Using machine learning-based analysis for behavioral differentiation between anxiety and depression

Thalia Richter,Andrey Markus,Hadas Okon-Singer,Gal Richter-Levin,Barak Fishbain

doi:10.1038/s41598-020-72289-9

Abstract

Anxiety and depression are distinct—albeit overlapping—psychiatric diseases, currently diagnosed by self-reported-symptoms. This research presents a new diagnostic methodology, which tests rigorously for differences in cognitive biases among subclinical anxious and depressed individuals. 125 participants were divided into four groups based on the levels of their anxiety and depression symptoms. A comprehensive behavioral test battery detected and quantified various cognitive–emotional biases. Advanced machine-learning tools, developed for this study, analyzed these results. These tools detect unique patterns that characterize anxiety versus depression to predict group membership. The prediction model for differentiating between symptomatic participants (i.e., high symptoms of depression, anxiety, or both) compared to the non-symptomatic control group revealed a 71.44% prediction accuracy for the former (sensitivity) and 70.78% for the latter (specificity). 68.07% and 74.18% prediction accuracy was obtained for a two-group model with high depression/anxiety, respectively. The analysis also disclosed which specific behavioral measures contributed to the prediction, pointing to key cognitive mechanisms in anxiety versus depression. These results lay the ground for improved diagnostic instruments and more effective and focused individually-based treatment.

Highlights

Anxiety and depression are distinct—albeit overlapping—psychiatric diseases, currently diagnosed by self-reported-symptoms
The current study sought to differentiate between subclinical levels of anxiety and depression by detecting if a unique pattern of biased reactions to emotional stimuli exists for each disorder, based on participants’ aggregated performance in several behavioral tasks
Data were analyzed by machine-learning tools that predicted the group membership of each participant according to his or her performance on the tasks

Summary

Introduction

Anxiety and depression are distinct—albeit overlapping—psychiatric diseases, currently diagnosed by self-reported-symptoms. The analysis disclosed which specific behavioral measures contributed to the prediction, pointing to key cognitive mechanisms in anxiety versus depression These results lay the ground for improved diagnostic instruments and more effective and focused individually-based treatment. The current study sought to examine the possibility of differentiating between anxiety and depression by detecting a unique pattern of biased reactions to emotional stimuli that characterize each disorder. This unique characterization is based on participants’ aggregated performance in several behavioral tasks, which target different cognitive functions known to be abnormal in anxiety and/or depression, and does not rely on self-report measures. The battery focuses on four of the most investigated bias categories: (a) attention biases: distinct sensitivity to threatening stimuli in the environment and difficulty in disregarding emotional d istractors[12,13,14]; (b) memory biases: enhanced remembering of content that is related to the disorder, relative to other information that was previously coded[15]; (c) interpretation biases: a tendency to interpret ambiguous stimuli as threatening or negative to the self[16]; and (d) expectancy biases: the tendency to expect an increased probability of negative events[17]

Methods

Results

Conclusion