levels of PAE. These measures included the (i) Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) antenatally, (ii) a retrospective alcohol questionnaire in the postnatal period at 3-6 weeks and/or 24 months testing age. At 6, 24 and 42 months, early neurodevelopmental outcomes were assessed using the Bayley-III Scales of Infant Development (BSID-III), the Kaufman Assessment Battery for Children (KABC-II) or the Peabody Picture Vocabulary Test (PPVT-IV). Conversational turn-taking in mother-child dyads was also assessed at 42 months testing age. Both univariate and multivariate analyses were used to analyse the data. Results: The findings of this thesisshowed that PAE was significantly associated with both fine motor (B=-3.30, 95%CI 0.06-0.46, p=0.001) and gross motor scores (B=-0.30, 95%CI 0.06-0.44 p=0.001) at 6 months (Chapter 4 Manuscript 2). Chapter 5 (Manuscript 3) showed that when accounting for the interaction between prenatal alcohol and tobacco smoking exposure, impaired fine motor functioning occurred up till 24 months (B=-12.59, 95%CI -21.98- -3.19, p=0.01), but these effects attenuated by 42 months. Significant interactions occurred between prenatal alcohol, including tobacco smoking exposure, and impaired receptive vocabulary (B=-2.49, 95%CI -5.24 -0.27, p=0.02) and cognitive functioning at 24 months (B=- 3.25, 95%CI -5.98- -0.52, p=0.02) (Chapter 5 Manuscript 3). Finally, when exploring conversational turn-taking in alcohol exposed mother-child dyads and unexposed dyads, PAE was significantly associated with conversational turn-taking i.e. child overlapping utterances (OR=3.25, CI 0.98-10.76, p=0.050) (Chapter 6 Manuscript 4). Conclusion: The associations of PAE with early neurodevelopmental outcomes shown here expand on the previous literature. Our findings reported that PAE may influence early neurodevelopmental outcomes, however, future studies should include additional longitudinal studies to replicate the findings, and ongoing follow-up of our own cohort may continue clarify the potential association of PAE and additional risk factors on later neurodevelopmental outcomes at school age and beyond. Effective alcohol programmes targeting pregnant women and interventions to address child developmental impairments in this vulnerable cohort are required.

Includes abstract. / Includes bibliographical references. / Second generation HIV surveillance surveys that collect biological and behavioural data from populations of interest is urgently needed to demonstrate accountability to domestic and international HIV funders, monitor trends in HIV and risk behaviours over time, and provide evidence of the effectiveness of HIV prevention efforts.

Background: Few forensic mental health facilities in South Africa use formal risk assessment instruments to monitor risk and recovery of forensic state patients. The study set out to examine the usefulness of the therapeutic alliance as a proxy measure of violent recidivism in a forensic state facility. The study proposed that the nature and strength of the therapeutic alliance is associated with risk of violence in a forensic population and that attachment security is a relevant factor in this relationship. Additional related factors were studied including demographic and contextual data and their influence on the alliance and risk. Design and method: A quantitative research method was used to sample both inpatients and outpatients (n=131) using a naturalistic, cross-sectional research design. Statistical analyses focussed on regression modelling and addressed mainly the statistical associations between ratings of the different variables. Various psychometric tests were administered and scored and entered into a database. It was hypothesised that a strong (positive) therapeutic relationship measured with the Dual- role Relationship Inventory-Revised (DRI-R) questionnaire is associated with low risk for violence using the Historical, Clinical, Risk Management (HCR-20) scale. Findings: Essentially there was a direct association between the therapeutic alliance and violent recidivism, that is, men with a strong therapeutic alliance have lower risk for violent behaviour. Key practitioner type was considered to be highly influential in establishing positive alliances and as a mediator of potential violent recidivism. An insecure attachment style was dominant in the study sample and insecure states of mind partially influenced current therapeutic alliances with an associated increased vulnerability for violence. Positive psychotic symptoms remained a high-risk factor for violence and criminal histories and antisocial behaviour may continue to present a risk for recidivism in the absence of psychosis. Conclusion: Violent recidivism can be adequately monitored by the DRI-R. Affiliation and control are not mutually exclusive in forensic mental health care. Addressing attachment deficits prevalent in this population may be useful in informing both risk and recovery. Symptom reduction remains an important aim in treatment and risk management. Ratings of the alliance by practitioners and how it concurs with risk is an area for further research. Key words: Therapeutic alliance, dual-role relationship, risk assessment, violence, attachment, state patients, recovery

Includes bibliographical references (p. 197-229). / The central hypothesis of this thesis was that HIV psychosis and mania are overlapping manifestations of the neuropathophysiological consequences of HIV characterized by symptoms suggestive of sub-acute delirium and cognitive impairment. It was also hypothesised that HIV-associated mania and psychosis are AIDS-defining features and should be indications for antiretroviral treatment.

Includes abstract. / Includes bibliographical references. / This interdisciplinary study, within the fields of anthropology and medicine, describes my qualitative research as an anthropologist into the mental illness histories of five young Xhosa-speaking men who, within a year of their circumcision, suffered psychotic breakdowns. The study took place in Valkenberg Hospital for the Mentally Unwell and the surrounding disadvantaged, largely Xhosa-speaking communities in Cape Town, South Africa. Numerous admissions were screened over a two-year period, but only these five men met the study criteria. Their case studies describe how they perceived their circumcision and the stresses that may have contributed to the onset of their psychotic illness.

Background: Adults who have an intellectual disability (ID) often continue to live with their parents long after their siblings have left home. While an increasing body of research has described positive parental experiences, research has also found that parents of adults who have ID and behavioural difficulties are more vulnerable to develop parental stress and depression. Aggression is one of the most difficult forms of problem behaviours to manage and could have a negative impact on the parent-child relationship, the child's social inclusivity and the psychological well-being of parents. Method: A case-based psychotherapy design was used to explore maternal experiences among mothers of adults with ID and aggression who access a specialised mental health service in Cape Town. Psychotherapy was used with six participants to attempt to reduce parental stress and other negative psychological states. In an area of research that has received scant attention in South Africa, the study extensively describes the psychotherapy process and the role of contextual factors in the lives of the participating mothers. The study used a mixed methods design which included psychometric measurements that were conducted at various intervals of intervention. Thematic analysis was used in all the case studies and interviews were scheduled before and after completing psychotherapy. External credibility was enhanced through the use of different qualitative strategies that included peer supervision and reflexivity. Findings: Besides elevated parental stress, the majority of participants presented with symptoms of depression and other mental health problems that varied according to their individual profiles. Although parental stress showed a discernible relationship with the child's behavioural difficulties, other significant life stressors contributed to maternal stress and depressive symptoms. Psychotherapy produced only modest improvement of parental stress among some of the participants. However, therapeutic input appeared to be more effective in reducing depressive symptoms among the majority of mothers. Critical reflection and discussion are centred on the clinical implications and meaning of findings on a psychological level.