Mental Disorders as Causal Systems

Abstract

Debates about posttraumatic stress disorder (PTSD) often turn on whether it is a timeless, cross-culturally valid natural phenomenon or a socially constructed idiom of distress. Most clinicians seem to favor the first view, differing only in whether they conceptualize PTSD as a discrete category or the upper end of a dimension of stress responsiveness. Yet both categorical and dimensional construals presuppose that PTSD symptoms are fallible indicators reflective of an underlying, latent variable. This presupposition has governed psychopathology research for decades, but it rests on problematic psychometric premises. In this article, we review an alternative, network perspective for conceptualizing mental disorders as causal systems of interacting symptoms, and we illustrate this perspective via analyses of PTSD symptoms reported by survivors of the Wenchuan earthquake in China. Finally, we foreshadow emerging computational methods that may disclose the causal structure of mental disorders.

PTSD symptomics: network analyses in the field of psychotraumatology

“…The symptom of feeling emotionally numb was identified as the most central symptom in the network, although the authors also state that the order of centrality estimates must be interpreted with some care due to limited power. Similar to prior studies (McNally et al, 2015) and also results of papers published in this special issue (Spiller et al, 2017), trauma-related amnesia was identified as the least central symptom. When adding covariates to the network, a strong negative connection emerged between social support and sleep problems, and female sex was strongly related to physiological cue reactivity, highlighting the importance of including non-symptom nodes in network models, and justifying the authors’ attempt to determine why and how covariates could be related PTSD symptomatology.…”

“…Symptomics research in the area of PTSD has been scarce so far, although several studies have focused on investigating the network structures of PTSD symptoms (Afzali et al, 2017; Armour, Fried, Deserno, Tsai, & Pietrzak, 2017; Bryant et al, 2017; Fried et al, 2017; McNally et al, 2015; Mitchell et al, 2017). The present special issue of EJPT adds to the literature by curating four additional PTSD network studies, each looking at a different aspect of PTSD.…”

Making connections: exploring the centrality of posttraumatic stress symptoms and covariates after a terrorist attack

“…A small number of studies have explored networks of posttraumatic stress symptoms (Armour, Fried, Deserno, Tsai, & Pietrzak, 2017; Bryant et al, 2016; McNally, 2016; McNally et al, 2015). There were some similarities across the findings of these studies.…”

A network analysis of anger, shame, proposed ICD-11 post-traumatic stress disorder, and different types of childhood trauma in foster care settings in a sample of adult survivors

“…From a network-informed treatment perspective, it may be assumed that the anger symptoms on meaningful past events are the trauma-related promoters that activate the anger rumination subnetwork. These ‘activating’ symptoms are especially important in a symptom network (Cramer et al, 2016) and can foster a beneficial therapeutic cascade when addressed as treatment targets (McNally et al, 2015). Owens, Chard, and Ann Cox (2008) found support for the importance of considering anger in the treatment of PTSD in a veteran sample, where low levels of pretreatment anger predicted low levels of post-treatment PTSD, even for people with higher levels of pretreatment PTSD.…”

“…This model assumes that psychological constructs, such as mental disorders, are reflective and not directly observable entities that can be measured only indirectly by the symptoms they cause (Borsboom & Cramer, 2013). Recent theoretical and empirical considerations have questioned this approach and have put forward another model, the network model, as an alternative basis for modeling psychological constructs and their interactions (Borsboom, 2017; Schmittmann et al, 2013), including PTSD (Armour, Fried, Deserno, Tsai, & Pietrzak, 2017; McNally et al, 2015). In the network model, the associations of variables can be visualized and investigated in an explorative manner on the item/symptom level (e.g.…”