Individuals with bipolar disorder often experience reduced quality of life (QoL). Transcranial direct current stimulation (tDCS) is a promising non-invasive treatment for bipolar depression that is portable, safe, and suitable for use at home. We developed a home-based tDCS protocol with real-time remote supervision and examined its effect on QoL in bipolar depression. In an open‐label design, 44 participants (31 women) with bipolar depression of at least a moderate severity received 21 sessions of home‐based tDCS (2 mA, 30 min, F3 anode/F4 cathode) over 6 weeks, with a follow-up visit conducted 5 months from baseline. QoL was assessed using the quality of life enjoyment and satisfaction questionnaire (Q-LES-Q) at baseline, week 2, end of treatment, and follow-up session. Baseline and post treatment scores were compared with healthy control participants (28 adults; 17 women). At baseline and at the end of treatment, bipolar participants showed a significantly lower Q-LES-Q score than healthy controls (p < 0.001). Within the bipolar group, there was a significant improvement in total Q-LES-Q scores (p < 0.001) and across multiple domains by week 6 and remained elevated at follow-up. Changes in Q-LES-Q were no longer significant after adjustment for depressive symptoms. A 6-week course of supervised home-based tDCS was associated with significant QoL improvements in bipolar depression, which appeared to be closely linked to reduction in depressive symptoms. Randomized, sham‐controlled trials are warranted to clarify the specific contribution of tDCS to improve QoL in bipolar depression.

Background:Functional impairments associated with mental health conditions are on the rise. Predicting functional outcomes may improve the targeting of preventive interventions. While prognostic models have primarily focused on psychosis, early recognition services require a transdiagnostic approach. Objective:This study aimed to predict global functioning within a 2-year follow-up using baseline clinical and structural magnetic resonance imaging (MRI) data in a population-based sample of young, help-seeking individuals presenting with affective and anxiety symptoms as well as attention-deficit hyperactivity disorder. Methods:We classified 357 help-seeking individuals aged 18-35 years recruited from 9 sites as "impaired" (Global Assessment of Functioning [GAF] ≤60; n=228) or "nonimpaired" (GAF>60; n=129) at year 1 and/or year 2 follow-up. GAF classification group status at follow-up was predicted using linear support vector machine (SVM), decision tree, and large language model (LLM) Llama-3 using clinical assessments and/or structural MRI. Leave-one-site-out (SVM) or external sample (LLM) was used for validation. Results:SVM achieved balanced accuracy of 69.2% using clinical features only. Items related to baseline occupational functioning, interpersonal relationships, cognitive functioning, psychotic and affective symptoms, as well as the presence of anxiety disorder, were most predictive. The decision tree further reduced the feature set to 5 predictive items, achieving balanced accuracy of 76.6%. Although amygdala and hippocampal subregions achieved balanced accuracy of 57.1%, structural MRI did not improve the overall prediction. Llama-3 performed comparably well to SVM (balanced accuracy of 72.6%). Conclusions:Machine learning demonstrated good performance in predicting global functioning. Interestingly, the out-of-the-box LLM performed comparably well without being trained or fine-tuned, highlighting the potential of leveraging free-text data for mental health prognosis.

BACKGROUND:Affective and psychotic disorders often emerge during adolescence and early adulthood. Early detection and timely treatment of individuals with mental disorders (particularly affective and psychotic disorders) can play a decisive role in improving the course of disease and treatment outcomes. To bridge the gap between early detection and evidence-based, integrated, and cross-setting treatment, there is a need for conceptual advancement and closer coordination of outpatient, day clinic, and inpatient care services, particularly during the transition between child and adolescent psychiatry and adult psychiatry. METHODS:The aim of this paper is to present the services offered by the "Young People" track at the Carl Gustav Carus University Hospital in Dresden as a potential best-practice example. It addresses established workflows, team structure, and the nature and frequency of interventions. In addition, descriptive data on all individuals seeking help at the Early Detection and Intervention Center between May 2018 and October 2025 as well as those treated in the day clinic between December 2019 and December 2024 are analyzed. RESULTS:Care is provided through a stepped pathway-starting with an initial general, low-threshold point of contact, moving on to specialized early detection of affective and psychotic disorders, and culminating in targeted referrals to established outpatient, day-care, and inpatient treatment programs for young people. From May 2018 to October 2025, 859 young people made their first visit to the Early Detection and Intervention Center (52.5% female; average age: 24 years). 31.4% did not meet the criteria for a mental disorder, 35.4% fulfilled the criteria of one, 20.6% of two, and 12.6% of at least three diagnoses. N = 63 met the risk criteria for developing bipolar disorder, n = 77 for developing psychosis. More than 100 young people were treated in the outpatient setting. The day clinic treated N = 283 patients (average duration: 7 weeks) with high treatment acceptance and significant improvements in symptom burden, self-management, and quality of life. Initial steps towards establishing a specialized inpatient treatment programme are currently underway with the allocation of six beds for young people, enabling continuous care even in times of increased treatment needs. DISCUSSION AND CONCLUSION:The services offered by the "Young People" track at Dresden University Hospital bridge potential gaps in care during the transition from adolescent to adult psychiatry. Through low-threshold early detection and risk-adapted, continuous, and cross-setting care, young people-with or without previous treatment experience-receive targeted support. Treatment tailored to their developmental needs facilitates inter alia processes of maturation, career orientation, and the gradual achievement of independence. The accompanying research ensures the continuous, evidence-based development of the integrative and patient-centered care concept, which, as a potential best-practice example in Germany, can represent a decisive step towards an effective and person-centered care system for young people with affective and psychotic disorders.

Importance Few pharmacotherapies are approved for treatment-resistant depression, and many patients do not achieve remission following treatment with those therapies. Objective To examine the efficacy and safety of single-day treatment with a synthetic formulation of inhaled mebufotenin (GH001) vs placebo in patients with treatment-resistant depression. Design, Setting, and Participants This was a 7-day, randomized, double-blind, placebo-controlled phase 2b trial with a 6-month open-label extension phase conducted at 16 sites in Europe from May 2023 to March 2025. Adult patients aged 18 to 64 years with treatment-resistant depression, defined as nonresponse to 2 to 5 oral antidepressant treatments, with current episode duration of up to 2 years were included. Of 128 assessed for eligibility, 81 were randomized and completed the placebo-controlled period of the trial. Interventions Patients were randomly assigned 1:1 to receive an individualized dosing regimen of up to 3 escalating doses of GH001 (6, 12, and 18 mg) or a placebo individualized dosing regimen on a single day (day 1). Main Outcomes and Measures The primary efficacy end point was the change from baseline to day 8 in Montgomery-Åsberg Depression Rating Scale total score (range, 0-60; higher scores indicate greater severity of depression), comparing GH001 with placebo. Secondary end points included remission (Montgomery-Åsberg Depression Rating Scale score ≤10) at day 8. Results Among the 81 patients randomized to GH001 (n = 40) or placebo (n = 41), the mean (SD) age was 41.6 (11.4) years and 43.9 (10.9) years and 24 (60.0%) and 22 (53.7%) were female, respectively. Change in Montgomery-Åsberg Depression Rating Scale score from baseline to day 8 was significantly greater for GH001 vs placebo (least squares mean difference [SE], −15.5 [1.7]; P < .001; effect size, −2.0). Day 8 remission rates were 23/40 (57.5%) with GH001 and 0/41 (0%) with placebo. No severe or serious adverse events were reported in the placebo-controlled period. Conclusions and Relevance In this study, an individualized dosing regimen of inhaled GH001 resulted in significant improvements in depression symptoms relative to placebo and was well tolerated, supporting its potential as a novel, rapid-acting treatment for treatment-resistant depression. Trial Registration ClinicalTrials.gov Identifier: NCT05800860

Abstract Objective To evaluate the validity and correlates of inaccurate clinically diagnosed moderate/severe ICD-10 major depressive disorder (MDD) and psychiatric comorbidities in adult inpatients. Methods The multicenter Patient Characteristics, Validity of Clinical Diagnoses and Outcomes Associated with Suicidality in Inpatients with Symptoms of Depression (OASIS-D) study enrolled 18- to 75-year-old inpatients with MDD who underwent assessments for clinical, illness, and treatment characteristics, including the Mini-International-Neuropsychiatric-Interview (MINI). False-positive clinically MDD diagnoses were characterized via MINI. Diagnostic agreement between clinical and MINI-based psychiatric comorbidities was assessed with kappa-statistic. Results Among 401 patients (median age = 33.0 years; female = 57.1%), 42 (10.5%) were false positive for MDD (i.e., different primary diagnosis). Reasons for misdiagnosing MDD exclusive to false-positive cases included bipolar disorder (45.2%), subsyndromal depression (16.7%), and physical illness-related/substance use disorder (SUD)-related depression (4.8%), whereas missed diagnoses that could also occur in true-positive, MINI-based MDD patients included primary SUDs (23.8%) and primary anxiety disorders (9.5%). Among comorbidities, greatest disagreement between clinical and MINI-based research diagnoses existed for antisocial personality disorder (0% vs. 5.2%, κ =0), followed by anxiety disorders (7.5% vs. 45.1%, κ =0.17), obsessive-compulsive disorder (3.0% vs. 8.7%, κ =0.27), SUDs (11.0% vs. 28.4%, κ =0.29), eating disorders (3.5% vs. 7.2%, κ =0.39), and posttraumatic stress disorder (10.7% vs. 23.7%, κ =0.40). Conclusions Bipolar depression and SUDs contributed to two thirds of misdiagnosed MDD. Missed comorbidities included mostly anxiety disorders, SUDs, and PTSD. Accurate diagnostic assessment, including longitudinal history, is essential to reduce misdiagnosis and ensure guideline-based treatment for psychiatric disorders in MDD inpatients.

Hintergrund Affektive und psychotische Störungen beginnen häufig im Jugend- und jungen Erwachsenenalter. Die Früherkennung und die frühzeitige Behandlung von Menschen mit psychischen Störungen (insbesondere mit affektiven und psychotischen Störungen) können hierbei entscheidend zur Begünstigung der Störungsverläufe und Behandlungsergebnisse beitragen. Um den Übergang von der Früherkennung zu einer evidenzbasierten, integrierten und setting-übergreifenden Behandlung zu schaffen, braucht es eine konzeptionelle Weiterentwicklung und engere Vernetzung ambulanter, tagesklinischer und stationärer Versorgungsangebote – insbesondere im Übergang zwischen Kinder- und Jugendpsychiatrie und Erwachsenenpsychiatrie. Methodik Ziel des Beitrags ist es, die Versorgungsangebote des Tracks „Junge Menschen“ am Universitätsklinikum Carl Gustav Carus Dresden als ein mögliches Best-Practice-Beispielen vorzustellen. Thematisiert werden dabei etablierte Arbeitsabläufe, Teamstruktur sowie Art und Frequenz der Interventionen. Ergänzend werden deskriptive Daten aller Hilfesuchenden am Früherkennungs- und Frühinterventionszentrum zwischen Mai 2018 und Oktober 2025 sowie der tagesklinisch Behandelten zwischen Dezember 2019 und Dezember 2024 analysiert. Ergebnisse Die Versorgung erfolgt entlang eines gestuften Pfades – ausgehend von einer ersten allgemeinen, niedrigschwelligen Anlaufstelle über die spezialisierte Früherkennung affektiver und psychotischer Störungen bis hin zur gezielten Weiterleitung in etablierte ambulante, tagesklinische und stationäre Behandlungsangebote für junge Menschen. Von Mai 2018 bis Oktober 2025 wurden 859 junge Menschen im Früherkennungs- und Frühinterventionszentrum erstvorstellig (52,5% weiblich; Durchschnittsalter: 24 Jahre). 31,4% erfüllten nicht die Kriterien für eine psychische Störung, 35,4% erfüllten die Kriterien für eine, 20,6% für zwei und 12,6% für mindestens drei Diagnosen. N = 63 erfüllten Risikokriterien für eine sich entwickelnde bipolare Störung, n = 77 für eine sich entwickelnde Psychose. In der ambulanten Versorgung wurden über 100 junge Menschen behandelt. Die Tagesklinik behandelte N = 283 Patient*innen (Durchschnittsdauer: 7 Wochen) mit hoher Therapieakzeptanz und signifikanten Verbesserungen in Symptomlast, Selbstmanagement und Lebensqualität. Erste Schritte zur Einrichtung eines spezialisierten stationären Behandlungsangebotes für junge Menschen erfolgen aktuell mit der Belegung von sechs Betten für junge Menschen, was eine kontinuierliche Versorgung auch bei erhöhter Behandlungsbedürftigkeit ermöglicht. Diskussion und Schlussfolgerung Die dargestellten Angebote des Tracks „Junge Menschen“ am Universitätsklinikum Dresden überbrücken mögliche Versorgungslücken im Übergang von der Jugend- zur Erwachsenenpsychiatrie. Durch niedrigschwellige Früherkennung, risikoadaptierte, kontinuierliche und setting-übergreifende Versorgung werden junge Menschen mit und ohne Behandlungsvorerfahrungen gezielt unterstützt. Die auf ihre Entwicklungsaufgaben abgestimmte Behandlung fördert u.a. Prozesse der Nachreifung, der beruflichen Orientierung und der schrittweisen Erlangung von Selbstständigkeit. Die enge Verzahnung mit Forschung sichert die kontinuierliche evidenzbasierte Weiterentwicklung des integrativen und patientenzentrierten Versorgungskonzepts, das als ein mögliches Best-Practice-Beispielen in Deutschland einen entscheidenden Schritt hin zu einem effektiven und personenzentrierten Versorgungssystem für junge Menschen mit affektiven und psychotischen Störungen darstellen kann.

Artificial intelligence is increasingly being used in medical practice to complete tasks that were previously completed by the physician, such as visit documentation, treatment plans and discharge summaries. As artificial intelligence becomes a routine part of medical care, physicians increasingly trust and rely on its clinical recommendations. However, there is concern that some physicians, especially those younger and less experienced, will become over-reliant on artificial intelligence. Over-reliance on it may reduce the quality of clinical reasoning and decision-making, negatively impact patient communications and raise the potential for deskilling. As artificial intelligence becomes a routine part of medical treatment, it is imperative that physicians recognise the limitations of artificial intelligence tools. These tools may assist with basic administrative tasks but cannot replace the uniquely human interpersonal and reasoning skills of physicians. The purpose of this feature article is to discuss the risks of physician deskilling based on increasing reliance on artificial intelligence.

Importance:Few pharmacotherapies are approved for treatment-resistant depression, and many patients do not achieve remission following treatment with those therapies. Objective:To examine the efficacy and safety of single-day treatment with a synthetic formulation of inhaled mebufotenin (GH001) vs placebo in patients with treatment-resistant depression. Design, Setting, and Participants:This was a 7-day, randomized, double-blind, placebo-controlled phase 2b trial with a 6-month open-label extension phase conducted at 16 sites in Europe from May 2023 to March 2025. Adult patients aged 18 to 64 years with treatment-resistant depression, defined as nonresponse to 2 to 5 oral antidepressant treatments, with current episode duration of up to 2 years were included. Of 128 assessed for eligibility, 81 were randomized and completed the placebo-controlled period of the trial. Interventions:Patients were randomly assigned 1:1 to receive an individualized dosing regimen of up to 3 escalating doses of GH001 (6, 12, and 18 mg) or a placebo individualized dosing regimen on a single day (day 1). Main Outcomes and Measures:The primary efficacy end point was the change from baseline to day 8 in Montgomery-Åsberg Depression Rating Scale total score (range, 0-60; higher scores indicate greater severity of depression), comparing GH001 with placebo. Secondary end points included remission (Montgomery-Åsberg Depression Rating Scale score ≤10) at day 8. Results:Among the 81 patients randomized to GH001 (n = 40) or placebo (n = 41), the mean (SD) age was 41.6 (11.4) years and 43.9 (10.9) years and 24 (60.0%) and 22 (53.7%) were female, respectively. Change in Montgomery-Åsberg Depression Rating Scale score from baseline to day 8 was significantly greater for GH001 vs placebo (least squares mean difference [SE], -15.5 [1.7]; P < .001; effect size, -2.0). Day 8 remission rates were 23/40 (57.5%) with GH001 and 0/41 (0%) with placebo. No severe or serious adverse events were reported in the placebo-controlled period. Conclusions and Relevance:In this study, an individualized dosing regimen of inhaled GH001 resulted in significant improvements in depression symptoms relative to placebo and was well tolerated, supporting its potential as a novel, rapid-acting treatment for treatment-resistant depression. Trial Registration:ClinicalTrials.gov Identifier: NCT05800860.

Psychomotor slowing (PS) is common across mental disorders and has been linked to the clinical course of major depressive disorder (MDD). Prospective evidence as a predictor of treatment outcomes remains limited. We hypothesized that baseline PS severity would be associated with negative treatment outcomes in patients with MDD including the severity of acute suicidality. We analyzed data from a 6-month prospective, multicenter observational study of inpatients with MDD and acute suicidality. Depression (Montgomery-Asberg Depression Rating Scale), suicidality (Sheehan-Suicide Tracking Scale), Personal and Social Performance Scale (PSP), Work Productivity and Activity Impairment (WPAI, item #6) and quality of life (European Quality of Life Group, 5-Dimension) were assessed at baseline (T1), inpatient discharge (T2), and 3-months (T3) and 6-months (T4). Linear mixed models were fitted to examine associations between baseline PS (item #15 of the Quick Inventory of Depressive Symptomatology-Self-Report) and outcomes across timepoints, using Bonferroni correction. Among 235 participants (age = 36.4 ± 13.6 years; women = 57.9%), across all timepoints, baseline PS was significantly associated with greater depression symptom severity (p < 0.001), lower quality of life (p = 0.042), reduced leisure time productivity (p = 0.002), and higher suicidality (p = 0.001). The association of baseline PS with social functioning was negative (p = 0.043). Only for PSP, the association varied over time, i.e., T4 > T1 and T2 > T1. This first longitudinal study of PS in acutely suicidal MDD patients showed that baseline PS predicted a broad range of clinical and functional outcomes over six months. PS may serve as a prognostic marker and help identify distinct clinical trajectories in MDD.

This expert opinion paper addresses the critical balance between lithium’s therapeutic efficacy in recurrent mood disorders and its potential renal side effects. The objective is to provide evidence-based guidelines to enhance clinical decision-making, prevent emergence of, and mitigate risks associated with lithium-induced renal impairment. An extensive review of epidemiological, observational, and experimental studies on lithium-induced renal impairment, focusing on its pathophysiology, clinical manifestations, and risk factors was conducted. Expert consensus and recent data were integrated to develop a management algorithm for renal monitoring and intervention. Lithium remains the gold standard for mood stabilization in bipolar disorders, with robust evidence supporting its role in recurrence prevention and suicide risk reduction. While mild to moderate renal impairment is recognized as a risk factor, newer studies show a lower incidence of severe outcomes, such as end-stage kidney disease, necessitating dialysis treatment and renal transplantation, especially with appropriate monitoring, as compared to older studies. This paper addresses pathophysiological mechanisms, including arginine vasopressin resistance and chronic interstitial nephritis, alongside risk factors like rapid initial decline in glomerular filtration rate, early age at treatment initiation, cumulative dosage, mean serum levels of lithium and episodes of lithium intoxication. Effective management strategies, including judicious dosing, routine monitoring, and early nephrology referral, can significantly improve outcomes. Lithium remains an invaluable treatment for recurrent mood disorders, and its benefits often outweigh the risks when managed appropriately. This paper provides a practical framework for clinicians to address renal concerns, emphasizing the importance of systematic monitoring and individualized care. The paper underscores the need for continued research and education of clinicians and patients to optimize lithium use while safeguarding patient health.

Prompt engineering has the potential to enhance large language models’ (LLM) ability to solve tasks through improved in-context learning. In clinical research, the use of LLMs has shown expert-level performance for a variety of tasks ranging from pathology slide classification to identifying suicidality. We introduce clickBrick, a modular prompt-engineering framework, and rigorously test its effectiveness. Here, we explore the effects of increasingly structuring prompts with the clickBrick framework for a comprehensive psychopathological assessment of 100 index patients from psychiatric electronic health records. We compare the performance of two locally-run LLMs against an expert-labeled ground truth for a variety of successively built-up prompts for the extraction of 12 transdiagnostic psychopathological criteria. Potential clinical value was explored by training linear support vector machines on outputs from the strongest and weakest prompts to predict discharge ICD-10 main diagnoses for a historical sample of 1692 patients. We could reliably extract information across 12 distinct psychopathological classification tasks from unstructured clinical text with balanced accuracies spanning 71% to 94%. Across tasks, we observed a substantially improved extraction accuracy (between +19% and +36%) using clickBrick for the most reactive model. The comparison unveiled great variations between prompts with a reasoning prompt performing best in 7 out of 12 domains. Clinical value and internal validity were approximated by downstream classification of eventual psychiatric diagnoses for 1,692 patients. Here, clickBrick led to an improvement in overall classification accuracy from 71% to 76%. ClickBrick prompt engineering, i.e., iterative, expert-led design and testing, is critical for unlocking LLMs’ clinical potential. The framework offers a reproducible, explainable pathway for deploying trustworthy generative AI across mental health and other clinical fields.

BACKGROUND:Lithium remains a first-line treatment for bipolar disorder, but practice varies regarding dosing frequency. Although once-daily and twice-daily regimens are similarly effective, they produce serum lithium profiles that differ over the day, and whether greater fluctuations with once-daily dosing are attenuated in the brain is unclear. The aim of this study was to evaluate whether brain lithium profiles differed across the day with once-daily versus twice-daily dosing regimens. METHODS:In this repeated-measures, cross-sectional imaging study, euthymic individuals (aged 18-50 years) with bipolar disorder type I or II receiving stable lithium carbonate treatment were recruited at four psychiatric outpatient departments in Dresden, Germany. In the week before assessment, participants received lithium at 2000 h (full once-daily dose or evening dose of the twice-daily regimen) and 0800 (twice-daily regimen only; withheld until after the 0800 h sampling on assessment day). The core assessment was a 1-day protocol in which participants underwent three consecutive 7Li MRI scans and matched serum sampling over a 10-h period (0800, 1400, and 1800h). Whole-brain lithium measurements across timepoints were compared according to regimen using the brain observable lithium threshold region-of-interest approach. Secondary analyses modelled tissue-specific profiles, derived from the Newcastle lithium image voxel evaluation pipeline, and examined the relationship between serum lithium concentration and brain 7Li MRI signal intensity using repeated-measures ANOVA, linear mixed-effects models, and population pharmacokinetics. This study was informed by people with lived experience of bipolar disorder and long-term lithium treatment through patient and public involvement and engagement activities. FINDINGS:Between April 19, 2022, and Dec 9, 2023, 44 eligible participants were recruited, of whom three withdrew before assessment. Therefore, 41 euthymic individuals (all White European; 21 [51%] female and 20 [49%] male; mean age 39·2 years [SD 9·8]) with bipolar disorder receiving stable lithium carbonate treatment (20 in the once-daily dosing group and 21 in the twice-daily dosing group) were evaluated. At 12 h post dose (0800 h), brain 7Li MRI signal intensity and serum lithium concentrations were equivalent across regimens. In the once-daily group, serum lithium declined across the day, whereas in the twice-daily group it increased after the morning dose and then decreased. Brain 7Li MRI signal intensity closely mirrored these regimen-specific serum profiles. Brain 7Li MRI signal intensity was higher in white matter than grey matter in both regimens and brain-to-serum ratios indicated more rapid equilibration in grey matter compared with cerebrospinal fluid and white matter. INTERPRETATION:Brain lithium concentrations followed serum concentrations closely across the day but aligned between regimens at the standard 12-h post-dose sampling timepoint. These findings directly inform ongoing discussions about the applicability of a common reference range across regimens. The difference in brain and serum profiles of lithium with once-daily and twice-daily dosing might inform regimen selection and therapeutic monitoring strategies in clinical practice. FUNDING:Baszucki Brain Research Foundation and Deutsche Forschungsgemeinschaft (German Research Foundation).

BACKGROUND:Depressive episodes are associated with a higher risk for the onset of bipolar disorder (BD). This study investigates the contribution of specific depressive symptoms using a multi-method approach in persons at-risk for bipolar disorders. METHODS:In the Early-BipoLife study, N = 1083 participants at risk for BD were examined over two years (baseline, 6, 12, 18, and 24 months). Out of 1083 participants, N = 57 (5.3%) transitioned to BD and/or were prescribed with a mood stabilizing medication (lithium, lamotrigine, valproic acid, carbamazepine, quetiapine). At baseline, N = 880 participants met the diagnostic criteria for lifetime major depression (lifetime MD) (SCID). Depressive symptoms were recorded using diagnostic interviews (EPIbipolar, SCID), a clinician rating (ICD-C) and a self-rating (QIDS-SR16). Binary logistic regressions were calculated to assess the prospective association between depressive symptoms and transition/prescription of a mood stabilizing medication. RESULTS:Compared to 'no lifetime MD', 'lifetime MD' were at higher risk for transition and/or prescription of a mood stabilizing medication (OR = 2.87, 95%CI: 1.03-8.04). A higher QIDS-SR16 symptom load was associated with transition and/or prescription of a mood stabilizing medication (OR = 1.07, 95%CI: 1.01-1.13). Consistently and across methods, it was shown that suicidality at baseline predicted transition to BD and/or prescription of a mood stabilizing medication. Further baseline symptoms assessed by the clinicians (IDS-C), such as decreased or increased appetite and quality of mood were also predictive for this outcome. CONCLUSION:Depressive symptom load and suicidality were robust predictors for BD. Clinician judgment can provide important information for early detection of BD.