While designing and conducting these studies, we became increasingly aware of the importance of an overarching more theoretical question: What would be a suitable view or approach to psychopathology to better understand the similarities and differences of symptoms connected to PGD, PTSD, and depression better? We are aware that there are many approaches to understanding psychopathology, some more statistical, others more conceptual. It is beyond the goal of the dissertation to present them all (Borsboom, 2008; Insel et al., 2010;Wakefield, 1992aWakefield, , 1992b. However, in the following section, we briefly summarize the approaches to psychopathology that we use in this dissertation. The latent construct approach: categorical viewThis approach is the most common understanding of mental disorders in current psychiatry. The latent construct approach assumes that a latent disease entity is responsible for a specific group of symptoms. A well-known example is the symptom 'spitting blood' in a patient with a lung tumor. In this example the lung tumor is the underlying reason for and the latent disease entity behind the fact that the patient spits blood (Borsboom, 2008;Borsboom & Cramer, 2013). In the case of PGD, symptoms like 'persistent yearning' , 'emotional pain' and 'grief-related avoidance' are considered to serve as indicators of the underlying disease/disorder PGD. The latent PGD entity is the common cause that these symptoms (i.e., 'yearning' , 'emotional pain' , and 'grief-related avoidance') are seen in a patient as a coherent syndrome (Robinaugh, LeBlanc, Vuletich, & McNally, 2014).The classification of mental disorders by the American Psychiatric Association in the Diagnostic and Statistical Manual of Mental Disorders uses this latent construct approach. In clinical research, the approach is used in the following ways. First, one can study the existence of the latent disease entities or disorders. This means that you divide a study population in participants who do experience the full disorder and those who do not experience the disorder. Following this, you could investigate which predictor is more common in one of the groups, for example: Dyregrov, Nordanger, and Dyregrov (2003);Saavedra Pérez et al. (2017). Secondly, one may study the separate symptoms and their correlation to the latent disease entity. This could be useful when examining the representativeness of a symptom for the latent disease entity, for example when validating a questionnaire or trying to define the cluster of symptoms that may represent the latent disease entity, for example: Boelen and van den Prigerson et al. (2009). Traumatic loss Chapter 2The prevalence of prolonged grief disorder in bereaved individuals following unnatural losses: Systematic review and meta-regression analysis