Reevaluating the Efficacy and Predictability of Antidepressant Treatments: A Symptom Clustering Approach

Interpersonal Trauma and Sexual Function and Satisfaction: The Mediating Role of Negative Affect Among Survivors of Military Sexual Trauma

Healthy sexual function among women service members/veterans (SM/Vs) is associated with higher quality of life, lower incidence and severity of mental health diagnoses, higher relationship satisfaction, and less frequent suicidal ideation. Although trauma exposure has been established as a predictor of poor sexual function and satisfaction in women SM/Vs, no study to date has examined whether specific trauma types, such as military sexual trauma (MST), increase risk for sexual issues. Moreover, the possible mechanisms of this association have not been explored. The current study examined whether posttraumatic stress disorder (PTSD) and depression symptom clusters mediated the association of trauma type and sexual function and satisfaction in 426 trauma-exposed women SM/Vs. Two hundred seventy participants (63.4%) identified MST as their index trauma. Path analyses demonstrated that MST was related to poorer sexual function and lower satisfaction relative to the other traumas (χ²[28, N = 426] = 43.3, p = 0.03, CFI = 1.00, TLI = 0.99, and RMSEA = 0.04), and this association was mediated by higher non-somatic depressive symptoms and PTSD symptom clusters of anhedonia and negative alterations in cognition and mood (NACM). Causality cannot be inferred due to the cross-sectional nature of the data. However, our findings suggest that interventions aimed at decreasing sexual issues among female SM/Vs with MST should target depressogenic symptoms, whether the origin is depression or PTSD. Longitudinal research exploring the etiological processes that contribute to sexual dysfunction among those with MST is needed.

Specific depression dimensions are associated with a faster rate of cognitive decline in older adults

Introduction

Understanding the relationship between different depression presentations and cognitive outcome may elucidate high-risk sub-groups for cognitive decline.

Methods

In this study we utilized longitudinal data from the National Alzheimer's Coordinating Center (NACC) on 16,743 initially not demented older adults followed every 12 months for an average of 5 years. Depression dimensions were defined based on the 15-item Geriatric Depression Scale (GDS-15), that is, dysphoric mood, Withdrawal-Apathy-Vigor (WAV), anxiety, hopelessness, and subjective memory complaint (SMC).

Results

After adjustment for sociodemographic and clinical covariates, SMC and hopelessness were associated with faster decline in global cognition and all cognitive domains and WAV with decline executive function. Dysphoric mood and anxiety were not associated with a faster cognitive decline in any of the cognitive domains.

Discussion

Different depression dimensions had different associations with the rate of cognitive decline, suggesting distinct pathophysiology and the need for more targeted interventions.

Association of a novel EEG metric of sleep depth/intensity with attention-deficit/hyperactivity, learning, and internalizing disorders and their pharmacotherapy in adolescence

STUDY OBJECTIVES

Psychiatric/learning disorders are associated with sleep disturbances, including those arising from abnormal cortical activity. The odds ratio product (ORP) is a standardized electroencephalogram metric of sleep depth/intensity validated in adults, while ORP data in youth are lacking. We tested ORP as a measure of sleep depth/intensity in adolescents with and without psychiatric/learning disorders.

METHODS

Four hundred eighteen adolescents (median 16 years) underwent a 9-hour, in-lab polysomnography. Of them, 263 were typically developing (TD), 89 were unmedicated, and 66 were medicated for disorders including attention-deficit/hyperactivity (ADHD), learning (LD), and internalizing (ID). Central ORP during non-rapid eye movement (NREM) sleep was the primary outcome. Secondary/exploratory outcomes included central and frontal ORP during NREM stages, in the 9-seconds following arousals (ORP-9), in the first and second halves of the night, during REM sleep and wakefulness.

RESULTS

Unmedicated youth with ADHD/LD had greater central ORP than TD during stage 3 and in central and frontal regions during stage 2 and the second half of the sleep period, while ORP in youth with ADHD/LD on stimulants did not significantly differ from TD. Unmedicated youth with ID did not significantly differ from TD in ORP, while youth with ID on antidepressants had greater central and frontal ORP than TD during NREM and REM sleep, and higher ORP-9.

CONCLUSIONS

The greater ORP in unmedicated youth with ADHD/LD, and normalized levels in those on stimulants, suggests ORP is a useful metric of decreased NREM sleep depth/intensity in ADHD/LD. Antidepressants are associated with greater ORP/ORP-9, suggesting these medications induce cortical arousability.

Affective and somatic symptom clusters in depression and their relationship to treatment outcomes in the STAR*D sample

BACKGROUND

The heterogenous nature of depression continues to stymie efforts to identify biomarkers or predict treatment response. Efforts leveraging large datasets to define more uniform subtypes of depression or subgroups of depressed patients have considered only small subsets of symptoms. We aimed to understand how inclusion of more diverse complaints would impact data-emergent symptom and patient clusters.

METHODS

We applied principal components analysis to baselineInventory of Depressive Symptomatology data from 1491 patients with major depressive disorder to derive naturally co-occurring symptom subsets before utilizing k-means clustering to divide patients into groups based on standardized residuals of each symptom subset score. We evaluated the clinical utility of our approach by comparing how cluster membership impacted response to citalopram.

RESULTS

Prinicpal components analysis identified nine naturally co-occurring symptom subsets: core affective symptoms, appetite/weight loss, anxiety, somatic symptoms, insomnia, negative intrusive thoughts, leaden paralysis/mood quality, diurnal mood variation, and irritability. Cluster analysis identified two patient groups, differing significantly in 7 of 9 c symptom subsets. Patients distinguished by the prominence of somatic versus core affective symptoms exhibited less reduction in depression severity with citalopram treatment.

LIMITATIONS

Results depend not only on raw data, but also parameter selection, and interpretation. Replication is indicated.

CONCLUSIONS

Findings are consistent with previous reports linking somatic symptoms to treatment resistance and demonstrating that SSRIs are most effective in treating affective symptoms. A novel distinction between physical somatic symptoms and psychic anxiety highlights the utility of assessing a broad spectrum of symptoms when exploring heterogeneity in depression and the need for treatments targeting physical somatic symptoms specifically.

Evaluating the efficacy and moderators of algorithm-guided antidepressant treatments of major depressive disorder

BACKGROUND

In spite of numerous options, the most efficacious treatment for major depressive disorder (MDD) remains elusive. Algorithm-guided treatments (AGTs) are proposed to address inadequate remission and optimize treatment delivery. This study aimed to evaluate the clinical benefit of AGTs for MDD, and to explore specific moderators of treatment outcomes for individual patients.

METHODS

The study recruited 987 patients with MDD across eight hospitals who were randomly assigned to AGT with escitalopram (AGT-E), AGT with mirtazapine (AGT-M), or treatment-as-usual (TAU). The outcomes were symptom remission, response rate, early improvement rate, subsymptom clusters improvement over time, the mean time to first remission, relapse rate at 6-months posttreatment follow-up, quality of life (QOL), and adverse events.

RESUTLS

No significant differences were observed across groups in outcome, except that TAU showed significantly poorer QOL, higher relapse rates at 6-months posttreatment follow-up, and marginally significantly worse maximal burden of adverse events than the AGT groups. After 6 weeks of treatment initiation, remission rate did not significantly increase with extended treatment. AGT-M outperformed the TAU and AGT-E in treating sleep symptoms. AGT-E was less effective than AGT-M and TAU in patients with severe depression and somatic symptoms (DSSS). The superiority of TAU over AGTs was observed in recurrent MDD patients.

CONCLUSION

Although the superiority of AGTs over TAU was limited by failure of alternative subsequent treatment, AGTs outperformed in QOL and relapse rate. Types of disease episode and DSSS were regarded as specific moderators in treatment of depression. These findings might contribute to future research on targeted antidepressant treatment.

Associations between depressive symptom clusters and care utilization and costs among community-dwelling older adults

OBJECTIVES

Whether and how symptom clusters are associated with care utilization remains understudied. This study aims to investigate the economic impact of symptom clusters.

METHODS

We conducted cross-sectional analyses of data collected from 3255 older adults aged 60 years and over in Hong Kong using the Patient Health Questionnaire-9 and the Client Service Receipt Inventory to measure depressive symptoms and service utilization to calculate 1-year care expenditure. Based on Research Domain Criteria framework, we categorized depressive symptoms into four clusters: Negative Valance Systems and Externalizing (NVSE; anhedonia and depression), Negative Valance Systems and Internalizing (guilt and self-harm), Arousal and Regulatory Systems (sleep, fatigue, and appetite), and Cognitive and Sensorimotor Systems (CSS; concentration and psychomotor). Two-part models were used with four symptom clusters to estimate economic impacts on care utilization.

RESULTS

Core affective symptoms had the largest economic impact on non-psychiatric care expenditure; a one-point increase in NVSE was associated with USD$ 571 additional non-psychiatric care expenditure. The economic impacts of CSS on non-psychiatric care expenditure was attenuated when the severity level of NVSE was higher.

CONCLUSIONS

Our findings highlight the importance of understanding economic impacts on care utilization based on symptom profiles with a particular emphasis on symptom combinations. Policymakers should optimize care allocation based on older adults' depressive symptom profiles rather than simply considering their depression sum-score or the severity defined by cut-off points.

Comprehensive analysis of endoplasmic reticulum stress and immune infiltration in major depressive disorder

BACKGROUND

Major depressive disorder (MDD) is a life-threatening, debilitating mental health condition. An important factor in the development of depression is endoplasmic reticulum stress (ERS). However, their roles in MDD have not yet been established. The goal of this study was to examine ERS and its underlying molecular mechanisms in MDD.

METHODS

We used data from two microarray datasets (GSE98793 and GSE39653) and the GeneCards database to examine the reticulum stress-related differentially expressed genes (ERSR-DEGs) associated with MDD. Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), Gene Set Enrichment Analysis (GSEA), and Gene Set Variation Analysis (GSVA) were used to further investigate the function and mechanism of ERS in MDD. Moreover, we constructed protein-protein interaction (PPI) networks to identify hub genes as well as the regulatory network of microRNAs (miRNAs), transcription factors (TFs), and potential drugs related to ERSR-DEGs. CIBERSORT was then used to evaluate the immune activity of MDD samples and conduct a correlation analysis between the hub genes and immune cells.

RESULTS

In total, 37 ERSR-DEGs and five hub genes were identified (NCF1, MAPK14, CASP1, CYBA, and TNF). Functional enrichment analysis revealed that ERSR-DEGs were predominantly enriched in inflammation-and immunity-related pathways, such as tumor necrosis factor signaling, NF-κB signaling, and Toll-like receptor signaling pathways. Additionally, 179 miRNAs, 25 TFs, and 15 potential drugs were tested for their interactions with the ERSR-DEGs. CIBERSORT found high proportions of Tregs, monocytes, and macrophages M0 in the MDD samples. Among these, hub genes showed a significant correlation with immune cell infiltration in patients with MDD.

CONCLUSIONS

NCF1, MAPK14, CASP1, CYBA, and TNF are potential ERS-related biomarkers for the diagnosis of MDD. Our research has revealed a significant correlation between immune cells and ERS-related genes with MDD. Not only did our study contribute to a better understanding of the regulatory mechanisms of ERS in underlying MDD pathology, but it also established a paradigm for future studies on ERS.

A Randomized Clinical Trial Comparing Two Treatment Strategies, Evaluating the Meaningfulness of HAM-D Rating Scale in Patients With Major Depressive Disorder

INTRODUCTION

Due to the complexity of symptoms in major depressive disorder (MDD), the majority of depression scales fall short of accurately assessing a patient's progress. When selecting the most appropriate antidepressant treatment in MDD, a multidimensional scale such as the Hamilton Depression Rating scale (HAM-D) may provide clinicians with more information especially when coupled with unidimensional analysis of some key factors such as depressed mood, altered sleep, psychic and somatic anxiety and suicidal ideation etc.

METHODS

HAM-D measurements were carried out in patients with MDD when treated with two different therapeutic interventions. The prespecified primary efficacy variables for the study were changes in score from baseline to the end of the 12 weeks on HAM-D scale (i.e., ≤ 8 or ≥50% response). The study involved three assessment points (baseline, 6 weeks and 12 weeks).

RESULTS

Evaluation of both the absolute HAM-D scores and four factors derived from the HAM-D (depressed mood, sleep, psychic and somatic anxiety and suicidal ideation) revealed that the latter showed a greater promise in gauging the anti-depressant responses.

CONCLUSION

The study confirms the assumption that while both drugs may improve several items on the HAM-D scale, the overall protocol may fall short of addressing the symptoms diversity in MDD and thus the analysis of factor (s) in question might be more relevant and meaningful.

Identifying subtypes of depression in clinician-annotated text: a retrospective cohort study

Current criteria for depression are imprecise and do not accurately characterize its distinct clinical presentations. As a result, its diagnosis lacks clinical utility in both treatment and research settings. Data-driven efforts to refine criteria have typically focused on a limited set of symptoms that do not reflect the disorder's heterogeneity. By contrast, clinicians often write about patients in depth, creating descriptions that may better characterize depression. However, clinical text is not commonly used to this end. Here we show that clinically relevant depressive subtypes can be derived from unstructured electronic health records. Five subtypes were identified amongst 18,314 patients with depression treated at a large mental healthcare provider by using unsupervised machine learning: severe-typical, psychotic, mild-typical, agitated, and anergic-apathetic. Subtypes were used to place patients in groups for validation; groups were found to be associated with future outcomes and characteristics that were consistent with the subtypes. These associations suggest that these categorizations are actionable due to their validity with respect to disease prognosis. Moreover, they were derived with automated techniques that might theoretically be widely implemented, allowing for future analyses in more varied populations and settings. Additional research, especially with respect to treatment response, may prove useful in further evaluation.

Intervening on psychopathology networks: Evaluating intervention targets through simulations

Identifying the different influences of symptoms in dynamic psychopathology models may hold promise for increasing treatment efficacy in clinical applications. Dynamic psychopathology models study the behavioral patterns of symptom networks, where symptoms mutually enforce each other. Interventions could be tailored to specific symptoms that are most effective at lowering symptom activity or that hinder the further development of psychopathology. Simulating interventions in psychopathology network models fits in a novel tradition where symptom-specific perturbations are used as in silico interventions. Here, we present the NodeIdentifyR algorithm (NIRA) to identify the projected most efficient, symptom-specific intervention target in a network model (i.e., the Ising model). We implemented NIRA in a freely available R package. The technique studies the projected effects of symptom-specific interventions by simulating data while symptom parameters (i.e., thresholds) are systematically altered. The projected effect of these interventions is defined in terms of the expected change in overall symptom activity across simulations. With this algorithm, it is possible to study (1) whether symptoms differ in their projected influence on the behavior of the symptom network and, if so, (2) which symptom has the largest projected effect in lowering or increasing overall symptom activation. As an illustration, we apply the algorithm to an empirical dataset containing Post-Traumatic Stress Disorder symptom assessments of participants who experienced the Wenchuan earthquake in 2008. The most important limitations of the method are discussed, as well as recommendations for future research, such as shifting towards modeling individual processes to validate these types of simulation-based intervention methods.

Effectiveness of common antidepressants: a post market release study

BACKGROUND

This study summarizes the experiences of patients, who have multiple comorbidities, with 15 mono-treated antidepressants.

METHODS

This is a retrospective, observational, matched case control study. The cohort was organized using claims data available through OptumLabs for depressed patients treated with antidepressants between January 1, 2001 and December 31, 2018. The cohort included patients from all states within United States of America. The analysis focused on 3,678,082 patients with major depression who had 10,221,145 antidepressant treatments. Using the robust, and large predictors of remission, and propensity to prescribe an antidepressant, the study created 16,770 subgroups of patients. The study reports the remission rate for the antidepressants within the subgroups. The overall impact of antidepressant on remission was calculated as the common odds ratio across the strata.

FINDINGS

The study accurately modelled clinicians' prescription patterns (cross-validated Area under the Receiver Operating Curve, AROC, of 82.0%, varied from 77% to 90%) and patients' remission (cross-validated AROC of 72.0%, varied from 69.5% to 78%). In different strata, contrary to published randomized studies, remission rates differed significantly and antidepressants were not equally effective. For example, in age and gender subgroups, the best antidepressant had an average remission rate of 50.78%, 1.5 times higher than the average antidepressant (30.30% remission rate) and 20 times higher than the worst antidepressant. The Breslow-Day chi-square test for homogeneity showed that across strata a homogenous common odds-ratio did not exist (alpha<0.0001). Therefore, the choice of the optimal antidepressant depended on the strata defined by the patient's medical history.

INTERPRETATION

Study findings may not be appropriate for specific patients. To help clinicians assess the transferability of study findings to specific patient, the web site http://hi.gmu.edu/ad assesses the patient's medical history, finds similar cases in our data, and recommends an antidepressant based on the experience of remission in our data. Patients can share this site's recommendations with their clinicians, who can then assess the appropriateness of the recommendations.

FUNDING

This project was funded by the Robert Wood Johnson foundation grant #76786. The development of related web site was supported by grant 247-02-20 from Virginia's Commonwealth Health Research Board.

Commentary: the ethical challenges of machine learning in psychiatry: a focus on data, diagnosis, and treatment

The clinical interview is the psychiatrist's data gathering procedure. However, the clinical interview is not a defined entity in the way that 'vitals' are defined as measurements of blood pressure, heart rate, respiration rate, temperature, and oxygen saturation. There are as many ways to approach a clinical interview as there are psychiatrists; and trainees can learn as many ways of performing and formulating the clinical interview as there are instructors (Nestler, 1990). Even in the same clinical setting, two clinicians might interview the same patient and conduct very different examinations and reach different treatment recommendations. From the perspective of data science, this mismatch is not one of personal style or idiosyncrasy but rather one of uncertain salience: neither the clinical interview nor the data thereby generated is operationalized and, therefore, neither can be rigorously evaluated, tested, or optimized.

Post-Stroke Depressive Symptoms: Varying Responses to Escitalopram by Individual Symptoms and Lesion Location

OBJECTIVE

The efficacy of antidepressants in post-stroke depressive symptoms (PSD) varies. We aimed to examine whether the effect of escitalopram on PSD differs according to individual depressive symptoms and stroke lesion location.

METHODS

This is a post hoc analysis of EMOTION (ClinicalTrials.gov, NCT01278498), a randomized, placebo-controlled, double-blind trial that examined the efficacy of escitalopram on depression in acute stroke patients (237 with placebo, 241 with escitalopram). Depressive symptoms were evaluated with the 10-item Montgomery-Åsberg Depression Rating Scale (MADRS). Changes in MADRS and individual item scores at 12 weeks were compared between the treatment groups and among the stroke lesion location groups. Stroke lesion locations were grouped according to the anatomical distribution of serotonin fibers that originate from the midbrain/pons and spread to the forebrain via subcortical structures: "Midbrain-Pons," "Frontal-Subcortical," and "Others." Least-squares means were calculated to demonstrate the independent effect of lesion location.

RESULTS

Total MADRS scores decreased more significantly in the escitalopram than in the placebo group, while a significant effect of escitalopram was observed in only 3 items: apparent sadness, reported sadness, pessimistic thoughts. In the lesion location analyses, escitalopram users in the Frontal-Subcortical group showed significant improvement in total MADRS scores (placebo [n = 130] vs. escitalopram [n = 148], least-square mean [95% CI]: -2.3 [-3.5 to -0.2] vs. -4.5 [-5.5 to -3.4], p = .005), while those in the Midbrain-Pons and Others groups did not.

CONCLUSIONS

The effect of escitalopram on PSD may be more prominent in patients with particular depressive symptoms and stroke lesion locations, suggesting the need for tailored treatment strategies.

Behavioral mediators of stress-related mood symptoms in adolescence & young adulthood

BACKGROUND

Stress is a risk factor for unipolar and bipolar mood disorders, but the mechanisms linking stress to specific symptoms remain elusive. Behavioral responses to stress, such as impulsivity and social withdrawal, may mediate the associations between stress and particular mood symptoms.

METHODS

This study evaluated behavioral mediators of the relationship between self-reported intensity of daily stress and mood symptoms over up to eight weeks of daily diary surveys. The sample included individuals with unipolar or bipolar disorders, or with no psychiatric history (n = 113, ages 15-25).

RESULTS

Results showed that higher daily stress was related to higher severity of mania, and this pathway was mediated by impulsive behaviors. Higher stress also predicted higher severity of anhedonic depression, and social withdrawal mediated this relationship. A k-means clustering analysis revealed six subgroups with divergent profiles of stress-behavior-symptom pathways.

LIMITATIONS

Given the observational study design, analyses cannot determine causal relationships amongst these variables. Further work is needed to determine how relationships between these variables may vary based on stressor type, at different timescales, and within different populations.

CONCLUSIONS

Findings support a theoretical model in which impulsivity and social withdrawal act as behavioral mediators of the relationship between stress and mood symptoms. Additionally, distinct patterns of reactivity distinguished subgroups of people vulnerable to particular types of mood symptoms. These results provide novel information about how stress-reactive behaviors relate to specific mood symptoms, which may have clinical relevance as targets of intervention.

Multi-omics driven predictions of response to acute phase combination antidepressant therapy: a machine learning approach with cross-trial replication

Combination antidepressant pharmacotherapies are frequently used to treat major depressive disorder (MDD). However, there is no evidence that machine learning approaches combining multi-omics measures (e.g., genomics and plasma metabolomics) can achieve clinically meaningful predictions of outcomes to combination pharmacotherapy. This study examined data from 264 MDD outpatients treated with citalopram or escitalopram in the Mayo Clinic Pharmacogenomics Research Network Antidepressant Medication Pharmacogenomic Study (PGRN-AMPS) and 111 MDD outpatients treated with combination pharmacotherapies in the Combined Medications to Enhance Outcomes of Antidepressant Therapy (CO-MED) study to predict response to combination antidepressant therapies. To assess whether metabolomics with functionally validated single-nucleotide polymorphisms (SNPs) improves predictability over metabolomics alone, models were trained/tested with and without SNPs. Models trained with PGRN-AMPS' and CO-MED's escitalopram/citalopram patients predicted response in CO-MED's combination pharmacotherapy patients with accuracies of 76.6% (p < 0.01; AUC: 0.85) without and 77.5% (p < 0.01; AUC: 0.86) with SNPs. Then, models trained solely with PGRN-AMPS' escitalopram/citalopram patients predicted response in CO-MED's combination pharmacotherapy patients with accuracies of 75.3% (p < 0.05; AUC: 0.84) without and 77.5% (p < 0.01; AUC: 0.86) with SNPs, demonstrating cross-trial replication of predictions. Plasma hydroxylated sphingomyelins were prominent predictors of treatment outcomes. To explore the relationship between SNPs and hydroxylated sphingomyelins, we conducted multi-omics integration network analysis. Sphingomyelins clustered with SNPs and metabolites related to monoamine neurotransmission, suggesting a potential functional relationship. These results suggest that integrating specific metabolites and SNPs achieves accurate predictions of treatment response across classes of antidepressants. Finally, these results motivate functional investigation into how sphingomyelins might influence MDD pathophysiology, antidepressant response, or both.

Dynamic time warp analysis of individual symptom trajectories in depressed patients treated with electroconvulsive therapy

BACKGROUND

Although electroconvulsive therapy (ECT) effectively improves severity scores of depression, its effects on its individual symptoms has scarcely been studied. We aimed to study which depressive symptom trajectories dynamically cluster together in individuals as well as groups of patients during ECT using Dynamic Time Warp (DTW) analysis.

METHODS

We analysed the standardized weekly scores on the 25-item abbreviated version of the Comprehensive Psychopathological Rating Scale (CPRS) in depressed patients before and during their first six weeks of ECT treatment. DTW analysis was used to analyse the (dis)similarity of time series of items scores at the patient level (300 'DTW distances' per patient) as well as on the group level. Hierarchical cluster, network, and Distatis analyses yielded symptom dimensions.

RESULTS

We included 133 patients, 64.7% female, with an average age of 60.4 years (SD 15.1). Individual DTW distance matrices and networks revealed marked differences in hierarchical and network clusters among patients. Based on cluster analyses of the aggregated matrices, four symptom clusters emerged. In patients who reached remission, the average DTW distance between their symptoms was significantly smaller than non-remitters, reflecting denser symptom networks in remitters than non-remitters (p=0.04).

LIMITATIONS

The assessments were done only weekly during the first six weeks of ECT treatment. The use of individual items of the abbreviated CPRS may have led to measurement error as well as floor and ceiling effects.

CONCLUSION

DTW offers an efficient new approach to analyse symptom trajectories within individuals as well as groups of patients, aiding personalized medicine of psychopathology.

Machine Learning-Based Definition of Symptom Clusters and Selection of Antidepressants for Depressive Syndrome

The current polythetic and operational criteria for major depression inevitably contribute to the heterogeneity of depressive syndromes. The heterogeneity of depressive syndrome has been criticized using the concept of language game in Wittgensteinian philosophy. Moreover, “a symptom- or endophenotype-based approach, rather than a diagnosis-based approach, has been proposed” as the “next-generation treatment for mental disorders” by Thomas Insel. Understanding the heterogeneity renders promise for personalized medicine to treat cases of depressive syndrome, in terms of both defining symptom clusters and selecting antidepressants. Machine learning algorithms have emerged as a tool for personalized medicine by handling clinical big data that can be used as predictors for subtype classification and treatment outcome prediction. The large clinical cohort data from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D), Combining Medications to Enhance Depression Outcome (CO-MED), and the German Research Network on Depression (GRND) have recently began to be acknowledged as useful sources for machine learning-based depression research with regard to cost effectiveness and generalizability. In addition, noninvasive biological tools such as functional and resting state magnetic resonance imaging techniques are widely combined with machine learning methods to detect intrinsic endophenotypes of depression. This review highlights recent studies that have used clinical cohort or brain imaging data and have addressed machine learning-based approaches to defining symptom clusters and selecting antidepressants. Potentially applicable suggestions to realize machine learning-based personalized medicine for depressive syndrome are also provided herein.

[Serotonin-norepinephrine reuptake inhibitors in psychiatry and neurology]

Antidepressants are one of the most important classes of psychotropic drugs and they are widely used in clinical practice, mainly in psychiatry and neurology. The main indications for the use of antidepressants are depression and anxiety disorders. First-line antidepressants are selective serotonin reuptake inhibitors, as well as serotonin-norepinephrine reuptake inhibitors which due to their dual pharmacological action have an additional effect on pain syndromes that determines their use in the treatment of neuropathic pain and fibromyalgia. A special place among the serotonin-norepinephrine reuptake inhibitors has duloxetine, which is characterized by proven efficacy in the treatment of depression, anxiety disorders, as well as isolated and comorbid pain. The optimal balance of efficacy and tolerability determines the possibility of safe use of duloxetine in patients with severe neurological disorders.

Evaluation of Changes in Depression, Anxiety, and Social Anxiety Using Smartphone Sensor Features: Longitudinal Cohort Study

BACKGROUND

The assessment of behaviors related to mental health typically relies on self-report data. Networked sensors embedded in smartphones can measure some behaviors objectively and continuously, with no ongoing effort.

OBJECTIVE

This study aims to evaluate whether changes in phone sensor-derived behavioral features were associated with subsequent changes in mental health symptoms.

METHODS

This longitudinal cohort study examined continuously collected phone sensor data and symptom severity data, collected every 3 weeks, over 16 weeks. The participants were recruited through national research registries. Primary outcomes included depression (8-item Patient Health Questionnaire), generalized anxiety (Generalized Anxiety Disorder 7-item scale), and social anxiety (Social Phobia Inventory) severity. Participants were adults who owned Android smartphones. Participants clustered into 4 groups: multiple comorbidities, depression and generalized anxiety, depression and social anxiety, and minimal symptoms.

RESULTS

A total of 282 participants were aged 19-69 years (mean 38.9, SD 11.9 years), and the majority were female (223/282, 79.1%) and White participants (226/282, 80.1%). Among the multiple comorbidities group, depression changes were preceded by changes in GPS features (Time: r=-0.23, P=.02; Locations: r=-0.36, P<.001), exercise duration (r=0.39; P=.03) and use of active apps (r=-0.31; P<.001). Among the depression and anxiety groups, changes in depression were preceded by changes in GPS features for Locations (r=-0.20; P=.03) and Transitions (r=-0.21; P=.03). Depression changes were not related to subsequent sensor-derived features. The minimal symptoms group showed no significant relationships. There were no associations between sensor-based features and anxiety and minimal associations between sensor-based features and social anxiety.

CONCLUSIONS

Changes in sensor-derived behavioral features are associated with subsequent depression changes, but not vice versa, suggesting a directional relationship in which changes in sensed behaviors are associated with subsequent changes in symptoms.

Artificial Intelligence for Mental Health Care: Clinical Applications, Barriers, Facilitators, and Artificial Wisdom

Artificial intelligence (AI) is increasingly employed in health care fields such as oncology, radiology, and dermatology. However, the use of AI in mental health care and neurobiological research has been modest. Given the high morbidity and mortality in people with psychiatric disorders, coupled with a worsening shortage of mental health care providers, there is an urgent need for AI to help identify high-risk individuals and provide interventions to prevent and treat mental illnesses. While published research on AI in neuropsychiatry is rather limited, there is a growing number of successful examples of AI's use with electronic health records, brain imaging, sensor-based monitoring systems, and social media platforms to predict, classify, or subgroup mental illnesses as well as problems such as suicidality. This article is the product of a study group held at the American College of Neuropsychopharmacology conference in 2019. It provides an overview of AI approaches in mental health care, seeking to help with clinical diagnosis, prognosis, and treatment, as well as clinical and technological challenges, focusing on multiple illustrative publications. Although AI could help redefine mental illnesses more objectively, identify them at a prodromal stage, personalize treatments, and empower patients in their own care, it must address issues of bias, privacy, transparency, and other ethical concerns. These aspirations reflect human wisdom, which is more strongly associated than intelligence with individual and societal well-being. Thus, the future AI or artificial wisdom could provide technology that enables more compassionate and ethically sound care to diverse groups of people.

Application of Machine Learning Methods on Patient Reported Outcome Measurements for Predicting Outcomes: A Literature Review

The field of patient-centred healthcare has, during recent years, adopted machine learning and data science techniques to support clinical decision making and improve patient outcomes. We conduct a literature review with the aim of summarising the existing methodologies that apply machine learning methods on patient-reported outcome measures datasets for predicting clinical outcomes to support further research and development within the field. We identify 15 articles published within the last decade that employ machine learning methods at various stages of exploiting datasets consisting of patient-reported outcome measures for predicting clinical outcomes, presenting promising research and demonstrating the utility of patient-reported outcome measures data for developmental research, personalised treatment and precision medicine with the help of machine learning-based decision-support systems. Furthermore, we identify and discuss the gaps and challenges, such as inconsistency in reporting the results across different articles, use of different evaluation metrics, legal aspects of using the data, and data unavailability, among others, which can potentially be addressed in future studies.

Functional connectivity of the hippocampus in predicting early antidepressant efficacy in patients with major depressive disorder

BAKGROUD

The hippocampus is involved in the pathophysiology of major depressive disorder (MDD), and its structure and function have been reported to be related to the antidepressant response. This study aimed to identify relationships between hippocampal functional connectivity (FC) and early improvement in patients with MDD and to further explore the ability of hippocampal FC to predict early efficacy.

METHODS

Thirty-six patients with nonpsychotic MDD were recruited and underwent resting-state functional magnetic resonance imaging scans at baseline. After two weeks of treatment with escitalopram, patients were divided into subgroups with early improved depression (EID, n= 19) and nonimproved depression (NID, n=17) . A voxelwise FC analysis was performed with the bilateral hippocampus as seeds, two-sample t-tests were used to compare hippocampal FC between groups. Receiver operating characteristic (ROC) curves were constructed to determine the best FC measures and optimal threshold for differentiating EID from END.

RESULTS

The EID group showed significantly higher FC between the left hippocampus and left inferior frontal gyrus and precuneus than the END group. And the left hippocampal FC of these two regions were positively correlated with the reduction ratio of the depressive symptom scores. The ROC curve analysis revealed that summed FC scores for these two regions exhibited the highest area under the curve, with a sensitivity of 0.947 and specificity of 0.882 at a summed score of 0.14.

LIMITATIONS

The sample used in this study was relatively small.

CONCLUSIONS

These findings demonstrated that FC of the left hippocampus can predict early efficacy of antidepressant.

The analysis of network structure among the depressive symptoms in a clinical sample of children and adolescents

The symptoms of depressive disorder in children and adolescents vary widely and have complex interconnections with each other. This study aimed to identify the network structures among individual depressive symptoms in clinically referred children and adolescents. A total of 464 children and adolescents who visited the outpatient psychiatry clinic in South Korea were enrolled. The Children's Depression Inventory (CDI) was used to assess depressive symptoms. To construct the network structure and estimate the centrality indices among individual symptoms, the Gaussian graphical model was utilized with the tuning parameter to minimize the extended Bayesian information criterion. Among all symptoms, self-hatred had the highest strength centrality, followed by crying and self-deprecation. Among 191 valid edges constituting the CDI symptom network, sadness-crying, school work difficulty-school performance decrement, disobedience-fights, misbehavior-low self-esteem, self-deprecation-self-blame, school dislike-lack of friendship, self-hatred-negative body image, anhedonia-social withdrawal, self-hatred-suicidal ideation, crying-irritability, and sadness-loneliness showed significantly higher weights than the other edges. The present study identified the network structure among depressive symptoms in children and adolescents. Future studies including more symptoms of depression are warranted to provide insights into the underlying mechanisms of child and adolescent depression.

Smartphones and the Neuroscience of Mental Health

Improvements in understanding the neurobiological basis of mental illness have unfortunately not translated into major advances in treatment. At this point, it is clear that psychiatric disorders are exceedingly complex and that, in order to account for and leverage this complexity, we need to collect longitudinal data sets from much larger and more diverse samples than is practical using traditional methods. We discuss how smartphone-based research methods have the potential to dramatically advance our understanding of the neuroscience of mental health. This, we expect, will take the form of complementing lab-based hard neuroscience research with dense sampling of cognitive tests, clinical questionnaires, passive data from smartphone sensors, and experience-sampling data as people go about their daily lives. Theory- and data-driven approaches can help make sense of these rich data sets, and the combination of computational tools and the big data that smartphones make possible has great potential value for researchers wishing to understand how aspects of brain function give rise to, or emerge from, states of mental health and illness.

Replication of machine learning methods to predict treatment outcome with antidepressant medications in patients with major depressive disorder from STAR*D and CAN-BIND-1

Objectives

Antidepressants are first-line treatments for major depressive disorder (MDD), but 40–60% of patients will not respond, hence, predicting response would be a major clinical advance. Machine learning algorithms hold promise to predict treatment outcomes based on clinical symptoms and episode features. We sought to independently replicate recent machine learning methodology predicting antidepressant outcomes using the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) dataset, and then externally validate these methods to train models using data from the Canadian Biomarker Integration Network in Depression (CAN-BIND-1) dataset.

Methods

We replicated methodology from Nie et al (2018) using common algorithms based on linear regressions and decision trees to predict treatment-resistant depression (TRD, defined as failing to respond to 2 or more antidepressants) in the STAR*D dataset. We then trained and externally validated models using the clinical features found in both datasets to predict response (≥50% reduction on the Quick Inventory for Depressive Symptomatology, Self-Rated [QIDS-SR]) and remission (endpoint QIDS-SR score ≤5) in the CAN-BIND-1 dataset. We evaluated additional models to investigate how different outcomes and features may affect prediction performance.

Results

Our replicated models predicted TRD in the STAR*D dataset with slightly better balanced accuracy than Nie et al (70%-73% versus 64%-71%, respectively). Prediction performance on our external methodology validation on the CAN-BIND-1 dataset varied depending on outcome; performance was worse for response (best balanced accuracy 65%) compared to remission (77%). Using the smaller set of features found in both datasets generally improved prediction performance when evaluated on the STAR*D dataset.

Conclusion

We successfully replicated prior work predicting antidepressant treatment outcomes using machine learning methods and clinical data. We found similar prediction performance using these methods on an external database, although prediction of remission was better than prediction of response. Future work is needed to improve prediction performance to be clinically useful.

Cluster Analysis of Care Pathways in Adults with Major Depressive Disorder with Acute Suicidal Ideation or Behavior in the USA

BACKGROUND AND OBJECTIVE

Suicidal ideation or behavior are core symptoms of major depressive disorder (MDD). This study aimed to understand heterogeneity among patients with MDD and acute suicidal ideation or behavior.

METHODS

Adults with a diagnosis of MDD on the same day or 6 months before a claim for suicidal ideation or behavior (index date) were identified in the MarketScan^® Databases (10/01/2014-04/30/2019). A mathematical algorithm was used to cluster patients on characteristics of care measured pre-index. Patient care pathways were described by cluster during the 12-month pre-index period and up to 12 months post-index.

RESULTS

Among 38,876 patients with MDD and acute suicidal ideation or behavior, three clusters were identified. Across clusters, pre-index exposure to mental healthcare was revealed as a key differentiator: Cluster 1 (N = 16,025) was least exposed, Cluster 2 (N = 5640) moderately exposed, and Cluster 3 (N = 17,211) most exposed. Patients whose MDD diagnosis was first observed during their index event comprised 86.0% and 72.8% of Clusters 1 and 2, respectively; in Cluster 3, all patients had an MDD diagnosis pre-index. Within 30 days post-index, in Clusters 1, 2, and 3, respectively, 79.3%, 85.2%, and 88.2% used mental health services, including outpatient visits for MDD. Within 12 months post-index, 61.5%, 91.5%, and 84.6% had one or more antidepressant claim, respectively. Per-patient index event costs averaged $5614, $6645, and $5853, respectively.

CONCLUSIONS

Patients with MDD and acute suicidal ideation or behavior least exposed to the healthcare system pre-index similarly received the least care post-index. An opportunity exists to optimize treatment and follow-up with mental health services.

Predictors and moderators of quality of life in patients with major depressive disorder: An AGTs-MDD study report

Effective and targeted interventions for improving quality of life (QOL) in addition to achieving 'clinical remission' are imperatives for patients with major depressive disorder (MDD). This study aimed to examine potential predictors and moderators of QOL in depression. Data were obtained from the Algorithm Guided Treatment Strategies for Major Depressive Disorder (AGTs-MDD) study, a multisite, randomized controlled trial composed of 980 depressed patients. Mixed Model Repeated Measures (MMRM) analyses were conducted to identify baseline characteristics associated with QOL overall (predictors) and their interaction effects (moderators). Severe core depressive, anxiety and pain symptoms were found to be independently associated with poor QOL over the 12-week acute phase treatment. Severe depression, severe anxiety or pain symptoms, or severe suicidal ideation predicted a larger improvement of QOL during acute phase treatment, whereas males showed less improvement. None of the putative moderators were identified except for the educational level. Patients with lower educational level showed a larger improvement of QOL in the AGT started with escitalopram (AGT-E) group and AGT started with mirtazapine (AGT-M) group compared to the treatment as usual (TAU) group. These findings may help to instruct informed decision-making for heterogeneous patients with MDD in the view of full recovery.

The promise of machine learning in predicting treatment outcomes in psychiatry

For many years, psychiatrists have tried to understand factors involved in response to medications or psychotherapies, in order to personalize their treatment choices. There is now a broad and growing interest in the idea that we can develop models to personalize treatment decisions using new statistical approaches from the field of machine learning and applying them to larger volumes of data. In this pursuit, there has been a paradigm shift away from experimental studies to confirm or refute specific hypotheses towards a focus on the overall explanatory power of a predictive model when tested on new, unseen datasets. In this paper, we review key studies using machine learning to predict treatment outcomes in psychiatry, ranging from medications and psychotherapies to digital interventions and neurobiological treatments. Next, we focus on some new sources of data that are being used for the development of predictive models based on machine learning, such as electronic health records, smartphone and social media data, and on the potential utility of data from genetics, electrophysiology, neuroimaging and cognitive testing. Finally, we discuss how far the field has come towards implementing prediction tools in real-world clinical practice. Relatively few retrospective studies to-date include appropriate external validation procedures, and there are even fewer prospective studies testing the clinical feasibility and effectiveness of predictive models. Applications of machine learning in psychiatry face some of the same ethical challenges posed by these techniques in other areas of medicine or computer science, which we discuss here. In short, machine learning is a nascent but important approach to improve the effectiveness of mental health care, and several prospective clinical studies suggest that it may be working already.

Prediction of short-term antidepressant response using probabilistic graphical models with replication across multiple drugs and treatment settings

Heterogeneity in the clinical presentation of major depressive disorder and response to antidepressants limits clinicians' ability to accurately predict a specific patient's eventual response to therapy. Validated depressive symptom profiles may be an important tool for identifying poor outcomes early in the course of treatment. To derive these symptom profiles, we first examined data from 947 depressed subjects treated with selective serotonin reuptake inhibitors (SSRIs) to delineate the heterogeneity of antidepressant response using probabilistic graphical models (PGMs). We then used unsupervised machine learning to identify specific depressive symptoms and thresholds of improvement that were predictive of antidepressant response by 4 weeks for a patient to achieve remission, response, or nonresponse by 8 weeks. Four depressive symptoms (depressed mood, guilt feelings and delusion, work and activities and psychic anxiety) and specific thresholds of change in each at 4 weeks predicted eventual outcome at 8 weeks to SSRI therapy with an average accuracy of 77% (p = 5.5E-08). The same four symptoms and prognostic thresholds derived from patients treated with SSRIs correctly predicted outcomes in 72% (p = 1.25E-05) of 1996 patients treated with other antidepressants in both inpatient and outpatient settings in independent publicly-available datasets. These predictive accuracies were higher than the accuracy of 53% for predicting SSRI response achieved using approaches that (i) incorporated only baseline clinical and sociodemographic factors, or (ii) used 4-week nonresponse status to predict likely outcomes at 8 weeks. The present findings suggest that PGMs providing interpretable predictions have the potential to enhance clinical treatment of depression and reduce the time burden associated with trials of ineffective antidepressants. Prospective trials examining this approach are forthcoming.

Symptom Cluster-Matching Antidepressant Treatment: A Case Series Pilot Study

Despite treating depression with antidepressants, their effectiveness is often insufficient. Comparative effectiveness studies and meta-analyses show the effectiveness of antidepressants; however, they do not provide clear indications as to the choice of a specific antidepressant. The rational choice of antidepressants may be based on matching their mechanisms of action to the symptomatic profiles of depression, reflecting the heterogeneity of symptoms in different patients. The authors presented a series of cases of patients diagnosed with depression in whom at least one previous antidepressant treatment was shown to be ineffective before drug targeted symptom cluster-matching treatment (SCMT). The presented pilot study shows for the first time the effectiveness of SCMT in the different clusters of depressive symptoms. All the described patients obtained recovery from depressive symptoms after introducing drug-targeted SCMT. Once validated in clinical trials, SCMT might become an effective and rational method of selecting an antidepressant according to the individual profile of depressive symptoms, the mechanism of their formation, and the mechanism of drug action. Although the study results are preliminary, SCMT can be a way to personalize treatment, increasing the likelihood of improvement even in patients who meet criteria for treatment-resistant depression.

Parsing the antidepressant effects of non-invasive brain stimulation and pharmacotherapy: A symptom clustering approach on ELECT-TDCS

BACKGROUND

Transcranial direct current stimulation (tDCS) presents small antidepressant efficacy at group level and considerable inter-individual variability of response. Its heterogeneous effects bring the need to investigate whether specific groups of patients submitted to tDCS could present comparable or larger improvement compared to pharmacotherapy. Aggregate measurements might be insufficient to address its effects.

OBJECTIVE

/Hypothesis: To determine the efficacy of tDCS, compared to pharmacotherapy and placebo, in depressive symptom clusters.

METHODS

Data from ELECT-TDCS (Escitalopram versus Electrical Direct-Current Therapy for Treating Depression Clinical Study, ClinicalTrials.gov, NCT01894815), in which antidepressant-free, depressed patients were randomized to receive 22 bifrontal tDCS (2 mA, 30 min) sessions (n = 94), escitalopram 20 mg/day (n = 91), or placebo (n = 60) over 10 weeks. Agglomerative hierarchical clustering identified "sleep/insomnia", "core depressive", "guilt/anxiety", and "atypical" clusters that were the dependent measure. Trajectories were estimated using linear mixed regression models. Effect sizes are expressed in raw HAM-D units. P-values were adjusted for multiple comparisons.

RESULTS

For core depressive symptoms, escitalopram was superior to tDCS (ES = -0.56; CI_95% = -0.94 to -0.17, p = .009), which was superior to placebo (ES = 0.49; CI_95% = 0.06 to 0.92, p = .042). TDCS but not escitalopram was superior to placebo in sleep/insomnia symptoms (ES = 0.87; CI_95% = 0.22 to 1.52, p = .015). Escitalopram but not tDCS was superior to placebo in guilt/anxiety symptoms (ES = 1.66; CI_95% = 0.58 to 2.75, p = .006). No active intervention was superior to placebo for atypical symptoms.

CONCLUSIONS

Pharmacotherapy and non-invasive brain stimulation produce distinct effects in depressive symptoms. TDCS or escitalopram could be chosen according to specific clusters of symptoms for a bigger response.

TRIAL REGISTRATION

ClinicalTrials.gov, NCT01894815.

AI in patient flow: applications of artificial intelligence to improve patient flow in NHS acute mental health inpatient units

Introduction

Growing demand for mental health services, coupled with funding and resource limitations, creates an opportunity for novel technological solutions including artificial intelligence (AI). This study aims to identify issues in patient flow on mental health units and align them with potential AI solutions, ultimately devising a model for their integration at service level.

Method

Following a narrative literature review and pilot interview, 20 semi-structured interviews were conducted with AI and mental health experts. Thematic analysis was then used to analyse and synthesise gathered data and construct an enhanced model.

Results

Predictive variables for length-of-stay and readmission rate are not consistent in the literature. There are, however, common themes in patient flow issues. An analysis identified several potential areas for AI-enhanced patient flow. Firstly, AI could improve patient flow by streamlining administrative tasks and optimising allocation of resources. Secondly, real-time data analytics systems could support clinician decision-making in triage, discharge, diagnosis and treatment stages. Finally, longer-term, development of solutions such as digital phenotyping could help transform mental health care to a more preventative, personalised model.

Conclusions

Recommendations were formulated for NHS trusts open to adopting AI patient flow enhancements. Although AI offers many promising use-cases, greater collaborative investment and infrastructure are needed to deliver clinically validated improvements. Concerns around data-use, regulation and transparency remain, and hospitals must continue to balance guidelines with stakeholder priorities. Further research is needed to connect existing case studies and develop a framework for their evaluation.

Distinct trajectories of antidepressant response to intravenous ketamine

BACKGROUND

The N-methyl-D-aspartate receptor antagonist ketamine is potentially effective in treatment resistant depression. However, its antidepressant efficacy is highly variable, and there is little information about predictors of response.

METHODS

We employed growth mixture modeling (GMM) analysis to examine specific response trajectories to intravenous (IV) ketamine (three infusions; mean dose 0.63 mg/kg, SD 0.28, range 0.30 - 2.98 mg/kg over 40 min) in 328 depressed adult outpatients referred to a community clinic. The Quick Inventory of Depressive Symptomatology-Self-Report (QIDS-SR) assessed depression severity at baseline and before each infusion, up to three infusions for four total observations.

RESULTS

GMM revealed three QIDS-SR response trajectories. There were two groups of severely depressed patients, with contrasting responses to ketamine. One group (n=135, baseline QIDS-SR=18.8) had a robust antidepressant response (final QIDS-SR=7.3); the other group (n=97, QIDS-SR=19.8) was less responsive (final QIDS-SR=15.6). A third group (n=96) was less severely depressed at baseline (QIDS-SR=11.7), with intermediate antidepressant response (final QIDS-SR=6.6). Comparisons of demographic and clinical characteristics between groups with severe baseline depression revealed higher childhood physical abuse in the group with robust ketamine response (p=0.01).

LIMITATIONS

This was a retrospective analysis on a naturalistic sample. Patients were unblinded and more heterogenous than those included in most controlled clinical trial samples. Information pertaining to traumatic events occurring after childhood and pre-existing or concurrent medical conditions that may have affected outcomes was not available.

CONCLUSIONS

Overall, ketamine's effect in patients with severe baseline depression and history of childhood maltreatment may be consistent with ketamine-induced blockade of behavioral sensitization.

Artificial intelligence and machine learning-aided drug discovery in central nervous system diseases: State-of-the-arts and future directions

Neurological disorders significantly outnumber diseases in other therapeutic areas. However, developing drugs for central nervous system (CNS) disorders remains the most challenging area in drug discovery, accompanied with the long timelines and high attrition rates. With the rapid growth of biomedical data enabled by advanced experimental technologies, artificial intelligence (AI) and machine learning (ML) have emerged as an indispensable tool to draw meaningful insights and improve decision making in drug discovery. Thanks to the advancements in AI and ML algorithms, now the AI/ML-driven solutions have an unprecedented potential to accelerate the process of CNS drug discovery with better success rate. In this review, we comprehensively summarize AI/ML-powered pharmaceutical discovery efforts and their implementations in the CNS area. After introducing the AI/ML models as well as the conceptualization and data preparation, we outline the applications of AI/ML technologies to several key procedures in drug discovery, including target identification, compound screening, hit/lead generation and optimization, drug response and synergy prediction, de novo drug design, and drug repurposing. We review the current state-of-the-art of AI/ML-guided CNS drug discovery, focusing on blood-brain barrier permeability prediction and implementation into therapeutic discovery for neurological diseases. Finally, we discuss the major challenges and limitations of current approaches and possible future directions that may provide resolutions to these difficulties.

Individual Differences in Response to Antidepressants: A Meta-analysis of Placebo-Controlled Randomized Clinical Trials

Importance

Antidepressants are commonly used to treat major depressive disorder (MDD). Antidepressant outcomes can vary based on individual differences; however, it is unclear whether specific factors determine this variability or whether it is at random.

Objective

To investigate the assumption of systematic variability in symptomatic response to antidepressants and to assess whether variability is associated with MDD severity, antidepressant class, or study publication year.

Data Sources

Data used were updated from a network meta-analysis of treatment with licensed antidepressants in adults with MDD. The Cochrane Central Register of Controlled Trials, CINAHL, Embase, LILACS database, MEDLINE, MEDLINE In-Process, and PsycInfo were searched from inception to March 21, 2019. Additional sources were international trial registries and sponsors, drug companies and regulatory agencies' websites, and reference lists of published articles. Data were analyzed between June 8, 2020, and June 13, 2020.

Study Selection

Analysis was restricted to double-blind, randomized placebo-controlled trials with depression scores available at the study's end point.

Data Extraction and Synthesis

Baseline means, number of participants, end point means and SDs of total depression scores, antidepressant type, and publication year were extracted.

Main Outcomes and Measures

Log SDs (bln σ̂) were derived for treatment groups (ie, antidepressant and placebo). A random-slope mixed-effects model was conducted to estimate the difference in bln σ̂ between treatment groups while controlling for end point mean. Secondary models determined whether differences in variability between groups were associated with baseline MDD severity; antidepressant class (selective serotonin reuptake inhibitors and other related drugs; serotonin and norepinephrine reuptake inhibitors; norepinephrine-dopamine reuptake inhibitors; noradrenergic agents; or other antidepressants); and publication year.

Results

In the 91 eligible trials (18 965 participants), variability in response did not differ significantly between antidepressants and placebo (bln σ̂, 1.02; 95% CI, 0.99-1.05; P = .19). This finding is consistent with a range of treatment effect SDs (up to 16.10), depending on the association between the antidepressant and placebo effects. Variability was not associated with baseline MDD severity or publication year. Responses to noradrenergic agents were 11% more variable than responses to selective serotonin reuptake inhibitors (bln σ̂, 1.11; 95% CI, 1.01-1.21; P = .02).

Conclusions and Relevance

Although this study cannot rule out the possibility of treatment effect heterogeneity, it does not provide empirical support for personalizing antidepressant treatment based solely on total depression scores. Future studies should explore whether individual symptom scores or biomarkers are associated with variability in response to antidepressants.

Reconciling Statistical and Clinicians' Predictions of Suicide Risk

Statistical models, including those based on electronic health records, can accurately identify patients at high risk for a suicide attempt or death, leading to implementation of risk prediction models for population-based suicide prevention in health systems. However, some have questioned whether statistical predictions can really inform clinical decisions. Appropriately reconciling statistical algorithms with traditional clinician assessment depends on whether predictions from these two methods are competing, complementary, or merely duplicative. In June 2019, the National Institute of Mental Health convened a meeting, "Identifying Research Priorities for Risk Algorithms Applications in Healthcare Settings to Improve Suicide Prevention." Here, participants of this meeting summarize key issues regarding the potential clinical application of suicide prediction models. The authors attempt to clarify the key conceptual and technical differences between traditional risk prediction by clinicians and predictions from statistical models, review the limited evidence regarding both the accuracy of and the concordance between these alternative methods of prediction, present a conceptual framework for understanding agreement and disagreement between statistical and clinician predictions, identify priorities for improving data regarding suicide risk, and propose priority questions for future research. Future suicide risk assessment will likely combine statistical prediction with traditional clinician assessment, but research is needed to determine the optimal combination of these two methods.

Childhood maltreatment history for guiding personalized antidepressant choice in major depressive disorder: Preliminary results from a systematic review

Childhood maltreatment (CM) is a predictor of poor outcome across treatments for major depressive disorder (MDD), while its potential role as a predictor of differential responses to specific antidepressants has received little attention. The present systematic review examined pharmacological studies (published up to June 30th, 2020) that included head-to-head comparisons of antidepressant treatments among adult MDD patients with a reported history of CM or no history to evaluate if CM may help clinicians choose antidepressants with greatest likelihood of successful outcome. Only three studies were included, providing limited and provisional results. These preliminary findings suggest that sustained-release bupropion (alone or in combination) or aripiprazole-augmentation as next-step intervention did not demonstrate differential outcome among MDD patients with or without a history of childhood adversity. Further, sertraline and the group of antidepressants with low affinity for the serotonin transporter may be less suitable for MDD patients with childhood abuse history than escitalopram, venlafaxine-XR, or antidepressants with high affinity for the serotonin transporter. The critical question of the most potentially efficacious treatment regimens for adult MDD with CM history requires further large-sample studies involving a greater number of medications, specifically designed to analyse the moderating effects of different types of CM, and possibly including objective biomarkers.

Applications of artificial intelligence to improve patient flow on mental health inpatient units - Narrative literature review

BACKGROUND

Despite a growing body of research into both Artificial intelligence and mental health inpatient flow issues, few studies adequately combine the two. This review summarises findings in the fields of AI in psychiatry and patient flow from the past 5 years, finds links and identifies gaps for future research.

METHODS

The OVID database was used to access Embase and Medline. Top journals such as JAMA, Nature and The Lancet were screened for other relevant studies. Selection bias was limited by strict inclusion and exclusion criteria.

RESEARCH

3,675 papers were identified in March 2020, of which a limited number focused on AI for mental health unit patient flow. After initial screening, 323 were selected and 83 were subsequently analysed. The literature review revealed a wide range of applications with three main themes: diagnosis (33%), prognosis (39%) and treatment (28%). The main themes that emerged from AI in patient flow studies were: readmissions (41%), resource allocation (44%) and limitations (91%). The review extrapolates those solutions and suggests how they could potentially improve patient flow on mental health units, along with challenges and limitations they could face.

CONCLUSION

Research widely addresses potential uses of AI in mental health, with some focused on its applicability in psychiatric inpatients units, however research rarely discusses improvements in patient flow. Studies investigated various uses of AI to improve patient flow across specialities. This review highlights a gap in research and the unique research opportunity it presents.

The intervention, the patient and the illness - Personalizing non-invasive brain stimulation in psychiatry

Current hypotheses on the therapeutic action of non-invasive brain stimulation (NIBS) in psychiatric disorders build on the abundant data from neuroimaging studies. This makes NIBS a very promising tool for developing personalized interventions within a precision medicine framework. NIBS methods fundamentally vary in their neurophysiological properties. They comprise repetitive transcranial magnetic stimulation (rTMS) and its variants (e.g. theta burst stimulation - TBS) as well as different types of transcranial electrical stimulation (tES), with the largest body of evidence for transcranial direct current stimulation (tDCS). In the last two decades, significant conceptual progress has been made in terms of NIBS targets, i.e. from single brain regions to neural circuits and to functional connectivity as well as their states, recently leading to brain state modulating closed-loop approaches. Regarding structural and functional brain anatomy, NIBS meets an individually unique constellation, which varies across normal and pathophysiological states. Thus, individual constitutions and signatures of disorders may be indistinguishable at a given time point, but can theoretically be parsed along course- and treatment-related trajectories. We address precision interventions on three levels: 1) the NIBS intervention, 2) the constitutional factors of a single patient, and 3) the phenotypes and pathophysiology of illness. With examples from research on depressive disorders, we propose solutions and discuss future perspectives, e.g. individual MRI-based electrical field strength as a proxy for NIBS dosage, and also symptoms, their clusters, or biotypes instead of disorder focused NIBS. In conclusion, we propose interleaved research on these three levels along a general track of reverse and forward translation including both clinically directed research in preclinical model systems, and biomarker guided controlled clinical trials. Besides driving the development of safe and efficacious interventions, this framework could also deepen our understanding of psychiatric disorders at their neurophysiological underpinnings.

Do self-criticism and somatic symptoms play a key role in chronic depression? Exploring the factor structure of Beck depression inventory-II in a sample of chronically depressed inpatients

BACKGROUND

The factor structure of depression differs for different sub-samples. The purpose of this study was to explore the factor structure of Beck Depression Inventory-II in patients with chronic depression presenting for inpatient treatment.

METHODS

Using exploratory structural equation modeling (ESEM), we explored whether a two-factor solution or a bifactor solution provided best model fit for a sample of 377 patients. For the best fitting model stability was assessed with tests for invariance across primary diagnosis (persistent depressive disorder v. recurrent major depressive disorder), and presence of comorbidity.

RESULTS

A bifactor solution with one general factor and two specific factors provided best model fit. Invariance analyses provided support for measurement invariance and stability of the factor solution.

LIMITATIONS

The naturalistic study design implies some uncertainty regarding possible systematic differences between the patients on demographic and clinical characteristics.

CONCLUSION

The factor structure in our sample was best explained by a general depression factor, one specific factor pertaining to self-criticism, and one consisting of the somatic items fatigue, disturbance of sleep, and appetite. Clinicians could benefit from paying special attention to the subfactors identified, as these findings may have implications for treatment choice for patients with chronic depression.

How machine-learning recommendations influence clinician treatment selections: the example of the antidepressant selection

Decision support systems embodying machine learning models offer the promise of an improved standard of care for major depressive disorder, but little is known about how clinicians' treatment decisions will be influenced by machine learning recommendations and explanations. We used a within-subject factorial experiment to present 220 clinicians with patient vignettes, each with or without a machine-learning (ML) recommendation and one of the multiple forms of explanation. We found that interacting with ML recommendations did not significantly improve clinicians' treatment selection accuracy, assessed as concordance with expert psychopharmacologist consensus, compared to baseline scenarios in which clinicians made treatment decisions independently. Interacting with incorrect recommendations paired with explanations that included limited but easily interpretable information did lead to a significant reduction in treatment selection accuracy compared to baseline questions. These results suggest that incorrect ML recommendations may adversely impact clinician treatment selections and that explanations are insufficient for addressing overreliance on imperfect ML algorithms. More generally, our findings challenge the common assumption that clinicians interacting with ML tools will perform better than either clinicians or ML algorithms individually.

Machine-learning-based knowledge discovery in rheumatoid arthritis-related registry data to identify predictors of persistent pain

Early detection of patients with chronic diseases at risk of developing persistent pain is clinically desirable for timely initiation of multimodal therapies. Quality follow-up registries may provide the necessary clinical data; however, their design is not focused on a specific research aim, which poses challenges on the data analysis strategy. Here, machine-learning was used to identify early parameters that provide information about a future development of persistent pain in rheumatoid arthritis (RA). Data of 288 patients were queried from a registry based on the Swedish Epidemiological Investigation of RA. Unsupervised data analyses identified the following 3 distinct patient subgroups: low-, median-, and high-persistent pain intensity. Next, supervised machine-learning, implemented as random forests followed by computed ABC analysis-based item categorization, was used to select predictive parameters among 21 different demographic, patient-rated, and objective clinical factors. The selected parameters were used to train machine-learned algorithms to assign patients pain-related subgroups (1000 random resamplings, 2/3 training, and 1/3 test data). Algorithms trained with 3-month data of the patient global assessment and health assessment questionnaire provided pain group assignment at a balanced accuracy of 70%. When restricting the predictors to objective clinical parameters of disease severity, swollen joint count and tender joint count acquired at 3 months provided a balanced accuracy of RA of 59%. Results indicate that machine-learning is suited to extract knowledge from data queried from pain- and disease-related registries. Early functional parameters of RA are informative for the development and degree of persistent pain.

Dissecting diagnostic heterogeneity in depression by integrating neuroimaging and genetics

Depression is a heterogeneous and etiologically complex psychiatric syndrome, not a unitary disease entity, encompassing a broad spectrum of psychopathology arising from distinct pathophysiological mechanisms. Motivated by a need to advance our understanding of these mechanisms and develop new treatment strategies, there is a renewed interest in investigating the neurobiological basis of heterogeneity in depression and rethinking our approach to diagnosis for research purposes. Large-scale genome-wide association studies have now identified multiple genetic risk variants implicating excitatory neurotransmission and synapse function and underscoring a highly polygenic inheritance pattern that may be another important contributor to heterogeneity in depression. Here, we review various sources of phenotypic heterogeneity and approaches to defining and studying depression subtypes, including symptom-based subtypes and biology-based approaches to decomposing the depression syndrome. We review "dimensional," "categorical," and "hybrid" approaches to parsing phenotypic heterogeneity in depression and defining subtypes using functional neuroimaging. Next, we review recent progress in neuroimaging genetics (correlating neuroimaging patterns of brain function with genetic data) and its potential utility for generating testable hypotheses concerning molecular and circuit-level mechanisms. We discuss how genetic variants and transcriptomic profiles may confer risk for depression by modulating brain structure and function. We conclude by highlighting several promising areas for future research into the neurobiological underpinnings of heterogeneity, including efforts to understand sexually dimorphic mechanisms, the longitudinal dynamics of depressive episodes, and strategies for developing personalized treatments and facilitating clinical decision-making.

Decision Models and Technology Can Help Psychiatry Develop Biomarkers

Why is psychiatry unable to define clinically useful biomarkers? We explore this question from the vantage of data and decision science and consider biomarkers as a form of phenotypic data that resolves a well-defined clinical decision. We introduce a framework that systematizes different forms of phenotypic data and further introduce the concept of decision model to describe the strategies a clinician uses to seek out, combine, and act on clinical data. Though many medical specialties rely on quantitative clinical data and operationalized decision models, we observe that, in psychiatry, clinical data are gathered and used in idiosyncratic decision models that exist solely in the clinician's mind and therefore are outside empirical evaluation. This, we argue, is a fundamental reason why psychiatry is unable to define clinically useful biomarkers: because psychiatry does not currently quantify clinical data, decision models cannot be operationalized and, in the absence of an operationalized decision model, it is impossible to define how a biomarker might be of use. Here, psychiatry might benefit from digital technologies that have recently emerged specifically to quantify clinically relevant facets of human behavior. We propose that digital tools might help psychiatry in two ways: first, by quantifying data already present in the standard clinical interaction and by allowing decision models to be operationalized and evaluated; second, by testing whether new forms of data might have value within an operationalized decision model. We reference successes from other medical specialties to illustrate how quantitative data and operationalized decision models improve patient care.

Combining machine learning algorithms for prediction of antidepressant treatment response

OBJECTIVES

Predictors for unfavorable treatment outcome in major depressive disorder (MDD) applicable for treatment selection are still lacking. The database of a longitudinal multicenter study on 1079 acutely depressed patients, performed by the German research network on depression (GRND), allows supervised and unsupervised learning to further elucidate the interplay of clinical and psycho-sociodemographic variables and their predictive impact on treatment outcome phenotypes.

EXPERIMENTAL PROCEDURES

Treatment response was defined by a change of HAM-D 17-item baseline score ≥50% and remission by the established threshold of ≤7, respectively, after up to eight weeks of inpatient treatment. After hierarchical symptom clustering and stratification by treatment subtypes (serotonin reuptake inhibitors, tricyclic antidepressants, antipsychotic, and lithium augmentation), prediction models for different outcome phenotypes were computed with random forest in a cross-center validation design. In total, 88 predictors were implemented.

RESULTS

Clustering revealed four distinct HAM-D subscores related to emotional, anxious, sleep, and appetite symptoms, respectively. After feature selection, classification models reached moderate to high accuracies up to 0.85. Highest accuracies were observed for the SSRI and TCA subgroups and for sleep and appetite symptoms, while anxious symptoms showed poor predictability.

CONCLUSION

Our results support a decisive role for machine learning in the management of antidepressant treatment. Treatment- and symptom-specific algorithms may increase accuracies by reducing heterogeneity. Especially, predictors related to duration of illness, baseline depression severity, anxiety and somatic symptoms, and personality traits moderate treatment success. However, prospectives application of machine learning models will be necessary to prove their value for the clinic.

Precision clinical trials: a framework for getting to precision medicine for neurobehavioural disorders

The goal of precision medicine (individually tailored treatments) is not being achieved for neurobehavioural conditions such as psychiatric disorders. Traditional randomized clinical trial methods are insufficient for advancing precision medicine because of the dynamic complexity of these conditions. We present a pragmatic solution: the precision clinical trial framework, encompassing methods for individually tailored treatments. This framework includes the following: (1) treatment-targeted enrichment, which involves measuring patients' response after a brief bout of an intervention, and then randomizing patients to a full course of treatment, using the acute response to predict long-term outcomes; (2) adaptive treatments, which involve adjusting treatment parameters during the trial to individually optimize the treatment; and (3) precise measurement, which involves measuring predictor and outcome variables with high accuracy and reliability using techniques such as ecological momentary assessment. This review summarizes precision clinical trials and provides a research agenda, including new biomarkers such as precision neuroimaging, transcranial magnetic stimulation-electroencephalogram digital phenotyping and advances in statistical and machine-learning models. Validation of these approaches - and then widespread incorporation of the precision clinical trial framework - could help achieve the vision of precision medicine for neurobehavioural conditions.

Understanding personalized dynamics to inform precision medicine: a dynamic time warp analysis of 255 depressed inpatients

BACKGROUND

Major depressive disorder (MDD) shows large heterogeneity of symptoms between patients, but within patients, particular symptom clusters may show similar trajectories. While symptom clusters and networks have mostly been studied using cross-sectional designs, temporal dynamics of symptoms within patients may yield information that facilitates personalized medicine. Here, we aim to cluster depressive symptom dynamics through dynamic time warping (DTW) analysis.

METHODS

The 17-item Hamilton Rating Scale for Depression (HRSD-17) was administered every 2 weeks for a median of 11 weeks in 255 depressed inpatients. The DTW analysis modeled the temporal dynamics of each pair of individual HRSD-17 items within each patient (i.e., 69,360 calculated "DTW distances"). Subsequently, hierarchical clustering and network models were estimated based on similarities in symptom dynamics both within each patient and at the group level.

RESULTS

The sample had a mean age of 51 (SD 15.4), and 64.7% were female. Clusters and networks based on symptom dynamics markedly differed across patients. At the group level, five dynamic symptom clusters emerged, which differed from a previously published cross-sectional network. Patients who showed treatment response or remission had the shortest average DTW distance, indicating denser networks with more synchronous symptom trajectories.

CONCLUSIONS

Symptom dynamics over time can be clustered and visualized using DTW. DTW represents a promising new approach for studying symptom dynamics with the potential to facilitate personalized psychiatric care.