Discovering the Causal Structure of the Hamilton Rating Scale for Depression Using Causal Discovery

Lu Wang, Mark H. Chignell, Haoyan Jiang, Sachinthya Lokuge, Geneva Mason, Kathryn Fotinos, Martin Katzman

Abstract: Major depressive disorder (MDD) causes many negative consequences including suicide and disability, and is one of the leading preventable causes of death in many countries. The Hamilton Rating Scale for Depression (HAM-D) used in this paper evaluates depression severity based on 17 symptoms of MDD (i.e., the HAM-D17, referred to below as the HAM-D). Studying the relationship amongst MDD symptoms not only provides insight into depressive symptoms but also helps identify subgroups of patients with depression. We employ causal discovery to discover the causal and correlational relationships among HAM-D symptoms. To the best of our knowledge, this is the first study to investigate HAM-D using causal discovery.