Constructing a Causal Model of Pathways from Poverty to Suicide in Nigeria Using Directed Acyclic Graphs and Electronic Health Records

Suicide in Nigeria is embedded in a complex socioeconomic causal web, yet the causal pathways connecting poverty to suicidal behavior have not been rigorously tested using modern epidemiological methods. This study constructs and tests a causal model of pathways from poverty to suicide using directed acyclic graphs (DAGs) and electronic health record (EHR) data from Nigerian tertiary hospitals. Phase one will develop a DAG of the poverty-suicide causal structure through a systematic review and expert consensus workshops with Nigerian epidemiologists, psychiatrists, and sociologists. Phase two will validate the DAG model using EHR data from 5,000 psychiatric patients at five hospitals in Lagos, Kano, Enugu, Abuja, and Benin City, covering a five-year period. Causal inference methods including propensity score matching, instrumental variable analysis, and mediation analysis will test the direct and mediated effects of poverty indicators (household income, food insecurity, unemployment) on suicidal ideation and attempts. Confounders including depression, substance use, and domestic violence will be pre-specified in the DAG. This study constitutes an original methodological contribution to suicide epidemiology in Nigeria, applying cutting-edge causal inference to a multi-site EHR dataset. Keywords: suicide, poverty, directed acyclic graphs, causal inference, Nigeria