Prediction of Antidepressant Treatment Outcome Using Event-Related Potential in Patients with Major Depressive Disorder

The Relationship Between Theta Power, Theta Asymmetry and the Effect of Escitalopram in the Treatment of Depression

Objective

Only about one-third of depressed patients respond to initial antidepressant treatment. Therefore, it is crucial to find effective predictors of antidepressants. The purpose of our study was to learn the relationship between EEG theta power, theta asymmetry, and the efficacy of escitalopram.

Methods

The study included 34 patients with depression. Before and after each patient's course of treatment, EEG data was gathered. Both the Hamilton Anxiety Scale (HAMA) and the 17-item Hamilton Depression Scale (HAMD-17) were evaluated simultaneously. The natural logarithm of right frontal theta power minus left frontal theta power was used to calculate inter-electrode theta asymmetry (AT).

Results

First, our study found no statistically significant difference between intra-electrode theta power and inter-electrode AT before and after treatment (P ≥ 0.05). When we later looked at the data regarding treatment effects, the findings revealed that patients (n = 9) who did not respond to treatment had lower baseline theta power at C4 [6.190 (2.000, 12.990) vs 15.800 (7.255, 22.330), z = -2.166, P = 0.030]. The two groups had no difference in other electrodes (P ≥ 0.05). The AT of C3/C4 in non-responders (n = 9) was lower [0.012 (0.795) vs 0.733 (0.539), t = -3.224, P = 0.005]. However, there was no difference in inter-electrode AT between the two groups in F3/F4 and F7/F8 (P ≥ 0.05). We finally show that the theta power at C4 was negatively correlated with HAMD scores before treatment (r = -0.346, P = 0.045).

Conclusion

Our findings determined that increased theta power and positive asymmetry in the right frontal-central area correlate with favourable escitalopram treatment, providing a basis for finding predictive markers for antidepressants.

AI-assisted prediction of differential response to antidepressant classes using electronic health records

Antidepressant selection is largely a trial-and-error process. We used electronic health record (EHR) data and artificial intelligence (AI) to predict response to four antidepressants classes (SSRI, SNRI, bupropion, and mirtazapine) 4 to 12 weeks after antidepressant initiation. The final data set comprised 17,556 patients. Predictors were derived from both structured and unstructured EHR data and models accounted for features predictive of treatment selection to minimize confounding by indication. Outcome labels were derived through expert chart review and AI-automated imputation. Regularized generalized linear model (GLM), random forest, gradient boosting machine (GBM), and deep neural network (DNN) models were trained and their performance compared. Predictor importance scores were derived using SHapley Additive exPlanations (SHAP). All models demonstrated similarly good prediction performance (AUROCs ≥ 0.70, AUPRCs ≥ 0.68). The models can estimate differential treatment response probabilities both between patients and between antidepressant classes for the same patient. In addition, patient-specific factors driving response probabilities for each antidepressant class can be generated. We show that antidepressant response can be accurately predicted from real-world EHR data with AI modeling, and our approach could inform further development of clinical decision support systems for more effective treatment selection.

Challenges for Artificial Intelligence in Recognizing Mental Disorders

Artificial Intelligence (AI) appears to be making important advances in the prediction and diagnosis of mental disorders. Researchers have used visual, acoustic, verbal, and physiological features to train models to predict or aid in the diagnosis, with some success. However, such systems are rarely applied in clinical practice, mainly because of the many challenges that currently exist. First, mental disorders such as depression are highly subjective, with complex symptoms, individual differences, and strong socio-cultural ties, meaning that their diagnosis requires comprehensive consideration. Second, there are many problems with the current samples, such as artificiality, poor ecological validity, small sample size, and mandatory category simplification. In addition, annotations may be too subjective to meet the requirements of professional clinicians. Moreover, multimodal information does not solve the current challenges, and within-group variations are greater than between-group characteristics, also posing significant challenges for recognition. In conclusion, current AI is still far from effectively recognizing mental disorders and cannot replace clinicians' diagnoses in the near future. The real challenge for AI-based mental disorder diagnosis is not a technical one, nor is it wholly about data, but rather our overall understanding of mental disorders in general.

Changes in Electric Brain Response to Affective Stimuli in the First Week of Antidepressant Treatment: An Exploratory Study

INTRODUCTION

Asymmetrical alpha and frontal theta activity have been discussed as neurobiological markers for antidepressant treatment response. While most studies focus on resting-state EEG, there is evidence that task-related activity assessed at multiple time points might be superior in detecting subtle early differences.

METHODS

This was a naturalistic study design assessing participants in a psychiatric in- and outpatient hospital setting. We investigated stimulus-related EEG asymmetry (frontal and occipital alpha-1 and alpha-2) and power (frontal midline theta) assessed at baseline and 1 week after initiation of pharmacological depression treatment while presenting affective stimuli. We then compared week 4 responders and nonresponders to antidepressant treatment.

RESULTS

Follow-up analyses of a significant group × emotion × time interaction (p < 0.04) for alpha-1 asymmetry showed that responders differed significantly at baseline in their asymmetry scores in response to sad compared to happy faces with a change in this pattern 1 week later. Nonresponders did not show this pattern. No significant results were found for alpha-2, occipital alpha-1, and occipital alpha-2 asymmetry or frontal midline theta power.

DISCUSSION

Our study addresses the gap in comparisons of task-related EEG activity changes measured at two time points and supports the potential value of this approach in detecting early differences in responders versus nonresponders to pharmacological treatment. Important limitations include the small sample size and the noncontrolled study design.

P300 event-related potentials in patients with different subtypes of depressive disorders

OBJECTIVE

To explore the differences in event-related potentials (ERPs) of the subclinical types of major depressive disorders (MDD): melancholic (MEL), atypical (ATY), and anxious (ANX).

METHODS

Patients with MDD treated in the Clinical Department of Shanghai Mental Health Center between September 2017 and December 2020 were prospectively included. This study was approved by the Ethics Committee of the Shanghai Mental Health Center. They were evaluated using the Mini-International Neuropsychiatric Interview (MINI), 17-item Hamilton Depression Scale (HAMD-17), 30-item Self-rated Inventory of Depressive Symptomatology (IDS-30SR), 16-item Quick Inventory of Negative Symptom Scale (QIDS-16SR), and auditory and visual P300 ERPs.

RESULTS

Finally, 27, 14, and 20 patients with MEL, ATY, and ANX MDD were included in this study, respectively. There were no significant differences in demographic characteristics and HAMD-17, IDS-30SR, and QIDS-16SR total scores among the three groups (all P > 0.05). On the C3 lead, the latency for patients with MEL MDD was the longest, and the latency for patients with ATY MDD was the shortest (MEL vs. ATY vs. ANX: 373.89 ± 6.60 vs. 344.79 ± 9.78 vs. 359.33 ± 7.62, P = 0.039). On the Pz lead, the latency for patients with MEL MDD was the longest, and the latency for patients with ATY MDD was the shortest (MEL vs. ATY vs. ANX: 376.14 ± 6.51 vs. 347.21 ± 9.42 vs. 362.22 ± 8.63, P = 0.047). There were no differences in visual P300 ERPs among the three groups.

CONCLUSION

There are significant differences in auditory C3 and Pz latency among MEL, ATY, and ANX MDD. These differences could help diagnose the subtype of MDD.

Mismatch negativity in patients with major depressive disorder: A meta-analysis

OBJECTIVE

Deficits of mismatch negativity (MMN), a general index of echoic memory function, have been documented in patients with schizophrenia. However, it remains controversial whether patients with major depressive disorder (MDD) demonstrate MMN defects compared with healthy controls (HC).

METHODS

After screening 41 potential studies identified in PubMed and Medline, 13 studies consisting of 343 HC and 339 patients with MDD were included in the present meta-analysis. The effect sizes (Hedges's g) with a random-effect and inverse-variance weighted model were estimated for the MMN amplitudes and latencies. The effects of different deviant types (i.e., frequency and duration) and of different illness stages (i.e., acute and chronic) on MMN were also examined.

RESULTS

We found that 1) MMN amplitudes (g = 1.273, p < 0.001) and latencies (g = 0.303, p = 0.027) to duration, but not frequency deviants, were significantly impaired in patients with MDD compared to HC; 2), acute patients exhibited lower MMN amplitudes (g = 1.735, p < 0.001) and prolonged MMN latencies (g = 0.461, p = 0.007) for the duration deviants compared to HC. Only the attenuated duration MMN amplitudes were detected in patients with chronic MDD (g = 0.822, p = 0.027); and 3) depressive symptoms did not significantly correlate with MMN responses.

CONCLUSIONS

Patients with MDD demonstrated abnormal MMN responses to duration deviants compared to HC.

SIGNIFICANCE

Duration MMN may constitute an electrophysiological indicator to differentiate HC from patients with MDD, particularly those in the acute stage.

A systematic review of the antidepressant effects of curcumin: Beyond monoamines theory

BACKGROUND

Depression is a severe, chronic, and recurring mental health disorder, which prevalence and morbimortality have increased in recent years. Several theories are proposed to elucidate the mechanisms of depression, such as the involvement of inflammation and the release of cytokines. Alternative treatments have been developed to improve outcomes of the commonly used drugs, and the use of Curcuma longa stands out. Its primary compound is named curcumin that exhibits antioxidant and anti-inflammatory effects.

AIMS

Several studies have shown that curcumin may play antidepressant actions and, therefore, this study aimed to perform a systematic review of the antidepressant effects of curcumin to evaluate the impact of this compound in the treatment of this condition.

METHODS

This systematic review has included studies available in MEDLINE-PubMed, EMBASE, and Cochrane databases, and the final selection included 10 randomized clinical trials.

CONCLUSION

Curcumin improves depressant and anxiety behavior in humans. It can increase monoamines and brain-derived neurotrophic factor levels and may inhibit the production of pro-inflammatory cytokines and neuronal apoptosis in the brain. Systemically, curcumin enhanced insulin sensitivity, reduced cortisol levels, and reversed metabolic abnormalities. Studies with larger samples and standardized dose and formulation are required to demonstrate the benefits of curcumin in depression treatment since there are many variations in this compound's use.