Revealing the dynamic network structure of the Beck Depression Inventory-II

Abstract

Background: Structured interviews and questionnaires are important tools to screen for major depressive disorder. Recent research suggests that, in addition to studying the mean level of total scores, researchers should focus on the dynamic relations among depressive symptoms as they unfold over time. Using network analysis, this paper is the first to investigate these patterns of short-term (i.e. session to session) dynamics for a widely used psychological questionnaire for depression - the Beck Depression Inventory (BDI-II).

Method: With the newly developed vector autoregressive (VAR) multilevel method we estimated the network of symptom dynamics that characterizes the BDI-II, based on repeated administrations of the questionnaire to a group of depressed individuals who participated in a treatment study of an average of 14 weekly assessments. Also the centrality of symptoms and the community structure of the network were examined.

Results: The analysis showed that all BDI-II symptoms are directly or indirectly connected through patterns of temporal influence. In addition, these influences are mutually reinforcing, 'loss of pleasure' being the most central item in the network. Community analyses indicated that the dynamic structure of the BDI-II involves two clusters, which is consistent with earlier psychometric analyses.

Conclusion: The network approach expands the range of depression research, making it possible to investigate the dynamic architecture of depression and opening up a whole new range of questions and analyses. Regarding clinical practice, network analyses may be used to indicate which symptoms should be targeted, and in this sense may help in setting up treatment strategies.

Keywords: vector autoregressive (VAR).