Biomarker-based subtyping of depression and anxiety disorders using Latent Class Analysis. A NESDA study

Towards Outcome-Driven Patient Subgroups: A Machine Learning Analysis Across Six Depression Treatment Studies

BACKGROUND

Major depressive disorder (MDD) is a heterogeneous condition; multiple underlying neurobiological and behavioral substrates are associated with treatment response variability. Understanding the sources of this variability and predicting outcomes has been elusive. Machine learning (ML) shows promise in predicting treatment response in MDD, but its application is limited by challenges to the clinical interpretability of ML models, and clinicians often lack confidence in model results. In order to improve the interpretability of ML models in clinical practice, our goal was to demonstrate the derivation of treatment-relevant patient profiles comprised of clinical and demographic information using a novel ML approach.

METHODS

We analyzed data from six clinical trials of pharmacological treatment for depression (total n = 5438) using the Differential Prototypes Neural Network (DPNN), a ML model that derives patient prototypes which can be used to derive treatment-relevant patient clusters while learning to generate probabilities for differential treatment response. A model classifying remission and outputting individual remission probabilities for five first-line monotherapies and three combination treatments was trained using clinical and demographic data. Prototypes were evaluated for interpretability by assessing differences in feature distributions (e.g. age, sex, symptom severity) and treatment-specific outcomes.

RESULTS

A 3-prototype model achieved an area under the receiver operating curve of 0.66 and an expected absolute improvement in remission rate for those receiving the best predicted treatment of 6.5% (relative improvement of 15.6%) compared to the population remission rate. We identified three treatment-relevant patient clusters. Cluster A patients tended to be younger, to have increased levels of fatigue, and more severe symptoms. Cluster B patients tended to be older, female, have less severe symptoms, and the highest remission rates. Cluster C patients had more severe symptoms, lower remission rates, more psychomotor agitation, more intense suicidal ideation, and more somatic genital symptoms.

CONCLUSION

It is possible to produce novel treatment-relevant patient profiles using ML models; doing so may improve interpretability of ML models and the quality of precision medicine treatments for MDD.

Person-centered methods to advance developmental psychopathology

Dante Cicchetti's remarkable contributions to the field of developmental psychopathology include the advancement of key principles such as the interplay of typical and atypical development, multifinality and equifinality, the dynamic processes of resilience, and the integration of multiple levels of analysis into developmental theories. In this paper we assert that person-centered data analytic methods are particularly well-suited to advancing these tenets of developmental psychopathology. We illustrate their utility with a brief novel empirical study focused on underlying patterns of childhood neuroendocrine regulation and prospective links with emerging adult functioning. Results indicate that a childhood neuroendocrine profile marked by high diurnal cortisol paired with low diurnal DHEA was uniquely associated with more adaptive functioning in emerging adulthood. We discuss these findings, and person-centered methods more broadly, within the future of developmental psychopathology.

A phenomenologically grounded specification of varieties of adolescent depression

Researchers are increasingly acknowledging that psychopathological conditions usually grouped together under the generic label "depression" are highly diverse. However, no differential therapeutic approach currently exists that is sensitive to the varieties of depression afflicting young people. In fact, the discussion is missing something much more fundamental: a specification of the types of adolescent depression. Recent research that has aimed to classify different kinds of depression has mainly studied adult populations and predominantly used technically complicated measurements of biological markers. The neglect of the potential particularities of dysphoric disorders affecting youths is unfortunate, and the exclusive focus on biological parameters unnecessarily restrictive. Moreover, this one-sidedness obfuscates more directly available sources of clinically relevant data that could orient conceptualization efforts in child and adolescent psychiatry. Particularly, clues for discriminating different types of adolescent depression may be obtained by analyzing personally articulated accounts of how affected young people experience changes in their relation to the world and to themselves. Thus, here we present and discuss the findings of a study that explored the possibility of specifying types of adolescent depression in a phenomenological way. The study investigated the association between these types and the vicissitudes of personality development. In accounts given by youths diagnosed with depression during semi-structured interviews, we identified themes and examined their phenomenological centrality. Specifically, our qualitative analyses aimed to determine the relative importance of certain themes with respect to the overall intelligibility of the described changes to the relational space. Based on the findings of these analyses, we differentiate three specifiers of adolescent depression and suggest an association between particular types of experiences and the trajectory of affected adolescents' personality development. To our knowledge, this is the first phenomenologically grounded specification of types of adolescent depression with potential therapeutic significance. Thus, based on this contribution, we propose that modes of scientific exploration that are close to phenomenological philosophy-which have been ignored in the context of developmental psychopathology-could offer a foundation to theories developed in the field of child and adolescent psychiatry.

Mental health profiles and the related socio-demographic predictors in Hong Kong university students under the COVID-19 pandemic: A latent class analysis

While the COVID-19 pandemic has brought about significant challenges to mental health of university students, there is limited research in this area. Particularly, few studies examined the person-centered mental health symptom profiles such as depression and anxiety and the related socio-demographic predictors. Using Latent Class Analysis (LCA), this study investigated the symptom profiles of depression and anxiety in university students in Hong Kong under the COVID-19 pandemic and the socio-demographic predictors. A total of 978 undergraduate students completed an online questionnaire including socio-demographic factors and measures of depression and anxiety during the summer of 2022. The LCA identified three latent classes: "normal" group, "moderate comorbid depression and anxiety" group and "severe comorbid depression and anxiety" group. Multinominal logistic regression showed that comparing with the "normal" group and the "moderate symptom" group, the "severe symptom" group had higher personal financial difficulties and individual/family member unemployment during the pandemic. In contrast, other socio-demographic factors (age, gender, year of study, living status, and COVID-19 infection status) had no significant association with group status. The study contributes to understanding of person-centered depression and anxiety symptom profiles and the risk role of personal financial difficulty in mental health of university students under the pandemic.

Comprehensive Sex-Stratified Genetic Analysis of 28 Blood Biomarkers and Depression Reveals a Significant Association between Depression and Low Levels of Total Protein in Females

Introduction

Major depression (MD) is more common amongst women than men, and MD episodes have been associated with fluctuations in reproductive hormones amongst women. To investigate biological underpinnings of heterogeneity in MD, the associations between depression, stratified by sex and including perinatal depression (PND), and blood biomarkers, using UK Biobank (UKB) data, were evaluated, and extended to include the association of depression with biomarker polygenic scores (PGS), generated as proxy for each biomarker.

Method

Using female (N = 39,761) and male (N = 38,821) UKB participants, lifetime MD and PND were tested for association with 28 blood biomarkers. A GWAS was conducted for each biomarker and genetic correlations with depression subgroups were estimated. Using independent data from the Australian Genetics of Depression Study, PGS were constructed for each biomarker, and tested for association with depression status (n [female cases/controls] = 9,006/6,442; n [male cases/controls] = 3,106/6,222). Regions of significant local genetic correlation between depression subgroups and biomarkers highlighted by the PGS analysis were identified.

Results

Depression in females was significantly associated with levels of twelve biomarkers, including total protein (OR = 0.90, CI = [0.86, 0.94], p = 3.9 × 10-6) and vitamin D (OR = 0.94, CI = [0.90, 0.97], p = 2.6 × 10-4), and PND with five biomarker levels, also including total protein (OR = 0.88, CI = [0.81, 0.96], p = 4.7 × 10-3). Depression in males was significantly associated with levels of eleven biomarkers. In the independent Australian Genetics of Depression Study, PGS analysis found significant associations for female depression and PND with total protein (female depression: OR = 0.93, CI = [0.88, 0.98], p = 3.6 × 10-3; PND: OR = 0.91, CI = [0.86, 0.96], p = 1.1 × 10-3), as well as with vitamin D (female depression: OR = 0.93, CI = [0.89, 0.97], p = 2.0 × 10-3; PND: OR = 0.92, CI = [0.87, 0.97], p = 1.4 × 10-3). The male depression sample did not report any significant results, and the point estimate of total protein (OR = 0.98, CI = [0.92-1.04], p = 4.7 × 10-1) did not indicate any association. Local genetic correlation analysis highlighted significant genetic correlation between PND and total protein, located in 5q13.3 (rG = 0.68, CI = [0.33, 1.0], p = 3.6 × 10-4).

Discussion and Conclusion

Multiple lines of evidence from genetic analysis highlight an association between total serum protein levels and depression in females. Further research involving prospective measurement of total protein and depressive symptoms is warranted.

A network analysis of depressive symptoms and metabolomics

BACKGROUND

Depression is associated with metabolic alterations including lipid dysregulation, whereby associations may vary across individual symptoms. Evaluating these associations using a network perspective yields a more complete insight than single outcome-single predictor models.

METHODS

We used data from the Netherlands Study of Depression and Anxiety (N = 2498) and leveraged networks capturing associations between 30 depressive symptoms (Inventory of Depressive Symptomatology) and 46 metabolites. Analyses involved 4 steps: creating a network with Mixed Graphical Models; calculating centrality measures; bootstrapping for stability testing; validating central, stable associations by extra covariate-adjustment; and validation using another data wave collected 6 years later.

RESULTS

The network yielded 28 symptom-metabolite associations. There were 15 highly-central variables (8 symptoms, 7 metabolites), and 3 stable links involving the symptoms Low energy (fatigue), and Hypersomnia. Specifically, fatigue showed consistent associations with higher mean diameter for VLDL particles and lower estimated degree of (fatty acid) unsaturation. These remained present after adjustment for lifestyle and health-related factors and using another data wave.

CONCLUSIONS

The somatic symptoms Fatigue and Hypersomnia and cholesterol and fatty acid measures showed central, stable, and consistent relationships in our network. The present analyses showed how metabolic alterations are more consistently linked to specific symptom profiles.

Network analysis of plasma proteomes in affective disorders

The conventional differentiation of affective disorders into major depressive disorder (MDD) and bipolar disorder (BD) has insufficient biological evidence. Utilizing multiple proteins quantified in plasma may provide critical insight into these limitations. In this study, the plasma proteomes of 299 patients with MDD or BD (aged 19-65 years old) were quantified using multiple reaction monitoring. Based on 420 protein expression levels, a weighted correlation network analysis was performed. Significant clinical traits with protein modules were determined using correlation analysis. Top hub proteins were determined using intermodular connectivity, and significant functional pathways were identified. Weighted correlation network analysis revealed six protein modules. The eigenprotein of a protein module with 68 proteins, including complement components as hub proteins, was associated with the total Childhood Trauma Questionnaire score (r = -0.15, p = 0.009). Another eigenprotein of a protein module of 100 proteins, including apolipoproteins as hub proteins, was associated with the overeating item of the Symptom Checklist-90-Revised (r = 0.16, p = 0.006). Functional analysis revealed immune responses and lipid metabolism as significant pathways for each module, respectively. No significant protein module was associated with the differentiation between MDD and BD. In conclusion, childhood trauma and overeating symptoms were significantly associated with plasma protein networks and should be considered important endophenotypes in affective disorders.

Prediction of Diagnosis and Treatment Response in Adolescents With Depression by Using a Smartphone App and Deep Learning Approaches: Usability Study

BACKGROUND

Lack of quantifiable biomarkers is a major obstacle in diagnosing and treating depression. In adolescents, increasing suicidality during antidepressant treatment further complicates the problem.

OBJECTIVE

We sought to evaluate digital biomarkers for the diagnosis and treatment response of depression in adolescents through a newly developed smartphone app.

METHODS

We developed the Smart Healthcare System for Teens At Risk for Depression and Suicide app for Android-based smartphones. This app passively collected data reflecting the social and behavioral activities of adolescents, such as their smartphone usage time, physical movement distance, and the number of phone calls and text messages during the study period. Our study consisted of 24 adolescents (mean age 15.4 [SD 1.4] years, 17 girls) with major depressive disorder (MDD) diagnosed with Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version and 10 healthy controls (mean age 13.8 [SD 0.6] years, 5 girls). After 1 week's baseline data collection, adolescents with MDD were treated with escitalopram in an 8-week, open-label trial. Participants were monitored for 5 weeks, including the baseline data collection period. Their psychiatric status was measured every week. Depression severity was measured using the Children's Depression Rating Scale-Revised and Clinical Global Impressions-Severity. The Columbia Suicide Severity Rating Scale was administered in order to assess suicide severity. We applied the deep learning approach for the analysis of the data. Deep neural network was employed for diagnosis classification, and neural network with weighted fuzzy membership functions was used for feature selection.

RESULTS

We could predict the diagnosis of depression with training accuracy of 96.3% and 3-fold validation accuracy of 77%. Of the 24 adolescents with MDD, 10 responded to antidepressant treatments. We predicted the treatment response of adolescents with MDD with training accuracy of 94.2% and 3-fold validation accuracy of 76%. Adolescents with MDD tended to move longer distances and use smartphones for longer periods of time compared to controls. The deep learning analysis showed that smartphone usage time was the most important feature in distinguishing adolescents with MDD from controls. Prominent differences were not observed in the pattern of each feature between the treatment responders and nonresponders. The deep learning analysis revealed that the total length of calls received as the most important feature predicting antidepressant response in adolescents with MDD.

CONCLUSIONS

Our smartphone app demonstrated preliminary evidence of predicting diagnosis and treatment response in depressed adolescents. This is the first study to predict the treatment response of adolescents with MDD by examining smartphone-based objective data with deep learning approaches.

Dissection of depression heterogeneity using proteomic clusters

BACKGROUND

The search for relevant biomarkers of major depressive disorder (MDD) is challenged by heterogeneity; biological alterations may vary in patients expressing different symptom profiles. Moreover, most research considers a limited number of biomarkers, which may not be adequate for tagging complex network-level mechanisms. Here we studied clusters of proteins and examined their relation with MDD and individual depressive symptoms.

METHODS

The sample consisted of 1621 subjects from the Netherlands Study of Depression and Anxiety (NESDA). MDD diagnoses were based on DSM-IV criteria and the Inventory of Depressive Symptomatology questionnaire measured endorsement of 30 symptoms. Serum protein levels were detected using a multi-analyte platform (171 analytes, immunoassay, Myriad RBM DiscoveryMAP 250+). Proteomic clusters were computed using weighted correlation network analysis (WGCNA).

RESULTS

Six proteomic clusters were identified, of which one was nominally significantly associated with current MDD (p = 9.62E-03, Bonferroni adj. p = 0.057). This cluster contained 21 analytes and was enriched with pathways involved in inflammation and metabolism [including C-reactive protein (CRP), leptin and insulin]. At the individual symptom level, this proteomic cluster was associated with ten symptoms, among which were five atypical, energy-related symptoms. After correcting for several health and lifestyle covariates, hypersomnia, increased appetite, panic and weight gain remained significantly associated with the cluster.

CONCLUSIONS

Our findings support the idea that alterations in a network of proteins involved in inflammatory and metabolic processes are present in MDD, but these alterations map predominantly to clinical symptoms reflecting an imbalance between energy intake and expenditure.

Latent class analysis of psychotic-affective disorders with data-driven plasma proteomics

Data-driven approaches to subtype transdiagnostic samples are important for understanding heterogeneity within disorders and overlap between disorders. Thus, this study was conducted to determine whether plasma proteomics-based clustering could subtype patients with transdiagnostic psychotic-affective disorder diagnoses. The study population included 504 patients with schizophrenia, bipolar disorder, and major depressive disorder and 160 healthy controls, aged 19 to 65 years. Multiple reaction monitoring was performed using plasma samples from each individual. Pathologic peptides were determined by linear regression between patients and healthy controls. Latent class analysis was conducted in patients after peptide values were stratified by sex and divided into tertile values. Significant demographic and clinical characteristics were determined for the latent clusters. The latent class analysis was repeated when healthy controls were included. Twelve peptides were significantly different between the patients and healthy controls after controlling for significant covariates. Latent class analysis based on these peptides after stratification by sex revealed two distinct classes of patients. The negative symptom factor of the Brief Psychiatric Rating Scale was significantly different between the classes (t = -2.070, p = 0.039). When healthy controls were included, two latent classes were identified, and the negative symptom factor of the Brief Psychiatric Rating Scale was still significant (t = -2.372, p = 0.018). In conclusion, negative symptoms should be considered a significant biological aspect for understanding the heterogeneity and overlap of psychotic-affective disorders.

Late-life depression, allostatic load, and risk of dementia: The AGES-Reykjavik study

BACKGROUND

The current study aimed to assess if the relation between depression and dementia could be explained by allostatic load (AL) profiles, as well as assessing their risk on incident all-cause dementia, Alzheimer's disease (AD), and non-AD dementias.

METHODS

The study included individuals without dementia at baseline from the population-based AGES-Reykjavik Study. Depressive symptoms assessed with the Geriatric Depression Scale-15 and AL markers were collected at baseline. Latent profile analysis (LPA) was performed on the AL markers. Incident dementia was measured during 12-years of follow-up. Cox regressions adjusted for AL profiles were performed to evaluate if AL could explain the relation between depressive symptoms and incident dementia. Additional Cox regressions exploring the interaction with depressive symptoms and AL profiles were also performed.

RESULTS

LPA revealed four profiles based on AL factors: 'Low cardiovascular dysregulation' (43 %), 'Average' (42 % prevalence), 'High cardiovascular dysregulation' (11 %), and 'Multisystem dysregulation' (4 %). Cox regression analyses found an increased risk for dementia in the 'Multisystem dysregulation' group (HR 1.72; 95 % CI 1.26-2.33), as well as for AD (HR 1.75; 95 % CI: 1.12-2.71) and non-AD dementias (HR 1.87; 95 % CI: 1.23-2.84). AL profiles did not mediate the risk of all-cause dementia with depressive symptoms; however, there was evidence of additive interaction with depressive symptoms and the 'Multisystem dysregulation' profile and all-cause dementia (RERI 0.15; 95 % CI 0.03-0.26).

CONCLUSION

AL profiles and depressive symptoms were independently related to dementia. Individuals with multisystem dysregulation could be more susceptible to the negative effects of depressive symptomology on incident dementia.

Metabolomics dissection of depression heterogeneity and related cardiometabolic risk

BACKGROUND

A recent hypothesis postulates the existence of an 'immune-metabolic depression' (IMD) dimension characterized by metabolic dysregulations. Combining data on metabolomics and depressive symptoms, we aimed to identify depressions associated with an increased risk of adverse metabolic alterations.

METHOD

Clustering data were from 1094 individuals with major depressive disorder in the last 6 months and measures of 149 metabolites from a ¹H-NMR platform and 30 depressive symptoms (IDS-SR30). Canonical correlation analyses (CCA) were used to identify main independent metabolite-symptom axes of variance. Then, for the replication, we examined the association of the identified dimensions with metabolites from the same platform and cardiometabolic diseases in an independent population-based cohort (n = 6572).

RESULTS

CCA identified an overall depression dimension and a dimension resembling IMD, in which symptoms such as sleeping too much, increased appetite, and low energy level had higher relative loading. In the independent sample, the overall depression dimension was associated with lower cardiometabolic risk, such as (i.e. per s.d.) HOMA-1B -0.06 (95% CI -0.09 - -0.04), and visceral adipose tissue -0.10 cm² (95% CI -0.14 - -0.07). In contrast, the IMD dimension was associated with well-known cardiometabolic diseases such as higher visceral adipose tissue 0.08 cm² (95% CI 0.04-0.12), HOMA-1B 0.06 (95% CI 0.04-0.09), and lower HDL-cholesterol levels -0.03 mmol/L (95% CI -0.05 - -0.01).

CONCLUSIONS

Combining metabolomics and clinical symptoms we identified a replicable depression dimension associated with adverse metabolic alterations, in line with the IMD hypothesis. Patients with IMD may be at higher cardiometabolic risk and may benefit from specific treatment targeting underlying metabolic dysregulations.

The cardiometabolic depression subtype and its association with clinical characteristics: The Maastricht Study

BACKGROUND

Individuals with depression often show an adverse cardiometabolic risk profile and might represent a distinct depression subtype. The aim of this study was to investigate whether a cardiometabolic depression subtype could be identified and to investigate its association with demographics and clinical characteristics (severity, symptomatology, anti-depressant use, persistence and cognitive functioning).

METHODS

We used data from The Maastricht Study, a population-based cohort in the southern part of The Netherlands. A total of 248 participants with major depressive disorder were included (mean [SD] age, 58.8 ± 8.5 years; 121 [48.8 %] were men). Major depressive disorder was assessed at baseline by the Mini-International Neuropsychiatric Interview. Cardiometabolic risk factors were defined as indicators of the metabolic syndrome according to the National Cholesterol Education Program Adult Treatment Panel III guidelines. We measured severity and persistence of depressive symptoms by use of the 9-item Patient Health Questionnaire.

RESULTS

Latent class analysis resulted in two subtypes, one with cardiometabolic depression (n = 145) and another with non-cardiometabolic depression (n = 103). The cardiometabolic depression subtype was characterized by being male, low education, more severe depressive symptoms, less symptoms of depressed mood and more symptoms of loss of energy, more use of antidepressant medication and lower cognitive functioning.

LIMITATIONS

No conclusions can be made about causality.

CONCLUSIONS

Latent class analysis suggested a distinct cardiometabolic depression subtype. Participants with cardiometabolic depression differed from participants with non-cardiometabolic depression in terms of demographics and clinical characteristics. The existence of a cardiometabolic depression subtype may indicate the need for prevention and treatment targeting cardiometabolic risk management.

Investigating data-driven biological subtypes of psychiatric disorders using specification-curve analysis

BACKGROUND

Cluster analyses have become popular tools for data-driven classification in biological psychiatric research. However, these analyses are known to be sensitive to the chosen methods and/or modelling options, which may hamper generalizability and replicability of findings. To gain more insight into this problem, we used Specification-Curve Analysis (SCA) to investigate the influence of methodological variation on biomarker-based cluster-analysis results.

METHODS

Proteomics data (31 biomarkers) were used from patients (n = 688) and healthy controls (n = 426) in the Netherlands Study of Depression and Anxiety. In SCAs, consistency of results was evaluated across 1200 k-means and hierarchical clustering analyses, each with a unique combination of the clustering algorithm, fit-index, and distance metric. Next, SCAs were run in simulated datasets with varying cluster numbers and noise/outlier levels to evaluate the effect of data properties on SCA outcomes.

RESULTS

The real data SCA showed no robust patterns of biological clustering in either the MDD or a combined MDD/healthy dataset. The simulation results showed that the correct number of clusters could be identified quite consistently across the 1200 model specifications, but that correct cluster identification became harder when the number of clusters and noise levels increased.

CONCLUSION

SCA can provide useful insights into the presence of clusters in biomarker data. However, SCA is likely to show inconsistent results in real-world biomarker datasets that are complex and contain considerable levels of noise. Here, the number and nature of the observed clusters may depend strongly on the chosen model-specification, precluding conclusions about the existence of biological clusters among psychiatric patients.

Major depressive disorder comorbid with general anxiety disorder: Associations among neuroticism, adult stress, and the inflammatory index

Accumulating evidence shows that higher neuroticism and adult stress may be potential risk factors for major depressive disorder (MDD) comorbid with generalized anxiety disorder (GAD). Studies have shown that anxious and depressed patients have significantly more neurobiological abnormalities than non-anxious depressed patients. However, the biological mechanism of comorbidity remains unknown. A study of serum markers allows a better understanding of the mechanism. This was a multi-centre, cross-sectional study. A total of 169 MDD patients (42 MDD patients with comorbid GAD and 127 MDD patients without comorbid GAD) were studied to analyse the risk factors for MDD with comorbid GAD. Twenty-four peripheral serum markers were measured. Path analysis was applied to test the association among neuroticism, adult stress, inflammatory markers, and psychopathology. After Bonferroni correction, MDD patients with comorbid GAD had lower levels of CCL2 (P = 0.001) and higher levels of α2M (P < 0.001) and TLR-1 (P = 0.001) than MDD patients without comorbid GAD (adjusted P < 0.002). In the path analyses of the association among adult stress, the inflammatory index, and psychopathology, neuroticism had a direct effect (β = 0.238, P = 0.003) and an indirect effect (β = 0.068, P = 0.004) on MDD and GAD comorbidity through adult stress and the inflammatory index. Our results suggest that MDD with comorbid GAD is associated with higher levels of inflammatory markers, stress factors and personality traits, which may provide some cues for early identification or more tailored and comprehensive treatment for MDD with comorbid GAD.

Subtyping late-life depression according to inflammatory and metabolic dysregulation: a prospective study

BACKGROUND

Inflammation and metabolic dysregulation are age-related physiological changes and are associated with depressive disorder. We tried to identify subgroups of depressed older patients based on their metabolic-inflammatory profile and examined the course of depression for these subgroups.

METHODS

This clinical cohort study was conducted in a sample of 364 depressed older (⩾60 years) patients according to DSM-IV criteria. Severity of depressive symptoms was monitored every 6 months and a formal diagnostic interview repeated at 2-year follow-up. Latent class analyses based on baseline metabolic and inflammatory biomarkers were performed. Adjusted for confounders, we compared remission of depression at 2-year follow-up between the metabolic-inflammatory subgroups with logistic regression and the course of depression severity over 2-years by linear mixed models.

RESULTS

We identified a 'healthy' subgroup (n = 181, 49.7%) and five subgroups characterized by different profiles of metabolic-inflammatory dysregulation. Compared to the healthy subgroup, patients in the subgroup with mild 'metabolic and inflammatory dysregulation' (n = 137, 37.6%) had higher depressive symptom scores, a lower rate of improvement in the first year, and were less likely to be remitted after 2-years [OR 0.49 (95% CI 0.26-0.91)]. The four smaller subgroups characterized by a more specific immune-inflammatory dysregulation profile did not differ from the two main subgroups regarding the course of depression.

CONCLUSIONS

Nearly half of the patients with late-life depressions suffer from metabolic-inflammatory dysregulation, which is also associated with more severe depression and a worse prognosis. Future studies should examine whether these depressed older patients benefit from a metabolic-inflammatory targeted treatment.

Development of medical device software for the screening and assessment of depression severity using data collected from a wristband-type wearable device: SWIFT study protocol

INTRODUCTION

Few biomarkers can be used clinically to diagnose and assess the severity of depression. However, a decrease in activity and sleep efficiency can be observed in depressed patients, and recent technological developments have made it possible to measure these changes. In addition, physiological changes, such as heart rate variability, can be used to distinguish depressed patients from normal persons; these parameters can be used to improve diagnostic accuracy. The proposed research will explore and construct machine learning models capable of detecting depressive episodes and assessing their severity using data collected from wristband-type wearable devices.

METHODS AND ANALYSIS

Patients with depressive symptoms and healthy subjects will wear a wristband-type wearable device for 7 days; data on triaxial acceleration, pulse rate, skin temperature, and ultraviolet light will be collected. On the seventh day of wearing, the severity of depressive episodes will be assessed using Structured Clinical Interview for DSM-5 (SCID-5), Hamilton Depression Rating Scale (HAMD), and other scales. Data for up to five 7-day periods of device wearing will be collected from each subject. Using wearable device data associated with clinical symptoms as supervisory data, we will explore and build a machine learning model capable of identifying the presence or absence of depressive episodes and predicting the HAMD scores for an unknown data set.

DISCUSSION

Our machine learning model could improve the clinical diagnosis and management of depression through the use of a wearable medical device.

CLINICAL TRIAL REGISTRATION

[https://jrct.niph.go.jp/latest-detail/jRCT1031210478], identifier [jRCT1031210478].

Psychometric Properties of the HADS Measure of Anxiety and Depression Among Multiple Sclerosis Patients in Croatia

Depression and anxiety are common complaints in patients with multiple sclerosis (MS). The study objective was to investigate the factor structure, internal consistency, and correlates of the Croatian version of the Hospital Anxiety and Depression Scale (HADS) in patients with MS. A total of 179 patients with MS and 999 controls were included in the online survey. All subjects completed the HADS and self-administered questionnaires capturing information of demographic, education level, disease-related variables, and the Multiple Sclerosis Impact Scale-29 (MSIS-29). Psychometric properties were examined by estimating the validity, reliability, and factor structure of the HADS in patients with MS. The two HADS subscales (anxiety and depression) had excellent internal consistencies (Cronbach’s α value 0.82–0.83), and factor analysis confirmed a two-factor structure. The convergent validity of the HADS subscales appeared to be good due to the significant correlations between HADS and MSIS-29. Receiver operating characteristic (ROC) analysis indicates that the HADS subscales have a significant diagnostic validity for group differentiation. Hierarchical regression analysis using MSIS-29 subscales as criterion variables showed consistent evidence for the incremental validity of the HADS. The HADS is a reliable and valid self-assessment scale in patients with MS and is suggested to be used in clinical monitoring of the psychiatric and psychological status of patients with MS.

The Anaesthetic Biobank of Cerebrospinal fluid: a unique repository for neuroscientific biomarker research

Background

The pathophysiology of numerous central nervous system disorders remains poorly understood. Biomarker research using cerebrospinal fluid (CSF) is a promising way to illuminate the neurobiology of neuropsychiatric disorders. CSF biomarker studies performed so far generally included patients with neurodegenerative diseases without an adequate control group. The Anaesthetic Biobank of Cerebrospinal fluid (ABC) was established to address this. The aims are to (I) provide healthy-control reference values for CSF-based biomarkers, and (II) to investigate associations between CSF-based candidate biomarkers and neuropsychiatric symptoms.

Methods

In this cross-sectional study, we collect and store CSF and blood from adult patients undergoing spinal anaesthesia for elective surgery. Blood (20.5 mL) is collected during intravenous cannulation and CSF (10 mL) is aspirated prior to intrathecal local anaesthetic injection. A portion of the blood and CSF is sent for routine laboratory analyses, the remaining material is stored at -80 °C. Relevant clinical, surgical and anaesthetic data are registered. A neurological examination and Montreal Cognitive Assessment (MoCA) are performed pre-operatively and a subset of patients fill in questionnaires on somatic and mental health (depression, anxiety and stress).

Results

Four-hundred-fifty patients (58% male; median age: 56 years) have been enrolled in the ABC. The planned spinal anaesthetic procedure was not attempted for various reasons in eleven patients, in fourteen patients the spinal puncture failed and in twelve patients CSF aspiration was unsuccessful. A mean of 9.3 mL CSF was obtained in the remaining 413 of patients. Most patients had a minor medical history and 60% scored in the normal range on the MoCA (median score: 26).

Conclusions

The ABC is an ongoing biobanking project that can contribute to CSF-based biomarker research. The large sample size with constant sampling methods and extensive patient phenotyping provide excellent conditions for future neuroscientific research.

Latent transition analysis of cardiac arrest patients treated in the intensive care unit

BACKGROUND AND OBJECTIVE

Post-cardiac arrest (CA) syndrome is heterogenous in their clinical presentations and outcomes. This study aimed to explore the transition and stability of subphenotypes (profiles) of CA treated in the intensive care unit (ICU).

PATIENTS AND METHODS

Clinical features of CA patients on day 1 and 3 after ICU admission were modeled by latent transition analysis (LTA) to explore the transition between subphenotypes over time. The association between different transition patterns and mortality outcome was explored using multivariable logistic regression.

RESULTS

We identified 848 eligible patients from the database. The LPA identified three distinct subphenotypes: Profile 1 accounted for the largest proportion (73%) and was considered as the baseline subphenotype. Profile 2 (13%) was characterized by brain injury and profile 3 (14%) was characterized by multiple organ dysfunctions. The same three subphenotypes were identified on day 3. The LTA showed consistent subphenotypes. A majority of patients in profile 2 (72%) and 3 (82%) on day 1 switched to profile 1 on day 3. In the logistic regression model, patients in profile 1 on day 1 transitioned to profile 3 had worse survival outcome than those continue to remain in profile 1 (OR: 20.64; 95% CI: 6.01 to 70.94; p < 0.001) and transitioned to profile 2 (OR: 8.42; 95% CI: 2.22 to 31.97; p = 0.002) on day 3.

CONCLUSION

The study identified three subphenotypes of CA, which was consistent on day 1 and 3 after ICU admission. Patients who transitioned to profile 3 on day 3 had significantly worse survival outcome than those remained in profile 1 or 2.

Peripheral Markers of Depression

Major Depressive Disorder (MDD) is a leading cause of disability worldwide, creating a high medical and socioeconomic burden. There is a growing interest in the biological underpinnings of depression, which are reflected by altered levels of biological markers. Among others, enhanced inflammation has been reported in MDD, as reflected by increased concentrations of inflammatory markers-C-reactive protein, interleukin-6, tumor necrosis factor-α and soluble interleukin-2 receptor. Oxidative and nitrosative stress also plays a role in the pathophysiology of MDD. Notably, increased levels of lipid peroxidation markers are characteristic of MDD. Dysregulation of the stress axis, along with increased cortisol levels, have also been reported in MDD. Alterations in growth factors, with a significant decrease in brain-derived neurotrophic factor and an increase in fibroblast growth factor-2 and insulin-like growth factor-1 concentrations have also been found in MDD. Finally, kynurenine metabolites, increased glutamate and decreased total cholesterol also hold promise as reliable biomarkers for MDD. Research in the field of MDD biomarkers is hindered by insufficient understanding of MDD etiopathogenesis, substantial heterogeneity of the disorder, common co-morbidities and low specificity of biomarkers. The construction of biomarker panels and their evaluation with use of new technologies may have the potential to overcome the above mentioned obstacles.

Metabolomic Biomarkers in Anxiety Disorders

Anxiety disorders range among the most prevalent psychiatric disorders and belong to the leading disorders in the study of the total global burden of disease. Anxiety disorders are complex conditions, with not fully understood etiological mechanisms. Numerous factors, including psychological, genetic, biological, and chemical factors, are thought to be involved in their etiology. Although the diagnosis of anxiety disorders is constantly evolving, diagnostic manuals rely on symptom lists, not on objective biomarkers and treatment effects are small to moderate. The underlying biological factors that drive anxiety disorders may be better suited to serve as biomarkers for guiding personalized medicine, as they are objective and can be measured externally. Therefore, the incorporation of novel biomarkers into current clinical methods might help to generate a classification system for anxiety disorders that can be linked to the underlying dysfunctional pathways. The study of metabolites (metabolomics) in a large-scale manner shows potential for disease diagnosis, for stratification of patients in a heterogeneous patient population, for monitoring therapeutic efficacy and disease progression, and for defining therapeutic targets. All of these are important properties for anxiety disorders, which is a multifactorial condition not involving a single-gene mutation. This review summarizes recent investigations on metabolomics studies in anxiety disorders.

Causes and Consequences of Diagnostic Heterogeneity in Depression: Paths to Discovering Novel Biological Depression Subtypes

Depression is a highly heterogeneous syndrome that bears only modest correlations with its biological substrates, motivating a renewed interest in rethinking our approach to diagnosing depression for research purposes and new efforts to discover subtypes of depression anchored in biology. Here, we review the major causes of diagnostic heterogeneity in depression, with consideration of both clinical symptoms and behaviors (symptomatology and trajectory of depressive episodes) and biology (genetics and sexually dimorphic factors). Next, we discuss the promise of using data-driven strategies to discover novel subtypes of depression based on functional neuroimaging measures, including dimensional, categorical, and hybrid approaches to parsing diagnostic heterogeneity and understanding its biological basis. The merits of using resting-state functional magnetic resonance imaging functional connectivity techniques for subtyping are considered along with a set of technical challenges and potential solutions. We conclude by identifying promising future directions for defining neurobiologically informed depression subtypes and leveraging them in the future for predicting treatment outcomes and informing clinical decision making.

Evaluating depression with multimodal wristband-type wearable device: screening and assessing patient severity utilizing machine-learning

OBJECTIVE

We aimed to develop a machine learning algorithm to screen for depression and assess severity based on data from wearable devices.

METHODS

We used a wearable device that calculates steps, energy expenditure, body movement, sleep time, heart rate, skin temperature, and ultraviolet light exposure. Depressed patients and healthy volunteers wore the device continuously for the study period. The modalities were compared hourly between patients and healthy volunteers. XGBoost was used to build machine learning models and 10-fold cross-validation was applied for the validation.

RESULTS

Forty-five depressed patients and 41 healthy controls participated, creating a combined 5,250 days' worth of data. Heart rate, steps, and sleep were significantly different between patients and healthy volunteers in some comparisons. Similar differences were also observed longitudinally when patients' symptoms improved. Based on seven days' data, the model identified symptomatic patients with 0.76 accuracy and predicted Hamilton Depression Rating Scale-17 scores with a 0.61 correlation coefficient. Skin temperature, sleep time-related features, and the correlation of those modalities were the most significant features in machine learning.

LIMITATIONS

The small number of subjects who participated in this study may have weakened the statistical significance of the study. There are differences in the demographic data among groups although we performed a correction for multiple comparisons. Validation in independent datasets was not performed, although 10-fold cross validation with the internal data was conducted.

CONCLUSION

The results indicated that utilizing wearable devices and machine learning may be useful in identifying depression as well as assessing severity.