The potential of predictive analytics to provide clinical decision support in depression treatment planning

Common polygenic variation in the early medication change (EMC) cohort affects disorder risk, but not the antidepressant treatment response

BACKGROUND

Given the great interest in identifying reliable predictors of the response to antidepressant drugs, the present study investigated whether polygenic scores (PGS) for Major Depressive Disorder (MDD) and antidepressant treatment response (ADR) were related to the complex trait of antidepressant response in the Early Medication Change (EMC) cohort.

METHODS

In this secondary analysis of the EMC trial (N = 889), 481 MDD patients were included and compared to controls from a population-based cohort. Patients were treated over eight weeks within a pre-defined treatment-algorithm. We investigated patients' genetic variation associated with MDD and ADR, using PGS and examined the association of PGS with treatment outcomes (early improvement, response, remission). Additionally, the influence of two cytochrome P450 drug-metabolizing enzymes (CYP2C19, CYP2D6) was determined.

RESULTS

PGS for MDD was significantly associated with disorder status (NkR2 = 2.48 %, p < 1*10-12), with higher genetic burden in EMC patients compared to controls. The PGS for ADR did not explain remission status. The PGS for MDD and ADR were also not associated with treatment outcomes. In addition, there were no effects of common CYP450 gene variants on ADR.

LIMITATIONS

The study was limited by variability in the outcome parameters due to differences in treatment and insufficient sample size in the used ADR genome-wide association study (GWAS).

CONCLUSIONS

The present study confirms a polygenic contribution to MDD burden in the EMC patients. Larger GWAS with homogeneity in antidepressant treatments are needed to explore the genetic variation associated with ADR and realize the potential of PGS to contribute to specific response subtypes.

Data-driven personalized medicine approaches to cognitive-behavioral therapy allocation in a large sample: A reanalysis of the ENRICHED study

BACKGROUND

Although effective treatments for common mental health problems are available, individual responses to treatments are difficult to predict. Treatment efficacy could be optimized by targeting interventions using individual predictions of treatment outcomes. The aim of this study was to develop a prediction algorithm using data from one of the largest randomized controlled trials on psychological interventions for common mental health problems.

METHODS

This is a secondary analysis of the Enhancing Recovery in Coronary Heart Disease study investigating the effectiveness of cognitive behavioral therapy (CBT) and care as usual (CAU) for depression and low perceived social support following acute myocardial infarction. 2481 participants were randomly assigned to CBT and CAU. Baseline social-demographics, depression characteristics, comorbid symptoms, and stress and adversity measures were used to build an algorithm predicting post-treatment depression severity using elastic net regularization. Performance and generalizability of this algorithm were determined in a hold-out sample (n = 1203).

RESULTS

Treatment matching based on predictions in the hold-out sample resulted in inconsistent and small effects (d = 0.15), that were more pronounced for individuals matched to CBT (d = 0.22). We identified a small subgroup of individuals for which CBT did not appear more efficacious than CAU.

LIMITATIONS

Limitations are a poorly defined CAU condition, a low-severity sample, specific exclusion criteria and unavailability of certain baseline variables.

CONCLUSIONS

Small matching effects are likely a realistic representation of the performance and generalizability of multivariable prediction algorithms based on clinical measures. Results indicate that future work and new approaches are needed.

Evening Chronotypes With Depression Report Poorer Outcomes of Selective Serotonin Reuptake Inhibitors: A Survey-Based Study of Self-Ratings

BACKGROUND

Preliminary evidence suggests that evening chronotype is related to poorer efficacy of selective serotonin reuptake inhibitors. It is unknown whether this is specific to particular medications, self-rated chronotype, or efficacy.

METHODS

In the Australian Genetics of Depression Study (n = 15,108; 75% women; 18-90 years; 68% with ≥1 other lifetime diagnosis), a survey recorded experiences with 10 antidepressants, and the reduced Morningness-Eveningness Questionnaire was used to estimate chronotype. A chronotype polygenic score was calculated. Age- and sex-adjusted regression models (Bonferroni-corrected) estimated associations among antidepressant variables (how well the antidepressant worked [efficacy], duration of symptom improvement, side effects, discontinuation due to side effects) and self-rated and genetic chronotypes.

RESULTS

The chronotype polygenic score explained 4% of the variance in self-rated chronotype (r = 0.21). Higher self-rated eveningness was associated with poorer efficacy of escitalopram (odds ratio [OR] = 1.04; 95% CI, 1.02 to 1.06; p = .000035), citalopram (OR = 1.03; 95% CI, 1.01 to 1.05; p = .004), fluoxetine (OR = 1.03; 95% CI, 1.01 to 1.05; p = .001), sertraline (OR = 1.02; 95% CI, 1.01 to 1.04; p = .0008), and desvenlafaxine (OR = 1.03; 95% CI, 1.01 to 1.05; p = .004), and a profile of increased side effects (80% of those recorded; ORs = 0.93-0.98), with difficulty getting to sleep the most common. Self-rated chronotype was unrelated to duration of improvement or discontinuation. The chronotype polygenic score was only associated with suicidal thoughts and attempted suicide (self-reported). While our measures are imperfect, and not of circadian phase under controlled conditions, the model coefficients suggest that dysregulation of the phenotypic chronotype relative to its genetic proxy drove relationships with antidepressant outcomes.

CONCLUSIONS

The idea that variation in circadian factors influences response to antidepressants was supported and encourages exploration of circadian mechanisms of depressive disorders and antidepressant treatments.

Predicting the outcome of psilocybin treatment for depression from baseline fMRI functional connectivity

BACKGROUND

Psilocybin is a serotonergic psychedelic drug under assessment as a potential therapy for treatment-resistant and major depression. Heterogeneous treatment responses raise interest in predicting the outcome from baseline data.

METHODS

A machine learning pipeline was implemented to investigate baseline resting-state functional connectivity measured with functional magnetic resonance imaging (fMRI) as a predictor of symptom severity in psilocybin monotherapy for treatment-resistant depression (16 patients administered two 5 mg capsules followed by 25 mg, separated by one week). Generalizability was tested in a sample of 22 patients who participated in a psilocybin vs. escitalopram trial for moderate-to-severe major depression (two separate doses of 25 mg of psilocybin 3 weeks apart plus 6 weeks of daily placebo vs. two separate doses of 1 mg of psilocybin 3 weeks apart plus 6 weeks of daily oral escitalopram). The analysis was repeated using both samples combined.

RESULTS

Functional connectivity of visual, default mode and executive networks predicted early symptom improvement, while the salience network predicted responders up to 24 weeks after treatment (accuracy≈0.9). Generalization performance was borderline significant. Consistent results were obtained from the combined sample analysis. Fronto-occipital and fronto-temporal coupling predicted early and late symptom reduction, respectively.

LIMITATIONS

The number of participants and differences between the two datasets limit the generalizability of the findings, while the lack of a placebo arm limits their specificity.

CONCLUSIONS

Baseline neurophysiological measurements can predict the outcome of psilocybin treatment for depression. Future research based on larger datasets should strive to assess the generalizability of these predictions.

Optimizing precision medicine for second-step depression treatment: a machine learning approach

BACKGROUND

Less than a third of patients with depression achieve successful remission with standard first-step antidepressant monotherapy. The process for determining appropriate second-step care is often based on clinical intuition and involves a protracted course of trial and error, resulting in substantial patient burden and unnecessary delay in the provision of optimal treatment. To address this problem, we adopt an ensemble machine learning approach to improve prediction accuracy of remission in response to second-step treatments.

METHOD

Data were derived from the Level 2 stage of the STAR*D dataset, which included 1439 patients who were randomized into one of seven different second-step treatment strategies after failing to achieve remission during first-step antidepressant treatment. Ensemble machine learning models, comprising several individual algorithms, were evaluated using nested cross-validation on 155 predictor variables including clinical and demographic measures.

RESULTS

The ensemble machine learning algorithms exhibited differential classification performance in predicting remission status across the seven second-step treatments. For the full set of predictors, AUC values ranged from 0.51 to 0.82 depending on the second-step treatment type. Predicting remission was most successful for cognitive therapy (AUC = 0.82) and least successful for other medication and combined treatment options (AUCs = 0.51-0.66).

CONCLUSION

Ensemble machine learning has potential to predict second-step treatment. In this study, predictive performance varied by type of treatment, with greater accuracy in predicting remission in response to behavioral treatments than to pharmacotherapy interventions. Future directions include considering more informative predictor modalities to enhance prediction of second-step treatment response.

Transdiagnostic and tailored internet intervention to improve mental health among university students: Research protocol for a randomized controlled trial

BACKGROUND

Emerging adulthood is often associated with mental health problems. About one in three university students report symptoms of depression and anxiety that can negatively affect their developmental trajectory concerning work, intimate relationships, and health. This can interfere with academic performance, as mood and anxiety disorders are key predictors of dropout from higher education. A treatment gap exists, where a considerable proportion of students do not seek help for mood and anxiety symptoms. Offering internet interventions to students with mental health problems could reduce the treatment gap, increase mental health, and improve academic performance. A meta-analysis on internet interventions for university students showed small effects for depression and none for anxiety. Larger trials are recommended to further explore effects of guidance, transdiagnostic approaches, and individual treatment components.

METHODS

This study will offer 1200 university students in Sweden participation in a three-armed randomized controlled trial (RCT) evaluating a guided or unguided transdiagnostic internet intervention for mild to moderate depression and anxiety, where the waitlist control group accesses the intervention at 6-month follow-up. Students reporting suicidal ideation/behaviors will be excluded and referred to treatment within the existing healthcare system. An embedded study within the trial (SWAT) will assess at week 3 of 8 whether participants in the guided and unguided groups are at higher risk of failing to benefit from treatment. Those at risk will be randomized to an adaptive treatment strategy, or to continue the treatment as originally randomized. Primary outcomes are symptoms of depression and anxiety. Follow-ups will occur at post-treatment and at 6-, 12-, and 24-month post-randomization. Between-group outcome analyses will be reported, and qualitative interviews about treatment experiences are planned.

DISCUSSION

This study investigates the effects of a transdiagnostic internet intervention among university students in Sweden, with an adaptive treatment strategy employed during the course of treatment to minimize the risk of treatment failure. The study will contribute knowledge about longitudinal trajectories of mental health and well-being following treatment, taking into account possible gender differences in responsiveness to treatment. With time, effective internet interventions could make treatment for mental health issues more widely accessible to the student group.

Which client with generalized anxiety disorder benefits from a mindfulness ecological momentary intervention versus a self-monitoring app? Developing a multivariable machine learning predictive model

Precision medicine methods (machine learning; ML) can identify which clients with generalized anxiety disorder (GAD) benefit from mindfulness ecological momentary intervention (MEMI) vs. self-monitoring app (SM). We used randomized controlled trial data of MEMI vs. SM for GAD (N = 110) and tested three ML models to predict one-month follow-up reliable improvement in GAD severity, perseverative cognitions (PC), trait mindfulness (TM), and executive function (EF). Eleven baseline predictors were tested regarding differential reliable change from MEMI vs. SM (age, sex, race, EF errors, inhibitory dyscontrol, set-shifting deficits, verbal fluency, working memory, GAD severity, TM, PC). The final top five prescriptive predictor models of all outcomes performed well (AUC = .752 .886). The following variables predicted better outcome from MEMI vs. SM: Higher GAD severity predicted more GAD improvement but less EF improvement. Elevated PC, inhibitory dyscontrol, and verbal dysfluency predicted better improvement in most outcomes. Greater set-shifting and TM predicted stronger improvements in GAD symptoms and TM. Older age predicted more alleviation of GAD and PC symptoms. Women exhibited more enhancements in trait mindfulness and EF than men. White individuals benefitted more than non-White. PC, TM, EF, and sociodemographic data might help predictive models optimize intervention selection for GAD.

A two-phase, prescriptive comparative effectiveness study to optimize the treatment of co-occurring insomnia and depression with digital interventions

BACKGROUND

Insomnia and depression frequently co-occur. Significant barriers preclude a majority of patients from receiving first line treatments for both disorders in a sequential treatment episode. Although digital versions of cognitive behavioral therapy for insomnia (CBTI) and for depression (CBTD) hold some promise to meet demand, especially when paired with human support, it is unknown whether heterogeneity of treatment effects exist, such that some patients would be optimally treated with single or sequential interventions.

OBJECTIVE

Describe the protocol for a two-phase, prescriptive comparative effectiveness study to develop and evaluate an individualized intervention rule (IIR) for prescribing the optimal digital treament of co-occurring insomnia and depression.

METHODS

The proposed sample size is 2300 U.S. military veterans with insomnia and depression recruited nationally (Phase 1 = 1500; Phase 2 = 800). In each phase, the primary endpoint will be remission of both depression and insomnia 3 months following a 12-week intervention period. Phase 1 is a 5-arm randomized trial: two single digital interventions (CBT-I or CBT-D); two sequenced interventions (CBT-I + D or CBT-D + I); and a mood monitoring control condition. A cutting-edge ensemble machine learning method will be used to develop the IIR. Phase 2 will evaluate the IIR by randomizing participants with equal allocation to either the IIR predicted optimal intervention for that individual or by randomization to one the four CBT interventions.

RESULTS

Study procedures are ongoing. Results will be reported in future manuscripts.

CONCLUSION

The study will generate evidence on the optimal scalable approach to treat co-occurring insomnia and depression.

Investigation of Active Ingredients Within Internet-Delivered Cognitive Behavioral Therapy for Depression: A Randomized Optimization Trial

Importance

There is limited understanding of how complex evidence-based psychological interventions such as cognitive behavioral therapy (CBT) for depression work. Identifying active ingredients may help to make therapy more potent, brief, and scalable.

Objective

To test the individual main effects and interactions of 7 treatment components within internet-delivered CBT for depression to investigate its active ingredients.

Design, Setting, and Participants

This randomized optimization trial using a 32-condition, balanced, fractional factorial optimization experiment (IMPROVE-2) recruited adults with depression (Patient Health Questionnaire-9 [PHQ-9] score ≥10) from internet advertising and the UK National Health Service Improving Access to Psychological Therapies service. Participants were randomized from July 7, 2015, to March 29, 2017, with follow-up for 6 months after treatment until December 29, 2017. Data were analyzed from July 2018 to April 2023.

Interventions

Participants were randomized with equal probability to 7 experimental factors within the internet CBT platform, each reflecting the presence vs absence of specific treatment components (activity scheduling, functional analysis, thought challenging, relaxation, concreteness training, absorption, and self-compassion training).

Main Outcomes and Measures

The primary outcome was depression symptoms (PHQ-9 score). Secondary outcomes include anxiety symptoms and work, home, and social functioning.

Results

Among 767 participants (mean age [SD] age, 38.5 [11.62] years; range, 18-76 years; 635 women [82.8%]), 506 (66%) completed the 6-month posttreatment follow-up. On average, participants receiving internet-delivered CBT had reduced depression (pre-to-posttreatment difference in PHQ-9 score, -7.79 [90% CI, -8.21 to -7.37]; 6-month follow-up difference in PHQ-9 score, -8.63 [90% CI, -9.04 to -8.22]). A baseline score-adjusted analysis of covariance model using effect-coded intervention variables (-1 or +1) found no main effect on depression symptoms for the presence vs absence of activity scheduling, functional analysis, thought challenging, relaxation, concreteness training, or self-compassion training (posttreatment: largest difference in PHQ-9 score [functional analysis], -0.09 [90% CI, -0.56 to 0.39]; 6-month follow-up: largest difference in PHQ-9 score [relaxation], -0.18 [90% CI, -0.61 to 0.25]). Only absorption training had a significant main effect on depressive symptoms at 6-month follow-up (posttreatment difference in PHQ-9 score, 0.21 [90% CI, -0.27 to 0.68]; 6-month follow-up difference in PHQ-9 score, -0.54, [90% CI, -0.97 to -0.11]).

Conclusions and Relevance

In this randomized optimization trial, all components of internet-delivered CBT except absorption training did not significantly reduce depression symptoms relative to their absence despite an overall average reduction in symptoms. The findings suggest that treatment benefit from internet-delivered CBT probably accrues from spontaneous remission, factors common to all CBT components (eg, structure, making active plans), and nonspecific therapy factors (eg, positive expectancy), with the possible exception of absorption focused on enhancing direct contact with positive reinforcers.

Trial Registration

isrctn.org Identifier: ISRCTN24117387.

Predicting heterogeneous treatment effects of an Internet-based depression intervention for patients with chronic back pain: Secondary analysis of two randomized controlled trials

Background

Depression is highly prevalent among individuals with chronic back pain. Internet-based interventions can be effective in treating and preventing depression in this patient group, but it is unclear who benefits most from this intervention format.

Method

In an analysis of two randomized trials (N = 504), we explored ways to predict heterogeneous treatment effects of an Internet-based depression intervention for patients with chronic back pain. Univariate treatment-moderator interactions were explored in a first step. Multilevel model-based recursive partitioning was then applied to develop a decision tree model predicting individualized treatment benefits.

Results

The average effect on depressive symptoms was d = -0.43 (95 % CI: -0.68 to -0.17; 9 weeks; PHQ-9). Using univariate models, only back pain medication intake was detected as an effect moderator, predicting higher effects. More complex interactions were found using recursive partitioning, resulting in a final decision tree with six terminal nodes. The model explained a large amount of variation (bootstrap-bias-corrected R2 = 45 %), with predicted subgroup-conditional effects ranging from di = 0.24 to -1.31. External validation in a pilot trial among patients on sick leave (N = 76; R2 = 33 %) pointed to the transportability of the model.

Conclusions

The studied intervention is effective in reducing depressive symptoms, but not among all chronic back pain patients. Predictions of the multivariate tree learning model suggest a pattern in which patients with moderate depression and relatively low pain self-efficacy benefit most, while no benefits arise when patients' self-efficacy is already high. If corroborated in further studies, the developed tree algorithm could serve as a practical decision-making tool.

Evaluation of randomized controlled trials: a primer and tutorial for mental health researchers

BACKGROUND

Considered one of the highest levels of evidence, results of randomized controlled trials (RCTs) remain an essential building block in mental health research. They are frequently used to confirm that an intervention "works" and to guide treatment decisions. Given their importance in the field, it is concerning that the quality of many RCT evaluations in mental health research remains poor. Common errors range from inadequate missing data handling and inappropriate analyses (e.g., baseline randomization tests or analyses of within-group changes) to unduly interpretations of trial results and insufficient reporting. These deficiencies pose a threat to the robustness of mental health research and its impact on patient care. Many of these issues may be avoided in the future if mental health researchers are provided with a better understanding of what constitutes a high-quality RCT evaluation.

METHODS

In this primer article, we give an introduction to core concepts and caveats of clinical trial evaluations in mental health research. We also show how to implement current best practices using open-source statistical software.

RESULTS

Drawing on Rubin's potential outcome framework, we describe that RCTs put us in a privileged position to study causality by ensuring that the potential outcomes of the randomized groups become exchangeable. We discuss how missing data can threaten the validity of our results if dropouts systematically differ from non-dropouts, introduce trial estimands as a way to co-align analyses with the goals of the evaluation, and explain how to set up an appropriate analysis model to test the treatment effect at one or several assessment points. A novice-friendly tutorial is provided alongside this primer. It lays out concepts in greater detail and showcases how to implement techniques using the statistical software R, based on a real-world RCT dataset.

DISCUSSION

Many problems of RCTs already arise at the design stage, and we examine some avoidable and unavoidable "weak spots" of this design in mental health research. For instance, we discuss how lack of prospective registration can give way to issues like outcome switching and selective reporting, how allegiance biases can inflate effect estimates, review recommendations and challenges in blinding patients in mental health RCTs, and describe problems arising from underpowered trials. Lastly, we discuss why not all randomized trials necessarily have a limited external validity and examine how RCTs relate to ongoing efforts to personalize mental health care.

A randomized controlled trial on using predictive algorithm to adapt level of psychological care for community college students: STAND triaging and adapting to level of care study protocol

BACKGROUND

There is growing interest in using personalized mental health care to treat disorders like depression and anxiety to improve treatment engagement and efficacy. This randomized controlled trial will compare a traditional symptom severity decision-making algorithm to a novel multivariate decision-making algorithm for triage to and adaptation of mental health care. The stratified levels of care include a self-guided online wellness program, coach-guided online cognitive behavioral therapy, and clinician-delivered psychotherapy with or without pharmacotherapy. The novel multivariate algorithm will be comprised of baseline (for triage and adaptation) and time-varying variables (for adaptation) in four areas: social determinants of mental health, early adversity and life stressors, predisposing, enabling, and need influences on health service use, and comprehensive mental health status. The overarching goal is to evaluate whether the multivariate algorithm improves adherence to treatment, symptoms, and functioning above and beyond the symptom-based algorithm.

METHODS/DESIGN

This trial will recruit a total of 1000 participants over the course of 5 years in the greater Los Angeles Metropolitan Area. Participants will be recruited from a highly diverse sample of community college students. For the symptom severity approach, initial triaging to level of care will be based on symptom severity, whereas for the multivariate approach, the triaging will be based on a comprehensive set of baseline measures. After the initial triaging, level of care will be adapted throughout the duration of the treatment, utilizing either symptom severity or multivariate statistical approaches. Participants will complete computerized assessments and self-report questionnaires at baseline and up to 40 weeks. The multivariate decision-making algorithm will be updated annually to improve predictive outcomes.

DISCUSSION

Results will provide a comparison on the traditional symptom severity decision-making and the novel multivariate decision-making with respect to treatment adherence, symptom improvement, and functional recovery. Moreover, the developed multivariate decision-making algorithms may be used as a template in other community college settings. Ultimately, findings will inform the practice of level of care triage and adaptation in psychological treatments, as well as the use of personalized mental health care broadly.

TRIAL REGISTRATION

ClinicalTrials.gov NCT05591937, submitted August 2022, published October 2022.

Pharmacomicrobiomics of Antidepressants in Depression: A Systematic Review

This systematic review evaluated the animal and human evidence for pharmacomicrobiomics (PMx) interactions of antidepressant medications. Studies of gut microbiota effects on functional and behavioral effects of antidepressants in human and animal models were identified from PubMed up to December 2022. Risk of bias was assessed, and results are presented as a systematic review following PRISMA guidelines. A total of 28 (21 animal, 7 human) studies were included in the review. The reviewed papers converged on three themes: (1) Antidepressants can alter the composition and metabolites of gut microbiota, (2) gut microbiota can alter the bioavailability of certain antidepressants, and (3) gut microbiota may modulate the clinical or modeled mood modifying effects of antidepressants. The majority (n = 22) of studies had at least moderate levels of bias present. While strong evidence is still lacking to understand the clinical role of antidepressant PMx in human health, there is evidence for interactions among antidepressants, microbiota changes, microbiota metabolite changes, and behavior. Well-controlled studies of the mediating and moderating effects of baseline and treatment-emergent changes in microbiota on therapeutic and adverse responses to antidepressants are needed to better establish a potential role of PMx in personalizing antidepressant treatment selection and response prediction.

Resting-state functional connectivity predicting clinical improvement following treatment in female adolescents with non-suicidal self-injury

BACKGROUND

Non-suicidal self-injury (NSSI) is highly prevalent among adolescents and predicts future psychopathology including suicide. To improve therapeutic decisions and clinical outcome of patients engaging in NSSI, it seems beneficial to determine neurobiological markers associated with treatment response. The present study investigated whether resting-state functional brain connectivity (RSFC) served to predict clinical improvements following treatment in adolescents engaging in NSSI.

METHODS

N = 27 female adolescents with NSSI took part in a baseline MRI exam and clinical outcome was assessed at follow-ups one, two and three years after baseline. During the follow-up period, patients received in- and/or outpatient treatment. Mixed-effects linear regression models were calculated to examine whether RSFC was associated with clinical improvement.

RESULTS

Patients' clinical outcome improved across time. Lower baseline RSFC between left paracentral gyrus and right anterior cingulate gyrus was associated with clinical improvement from baseline to one-year and from two-year to three-year follow-up. Lower and higher baseline RSFC in several inter- and intrahemispheric cortico-cortical and cortico-subcortical connections of interest were associated with clinical symptomatology and its severity, independent from time.

LIMITATIONS

A relatively small sample size constrains the generalizability of our findings. Further, no control group not receiving treatment was recruited, therefore clinical changes across time cannot solely be attributed to treatment.

CONCLUSIONS

While there was some evidence that RSFC was associated with clinical improvement following treatment, our findings suggest that functional connectivity is more predictive of severity of psychopathology and global functioning independent of time and treatment. We thereby add to the limited research on neurobiological markers as predictors of clinical outcome after treatment.

Behavioural and nondirective parent training for children with externalising disorders: First steps towards personalised treatment recommendations

For children with externalising disorders, parent training programmes with different theoretical foundations are available. Currently, there is little knowledge concerning which programme should be recommended to a family based on their individual needs (e.g., single parenthood). The personalised advantage index (PAI) indicates the predicted treatment advantage of one treatment over another. The aim of the present study was to examine the usefulness of this score in providing individualised treatment recommendations. The analysis considered 110 parents (per-protocol sample) of children (4-11 years) with attention-deficit/hyperactivity (ADHD) or oppositional defiant disorder (ODD), randomised to either a behavioural or a nondirective telephone-assisted self-help parent training. In multiple moderator analyses with four different regression algorithms (linear, ridge, k-nearest neighbors, and tree), the linear model was preferred for computing the PAI. For ODD, families randomised to their PAI-predicted optimal intervention showed a treatment advantage of d = 0.54, 95% CI [0.17, 0.97]; for ADHD, the advantage was negligible at d = 0.35, 95% CI [-0.01, 0.78]. For children with conduct problems, it may be helpful if the PAI includes the treatment moderators single parent status and ODD baseline symptoms when providing personalised treatment recommendations for the selection of behavioural versus nondirective parent training. TRIAL REGISTRATION: The study was registered prospectively with ClinicalTrials.gov (Identifier NCT01350986).

The Precision in Psychiatry (PIP) study: Testing an internet-based methodology for accelerating research in treatment prediction and personalisation

BACKGROUND

Evidence-based treatments for depression exist but not all patients benefit from them. Efforts to develop predictive models that can assist clinicians in allocating treatments are ongoing, but there are major issues with acquiring the volume and breadth of data needed to train these models. We examined the feasibility, tolerability, patient characteristics, and data quality of a novel protocol for internet-based treatment research in psychiatry that may help advance this field.

METHODS

A fully internet-based protocol was used to gather repeated observational data from patient cohorts receiving internet-based cognitive behavioural therapy (iCBT) (N = 600) or antidepressant medication treatment (N = 110). At baseline, participants provided > 600 data points of self-report data, spanning socio-demographics, lifestyle, physical health, clinical and other psychological variables and completed 4 cognitive tests. They were followed weekly and completed another detailed clinical and cognitive assessment at week 4. In this paper, we describe our study design, the demographic and clinical characteristics of participants, their treatment adherence, study retention and compliance, the quality of the data gathered, and qualitative feedback from patients on study design and implementation.

RESULTS

Participant retention was 92% at week 3 and 84% for the final assessment. The relatively short study duration of 4 weeks was sufficient to reveal early treatment effects; there were significant reductions in 11 transdiagnostic psychiatric symptoms assessed, with the largest improvement seen for depression. Most participants (66%) reported being distracted at some point during the study, 11% failed 1 or more attention checks and 3% consumed an intoxicating substance. Data quality was nonetheless high, with near perfect 4-week test retest reliability for self-reported height (ICC = 0.97).

CONCLUSIONS

An internet-based methodology can be used efficiently to gather large amounts of detailed patient data during iCBT and antidepressant treatment. Recruitment was rapid, retention was relatively high and data quality was good. This paper provides a template methodology for future internet-based treatment studies, showing that such an approach facilitates data collection at a scale required for machine learning and other data-intensive methods that hope to deliver algorithmic tools that can aid clinical decision-making in psychiatry.

Routinely accessible parameters of mineralocorticoid receptor function, depression subtypes and response prediction: a post-hoc analysis from the early medication change trial in major depressive disorder

OBJECTIVES

Previous studies indicated a relationship between aldosterone, the mineralocorticoid receptor (MR), and antidepressant treatment outcome. Physiological indicators of MR function (blood pressure and electrolytes) are easily accessible and may therefore serve as useful predictors. Thus, our aim was to investigate the predictive value of peripheral MR-related markers for antidepressant treatment outcomes.

METHODS

826 MDD patients who had participated in the randomised-controlled Early Medication Change (EMC) trial were analysed. Depression severity and MR-related markers were assessed weekly. In 562 patients, genetic variation of five MR-related genes was determined.

RESULTS

Patients with blood pressure <120mmHg showed higher depression severity (p = 0.005) than patients with blood pressure ≥120mmHg. Patients with a melancholic subtype had significantly lower blood pressures (p = 0.004). Na⁺/K⁺ ratio was positively and K⁺-concentration was negatively correlated to depression severity and to relative changes in HAMD from baseline to day 14, and 56 respectively (p < 0.001). For none of the MR-related genes, genetic variation was associated with treatment outcomes.

CONCLUSIONS

We confirmed early observations of an altered peripheral MR sensitivity, reflected by lower blood pressure, low K⁺ or high Na⁺/K⁺ ratio in patients with more severe depression. These routinely collected biomarkers may potentially be useful for risk stratification in an early stage of treatment. Trial Registration: clinicaltrials.gov Identifier: NCT00974155; https://www.clinicaltrials.gov/ct2/results?term=NCT00974155.

Epigenetic signatures in antidepressant treatment response: a methylome-wide association study in the EMC trial

Although the currently available antidepressants are well established in the treatment of the major depressive disorder (MDD), there is strong variability in the response of individual patients. Reliable predictors to guide treatment decisions before or in an early stage of treatment are needed. DNA-methylation has been proven a useful biomarker in different clinical conditions, but its importance for mechanisms of antidepressant response has not yet been determined. 80 MDD patients were selected out of >500 participants from the Early Medication Change (EMC) cohort with available genetic material based on their antidepressant response after four weeks and stratified into clear responders and age- and sex-matched non-responders (N = 40, each). Early improvement after two weeks was analyzed as a secondary outcome. DNA-methylation was determined using the Illumina EPIC BeadChip. Epigenome-wide association studies were performed and differentially methylated regions (DMRs) identified using the comb-p algorithm. Enrichment was tested for hallmark gene-sets and in genome-wide association studies of depression and antidepressant response. No epigenome-wide significant differentially methylated positions were found for treatment response or early improvement. Twenty DMRs were associated with response; the strongest in an enhancer region in SORBS2, which has been related to cardiovascular diseases and type II diabetes. Another DMR was located in CYP2C18, a gene previously linked to antidepressant response. Results pointed towards differential methylation in genes associated with cardiac function, neuroticism, and depression. Linking differential methylation to antidepressant treatment response is an emerging topic and represents a step towards personalized medicine, potentially facilitating the prediction of patients' response before treatment.

What Works for Whom: Patients' Psychological Resources and Vulnerabilities as Common and Specific Predictors of Working Alliance in Different Psychotherapies

Background

Across different types of psychotherapy, one of the most robust predictors of better therapeutic outcomes is a good working alliance between patient and therapist. Yet there is little comparative research on whether particular patients more likely achieve a better alliance in certain treatments which represent particular therapeutic approaches or durations.

Methods

326 patients suffering from depressive and/or anxiety disorder were randomized into two short-term (solution-focused or psychodynamic) and one long-term (psychodynamic) therapy models. Treatments lasted ~7 and 36 months, respectively. Before randomization, patients were assessed with the interview-based Suitability for Psychotherapy Scale and filled Childhood Family Atmosphere and Life Orientation Test questionnaires. Patients filled Working Alliance Inventory after 3rd therapy session and at end of treatment; the long-term therapy patients, additionally, at 7 months' time point. Linear regression models were used.

Results

Greater psychological resources (e.g., capacity for self-reflection, affect regulation, flexible interaction) had little effect on alliance during the course of the short-term therapies. However, they did predict better working alliances at end of long-term as opposed to short-term therapy. Childhood adversities impacted alliances already at 7 months.

Conclusions

Although patients with certain qualities achieve better alliances in long-term as opposed to short-term therapies, apparently the theoretical orientation of therapy makes little difference. For patients with childhood adversities, differences between long-term (psychodynamic) treatment vs. various brief therapy models may be particularly salient.

Study protocol for pragmatic trials of Internet-delivered guided and unguided cognitive behavior therapy for treating depression and anxiety in university students of two Latin American countries: the Yo Puedo Sentirme Bien study

BACKGROUND

Major depressive disorder (MDD) and generalized anxiety disorder (GAD) are highly prevalent among university students and predict impaired college performance and later life role functioning. Yet most students do not receive treatment, especially in low-middle-income countries (LMICs). We aim to evaluate the effects of expanding treatment using scalable and inexpensive Internet-delivered transdiagnostic cognitive behavioral therapy (iCBT) among college students with symptoms of MDD and/or GAD in two LMICs in Latin America (Colombia and Mexico) and to investigate the feasibility of creating a precision treatment rule (PTR) to predict for whom iCBT is most effective.

METHODS

We will first carry out a multi-site randomized pragmatic clinical trial (N = 1500) of students seeking treatment at student mental health clinics in participating universities or responding to an email offering services. Students on wait lists for clinic services will be randomized to unguided iCBT (33%), guided iCBT (33%), and treatment as usual (TAU) (33%). iCBT will be provided immediately whereas TAU will be whenever a clinic appointment is available. Short-term aggregate effects will be assessed at 90 days and longer-term effects 12 months after randomization. We will use ensemble machine learning to predict heterogeneity of treatment effects of unguided versus guided iCBT versus TAU and develop a precision treatment rule (PTR) to optimize individual student outcome. We will then conduct a second and third trial with separate samples (n = 500 per arm), but with unequal allocation across two arms: 25% will be assigned to the treatment determined to yield optimal outcomes based on the PTR developed in the first trial (PTR for optimal short-term outcomes for Trial 2 and 12-month outcomes for Trial 3), whereas the remaining 75% will be assigned with equal allocation across all three treatment arms.

DISCUSSION

By collecting comprehensive baseline characteristics to evaluate heterogeneity of treatment effects, we will provide valuable and innovative information to optimize treatment effects and guide university mental health treatment planning. Such an effort could have enormous public-health implications for the region by increasing the reach of treatment, decreasing unmet need and clinic wait times, and serving as a model of evidence-based intervention planning and implementation.

TRIAL STATUS

IRB Approval of Protocol Version 1.0; June 3, 2020. Recruitment began on March 1, 2021. Recruitment is tentatively scheduled to be completed on May 30, 2024.

TRIAL REGISTRATION

ClinicalTrials.gov NCT04780542 . First submission date: February 28, 2021.

Patterns and correlates of patient-reported helpfulness of treatment for common mental and substance use disorders in the WHO World Mental Health Surveys

Patient-reported helpfulness of treatment is an important indicator of quality in patient-centered care. We examined its pathways and predictors among respondents to household surveys who reported ever receiving treatment for major depression, generalized anxiety disorder, social phobia, specific phobia, post-traumatic stress disorder, bipolar disorder, or alcohol use disorder. Data came from 30 community epidemiological surveys - 17 in high-income countries (HICs) and 13 in low- and middle-income countries (LMICs) - carried out as part of the World Health Organization (WHO)'s World Mental Health (WMH) Surveys. Respondents were asked whether treatment of each disorder was ever helpful and, if so, the number of professionals seen before receiving helpful treatment. Across all surveys and diagnostic categories, 26.1% of patients (N=10,035) reported being helped by the very first professional they saw. Persisting to a second professional after a first unhelpful treatment brought the cumulative probability of receiving helpful treatment to 51.2%. If patients persisted with up through eight professionals, the cumulative probability rose to 90.6%. However, only an estimated 22.8% of patients would have persisted in seeing these many professionals after repeatedly receiving treatments they considered not helpful. Although the proportion of individuals with disorders who sought treatment was higher and they were more persistent in HICs than LMICs, proportional helpfulness among treated cases was no different between HICs and LMICs. A wide range of predictors of perceived treatment helpfulness were found, some of them consistent across diagnostic categories and others unique to specific disorders. These results provide novel information about patient evaluations of treatment across diagnoses and countries varying in income level, and suggest that a critical issue in improving the quality of care for mental disorders should be fostering persistence in professional help-seeking if earlier treatments are not helpful.

Rehman A, Morris DW. Major Depressive Disorder: Existing Hypotheses about Pathophysiological Mechanisms and New Genetic Findings

Major depressive disorder (MDD) is a common mental disorder generally characterized by symptoms associated with mood, pleasure and effectiveness in daily life activities. MDD is ranked as a major contributor to worldwide disability. The complex pathogenesis of MDD is not yet understood, and this is a major cause of failure to develop new therapies and MDD recurrence. Here we summarize the literature on existing hypotheses about the pathophysiological mechanisms of MDD. We describe the different approaches undertaken to understand the molecular mechanism of MDD using genetic data. Hundreds of loci have now been identified by large genome-wide association studies (GWAS). We describe these studies and how they have provided information on the biological processes, cell types, tissues and druggable targets that are enriched for MDD risk genes. We detail our understanding of the genetic correlations and causal relationships between MDD and many psychiatric and non-psychiatric disorders and traits. We highlight the challenges associated with genetic studies, including the complexity of MDD genetics in diverse populations and the need for a study of rare variants and new studies of gene-environment interactions.

Polysaccharide Regulation of Intestinal Flora: A Viable Approach to Maintaining Normal Cognitive Performance and Treating Depression

The intestinal tract of a healthy body is home to a large variety and number of microorganisms that will affect every aspect of the host’s life. In recent years, polysaccharides have been found to be an important factor affecting intestinal flora. Polysaccharides are widely found in nature and play a key role in the life activities of living organisms. In the intestinal tract of living organisms, polysaccharides have many important functions, such as preventing the imbalance of intestinal flora and maintaining the integrity of the intestinal barrier. Moreover, recent studies suggest that gut microbes can influence brain health through the brain-gut axis. Therefore, maintaining brain health through polysaccharide modulation of gut flora deserves further study. In this review, we outline the mechanisms by which polysaccharides maintain normal intestinal flora structure, as well as improving cognitive function in the brain via the brain-gut axis by virtue of the intestinal flora. We also highlight the important role that gut microbes play in the pathogenesis of depression and the potential for treating depression through the use of polysaccharides to modulate the intestinal flora.

Predicting patients who will drop out of out-patient psychotherapy using machine learning algorithms

BACKGROUND

About 30% of patients drop out of cognitive-behavioural therapy (CBT), which has implications for psychiatric and psychological treatment. Findings concerning drop out remain heterogeneous.

AIMS

This paper aims to compare different machine-learning algorithms using nested cross-validation, evaluate their benefit in naturalistic settings, and identify the best model as well as the most important variables.

METHOD

The data-set consisted of 2543 out-patients treated with CBT. Assessment took place before session one. Twenty-one algorithms and ensembles were compared. Two parameters (Brier score, area under the curve (AUC)) were used for evaluation.

RESULTS

The best model was an ensemble that used Random Forest and nearest-neighbour modelling. During the training process, it was significantly better than generalised linear modelling (GLM) (Brier score: d = -2.93, 95% CI (-3.95, -1.90)); AUC: d = 0.59, 95% CI (0.11 to 1.06)). In the holdout sample, the ensemble was able to correctly identify 63.4% of cases of patients, whereas the GLM only identified 46.2% correctly. The most important predictors were lower education, lower scores on the Personality Style and Disorder Inventory (PSSI) compulsive scale, younger age, higher scores on the PSSI negativistic and PSSI antisocial scale as well as on the Brief Symptom Inventory (BSI) additional scale (mean of the four additional items) and BSI overall scale.

CONCLUSIONS

Machine learning improves drop-out predictions. However, not all algorithms are suited to naturalistic data-sets and binary events. Tree-based and boosted algorithms including a variable selection process seem well-suited, whereas more advanced algorithms such as neural networks do not.

Towards optimal treatment selection for borderline personality disorder patients (BOOTS): a study protocol for a multicenter randomized clinical trial comparing schema therapy and dialectical behavior therapy

BACKGROUND

Specialized evidence-based treatments have been developed and evaluated for borderline personality disorder (BPD), including Dialectical Behavior Therapy (DBT) and Schema Therapy (ST). Individual differences in treatment response to both ST and DBT have been observed across studies, but the factors driving these differences are largely unknown. Understanding which treatment works best for whom and why remain central issues in psychotherapy research. The aim of the present study is to improve treatment response of DBT and ST for BPD patients by a) identifying patient characteristics that predict (differential) treatment response (i.e., treatment selection) and b) understanding how both treatments lead to change (i.e., mechanisms of change). Moreover, the clinical effectiveness and cost-effectiveness of DBT and ST will be evaluated.

METHODS

The BOOTS trial is a multicenter randomized clinical trial conducted in a routine clinical setting in several outpatient clinics in the Netherlands. We aim to recruit 200 participants, to be randomized to DBT or ST. Patients receive a combined program of individual and group sessions for a maximum duration of 25 months. Data are collected at baseline until three-year follow-up. Candidate predictors of (differential) treatment response have been selected based on the literature, a patient representative of the Borderline Foundation of the Netherlands, and semi-structured interviews among 18 expert clinicians. In addition, BPD-treatment-specific (ST: beliefs and schema modes; DBT: emotion regulation and skills use), BPD-treatment-generic (therapeutic environment characterized by genuineness, safety, and equality), and non-specific (attachment and therapeutic alliance) mechanisms of change are assessed. The primary outcome measure is change in BPD manifestations. Secondary outcome measures include functioning, additional self-reported symptoms, and well-being.

DISCUSSION

The current study contributes to the optimization of treatments for BPD patients by extending our knowledge on "Which treatment - DBT or ST - works the best for which BPD patient, and why?", which is likely to yield important benefits for both BPD patients (e.g., prevention of overtreatment and potential harm of treatments) and society (e.g., increased economic productivity of patients and efficient use of treatments).

TRIAL REGISTRATION

Netherlands Trial Register, NL7699 , registered 25/04/2019 - retrospectively registered.

Cardiac autonomic functioning and clinical outcome in adolescent borderline personality disorder over two years

The present study aimed to expand on previous findings that pre-treatment autonomic nervous system (ANS) functioning serves as a predictor of clinical outcome in adolescent borderline personality disorder (BPD), while examining whether the relationship between ANS functioning and treatment outcome may vary as a function of early life maltreatment (ELM). ANS stress response was examined considering changes in heart rate (HR) and vagally-mediated heart rate variability (vmHRV) over different conditions of the Montreal Imaging Stress Task (MIST) in a clinical sample of N = 27 adolescents across the spectrum of BPD severity. Participants received in- and/or outpatient treatment, while clinical data was assessed at routine follow-ups. Clinical outcome was defined by change in the number of fulfilled BPD criteria (as measured using the SCID-II), severity of psychopathology (CGI-S), and global level of functioning (GAF), measured 12 and 24 months after baseline assessments. Mixed-effects (random-intercept/random slope) linear regression models were calculated to examine markers of ANS function as potential predictors of clinical outcome. Irrespective of the presence of ELM exposure, both vmHRV resting-state and stress recovery measures were identified as significant predictors of clinical outcome over time. This study adds to the existing literature by replicating and expanding on preliminary findings, considering also physiological reactivity and recovery in addition to resting-state measures of ANS functioning. The present results further highlight the potential of markers of ANS functioning to serve as objective measures in the process of monitoring patient progress and to make predictions regarding treatment outcome in psychiatry research.

Machine learning in the prediction of depression treatment outcomes: a systematic review and meta-analysis

BACKGROUND

Multiple treatments are effective for major depressive disorder (MDD), but the outcomes of each treatment vary broadly among individuals. Accurate prediction of outcomes is needed to help select a treatment that is likely to work for a given person. We aim to examine the performance of machine learning methods in delivering replicable predictions of treatment outcomes.

METHODS

Of 7732 non-duplicate records identified through literature search, we retained 59 eligible reports and extracted data on sample, treatment, predictors, machine learning method, and treatment outcome prediction. A minimum sample size of 100 and an adequate validation method were used to identify adequate-quality studies. The effects of study features on prediction accuracy were tested with mixed-effects models. Fifty-four of the studies provided accuracy estimates or other estimates that allowed calculation of balanced accuracy of predicting outcomes of treatment.

RESULTS

Eight adequate-quality studies reported a mean accuracy of 0.63 [95% confidence interval (CI) 0.56-0.71], which was significantly lower than a mean accuracy of 0.75 (95% CI 0.72-0.78) in the other 46 studies. Among the adequate-quality studies, accuracies were higher when predicting treatment resistance (0.69) and lower when predicting remission (0.60) or response (0.56). The choice of machine learning method, feature selection, and the ratio of features to individuals were not associated with reported accuracy.

CONCLUSIONS

The negative relationship between study quality and prediction accuracy, combined with a lack of independent replication, invites caution when evaluating the potential of machine learning applications for personalizing the treatment of depression.

Early Enhancement of Neuroplasticity Index, the Ratio of Serum Brain-Derived Neurotrophic Factor Level to HAMD-24 Score, in Predicting the Long-Term Antidepressant Efficacy

Background: Current mainstream treatment of major depressive disorder (MDD) has a disadvantage in delayed onset of efficacy, making detection of early signatures predicative of the long-term treatment efficacy urgent.

Methods: MDD patients were scored with HAMD-24 and serum brain-derived neurotrophic factor (BDNF) levels were measured at different times in two independent trials: a single-arm observation of Yueju pill, a clinically approved traditional multiherbal medicine, and a two-arm random placebo-controlled trial for Yueju vs escitalopram. The ratio of the BDNF level to HAMD-24 score, or neuroplasticity index (NI), and its derived parameters were used for correlation analysis and receiver operating characteristic (ROC) analysis.

Results: On both the early (4th) and final (28th) days, Yueju and escitalopram significantly reduced HAMD-24 scores, compared to baselines, but only Yueju increased BDNF at both times. For either Yueju or escitalopram treatment, NI, but not BDNF, at baseline was correlated to NIs at the early or final treatment day. NI at early time was significantly correlated to early NI enhancement from the baseline for both Yueju and escitalopram, and to final NI enhancement from the baseline for Yueju in both trials. ROC analysis supported the predictability of Yueju’s final treatment efficacy from early NI enhancement.

Limitations: The small sample size and 28 days of treatment time may lead to the impossibility of ROC analysis of escitalopram.

Conclusion: Early NI enhancement is useful for prediction of long-term efficacy of Yueju and presumably some other antidepressants.

Clinical Trial Registration: [www.ClinicalTrials.gov], identifier [ChiCTR1900021114].

The development and validation of a prognostic model to PREDICT Relapse of depression in adult patients in primary care: protocol for the PREDICTR study

BACKGROUND

Most patients who present with depression are treated in primary care by general practitioners (GPs). Relapse of depression is common (at least 50% of patients treated for depression will relapse after a single episode) and leads to considerable morbidity and decreased quality of life for patients. The majority of patients will relapse within 6 months, and those with a history of relapse are more likely to relapse in the future than those with no such history. GPs see a largely undifferentiated case-mix of patients, and once patients with depression reach remission, there is limited guidance to help GPs stratify patients according to risk of relapse. We aim to develop a prognostic model to predict an individual's risk of relapse within 6-8 months of entering remission. The long-term objective is to inform the clinical management of depression after the acute phase.

METHODS

We will develop a prognostic model using secondary analysis of individual participant data drawn from seven RCTs and one longitudinal cohort study in primary or community care settings. We will use logistic regression to predict the outcome of relapse of depression within 6-8 months. We plan to include the following established relapse predictors in the model: residual depressive symptoms, number of previous depressive episodes, co-morbid anxiety and severity of index episode. We will use a "full model" development approach, including all available predictors. Performance statistics (optimism-adjusted C-statistic, calibration-in-the-large, calibration slope) and calibration plots (with smoothed calibration curves) will be calculated. Generalisability of predictive performance will be assessed through internal-external cross-validation. Clinical utility will be explored through net benefit analysis.

DISCUSSION

We will derive a statistical model to predict relapse of depression in remitted depressed patients in primary care. Assuming the model has sufficient predictive performance, we outline the next steps including independent external validation and further assessment of clinical utility and impact.

STUDY REGISTRATION

ClinicalTrials.gov ID: NCT04666662.

Early onset of depression and treatment outcome in patients with major depressive disorder

Major depressive disorder (MDD) is a highly heterogeneous disorder, which may partly explain why treatment outcome using antidepressants is unsatisfactory. We investigated the onset of depression as a possible clinical marker for therapy response prediction in the context of somatic biomarkers blood pressure and plasma electrolyte concentration. 889 MDD patients were divided into early (EO, n = 226), intermediate (IO, n = 493), and late onset (LO, n = 169) patients and were analyzed for differences in socio-demographic and clinical parameters, comorbidities and treatment outcome as well as systolic blood pressure and electrolytes. EO patients more often suffered from a recurrent depression, had more previous depressive episodes, a higher rate of comorbid axis I and II disorders, and more often reported of suicidality (p < 0.001) compared to IO and LO patients. Treatment outcome was not different from IO and LO patients, although LO patients responded faster. EO patients who showed an early non-improvement of depression after 2 weeks of therapy (<20% improvement) had a 4.3-fold higher likelihood to become non-remitter as compared to LO patients with an early improvement. EO patients had significantly lower systolic blood pressure than patients with IO or LO and electrolytes in EO patients were significantly correlated with depression severity. Our results confirm other studies showing an association of an early onset of depression with a slower treatment response. The worse treatment outcome in patients with an additional early non-improvement to antidepressant therapy opens perspectives to develop and test individualized treatment approaches for EO and LO patients in the future, which may be based on differences in autonomic regulation.

Machine learning to predict clinical remission in depressed patients after acute phase selective serotonin reuptake inhibitor treatment

OBJECTIVE

Selective serotonin reuptake inhibitors (SSRIs) are suggested as the first-line treatment for patients with major depressive disorder (MDD), but the remission rate is unsatisfactory. We aimed to establish machine learning models and explore variables available at baseline to predict the 8-week outcome among patients taking SSRIs.

METHODS

Data from 400 patients were used to build machine learnings. The last observation carried forward approach was used to determine the remitter/non-remitter status of the patients at week 8. Using least absolute shrinkage and selection operator (LASSO) to select features, we built 4 different machine learning algorithms including gradient boosting decision tree, support vector machine (SVM), random forests, and logistic regression with five-fold cross-validation. Then, we adopted Shapley additive explanations (SHAP) values to interpret the model output.

RESULTS

The remission rate is 67.8%. We obtained 78 features from the baseline characteristics, including 25 sociodemographic characteristics, 31 clinical features, 15 psychological traits and 7 neurocognitive functions, and 13 of these features were selected to establish SVM. The accuracy of the SVM prediction is 74.49%, reaching an average area under the curve of 0.734±0.043. The sensitivity is 0.899±0.038 with a positive predictive value of 0.776±0.028. The specificity is 0.422±0.091 with a negative predictive value of 0.674±0.086. According to the SHAP values, neurocognitive functions and anxiety and hypochondriasis symptoms were important predictors.

CONCLUSION

Our study supports the utilization of machine learning approaches with inexpensive and highly accessible variables to accurately predict the 8-week treatment outcome of SSRIs in patients with MDD.

Is social support pre-treatment associated with prognosis for adults with depression in primary care?

OBJECTIVE

Depressed patients rate social support as important for prognosis, but evidence for a prognostic effect is lacking. We aimed to test the association between social support and prognosis independent of treatment type, and the severity of depression, and other clinical features indicating a more severe illness.

METHODS

Individual patient data were collated from all six eligible RCTs (n = 2858) of adults seeking treatment for depression in primary care. Participants were randomized to any treatment and completed the same baseline assessment of social support and clinical severity factors. Two-stage random effects meta-analyses were conducted.

RESULTS

Social support was associated with prognosis independent of randomized treatment but effects were smaller when adjusting for depressive symptoms and durations of depression and anxiety, history of antidepressant treatment, and comorbid panic disorder: percentage decrease in depressive symptoms at 3-4 months per z-score increase in social support = -4.14(95%CI: -6.91 to -1.29). Those with a severe lack of social support had considerably worse prognoses than those with no lack of social support: increase in depressive symptoms at 3-4 months = 14.64%(4.25% to 26.06%).

CONCLUSIONS

Overall, large differences in social support pre-treatment were associated with differences in prognostic outcomes. Adding the Social Support scale to clinical assessments may be informative, but after adjusting for routinely assessed clinical prognostic factors the differences in prognosis are unlikely to be of a clinically important magnitude. Future studies might investigate more intensive treatments and more regular clinical reviews to mitigate risks of poor prognosis for those reporting a severe lack of social support.

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.

A personalized index to inform selection of a trauma-focused or non-trauma-focused treatment for PTSD

PTSD treatment guidelines recommend several treatments with extensive empirical support, including Prolonged Exposure (PE), a trauma-focused treatment and Present-Centered Therapy (PCT), a non-trauma-focused therapy. Research to inform treatment selection has yielded inconsistent findings with single prognostic variables that are difficult to integrate into clinical decision-making. We examined whether a combination of prognostic factors can predict different benefits in a trauma-focused vs. a non-trauma-focused psychotherapy. We applied a multi-method variable selection procedure and developed a prognostic index (PI) with a sample of 267 female veterans and active-duty service members (mean age 45; SD = 9.37; 53% White) with current PTSD who began treatment in a randomized clinical trial comparing PE and PCT. We conducted linear regressions predicting outcomes (Clinician-Administered PTSD Scale score) with treatment condition, the PI, and the interaction between the PI and treatment condition. The interaction between treatment type and PI moderated treatment response, moderated post-treatment symptom severity, b = 0.30, SE_b = 0.15 [95% CI: 0.01, 0.60], p = .049. For the 64% of participants with the best prognoses, PE resulted in better post-treatment outcomes; for the remainder, there was no difference. Use of a PI may lead to optimized patient outcomes and greater confidence when selecting trauma-focused treatments.

An individualized treatment rule to optimize probability of remission by continuation, switching, or combining antidepressant medications after failing a first-line antidepressant in a two-stage randomized trial

BACKGROUND

There is growing interest in using composite individualized treatment rules (ITRs) to guide depression treatment selection, but best approaches for doing this are not widely known. We develop an ITR for depression remission based on secondary analysis of a recently published trial for second-line antidepression medication selection using a cutting-edge ensemble machine learning method.

METHODS

Data come from the SUN(^_^)D trial, an open-label, assessor blinded pragmatic trial of previously-untreated patients with major depressive disorder from 48 clinics in Japan. Initial clinic-level randomization assigned patients to 50 or 100 mg/day sertraline. We focus on the 1549 patients who failed to remit within 3 weeks and were then rerandomized at the individual-level to continuation with sertraline, switching to mirtazapine, or combining mirtazapine with sertraline. The outcome was remission 9 weeks post-baseline. Predictors included socio-demographics, clinical characteristics, baseline symptoms, changes in symptoms between baseline and week 3, and week 3 side effects.

RESULTS

Optimized treatment was associated with significantly increased cross-validated week 9 remission rates in both samples [5.3% (2.4%), p = 0.016 50 mg/day sample; 5.1% (2.7%), p = 0.031 100 mg/day sample] compared to randomization (30.1-30.8%). Optimization was also associated with significantly increased remission in both samples compared to continuation [24.7% in both: 11.2% (3.8%), p = 0.002 50 mg/day sample; 11.7% (3.9%), p = 0.001 100 mg/day sample]. Non-significant gains were found for optimization compared to switching or combining.

CONCLUSIONS

An ITR can be developed to improve second-line antidepressant selection, but replication in a larger study with more comprehensive baseline predictors might produce stronger and more stable results.

Major Depressive Disorder: Advances in Neuroscience Research and Translational Applications

Major depressive disorder (MDD), also referred to as depression, is one of the most common psychiatric disorders with a high economic burden. The etiology of depression is still not clear, but it is generally believed that MDD is a multifactorial disease caused by the interaction of social, psychological, and biological aspects. Therefore, there is no exact pathological theory that can independently explain its pathogenesis, involving genetics, neurobiology, and neuroimaging. At present, there are many treatment measures for patients with depression, including drug therapy, psychotherapy, and neuromodulation technology. In recent years, great progress has been made in the development of new antidepressants, some of which have been applied in the clinic. This article mainly reviews the research progress, pathogenesis, and treatment of MDD.

Inclusion of genetic variants in an ensemble of gradient boosting decision trees does not improve the prediction of citalopram treatment response

Identifying in advance who is unlikely to respond to a specific antidepressant treatment is crucial to precision medicine efforts. The current work leverages genome-wide genetic variation and machine learning to predict response to the antidepressant citalopram using data from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial (n = 1257 with both valid genomic and outcome data). A confirmatory approach selected 11 SNPs previously reported to predict response to escitalopram in a sample different from the current study. A novel exploratory approach selected SNPs from across the genome using nested cross-validation with elastic net logistic regression with a predominantly lasso penalty (alpha = 0.99). SNPs from each approach were combined with baseline clinical predictors and treatment response outcomes were predicted using a stacked ensemble of gradient boosting decision trees. Using pre-treatment clinical and symptom predictors only, out-of-fold prediction of a novel treatment response definition based on STAR*D treatment guidelines was acceptable, AUC = .659, 95% CI [0.629, 0.689]. The inclusion of SNPs using confirmatory or exploratory selection methods did not improve the out-of-fold prediction of treatment response (AUCs were .662, 95% CI [0.632, 0.692] and .655, 95% CI [0.625, 0.685], respectively). A similar pattern of results were observed for the secondary outcomes of the presence or absence of distressing side effects regardless of treatment response and achieving remission or satisfactory partial response, assuming medication tolerance. In the current study, incorporating SNP variation into prognostic models did not enhance the prediction of citalopram response in the STAR*D sample.

The clinical characterization of the patient with primary psychosis aimed at personalization of management

The current management of patients with primary psychosis worldwide is often remarkably stereotyped. In almost all cases an antipsychotic medica-tion is prescribed, with second-generation antipsychotics usually preferred to first-generation ones. Cognitive behavioral therapy is rarely used in the vast majority of countries, although there is evidence to support its efficacy. Psychosocial interventions are often provided, especially in chronic cases, but those applied are frequently not validated by research. Evidence-based family interventions and supported employment programs are seldom implemented in ordinary practice. Although the notion that patients with primary psychosis are at increased risk for cardiovascular diseases and diabetes mellitus is widely shared, it is not frequent that appropriate measures be implemented to address this problem. The view that the management of the patient with primary psychosis should be personalized is endorsed by the vast majority of clinicians, but this personalization is lacking or inadequate in most clinical contexts. Although many mental health services would declare themselves "recovery-oriented", it is not common that a focus on empowerment, identity, meaning and resilience is ensured in ordinary practice. The present paper aims to address this situation. It describes systematically the salient domains that should be considered in the characterization of the individual patient with primary psychosis aimed at personalization of management. These include positive and negative symptom dimensions, other psychopathological components, onset and course, neurocognition and social cognition, neurodevelopmental indicators; social functioning, quality of life and unmet needs; clinical staging, antecedent and concomitant psychiatric conditions, physical comorbidities, family history, history of obstetric complications, early and recent environmental exposures, protective factors and resilience, and internalized stigma. For each domain, simple assessment instruments are identified that could be considered for use in clinical practice and included in standardized decision tools. A management of primary psychosis is encouraged which takes into account all the available treatment modalities whose efficacy is supported by research evidence, selects and modulates them in the individual patient on the basis of the clinical characterization, addresses the patient's needs in terms of employment, housing, self-care, social relationships and education, and offers a focus on identity, meaning and resilience.

Precision medicine for long-term depression outcomes using the Personalized Advantage Index approach: cognitive therapy or interpersonal psychotherapy?

BACKGROUND

Psychotherapies for depression are equally effective on average, but individual responses vary widely. Outcomes can be improved by optimizing treatment selection using multivariate prediction models. A promising approach is the Personalized Advantage Index (PAI) that predicts the optimal treatment for a given individual and the magnitude of the advantage. The current study aimed to extend the PAI to long-term depression outcomes after acute-phase psychotherapy.

METHODS

Data come from a randomized trial comparing cognitive therapy (CT, n = 76) and interpersonal psychotherapy (IPT, n = 75) for major depressive disorder (MDD). Primary outcome was depression severity, as assessed by the BDI-II, during 17-month follow-up. First, predictors and moderators were selected from 38 pre-treatment variables using a two-step machine learning approach. Second, predictors and moderators were combined into a final model, from which PAI predictions were computed with cross-validation. Long-term PAI predictions were then compared to actual follow-up outcomes and post-treatment PAI predictions.

RESULTS

One predictor (parental alcohol abuse) and two moderators (recent life events; childhood maltreatment) were identified. Individuals assigned to their PAI-indicated treatment had lower follow-up depression severity compared to those assigned to their PAI-non-indicated treatment. This difference was significant in two subsets of the overall sample: those whose PAI score was in the upper 60%, and those whose PAI indicated CT, irrespective of magnitude. Long-term predictions did not overlap substantially with predictions for acute benefit.

CONCLUSIONS

If replicated, long-term PAI predictions could enhance precision medicine by selecting the optimal treatment for a given depressed individual over the long term.

The clinical characterization of the adult patient with depression aimed at personalization of management

Depression is widely acknowledged to be a heterogeneous entity, and the need to further characterize the individual patient who has received this diagnosis in order to personalize the management plan has been repeatedly emphasized. However, the research evidence that should guide this personalization is at present fragmentary, and the selection of treatment is usually based on the clinician's and/or the patient's preference and on safety issues, in a trial-and-error fashion, paying little attention to the particular features of the specific case. This may be one of the reasons why the majority of patients with a diagnosis of depression do not achieve remission with the first treatment they receive. The predominant pessimism about the actual feasibility of the personalization of treatment of depression in routine clinical practice has recently been tempered by some secondary analyses of databases from clinical trials, using approaches such as individual patient data meta-analysis and machine learning, which indicate that some variables may indeed contribute to the identification of patients who are likely to respond differently to various antidepressant drugs or to antidepressant medication vs. specific psychotherapies. The need to develop decision support tools guiding the personalization of treatment of depression has been recently reaffirmed, and the point made that these tools should be developed through large observational studies using a comprehensive battery of self-report and clinical measures. The present paper aims to describe systematically the salient domains that should be considered in this effort to personalize depression treatment. For each domain, the available research evidence is summarized, and the relevant assessment instruments are reviewed, with special attention to their suitability for use in routine clinical practice, also in view of their possible inclusion in the above-mentioned comprehensive battery of measures. The main unmet needs that research should address in this area are emphasized. Where the available evidence allows providing the clinician with specific advice that can already be used today to make the management of depression more personalized, this advice is highlighted. Indeed, some sections of the paper, such as those on neurocognition and on physical comorbidities, indicate that the modern management of depression is becoming increasingly complex, with several components other than simply the choice of an antidepressant and/or a psychotherapy, some of which can already be reliably personalized.

Findings From World Mental Health Surveys of the Perceived Helpfulness of Treatment for Patients With Major Depressive Disorder

IMPORTANCE

The perceived helpfulness of treatment is an important patient-centered measure that is a joint function of whether treatment professionals are perceived as helpful and whether patients persist in help-seeking after previous unhelpful treatments.

OBJECTIVE

To examine the prevalence and factors associated with the 2 main components of perceived helpfulness of treatment in a representative sample of individuals with a lifetime history of DSM-IV major depressive disorder (MDD).

DESIGN, SETTING, AND PARTICIPANTS

This study examined the results of a coordinated series of community epidemiologic surveys of noninstitutionalized adults using the World Health Organization World Mental Health surveys. Seventeen surveys were conducted in 16 countries (8 surveys in high-income countries and 9 in low- and middle-income countries). The dates of data collection ranged from 2002 to 2003 (Lebanon) to 2016 to 2017 (Bulgaria). Participants included those with a lifetime history of treated MDD. Data analyses were conducted from April 2019 to January 2020. Data on socioeconomic characteristics, lifetime comorbid conditions (eg, anxiety and substance use disorders), treatment type, treatment timing, and country income level were collected.

MAIN OUTCOMES AND MEASURES

Conditional probabilities of helpful treatment after seeing between 1 and 5 professionals; persistence in help-seeking after between 1 and 4 unhelpful treatments; and ever obtaining helpful treatment regardless of number of professionals seen.

RESULTS

Survey response rates ranged from 50.4% (Poland) to 97.2% (Medellín, Columbia), with a pooled response rate of 68.3% (n = 117 616) across surveys. Mean (SE) age at first depression treatment was 34.8 (0.3) years, and 69.4% were female. Of 2726 people with a lifetime history of treatment of MDD, the cumulative probability (SE) of all respondents pooled across countries of helpful treatment after seeing up to 10 professionals was 93.9% (1.2%), but only 21.5% (3.2%) of patients persisted that long (ie, beyond 9 unhelpful treatments), resulting in 68.2% (1.1%) of patients ever receiving treatment that they perceived as helpful. The probability of perceiving treatment as helpful increased in association with 4 factors: older age at initiating treatment (adjusted odds ratio [AOR], 1.02; 95% CI, 1.01-1.03), higher educational level (low: AOR, 0.48; 95% CI, 0.33-0.70; low-average: AOR, 0.62; 95% CI, 0.44-0.89; high average: AOR, 0.67; 95% CI, 0.49-0.91 vs high educational level), shorter delay in initiating treatment after first onset (AOR, 0.98; 95% CI, 0.97-0.99), and medication received from a mental health specialist (AOR, 2.91; 95% CI, 2.04-4.15). Decomposition analysis showed that the first 2 of these 4 factors were associated with only the conditional probability of an individual treatment professional being perceived as helpful (age at first depression treatment: AOR, 1.02; 95% CI, 1.01-1.02; educational level: low: AOR, 0.48; 95% CI, 0.33-0.70; low-average: AOR, 0.62; 95% CI, 0.44-0.89; high-average: AOR, 0.67; 95% CI, 0.49-0.91 vs high educational level), whereas the latter 2 factors were associated with only persistence (treatment delay: AOR, 0.98; 95% CI, 0.97-0.99; treatment type: AOR, 3.43; 95% CI, 2.51-4.70).

CONCLUSIONS AND RELEVANCE

The probability that patients with MDD obtain treatment that they consider helpful might increase, perhaps markedly, if they persisted in help-seeking after unhelpful treatments with up to 9 prior professionals.

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

IMPORTANCE

Antidepressants are commonly used worldwide to treat major depressive disorder. Symptomatic response to antidepressants can vary depending on differences between individuals; however, this variability may reflect nonspecific or random factors.

OBJECTIVES

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

DATA SOURCES

Data used were from a recent network meta-analysis of acute treatment with licensed antidepressants in adults with major depressive disorder. The following databases were searched from inception to January 8, 2016: the Cochrane Central Register of Controlled Trials, CINAHL, Embase, LILACS database, MEDLINE, MEDLINE In-Process, and PsycINFO. Additional sources were international trial registries, drug approval agency websites, and key scientific journals.

STUDY SELECTION

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

DATA EXTRACTION AND SYNTHESIS

Baseline and end point means, SDs, number of participants in each group, antidepressant class, and publication year were extracted. The data were analyzed between August 14 and November 18, 2019.

MAIN OUTCOMES AND MEASURES

With the use of validated methods, coefficients of variation were derived for antidepressants and placebo, and their ratios were calculated to compare outcome variability between antidepressant and placebo. Ratios were entered into a random-effects model, with the expectation that response to antidepressants would be more variable than response to placebo. Analysis was repeated after stratifying by baseline severity of depression, antidepressant class (selective serotonin reuptake inhibitors: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline, and vilazodone; serotonin and norepinephrine reuptake inhibitors: desvenlafaxine and venlafaxine; norepinephrine-dopamine reuptake inhibitor: bupropion; noradrenergic agents: amitriptyline and reboxetine; and other antidepressants: agomelatine, mirtazapine, and trazodone), and publication year.

RESULTS

In the 87 eligible randomized placebo-controlled trials (17 540 unique participants), there was significantly more variability in response to antidepressants than to placebo (coefficients of variation ratio, 1.14; 95% CI, 1.11-1.17; P < .001). Baseline severity of depression did not moderate variability in response to antidepressants. Variability in response to selective serotonin reuptake inhibitors was lower than variability in response to noradrenergic agents (coefficients of variation ratio, 0.88; 95% CI, 0.80-0.97; P = .01), as was the variability in response to other antidepressants compared with noradrenergic agents (coefficients of variation ratio, 0.87; 95% CI, 0.79-0.97; P = .001). Variability also tended to be lower in studies that were published more recently, with coefficients of variation changing by a value of 0.005 (95% CI, 0.002-0.008; P = .003) for every year a study is more recent.

CONCLUSIONS AND RELEVANCE

Individual differences may be systematically associated with responses to antidepressants in major depressive disorder beyond placebo effects or statistical factors. This study provides empirical support for identifying moderators and personalizing antidepressant treatment.

What factors indicate prognosis for adults with depression in primary care? A protocol for meta-analyses of individual patient data using the Dep-GP database

Background: Pre-treatment severity is a key indicator of prognosis for those with depression. Knowledge is limited on how best to encompass severity of disorders. A number of non-severity related factors such as social support and life events are also indicators of prognosis. It is not clear whether this holds true after adjusting for pre-treatment severity as a) a depressive symptom scale score, and b) a broader construct encompassing symptom severity and related indicators: "disorder severity". In order to investigate this, data from the individual participants of clinical trials which have measured a breadth of "disorder severity" related factors are needed. Aims: 1) To assess the association between outcomes for adults seeking treatment for depression and the severity of depression pre-treatment, considered both as i) depressive symptom severity only and ii) "disorder severity" which includes depressive symptom severity and comorbid anxiety, chronicity, history of depression, history of previous treatment, functional impairment and health-related quality of life. 2) To determine whether i) social support, ii) life events, iii) alcohol misuse, and iv) demographic factors (sex, age, ethnicity, marital status, employment status, level of educational attainment, and financial wellbeing) are prognostic indicators of outcomes, independent of baseline "disorder severity" and the type of treatment received. Methods: Databases were searched for randomised clinical trials (RCTs) that recruited adults seeking treatment for depression from their general practitioners and used the same diagnostic and screening instrument to measure severity at baseline - the Revised Clinical Interview Schedule; outcome measures could differ between studies. Chief investigators of all studies meeting inclusion criteria were contacted and individual patient data (IPD) were requested. Conclusions: In total 15 RCTs met inclusion criteria. The Dep-GP database will include the 6271 participants from the 13 studies that provided IPD. This protocol outlines how these data will be analysed. Registration: PROSPERO CRD42019129512 (01/04/2019).