My overarching research goal is to characterize the interrelationships between mental health, economic status and HIV intervention participation and their affect on reproductive health related behaviors, understanding these relationships will allow me to plan and evaluate effective interventions in the future. In my dissertation research I propose to examine the role of economic status and mental health symptoms in The SHAZ! Project, a two year randomized controlled trial of a combined life skills, vocational training and social support intervention, with the following specific aims: 1. To determine the effect of different patterns of program adherence on the incidence of mental health symptoms over time. 2. To determine the effect of economic status on mental health symptoms over time. 3. To determine whether mental health symptoms at each time point reduce either adherence to the study intervention in subsequent follow up intervals and/or successful completion of vocational training programs. 4. To determine the effect of mental health symptoms at each time point on sexual risk behaviors in subsequent follow up intervals. 5. To determine the effect of mental health symptoms over the follow up period on specific adverse outcomes: unintended pregnancy, HIV and HSV-2. Mental Health Symptoms will be measured using the Shona Symptom Questionnaire(SSQ), a 14-item indigenous measure of non-psychotic, psychological morbidity. The SSQ has been used widely to measure psychological morbidity. In previous work 79% of women with high SSQ met the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria for clinical depression. It has been validated in Zimbabwe and found to have high internal consistency (r = 0.85). Baseline analysis will include regression analysis of economic status and psychosocial factors on mental health symptoms and will estimate population intervention parameters (as developed by Hubbard and Van der Laan) to rank potential effects of intervention. This parameter can be interpreted as the difference in the mean score of depressive symptoms in a population where all subjects have the ideal level of the risk factor versus the mean mental health symptoms score in the observed population. History adjusted marginal structural modeling (HA-MSM) will be used to analyze prospective data to examine effects of mental health and intervention participation corrected for time dependant confounding and past history of mental health and sexual behavior. Econometric models will implemented and compared to HA-MSMs. Health Relevance: Understanding these effects will provide insight into the success or failure of prevention programs and the extent to which individuals may require mental health services in order to benefit from reproductive health interventions.
