Statistical methods and issues in the study of suicide

Suicide is a major public health problem that both reflects and creates considerable human suffering. Suicide and suicide attempts are leading causes of death and morbidity worldwide at all ages. In the United States, suicide is the 11th leading cause of death, accounting for over 30,000 deaths per year. An even greater number of people attempt suicide. Because suicide is a very complex, multicausal human behavior with many "causes" involving biological as well as psychosocial and cultural components, coupled with extremely low base-rate of suicide behavior in the general population, it is quite a challenging task to model the multiple risk and protective factors and their interactions in an integrated pathway to suicide and suicide attempt. This challenge is further compounded by the limited analytic tools that have been used in the extant literature. In this chapter, we first provide a brief review of conceptually sound as well as accepted models for suicide and suicide attempt and discuss the limitations of statistical methods currently used for studies in this field. We then introduce the structural equation model as a way to integrate multi-faceted and -dimensional outcomes to provide a powerful framework for the complex pathways of suicide and suicide attempt. We illustrate this methodology by applying it to a simulated study sample with the underlying pathway from depression to suicide ideation derived based on the popular diathesis-stress model for pathways of suicide.