Search

Background: Major depressive disorder (MDD) is a highly prevalent condition with a substantial incidence of relapse or treatment resistance. A subset of patients show evidence of low-grade inflammation, with these patients having a higher likelihood of more severe or difficult to treat courses of illness. Anti-inflammatory treatment of MDD has been investigated with mixed results, and no known studies have included assessments beyond cessation of the anti-inflammatory agent, meaning it remains unknown if any benefit from treatment persists. The objective of the present study was to investigate treatment outcomes up to 29 weeks post-cessation of celecoxib or placebo augmentation of an antidepressant, and how concentrations of selected inflammatory markers change over the same period., Methods: The PREDDICT parallel-group, randomised, double-blind, placebo-controlled trial (University of Adelaide, Australia) ran from December 2017 to April 2020. Participants with MDD were stratified into normal range or elevated inflammation strata according to screening concentrations of high sensitivity C-reactive protein (hsCRP). Participants were randomised to treatment with vortioxetine and celecoxib or vortioxetine and placebo for six weeks, and vortioxetine alone for an additional 29 weeks (35 total weeks). Following a previous publication of results from the six-week RCT phase, exploratory analyses were performed on Montgomery-Åsberg Depression Rating Scale (MADRS) scores, response and remission outcomes, and selected peripheral inflammatory markers across the entire study duration up to week 35., Results: Participants retained at each observation were baseline N=119, week 2 N=115, week 4 N=103, week 6 N=104, week 8 N=98, week 22 N=81, and week 35 N=60. Those in the elevated hsCRP celecoxib-augmented group had a statistically significantly greater reduction in MADRS score from baseline to week 35 compared to all other groups, demonstrating the greatest clinical improvement long-term, despite no group or strata differences at preceding time points. Response and remission outcomes did not differ by treatment group or hsCRP strata at any time point. Changes in hsCRP between baseline and week 35 and Tumour Necrosis Factor-α (TNF-α) concentrations between baseline and week 6 and baseline and week 35 were statistically significantly associated with MADRS scores observed at week 6 and week 35 respectively, with reducing TNF-α concentrations associated with reducing MADRS scores and vice versa in each case. A post-hoc stratification of the participant cohort by baseline TNF-α concentrations led to significant prediction by the derived strata on clinical response at weeks 6, 8 and 35, with participants with elevated baseline TNF-α less likely to achieve clinical response., Interpretation: The present analysis suggests for the first time a possible longer-term clinical benefit of celecoxib augmentation of vortioxetine in inflammation-associated MDD treatment. However, further research is needed to confirm the finding and to ascertain the reason for such a delayed effect. Furthermore, the trial suggests that TNF-α may have a stronger relationship with anti-inflammatory MDD treatment outcomes than hsCRP, and should be investigated further for potential predictive utility., Clinical Trials Registration: Australian New Zealand Clinical Trials Registry (ANZCTR), ACTRN12617000527369p. Registered on 11 April 2017, http://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?ACTRN=12617000527369p., (Copyright © 2024. Published by Elsevier Inc.)

Background: Lithium (Li) remains the treatment of choice for bipolar disorders (BP). Its mood-stabilizing effects help reduce the long-term burden of mania, depression and suicide risk in patients with BP. It also has been shown to have beneficial effects on disease-associated conditions, including sleep and cardiovascular disorders. However, the individual responses to Li treatment vary within and between diagnostic subtypes of BP (e.g. BP-I and BP-II) according to the clinical presentation. Moreover, long-term Li treatment has been linked to adverse side-effects that are a cause of concern and non-adherence, including the risk of developing chronic medical conditions such as thyroid and renal disease. In recent years, studies by the Consortium on Lithium Genetics (ConLiGen) have uncovered a number of genetic factors that contribute to the variability in Li treatment response in patients with BP. Here, we leveraged the ConLiGen cohort (N = 2064) to investigate the genetic basis of Li effects in BP. For this, we studied how Li response and linked genes associate with the psychiatric symptoms and polygenic load for medical comorbidities, placing particular emphasis on identifying differences between BP-I and BP-II., Results: We found that clinical response to Li treatment, measured with the Alda scale, was associated with a diminished burden of mania, depression, substance and alcohol abuse, psychosis and suicidal ideation in patients with BP-I and, in patients with BP-II, of depression only. Our genetic analyses showed that a stronger clinical response to Li was modestly related to lower polygenic load for diabetes and hypertension in BP-I but not BP-II. Moreover, our results suggested that a number of genes that have been previously linked to Li response variability in BP differentially relate to the psychiatric symptomatology, particularly to the numbers of manic and depressive episodes, and to the polygenic load for comorbid conditions, including diabetes, hypertension and hypothyroidism., Conclusions: Taken together, our findings suggest that the effects of Li on symptomatology and comorbidity in BP are partially modulated by common genetic factors, with differential effects between BP-I and BP-II., (© 2024. The Author(s).)

Background: Multimodal modeling that combines biological and clinical data shows promise in predicting transition to psychosis in individuals who are at ultra-high risk. Individuals who transition to psychosis are known to have deficits at baseline in cognitive function and reductions in gray matter volume in multiple brain regions identified by magnetic resonance imaging., Methods: In this study, we used Cox proportional hazards regression models to assess the additive predictive value of each modality-cognition, cortical structure information, and the neuroanatomical measure of brain age gap-to a previously developed clinical model using functioning and duration of symptoms prior to service entry as predictors in the Personal Assessment and Crisis Evaluation (PACE) 400 cohort. The PACE 400 study is a well-characterized cohort of Australian youths who were identified as ultra-high risk of transitioning to psychosis using the Comprehensive Assessment of At Risk Mental States (CAARMS) and followed for up to 18 years; it contains clinical data (from N = 416 participants), cognitive data (n = 213), and magnetic resonance imaging cortical parameters extracted using FreeSurfer (n = 231)., Results: The results showed that neuroimaging, brain age gap, and cognition added marginal predictive information to the previously developed clinical model (fraction of new information: neuroimaging 0%-12%, brain age gap 7%, cognition 0%-16%)., Conclusions: In summary, adding a second modality to a clinical risk model predicting the onset of a psychotic disorder in the PACE 400 cohort showed little improvement in the fit of the model for long-term prediction of transition to psychosis., (Copyright © 2023 Society of Biological Psychiatry. Published by Elsevier Inc. All rights reserved.)

The CERT-D program offers a new treatment approach addressing disturbed cognitive and psychosocial functioning in major depressive disorder (MDD). The current analysis of a randomised controlled trial (RCT) comprises two objectives: Firstly, evaluating the program's efficacy of a personalised versus standard treatment and secondly, assessing the treatment's persistence longitudinally. Participants (N = 112) were randomised into a personalised or standard treatment group. Both groups received 8 weeks of cognitive training, followed by a three-month follow-up without additional training. The type of personalised training was determined by pre-treatment impairments in the domains of cognition, emotion-processing and social-cognition. Standard training addressed all three domains equivalent. Performance in these domains was assessed repeatedly during RCT and follow-up. Treatment comparisons during the RCT-period showed benefits of personalised versus standard treatment in certain aspects of social-cognition. Conversely, no benefits in the remaining domains were found, contradicting a general advantage of personalisation. Exploratory follow-up analysis on persistence of the program's effects indicated sustained intervention outcomes across the entire sample. A subsequent comparison of clinical outcomes between personalised versus standard treatment over a three-month follow-up period showed similar results. First evidence suggests that existing therapies for MDD could benefit from an adjunct administration of the CERT-D program., Competing Interests: Declaration of Competing Interest BB received speaker/consultation fees from: AstraZeneca, Lundbeck, Pfizer, Takeda, Servier, Bristol Myers Squibb, Otsuka, LivaNova, Boehringer-Ingelheim, Biogen and Janssen-Cilag. All other authors have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier B.V. All rights reserved.)

Background: The efficacy of digital meditation is well established. However, the extent to which the benefits remain after 12 weeks in real-world settings remains unknown. Additionally, findings related to dosage and practice habits have been mixed, and the studies were conducted on small and homogeneous samples and used a limited range of analytical procedures and meditation techniques. Findings related to the predictors of adherence are also lacking and may help inform future meditators and meditation programs on how to best structure healthy sustainable practices., Objective: This study aimed to measure outcome change across a large and globally diverse population of meditators and meditations in their naturalistic practice environments, assess the dose-response relationships between practice habits and outcome change, and identify predictors of adherence., Methods: We used ecological momentary assessment to assess participants' well-being over a 14-month period. We engineered outcomes related to the variability of change over time (equanimity) and recovery following a drop in mood (resilience) and established the convergent and divergent validity of these outcomes using a validated scale. Using linear mixed-effects and generalized additive mixed-effects models, we modeled outcome changes and patterns of dose-response across outcomes. We then used logistic regression to study the practice habits of participants in their first 30 sessions to derive odds ratios of long-term adherence., Results: Significant improvements were observed in all outcomes (P<.001). Generalized additive mixed models revealed rapid improvements over the first 50-100 sessions, with further improvements observed until the end of the study period. Outcome change corresponded to 1 extra day of improved mood for every 5 days meditated and half-a-day-faster mood recovery compared with baseline. Overall, consistency of practice was associated with the largest outcome change (4-7 d/wk). No significant differences were observed across session lengths in linear models (mood: P=.19; equanimity: P=.10; resilience: P=.29); however, generalized additive models revealed significant differences over time (P<.001). Longer sessions (21-30 min) were associated with the largest magnitude of change in mood from the 20th session onward and fewer sessions to recovery (increased resilience); midlength sessions (11-20 min) were associated with the largest decreases in recovery; and mood stability was similar across session lengths (equanimity). Completing a greater variety of practice types was associated with significantly greater improvements across all outcomes. Adhering to a long-term practice was best predicted by practice consistency (4-7 d/wk), a morning routine, and maintaining an equal balance between interoceptive and exteroceptive meditations., Conclusions: Long-term real-world digital meditation practice is effective and associated with improvements in mood, equanimity, and resilience. Practice consistency and variety rather than length best predict improvement. Long-term sustainable practices are best predicted by consistency, a morning routine, and a practice balanced across objects of focus that are internal and external to the body., (©Micah Cearns, Scott R Clark. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 19.09.2023.)

Background: Validated clinical prediction models of short-term remission in psychosis are lacking. Our aim was to develop a clinical prediction model aimed at predicting 4-6-week remission following a first episode of psychosis., Method: Baseline clinical data from the Athens First Episode Research Study was used to develop a Support Vector Machine prediction model of 4-week symptom remission in first-episode psychosis patients using repeated nested cross-validation. This model was further tested to predict 6-week remission in a sample of two independent, consecutive Danish first-episode cohorts., Results: Of the 179 participants in Athens, 120 were male with an average age of 25.8 years and average duration of untreated psychosis of 32.8 weeks. 62.9% were antipsychotic-naïve. Fifty-seven percent attained remission after 4 weeks. In the Danish cohort, 31% attained remission. Eleven clinical scale items were selected in the Athens 4-week remission cohort. These included the Duration of Untreated Psychosis, Personal and Social Performance Scale, Global Assessment of Functioning and eight items from the Positive and Negative Syndrome Scale. This model significantly predicted 4-week remission status (area under the receiver operator characteristic curve (ROC-AUC) = 71.45, P < .0001). It also predicted 6-week remission status in the Danish cohort (ROC-AUC = 67.74, P < .0001), demonstrating reliability., Conclusions: Using items from common and validated clinical scales, our model significantly predicted early remission in patients with first-episode psychosis. Although replicated in an independent cohort, forward testing between machine learning models and clinicians' assessment should be undertaken to evaluate the possible utility as a routine clinical tool., (© The Author(s) 2021. Published by Oxford University Press on behalf of the Maryland Psychiatric Research Center.All rights reserved. For permissions, please email: journals.permissions@oup.com.)

Background: Suppression of the rostral anterior cingulate cortex (rACC) has shown promise as a prognostic biomarker for depression. We aimed to use machine learning to characterise its ability to predict depression remission., Methods: Data were obtained from 81 15- to 25-year-olds with a major depressive disorder who had participated in the YoDA-C trial, in which they had been randomised to receive cognitive behavioural therapy plus either fluoxetine or placebo. Prior to commencing treatment patients performed a functional magnetic resonance imaging (fMRI) task to assess rACC suppression. Support vector machines were trained on the fMRI data using nested cross-validation, and were similarly trained on clinical data. We further tested our fMRI model on data from the YoDA-A trial, in which participants had completed the same fMRI paradigm., Results: Thirty-six of 81 (44%) participants in the YoDA-C trial achieved remission. Our fMRI model was able to predict remission status (AUC = 0.777 [95% confidence interval (CI) 0.638-0.916], balanced accuracy = 67%, negative predictive value = 74%, p < 0.0001). Clinical models failed to predict remission status at better than chance levels. Testing the model on the alternative YoDA-A dataset confirmed its ability to predict remission (AUC = 0.776, balanced accuracy = 64%, negative predictive value = 70%, p < 0.0001)., Conclusions: We confirm that rACC activity acts as a prognostic biomarker for depression. The machine learning model can identify patients who are likely to have difficult-to-treat depression, which might direct the earlier provision of enhanced support and more intensive therapies.

Lithium is the gold standard therapy for Bipolar Disorder (BD) but its effectiveness differs widely between individuals. The molecular mechanisms underlying treatment response heterogeneity are not well understood, and personalized treatment in BD remains elusive. Genetic analyses of the lithium treatment response phenotype may generate novel molecular insights into lithium's therapeutic mechanisms and lead to testable hypotheses to improve BD management and outcomes. We used fixed effect meta-analysis techniques to develop meta-analytic polygenic risk scores (MET-PRS) from combinations of highly correlated psychiatric traits, namely schizophrenia (SCZ), major depression (MD) and bipolar disorder (BD). We compared the effects of cross-disorder MET-PRS and single genetic trait PRS on lithium response. For the PRS analyses, we included clinical data on lithium treatment response and genetic information for n = 2283 BD cases from the International Consortium on Lithium Genetics (ConLi + Gen; www.ConLiGen.org ). Higher SCZ and MD PRSs were associated with poorer lithium treatment response whereas BD-PRS had no association with treatment outcome. The combined MET2-PRS comprising of SCZ and MD variants (MET2-PRS) and a model using SCZ and MD-PRS sequentially improved response prediction, compared to single-disorder PRS or to a combined score using all three traits (MET3-PRS). Patients in the highest decile for MET2-PRS loading had 2.5 times higher odds of being classified as poor responders than patients with the lowest decile MET2-PRS scores. An exploratory functional pathway analysis of top MET2-PRS variants was conducted. Findings may inform the development of future testing strategies for personalized lithium prescribing in BD., (© 2021. The Author(s).)

Bipolar affective disorder (BD) is a severe psychiatric illness, for which lithium (Li) is the gold standard for acute and maintenance therapies. The therapeutic response to Li in BD is heterogeneous and reliable biomarkers allowing patients stratification are still needed. A GWAS performed by the International Consortium on Lithium Genetics (ConLiGen) has recently identified genetic markers associated with treatment responses to Li in the human leukocyte antigens (HLA) region. To better understand the molecular mechanisms underlying this association, we have genetically imputed the classical alleles of the HLA region in the European patients of the ConLiGen cohort. We found our best signal for amino-acid variants belonging to the HLA-DRB1*11:01 classical allele, associated with a better response to Li (p < 1 × 10 -3 ; FDR < 0.09 in the recessive model). Alanine or Leucine at position 74 of the HLA-DRB1 heavy chain was associated with a good response while Arginine or Glutamic acid with a poor response. As these variants have been implicated in common inflammatory/autoimmune processes, our findings strongly suggest that HLA-mediated low inflammatory background may contribute to the efficient response to Li in BD patients, while an inflammatory status overriding Li anti-inflammatory properties would favor a weak response., (© 2021. The Author(s).)

We currently observe a disconcerting phenomenon in machine learning studies in psychiatry: While we would expect larger samples to yield better results due to the availability of more data, larger machine learning studies consistently show much weaker performance than the numerous small-scale studies. Here, we systematically investigated this effect focusing on one of the most heavily studied questions in the field, namely the classification of patients suffering from Major Depressive Disorder (MDD) and healthy controls based on neuroimaging data. Drawing upon structural MRI data from a balanced sample of N = 1868 MDD patients and healthy controls from our recent international Predictive Analytics Competition (PAC), we first trained and tested a classification model on the full dataset which yielded an accuracy of 61%. Next, we mimicked the process by which researchers would draw samples of various sizes (N = 4 to N = 150) from the population and showed a strong risk of misestimation. Specifically, for small sample sizes (N = 20), we observe accuracies of up to 95%. For medium sample sizes (N = 100) accuracies up to 75% were found. Importantly, further investigation showed that sufficiently large test sets effectively protect against performance misestimation whereas larger datasets per se do not. While these results question the validity of a substantial part of the current literature, we outline the relatively low-cost remedy of larger test sets, which is readily available in most cases.

Cognitive and emotional remediation training for depression (CERT-D): a randomised controlled trial to improve cognitive, emotional and functional outcomes in depression The aim of the current study was to evaluate an experimental treatment designed to improve psychosocial function in patients with Major Depressive Disorder (MDD) by reinforcing cognitive, emotional, and social-cognitive abilities. Participants (N = 112) with current or lifetime MDD were recruited to participate in a randomised, blinded, controlled trial. Exclusion criteria included diagnosis of a substance abuse disorder, bipolar disorder organic, eating disorders, or illness which affect cognitive function. The treatment involved repeated cognitive training designed to improve cognitive, emotional, and social-cognitive abilities. In training sessions, the principles of cognitive training were applied across cognitive, emotional, and social domains, with participants completing repeated mental exercises. Exercises included critically analysing interpretations of social interactions (e.g., body language), exploring emotional reactions to stimuli, and completing game-like cognitive training tasks. Training sessions placed great emphasis on the application of trained cognitive, emotional, and social cognitive skills to psychosocial outcomes. Outcomes demonstrated significant improvement in psychosocial function, symptom severity, self-reported cognition, and social-cognition. Our findings demonstrate the efficacy of multi-domain cognitive training to improve psychosocial functioning in individuals with MDD. We suggest that the present treatment could be deployed at a lower cost and with minimal training in comparison to established psychological therapies., (Copyright © 2021. Published by Elsevier B.V.)

It is imperative to understand the specific and shared etiologies of major depression and cardio-metabolic disease, as both traits are frequently comorbid and each represents a major burden to society. This study examined whether there is a genetic association between major depression and cardio-metabolic traits and if this association is stratified by age at onset for major depression. Polygenic risk scores analysis and linkage disequilibrium score regression was performed to examine whether differences in shared genetic etiology exist between depression case control status (N cases = 40,940, N controls = 67,532), earlier (N = 15,844), and later onset depression (N = 15,800) with body mass index, coronary artery disease, stroke, and type 2 diabetes in 11 data sets from the Psychiatric Genomics Consortium, Generation Scotland, and UK Biobank. All cardio-metabolic polygenic risk scores were associated with depression status. Significant genetic correlations were found between depression and body mass index, coronary artery disease, and type 2 diabetes. Higher polygenic risk for body mass index, coronary artery disease, and type 2 diabetes was associated with both early and later onset depression, while higher polygenic risk for stroke was associated with later onset depression only. Significant genetic correlations were found between body mass index and later onset depression, and between coronary artery disease and both early and late onset depression. The phenotypic associations between major depression and cardio-metabolic traits may partly reflect their overlapping genetic etiology irrespective of the age depression first presents., (© 2020 The Authors. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics published by Wiley Periodicals LLC.)

Objective: The increasing challenges facing the training of future neurosurgeons have led to continued development of simulation-based training, particularly for neurosurgical subspecialties. The simulators must be scientifically validated to fully assess their benefit and determine their educational effects. In this second part, we aim to identify the available simulators for spine, pediatric neurosurgery, interventional neuroradiology, and nontechnical skills, assess their validity, and determine their effectiveness., Methods: Both Medline and Embase were searched for English language articles that validate simulation models for neurosurgery. Each study was screened according to the Messick validity framework, and rated in each domain. The McGaghie model of translational outcomes was then used to determine a level of effectiveness for each simulator or training course., Results: Overall, 114 articles for 108 simulation-based training models or courses were identified. These articles included 24 for spine simulators, 3 for nontechnical skills, 10 for 9 pediatric neurosurgery simulators, and 12 for 11 interventional neuroradiology simulators. Achieving the highest rating for each validity domain were 3 models for content validity; 16 for response processes; 1 for internal structure; 2 for relations to other variables; and only 1 for consequences. For translational outcomes, 2 training courses achieved a level of effectiveness of >2, showing skills transfer beyond the simulator environment., Conclusions: With increasing simulators, there is a need for more validity studies and attempts to investigate translational outcomes to the operating theater when using these simulators. Nontechnical skills training is notably lacking, despite demand within the field., (Copyright © 2019 Elsevier Inc. All rights reserved.)

Objective: The recent emphasis on simulation-based training in neurosurgery has led to the development of many simulation models and training courses. We aim to identify the currently available simulators and training courses for neurosurgery, assess their validity, and determine their effectiveness., Methods: Both MEDLINE and Embase were searched for English language articles which validate simulation models for neurosurgery. Each study was screened according to the Messick validity framework and rated in each domain. The McGaghie model of translational outcomes was then used to determine a level of effectiveness (LoE) for each simulator or training course., Results: On screening of 6006 articles, 114 were identified to either validate or determine an LoE for 108 simulation-based training models or courses. Achieving the highest rating for each validity domain were 6 models and training courses for content validity, 12 for response processes, 4 for internal structure, 14 for relations to other variables, and none for consequences. For translational outcomes, 6 simulators or training achieved an LoE >2 and thus showed skills transfer beyond the simulation setting., Conclusions: With the advent of increasing neurosurgery simulators and training tools, there is a need for more validity studies. Further attempts to investigate translational outcomes to the operating theater when using these simulators is particularly warranted. More training tools incorporating full-immersion simulation and nontechnical skills training are recommended., (Copyright © 2019 Elsevier Inc. All rights reserved.)

Machine learning methods show promise to translate univariate biomarker findings into clinically useful multivariate decision support systems. At current, works in major depressive disorder have predominantly focused on neuroimaging and clinical predictor modalities, with genetic, blood-biomarker, and cardiovascular modalities lacking. In addition, the prediction of rehospitalization after an initial inpatient major depressive episode is yet to be explored, despite its clinical importance. To address this gap in the literature, we have used baseline clinical, structural imaging, blood-biomarker, genetic (polygenic risk scores), bioelectrical impedance and electrocardiography predictors to predict rehospitalization within 2 years of an initial inpatient episode of major depression. Three hundred and eighty patients from the ongoing 12-year Bidirect study were included in the analysis (rehospitalized: yes = 102, no = 278). Inclusion criteria was age ≥35 and <66 years, a current or recent hospitalisation for a major depressive episode and complete structural imaging and genetic data. Optimal performance was achieved with a multimodal panel containing structural imaging, blood-biomarker, clinical, medication type, and sleep quality predictors, attaining a test AUC of 67.74 (p = 9.99 -05 ). This multimodal solution outperformed models based on clinical variables alone, combined biomarkers, and individual data modality prognostication for rehospitalization prediction. This finding points to the potential of predictive models that combine multimodal clinical and biomarker data in the development of clinical decision support systems.

Background: Preliminary research suggests that major depressive disorder (MDD) is associated with structural alterations in the brain; as well as with low-grade peripheral inflammation. However, even though a link between inflammatory processes and altered brain structural integrity has been purported by experimental research, well-powered studies to confirm this hypothesis in patients with MDD have been lacking. We aimed to investigate the potential association between structural brain alterations and low-grade inflammation as interrelated biological correlates of MDD., Methods: In this cross-sectional study, 514 patients with MDD and 359 healthy controls underwent structural MRI. We used voxel-based morphometry to study local differences in grey matter volume. We also assessed serum levels of high-sensitivity C-reactive protein (hsCRP) in each participant., Results: Compared with healthy controls (age [mean ± standard deviation] 52.57 ± 7.94 yr; 50% male), patients with MDD (49.14 ± 7.28 yr, 39% male) exhibited significantly increased hsCRP levels (Z = −5.562, p < 0.001) and significantly decreased grey matter volume in the prefrontal cortex and the insula. Prefrontal grey matter volume reductions were significantly associated with higher hsCRP levels in patients with MDD (x = 50, y = 50, z = 8; t1,501 = 5.15; k = 92; pFWE < 0.001). In the MDD sample, the significant negative association between hsCRP and grey matter appeared independent of age, sex, body mass index, current smoking status, antidepressant load, hospitalization and medical comorbidities., Limitations: This study had a cross-sectional design., Conclusion: The present study highlights the role of reduced grey matter volume and low-grade peripheral inflammation as interrelated biological correlates of MDD. The reported inverse association between peripheral low-grade inflammation and brain structural integrity in patients with MDD translates current knowledge from experimental studies to the bedside., Competing Interests: B. Baune is member of advisory boards, received funding and/or gave presentations for AstraZeneca, Lundbeck, Janssen, Pfizer, Servier, and Wyeth, outside the submitted work. No other competing interests declared., (© 2019 Joule Inc. or its licensors)

Machine learning methods hold promise for personalized care in psychiatry, demonstrating the potential to tailor treatment decisions and stratify patients into clinically meaningful taxonomies. Subsequently, publication counts applying machine learning methods have risen, with different data modalities, mathematically distinct models, and samples of varying size being used to train and test models with the promise of clinical translation. Consequently, and in part due to the preliminary nature of such works, many studies have reported largely varying degrees of accuracy, raising concerns over systematic overestimation and methodological inconsistencies. Furthermore, a lack of procedural evaluation guidelines for non-expert medical professionals and funding bodies leaves many in the field with no means to systematically evaluate the claims, maturity, and clinical readiness of a project. Given the potential of machine learning methods to transform patient care, albeit, contingent on the rigor of employed methods and their dissemination, we deem it necessary to provide a review of current methods, recommendations, and future directions for applied machine learning in psychiatry. In this review we will cover issues of best practice for model training and evaluation, sources of systematic error and overestimation, model explainability vs. trust, the clinical implementation of AI systems, and finally, future directions for our field.

The St John Ambulance organization in Alberta issues over 100,000 first aid certificates each year (an 80% market share). The success of the organization is attributed to a province-wide infrastructure of volunteer public first aid providers and over 2000 volunteer instructors in industry and other spheres. In 1998 St John Ambulance piloted the use of CD-ROMs for teaching, producing two courses. One was a first aid course and the other was a babysitter course. The experience of the pilot programmes will allow the digital material to be transferred to our Website. Once this is done people around the world will be able to receive the self-paced learning components. Hands-on evaluation of skills, however, is mandatory to certification. Mass training where certification is required therefore requires a mixture of distance learning and hands-on testing, reinforcing the need for trained evaluators to be available to the population.