Parsing the heterogeneity of Major Depression: Biological subtyping and other statistical approaches to unravel the causes of Major Depression

Major Depression (MD) is the largest contributor to the global burden of disease. Unfortunately, standard pharmacological treatments are not always effective. Combined with the heterogeneity of the patient population, this indicates that there is likely no biological disturbance (e.g., impaired serotonin) underlying depression in all MD patients. Indeed, many potential risk factors for MD have been identified, ranging from genetic and environmental variables to different types of biological disturbances. The first part of this thesis provided more insight into the etiology of MD by using rich datasets and novel methodology to identify the most important predictors of MD. Family history of depression and anxiety was one of the most important predictors of both onset and recurrence of MD. This thesis also showed that the gender gap in MD prevalence arises early in life and remains stable over the lifetime. The second part of this thesis addressed the heterogeneity of the MD population by investigating if and how well studies based on biological data might enable the discovery of more homogeneous subtypes of MD. The results indicated that although subtyping based on etiology and pathophysiology is a promising research avenue, no definitive conclusions can be drawn as of yet. Importantly, the results showed that this kind of subtyping research is a very complex endeavor that requires elaborate and costly data collection as well as intricate research designs that enable the evaluation of the robustness of the model results.