Inhibitory control is a core cognitive function whose competence varies across the population, with impairments often observed in psychiatric conditions such as attention deficit hyperactivity disorder (ADHD). The Stop Signal Task (SST) is a widely used paradigm for assessing this ability. However, conventional formalizations of SST performance, such as the independent race model, rely on assumptions that are frequently violated in modern experimental designs. Furthermore, they typically fit only mean reaction times, overlooking crucial trial-by-trial dynamics. To address these limitations, we formalize the SST as a partially observable Markov decision process (POMDP). This framework characterizes inhibitory control with two components: noisy perceptual inference regarding stimuli and optimal control balanced against potential costs. To fit this model to the Adolescent Brain Cognitive Development (ABCD) study baseline cohort (N = 3,567), we introduce Transformer-encoded Simulation-Based Inference (TeSBI). This end-to-end architecture learns compact, sequence-aware embeddings from raw behavioral data. It enables efficient, amortized inference of individual-level posteriors. Extensive validation confirms it extracts reliable and identifiable parameters. We identify distinct latent computational attributes associated with scores on ADHD questionnaires. Controlling for sex, IQ, and medication status, children with higher ADHD scores exhibit subtle but robust shifts in computational attributes. They show a reduction in go cue directional precision, alongside a blunted sensitivity to stop error, go error and time costs. The learned embedding space reveals a continuous manifold in which children with higher ADHD scores are heterogeneously distributed, rather than forming distinct disorder clusters. This indicates that similar clinical characteristics can emerge from diverse combinations of computational mechanisms, supporting a dimensional perspective on neurodiversity. Our end-to-end framework can be extended to a broader range of cognitive tasks. It offers a scalable, theory-driven solution for analyzing large-scale behavioral data.

Serial cognitive assessment in multiple sclerosis can help identify patients at higher risk of disability progression, but high dimensionality of reaction-time data makes analysis challenging. We developed a method to convert longitudinal reaction-time data into images and applied deep survival modelling to predict disability progression in relapsing-remitting MS (RRMS). In this cohort study, clinical data were obtained from the MSBase registry and cognitive data from the MSReactor computerised cognitive battery between February 2016 and September 2022, with a median follow-up of 3.2 years. The serial reaction-time data from three tasks (psychomotor function [R], attention [G], and working memory [B]) were converted into multicolour images using RGB encoding. We extracted the key features from the images using a convolutional neural network and combined them with clinical variables in a Transformer-based survival model (MS-TranSurv) to predict time to confirmed disability progression. Discrimination, calibration, and accuracy were assessed using the C-index, integrated Brier score (iBS), and time-dependent area under the receiver operating characteristic (tAUROC). Performance was compared with other machine learning models, including Dynamic DeepHit, Recurrent Deep Survival Machines, a Cox proportional hazards model in traditional statistics using clinical variables only, and a version of MS-TransSurv using mean test values. A total of 746 RRMS patients were included. MS-TranSurv showed slightly higher discrimination compared with benchmark models, with a C-index of 0.61 (95% CI 0.54-0.68) and tAUROC of 0.74 (95% CI 0.63-0.85), as well as comparable calibration, with an iBS of 0.24 (95% CI 0.16-0.32). Using individual test-level data provided slightly better performance than using summary measures. Longitudinal cognitive reaction-time data can be used for survival-based prediction of disability progression in RRMS. This framework supports remote cognitive monitoring and may improve risk stratification and clinical trial enrichment, with potential applicability to other high-dimensional digital biomarkers.

Clinical supervision is a natural opportunity for promoting therapist use of clinical practices supported by research in community mental health services. However, the research on thespecific strategies in supervision that are likely to have a positive impact is limited. One model from the organizational management literature suggests supervision behaviors that promote a sense of accountability-i.e., a perceived expectation to fulfill duties or obligations-may represent key strategies by which supervision promotes use of clinical practices supported by research. The present study aimed to identify the presence of "accountability strategies" during clinical supervision and investigate the association between strategies and therapist use of a clinical procedure supported by research during subsequent psychotherapy sessions. Transcripts of supervision sessions (n = 196) and subsequent treatment sessions (n = 196) were observationally coded for supervision accountability strategies and therapist follow-through use of an identified clinical procedure. Binary, multi-level logistic regression models were used to assess the association between accountability strategies during supervision and therapist use of a planned (i.e., "expected") practice during the subsequent therapy session. Results indicated explicit selection of the practice the therapist will use in their next treatment session, preparation through activities such as consulting manuals and practicing role plays, and review of how use of the practice went in the past increased the odds of therapist use of the expected practice during the next treatment session. These findings are notable as they demonstrate that supervision can directly influence therapist behavior through three specific strategies.

Artificial intelligence (AI) is reshaping healthcare by enhancing clinical decision making, communication, and patient monitoring systems. Still, it is essential that these technologies lessen, rather than deepen, feelings of loneliness. This study centers the perspectives of individuals with serious mental illnesses, such as schizophrenia spectrum disorder, bipolar disorder, and major depressive disorder, to explore how they interpret and engage with conversational and companion-based AI in healthcare settings. By examining their perspectives, expectations, and concerns, the study highlights design factors that shape the perceived trustworthiness of AI tools. A secondary analysis of three focus groups was conducted across three supportive housing facilities in New York City, where qualitative data was subsequently collected, transcribed, and analyzed using RADar analysis. This study determined that participants expressed strong concerns about data privacy and misuse, but also recognized AI's potential to support reminders, mental health, and daily functioning. Trust in AI increased when tools were personalized, human-centered, and co-developed with people who understand lived experiences of serious mental illnesses. However, concerns remained about loss of critical thinking and overreliance. By emphasizing the lived experiences of adults with serious mental illness, this study identifies the factors that influence how they trust, use, and engage with AI in mental health care. These insights can guide the development of AI tools for individuals with serious mental illnesses.

Adolescents with mental illness (MI) often experience public stigma and self-stigma and struggle with the dilemma whether to disclose their MI, which may reduce their quality of life (QoL). Disclosure decisions are central in coping with MI and stigma, as disclosure may improve well-being and facilitate help-seeking, but can be risky. The peer-led Honest, Open, Proud (HOP) group program was developed to support disclosure decisions and to reduce self-stigma. In this two-arm 2:1-randomized controlled trial, 113 adolescents with MI aged 13-21 years and at least moderate disclosure-related distress were recruited from three Departments of Child and Adolescent Psychiatry and an outpatient practice for child and adolescent psychiatry in Southern Germany. Participants were assigned to HOP plus treatment as usual (TAU) or to TAU alone. Primary outcomes (stigma stress and QoL, both at T1) and secondary outcomes were assessed at baseline (T0), post-intervention (T1, 5 weeks after T0) and at follow-up 6 months after baseline (T2) and analyzed with mixed-model repeated measures. Compared to control, HOP significantly reduced stigma stress at T1 and T2. A significant increase in QoL was observed at T2, not T1. HOP led to improvements in attitudes to disclosure at school, university or work at T1, self-stigma and disclosure-related distress at T2, and empowerment at T1 and T2. Disclosure decisions at baseline did not moderate HOP effects. Since HOP reduces stigma stress and leads to broader improvements over time, it should be offered as a peer-led program to support adolescents in their coping with MI stigma.

To evaluate sociodemographic and clinical characteristics associated with a subsequent suicide attempt among youth with a positive suicide risk screen in the emergency department (ED).

This secondary data analysis used the 2015 to 2019 Pediatric Emergency Care Applied Research Network ED Screen for Teens at Risk for Suicide data set, a prospective cohort of adolescents aged 12 to 17 years in the ED. Adolescents with positive Ask Suicide-Screening Questions, a validated suicide risk screen, were included. Multivariable regression models assessed sociodemographic and clinical characteristics associated with suicide attempt, and mental health-related ED revisit/hospitalization, within 3 months after index ED visit.

Among 2,085 adolescents with a positive Ask Suicide-Screening Questions (40.7% age 14 to 15 years; 65.5% female, 49.0% Non-Hispanic White), 11.6% attempted suicide and 19.9% had an ED revisit/hospitalization within 3 months. Risk factors of suicide attempt included suicide attempt in the past month (adjusted odds ratio [aOR] 2.52, 95% confidence interval [CI] 1.84 to 3.45), prior mental health hospitalization (aOR 1.39, 95% CI 1.02 to 1.90), past nonsuicidal self-injury (aOR 2.06, 95% CI 1.41 to 3.01), and hopelessness (aOR 1.30, 95% CI 1.04 to 1.63). Risk factors of ED revisit/hospitalization within 3 months included prior mental health hospitalization (aOR 2.23, 95% CI 1.74 to 2.84), suicide attempt in the past month (aOR 1.63, 95% CI 1.24 to 2.14), and hopelessness (aOR 1.44, 95% CI 1.21 to 1.72).

Among youth with positive suicide risk screens in the ED, multiple sociodemographic and clinical factors were associated with subsequent suicide attempt and mental health-related ED revisits/hospitalizations. Following a positive suicide risk screen, these factors may be important for consideration when assessing risk and determining next steps in care.

ObjectivePeople in prison have high rates of self-harm and suicide but little is known about the patterns of occurrence over time in custodial settings. Rates of psychiatric diagnoses in custodial settings are also high, and strong associations have been found between psychiatric diagnoses and risk of self-harm and suicidality.MethodsIn this short report, we present a study examining 3 years of data (2020-2023) from a mental health unit within a prison in New South Wales, Australia. We describe the characteristics of the cohort and report rates of self-harm and suicidal behaviours during the admittance.ResultsOverall, 20% of the cohort reported thoughts of self-harm at least once during their stay, and 9% of the cohort had at least one recorded self-harm incident. Multivariable binary logistic regression analyses revealed greater odds of self-harm for individuals with a recorded history of self-harm or suicidal behaviour and for those who reported thoughts of self-harm to staff during their stay.ConclusionsFindings from our study highlight the high rates of self-harm behaviour and ideation among people with mental illness in a custodial setting, supporting the need for targeted prevention and aftercare strategies.

Mental health-focused chatbots have emerged as scalable interventions that deliver therapeutic content through conversational interfaces. This meta-analysis of 35 studies (N = 1,780) indicates that chatbots can reduce depression (g = -.53, 95% CI [-.67,-.40]) and anxiety (g = -.36, 95% CI [-.46,-.27]) symptoms over time. However, wide prediction intervals reveal substantial heterogeneity, suggesting that while some interventions are associated with symptom reduction, others may increase them. Moderator analyses revealed that therapeutic alliance significantly predicted reductions in both depression and anxiety symptoms. Task and goal alliance significantly predicted a reduction in depression symptoms, while bond alliance did not. Similarly, treatment acceptability significantly moderated depression symptom change, noting the importance of perceived intervention appropriateness. Mental health chatbots designed using cognitive behavioral therapy and interventions delivered over longer durations were associated with larger symptom reductions, particularly for anxiety. In addition, populations with both clinical and self-reported diagnoses were associated with larger reductions in depression symptoms compared to interventions recruiting general populations. Findings also reveal considerable gaps in mental health chatbot development and evaluation, as few studies have examined key communication-related moderators, such as therapeutic alliance and acceptability, despite their demonstrated importance in the mental health communication and therapeutic literature.

Although the Danish health registers provide rich data for epidemiological mental health research, the comparability of findings is hindered by different definitions of mental disorders across studies. We aimed to describe different approaches taken by researchers to define mental disorders using Danish register data and quantify the impact of selecting different definitions on estimates of prevalence and incidence.

We undertook a scoping review to identify the approaches to define mental disorders using Danish register data and produced a narrative synthesis. We compiled tables of potential definitions for any mental disorder and specific types of mental disorders using nationwide Danish registers. Using these definitions, we quantified (i) the number of individuals fulfilling each definition during 2018 and 2018-2022 and (ii) the estimated cumulative incidence.

Our search identified 2185 papers, 1155 of which used Danish register data to define mental disorders to describe the study population, an exposure, or an outcome. Our estimates of the number of individuals fulfilling each definition and the estimated cumulative incidences indicated considerable variation with the data sources and variables used to define mental disorders within the Danish national registers. For example, the cumulative incidence of any mental disorder at 80 years among females varied between 29.2% when only psychiatric unit patient register data were used and 62.1% when somatic patient, prescription, cause of death, and health services register data were added.

The different data sources and variables used to define mental disorders can result in substantial variation in the numbers of identified cases. Studies should clearly document all variables and values used to define their disorders of interest, as well as describe the disorder they aim to capture, discussing the strengths and limitations of their approach and the likely impacts on the sensitivity and specificity of including or omitting data from other available registers.

The objective of this study was to determine whether genetic ancestry is associated with differences in the clinical course of multiple sclerosis (MS).

Participants with MS living in the United Kingdom >18 years old were recruited from 2021 to 2025 and genotyped from saliva using a commercial array. Genetic ancestry was inferred using a random forest classifier trained on a diverse reference dataset. Participant-reported MS characteristics including the Age-Related Multiple Sclerosis Severity Score (ARMSS) and age at MS diagnosis were ascertained at recruitment.

We analyzed data from 816 people with MS (pwMS), comprising 260 people of South Asian ancestry (31.9%), 163 of African ancestry (20.0%), 316 of European ancestry (38.7%), and 77 participants from other backgrounds. The included cohort was predominantly female (72.3% women), diagnosed with MS at a median age of 32 years (interquartile range [IQR] = 14.0) and recruited at a median age of 44.6 years (IQR = 18.2). Neither South Asian nor African ancestry was consistently associated with higher MS severity compared with European-ancestry participants. No genetic variants were associated with MS severity at genomewide statistical significance.

In this cross-sectional, diverse, UK cohort, neither South Asian nor African ancestry is associated with greater MS-related disability. These findings argue against a genetic basis for the previously observed differences between MS outcomes when contrasting ethnic groups. ANN NEUROL 2026.