Using language in social media posts to study the network dynamics of depression longitudinally

Abstract

Network theory of mental illness posits that causal interactions between symptoms give rise to mental health disorders. Increasing evidence suggests that depression network connectivity may be a risk factor for transitioning and sustaining a depressive state. Here we analysed social media (Twitter) data from 946 participants who retrospectively self-reported the dates of any depressive episodes in the past 12 months and current depressive symptom severity. We construct personalised, within-subject, networks based on depression-related linguistic features. We show an association existed between current depression severity and 8 out of 9 text features examined. Individuals with greater depression severity had higher overall network connectivity between depression-relevant linguistic features than those with lesser severity. We observed within-subject changes in overall network connectivity associated with the dates of a self-reported depressive episode. The connectivity within personalized networks of depression-associated linguistic features may change dynamically with changes in current depression symptoms.

Possible Futures for Network Psychometrics

“…However, the broader notion of conceptualizing constructs as networks invites the use of other modes of observation, which generate other kinds of data and require other kinds of models. An early extension of this type concerns the analysis of time series of momentary mood states (e.g., experience sampling and other forms of mobile assessment; (Bringmann et al, 2013), but other examples include results from activity monitoring (e.g., actigraphy data, location data), posts on social media (Kelley & Gillan, 2022;Golino et al, 2022), intervention data (Blanken et al, 2019;Waldorp et al, 2021), data on interactions with other individuals (Bodner et al, 2021), and data gathered at different levels of observation (e.g., registrations of brain activity, genetic data, social network data; Blanken et al, 2021).…”

“…Bodner et al (2022) contribute new approaches to the analysis of categorical time series, and Golino et al (2022) extend exploratory graph analysis methods to analyze social media data. Golino et al's (2022) approach yields a highly interesting new way of using network analysis on qualitative bodies of data that I imagine could potentially have many applications, such as attempts to assess network structure based on the analysis of features of intra-individual time series of Twitter posts (Kelley & Gillan, 2022), and could also serve to bridge the traditionally deep divide between qualitative and quantitative data.…”

Association between depressive symptoms and diagnosis of diabetes and its complications: A network analysis in electronic health records

“…We used the Ising models with the extended Bayesian information criterion from the R package bootnet (43) to construct and estimate the centrality of depressive symptoms before and after the diagnosis of each T2DM complication. We set the hyper-parameter tuning to 0 to estimate more connections (28). The weighted networks were illustrated using the R package qgraph (44).…”

“…Based on the network structure of depressive symptoms, the associations between symptoms and disease can be estimated from a part-whole perspective (26). Increases in network connectivity are associated with the severity of depression and persistent depressive symptoms (27,28). Therefore, network analysis has recently been used as an alternative approach to assessing the severity of depressive symptoms.…”

Using Remotely Supervised At-Home TES for Enhancing Mental Resilience

“…This data stream can be recorded and interpreted by dedicated artificial intelligence (AI) systems, with the aim of signaling a potential transition from a healthy to a depressed mental state. Such a virtual mental health service, combining home-use tDCS treatment with a digital sensor and mental monitoring technology, could be used to index a potential decline in mental wellbeing and signal a future relapse episode to the user and their healthcare team (Gillan and Rutledge, 2021 ; Kelley and Gillan, 2022 ), indicating TES-based intervention at the earliest possible time point. Looking further ahead, there is room for optimization in the personalizing of stimulation parameters to make treatment more effective.…”