The Association between Mood, Inhibitory Control and Depressive Symptoms: An Ecological Momentary Assessment Study

Introduction Cognitive models of depression highlight the role of inhibitory control - the cognitive control ability which supports our goal directed behavior – as key and even causal feature of the disorder. According to these models, deficits in inhibitory control prevent the exclusion of irrelevant negative information, leading to rumination and sustained negative mood which result in depressive episodes. However, the scientific evidence linking deficits in inhibitory control to depression is thus far mixed. Moreover, although one’s inhibitory control ability may fluctuate, it is often assessed using a single-time measurement in the lab. Objectives Here we aimed to assess the association between intra-individual fluctuations in inhibitory control measured in ecological settings, daily mood states, and depressive symptoms. Methods N=106 participants (Mean age: 38 ± 10 years; range: 19-62 years; 68% female) reported their depressive symptoms (using the PHQ-9 scale) and completed a mobile version of the Go-NoGo inhibition task at baseline. They then completed a 5-day ecological-momentary-assessment (EMA) protocol, in which they reported their current mood (using the IMS-12 scale) and performed a shortened version of the Go-NoGo task twice/day using a mobile application. Depressive symptoms were assessed again following the 5-day EMA. Hierarchical-linear-modeling (HLM) was applied to examine the association between momentary IC and mood, with post-EMA depressive symptoms as a moderator. Inhibitory control was included as a time-varying predictor for mood in the 1st step, and depressive symptoms post-EMA and their interaction with inhibition were included in the 2nd step. Results At baseline, there were no correlations between depressive symptoms and inhibitory control (rp = .035, n.s). However, individuals with elevated depressive symptoms demonstrated worse and more variable inhibition performance over time (rp = .29, p = .002), as captured in the EMA measures. In addition, participants with more variable inhibitory control performance over time also reported more depressive symptoms at the end of the 5-day period (rp = .27, p = .006). Finally, post-EMA depressive symptoms moderated the association between momentary inhibitory control and daily mood, such that reduced inhibition was associated with more negative mood only for those with lower, but not with higher, depressive symptoms (Figure 1). Image: Conclusions Variable, rather than mere reduced inhibitory control is related to depressive symptoms. Moreover, the role of inhibition in modulating mood differs in non-depressed vs. depressed individuals. These findings contribute to our understanding of inhibition and mood in real life and help account for some of the discrepant findings related to cognitive control models of depression. Future investigations should examine the validity of these outcomes in other, clinical samples. Disclosure of Interest None Declared


EPP0602
The Association between Mood, Inhibitory Control and Depressive Symptoms: An Ecological Momentary Assessment Study M. Nahum1 *, N. Yitzhak 1 , O. Shimony-Mazar 1 , N. Oved2 and O. Bonne3 Introduction: Cognitive models of depression highlight the role of inhibitory control -the cognitive control ability which supports our goal directed behavioras key and even causal feature of the disorder.According to these models, deficits in inhibitory control prevent the exclusion of irrelevant negative information, leading to rumination and sustained negative mood which result in depressive episodes.However, the scientific evidence linking deficits in inhibitory control to depression is thus far mixed.Moreover, although one's inhibitory control ability may fluctuate, it is often assessed using a single-time measurement in the lab.
Objectives: Here we aimed to assess the association between intraindividual fluctuations in inhibitory control measured in ecological settings, daily mood states, and depressive symptoms.Methods: N=106 participants (Mean age: 38 AE 10 years; range: 19-62 years; 68% female) reported their depressive symptoms (using the PHQ-9 scale) and completed a mobile version of the Go-NoGo inhibition task at baseline.They then completed a 5-day ecological-momentary-assessment (EMA) protocol, in which they reported their current mood (using the IMS-12 scale) and performed a shortened version of the Go-NoGo task twice/day using a mobile application.Depressive symptoms were assessed again following the 5-day EMA.Hierarchical-linear-modeling (HLM) was applied to examine the association between momentary IC and mood, with post-EMA depressive symptoms as a moderator.Inhibitory control was included as a time-varying predictor for mood in the 1 st step, and depressive symptoms post-EMA and their interaction with inhibition were included in the 2 nd step.Results: At baseline, there were no correlations between depressive symptoms and inhibitory control (r p = .035,n.s).However, individuals with elevated depressive symptoms demonstrated worse and more variable inhibition performance over time (r p = .29,p = .002),as captured in the EMA measures.In addition, participants with more variable inhibitory control performance over time also reported more depressive symptoms at the end of the 5-day period (r p = .27,p = .006).Finally, post-EMA depressive symptoms moderated the association between momentary inhibitory control and daily mood, such that reduced inhibition was associated with more negative mood only for those with lower, but not with higher, depressive symptoms (Figure 1).

Image:
Conclusions: Variable, rather than mere reduced inhibitory control is related to depressive symptoms.Moreover, the role of inhibition in modulating mood differs in non-depressed vs. depressed individuals.These findings contribute to our understanding of inhibition and mood in real life and help account for some of the discrepant findings related to cognitive control models of depression.Future investigations should examine the validity of these outcomes in other, clinical samples.

Comparison of Machine Learning Algorithms For Beck Depression Inventory Measured Depression Status Classification
Introduction: Depression is a psychiatric disorder characterized by low mood and anhedonia.The diagnosis of depression and the initiation of treatment is important for improving quality of life and avoiding disability.Machine learning (ML), can be used for solving classification and regression problems.In this study, scores of multiple psychiatric scales were used to detect depression and different ML algorithms were used to study if they can help for diagnosing depression accurately.Objectives: The purpose of the study is to detect with high accuracy whether people are depressed or not by using widely used ML algorithms.It is also aimed to compare the algorithms used to predict depression with each other.
Methods: Data were collected from 96 university students.Beck Depression Inventory (BDI), Beck Anxiety Inventory, Neo Personality Inventory, Chronic Stress Scale (CSS), Perceived Stress Scale (PSS), Childhood Trauma Questionnaire, Post-Traumatic Stress Disorder Checklist (PTSD), SHAPS, Relationship Scales Questionnaire and Dissociative Events Scale were applied.14 points from the BDI was accepted as the cut-off value as depressed.Total scores of each scale was used as the dependent variable in the Xgboost (XGB) to classify the depression.By XGB, the most important 4 of these surveys and scales were selected to use in the Non-Linear (NL) models such as XGB, Decision Trees (DT), Support Vector Machines (SVM), K-Nearest Neighbor (KNN).Lastly, a linear model as a Logistic Regression (LR) model was also used to compare with the NL algorithms.The success of the models was measured with the Cross Validation method, which is the gold standard in ML.
Results: In the model in which all measurements are used as Independent Variables (IV), the XGB highlighted 4 scale scores: these are CSS, PSS, SHAPS and PTSD.All scale scores were used as IV, both XGB and DT classified depression with a success of 87.5%, while this score increased to 89.6% in both models when 4 prominent scales' scores were used as IV.In the KNN, the classification made with prominent scales increased the success from 83% to 86%.The variance explanation rate of the LR model using 4 prominent scales remained at 58%.Conclusions: With ML's ability to solve NL relationships and dimensional reduction ability, models in which a large number of variables are input and there is no high correlation between dependent variables and IV can be classified with high success.Also, the success of the models was increased by choosing the most importants of the many IV and the variables that contributed negatively to the model could be excluded.The use of ML can yield promising results in fields such as psychiatry where linear relationships cannot be observed much.Introduction: Treatment Resistant Depression (TRD) occurs in up to 30% of patients with Major Depressive Disorder (MDD).New treatments are clearly needed and there is a burgeoning interest in novel agents including ketamine.While ketamine in various formulations has been demonstrated to have a robust antidepressant effect there is a lack of evidence-based psychotherapies specifically designed for combination use.