A Dimensional Approach for Multivariate Discrete Time Survival Data

This study presents a comparative analysis of traditional Logistic Regression (LR) and Reduced Rank Regression (RRR) methods in the context of multivariate discrete time survival data. The research investigates the impact of childhood adversities (CAs) on the development of adult psychiatric disorders. While the traditional LR method does not account for potential correlations among dependent variables, the RRR approach considers these correlations, offering a more in-depth analysis. The results indicate that childhood adversities, such as parental mental illness and physical illness, have a significant impact, particularly on the development of posttraumatic stress disorder (PTSD). The RRR method allows for a dimensional evaluation of these effects, providing a clearer understanding of how each disorder is affected.