BackgroundAnxiety disorders and depressive disorders are the most prevalent mental disorders worldwide. Their diagnosis has long relied on clinical symptom assessment, and objective blood−based biomarkers remain lacking. Sex is a critical risk factor for these disorders; however, sex−specific divergence in blood biochemical profiles has yet to be systematically characterized.MethodsThis retrospective study enrolled 778 patients diagnosed with anxiety−depressive state at China−Japan Friendship Hospital. Demographic data, complete blood count parameters, and blood biochemical parameters were collected. Following missing value processing and multiple imputation, Mann–Whitney U tests were applied to identify sex−differentially expressed biomarkers. A random forest classifier was constructed to evaluate the discriminative capacity of combined multi−marker panels, with model performance comprehensively assessed through receiver operating characteristic curve analysis, SHAP−based explainability analysis, and multi−classifier probability projection. Age−stratified analyses were performed with a threshold of 50 years to explore the potential modifying effect of age on sex differences.ResultsSeveral biomarkers exhibiting significant differences between males and females were identified (FDR < 0.05), among which creatinine, hemoglobin, hematocrit, red blood cell count, and uric acid demonstrated the largest effect sizes. The random forest model achieved an area under the receiver operating characteristic curve of 0.902 on the independent test set. Multi−classifier probability projection following hyperparameter tuning yielded a Silhouette coefficient of 0.464 in the two−dimensional space, with permutational multivariate analysis of variance confirming highly significant centroid differences between groups (p < 0.001). Age−stratified analysis using hemoglobin as an example revealed that levels in males were significantly higher than those in females across both age strata, with the magnitude of the sex difference attenuated in the ≥50−year group compared with the <50−year group.ConclusionsRobust sex−related signals are embedded in routine blood biochemical markers. Although complete separation is difficult to achieve under unsupervised dimensionality reduction, these signals can be efficiently integrated through ensemble learning algorithms. This study provides a molecular phenotypic basis related to sex in patients with anxiety−depressive state and underscores the importance of fully considering sex as a variable in clinical laboratory testing.
A comparison between seven scales of neuropsychological assessments for cognitive impairment screening in Chinese older population: a cross-sectional study in Chongqing, China
BackgroundThe Clinical Dementia Rating Scale (CDR), the Ascertain Dementia 8 (AD8), the Mini-Cog, the Verbal Fluency Test (VFT), the Community Screener for Dementia (CSI-D), the Rey Auditory Verbal Learning Test (RAVLT), and the Activities of Daily Living Scale (ADL) represent seven commonly employed methods for community cognitive impairment screening in China that have garnered limited attention. This study aimed to assess the discriminatory power of these seven tests when administered concurrently in a single screening to detect cognitive impairment in the absence of a gold standard.MethodsWe conducted a cross-sectional survey among 1,506 elderly people aged 60 and above in the community. The cognitive and social functions of the elderly population were evaluated by using the Seven scales. The characteristics related to demographics and health were collected through questionnaire surveys, and the correlations and consistencies of the Seven scales with cognitive impairment were analyzed respectively. Multifactor logistic regression model was used to analyze the risk factors of cognitive impairment in each scale.ResultsThe positive screening proportions for the seven scales—CDR, ADL, AD8, CSI-D, Mini-Cog, VFT, and RAVLT—were 61.1%, 33.7%, 38.0%, 37.1%, 53.7%, 39.3%, and 52.9%, respectively. Agreement on cognitive impairment risk between each pair of scales was fair-to-moderate (Kappa: 0.32–0.645, all p < 0.001). Screening results differed significantly across the seven scales by age, educational, marital status, epilepsy, cerebral infarction, brain atrophy, and depression (all P < 0.001). Notably, Screen-positive proportions for cognitive impairment rose significantly with increasing age (p < 0.001); the ≥ 80-year-old group showed the highest proportions across scales. Conversely, higher educational attainment was associated with lower risk of screening positive for cognitive impairment (p < 0.001).ConclusionsOur study identified a relatively high screening positivity proportion for cognitive impairment in Chongqing, Age, Sex, Education, Marital status, Brain atrophy, Anxiety, and Depression are risk factors for screening positivity. The identified screening positivity for cognitive impairment varied across different neuropsychological assessment methods, indicating that assessment choice should be tailored to population characteristics rather than applying a one-size-fits-all approach. Selecting the appropriate tool can improve sensitivity and reduce missed diagnoses.
Changes in depression, anxiety, and post-traumatic stress symptoms among children and adolescents exposed to adverse childhood experiences following participation in the PROACT intervention in Nairobi, Kenya
IntroductionGlobally, children and adolescents exposed to Adverse Childhood Experiences (ACEs) face an increased risk of developing mental health disorders. The prevalence of these mental disorders is further amplified by the lack of access to specialised mental health treatment especially in low resource settings. There is an urgent need for scalable mental health interventions that can effectively address the needs of these vulnerable populations. Non-specialist-delivered interventions, such as PROACT (Psychoeducation, Relaxation, Problem-solving, Activation, and Cognitive Coping Therapy), represent a promising scalable approach that could help bridge the existing mental health treatment gap in low-resource settings.ObjectivesThis study aimed to assess changes in depression, anxiety, and post-traumatic stress symptoms among children and adolescents exposed to adverse childhood experiences following participation in the PROACT intervention delivered by trained social workers in Nairobi, Kenya.MethodologyMixed-methods pre-post study design was employed. Twenty purposively selected sites across Nairobi County each contributed one social worker (N = 20), who received training to deliver the intervention. A total of 40 children participated and received 4–6 PROACT sessions. Quantitative data were analysed using STATA version 17. Paired t-tests were used to compare baseline and endline scores, while mixed-effects linear regression models with participant ID as a random effect were fitted to estimate changes in outcomes over time and account for repeated measures. Statistically significant improvements were observed across all mental health outcomes. Mean anxiety scores decreased from 6.2 at baseline to 2.7 at endline (mean difference: −3.5; 95% CI: −4.7 to −2.2; p < 0.001), while mean depression scores decreased from 6.4 to 2.9 (mean difference: −3.6; 95% CI: −4.9 to −2.3; p < 0.001). Mean PTSD scores decreased from 16.7 (95% CI: 12.8–20.5) at baseline to 7.3 (95% CI: 4.4–10.1) at endline (mean difference: −9.4; 95% CI: −14.6 to −9.3; p < 0.001). Mixed-effects linear regression analyses corroborated these findings, demonstrating significant reductions in PTSD (β = −9.44), anxiety (β = −3.48), and depression (β = −3.59) symptoms (all p < 0.001).ConclusionThe PROACT intervention was feasible and acceptable when delivered by social workers in Nairobi primary healthcare facilities and was associated with improvements in mental health outcomes among children and adolescents. These findings highlight the potential of task-sharing approaches to expand access to mental healthcare in low- and middle-income countries (LMICs) and warrant further evaluation in controlled studies.
Child Mind Institute Launches Research Initiative To Inform Safer AI Systems for Youth
New project will develop tools to study youth mental health during AI chatbot use
New York, NY — Today, the Child Mind Institute announced a new research initiative, launched with support from the OpenAI Foundation, to build the research infrastructure needed to measure and better understand youth mental health while using AI chatbots and over time. The effort aims to identify valid markers that can inform the design and testing of safer AI systems.
Millions of children in the United States are struggling with mental health or learning challenges, yet there continues to be a dire shortage of qualified mental health professionals. That, along with stigma and misinformation about mental health, is pushing young people to explore using readily available, often free AI tools for support, including general-purpose chatbots, digital companions, or “therapy bots.”
Novel and emerging digital technologies are shaping mental health faster than the pace of science and the development of evidence-based care can keep up. This new initiative will help advance research studying the impact of these tools and better support youth mental health.
“As the leading nonprofit dedicated to improving youth mental health through science, education, and care, the Child Mind Institute is well positioned to help tackle the key issues affecting the well-being of young people, including technology and the proliferation of AI. It is our belief that with appropriate safeguards and evidence, digital tools may complement care from trained clinicians,” said Harold S. Koplewicz, MD, president and medical director at the Child Mind Institute. “We are excited about filling a gap that currently exists in the research of AI tools and to work toward creating a safer online experience for youth around the globe.”
Using clinical assessments, digital journals, de-identified AI chat histories, real-time behavioral measures, and existing youth mental health datasets, the Child Mind Institute will build an infrastructure to better understand interactions between youth users and AI conversational platforms. With the goal of strengthening our understanding of the mental state in AI chatbot interactions, this initial one-year project will enable the research team to identify and begin scaling the signals needed to meaningfully assess and monitor youth mental health alongside their use of AI tools.
The Child Mind Institute has long been committed to advancing technologies that make mental health research and care more precise, measurable, and connected to real-world settings. Grounded in science and focused on impact, the organization brings together scientific rigor, clinical expertise, and product development experience to build digital tools that accelerate research, improve care, and expand access without compromising safety, quality, or accountability.
“Many of our youth are turning to AI chatbots for important areas of their lives — and mental health is no exception. It is our responsibility to more holistically comprehend the impact of these tools on mental health, both in the long and short term,” said Michael P. Milham, MD, PhD, chief science officer at the Child Mind Institute. “Digital platforms provide an opportunity to rethink the way we conduct mental health research, but they also introduce new challenges and risks. It is critical to better understand the relationship between AI use and youth mental health, and to examine whether, and under what conditions, AI tools can strengthen evidence-based care, support clinician training, and expand access to high-quality mental health services.”
This research initiative is being independently developed and solely executed by the Child Mind Institute. To ensure the safety and security of all study participants, it will be conducted with strong privacy protections, informed consent, ethical oversight, and careful data governance. The project will be co-designed by its principal investigators, Gregory Kiar, PhD, Arno Klein, PhD, and Dr. Milham, who bring expertise in computational methods, digital measurement, clinical science, and youth mental health. Consistent with the Child Mind Institute’s open science philosophy, all data will be shared to help fuel discovery across the field.
About the Child Mind Institute
The Child Mind Institute is an independent nonprofit organization dedicated to transforming the lives of children and families struggling with mental health and learning disorders. Through cutting-edge research, evidence-based clinical care, and public education, the Child Mind Institute builds open science platforms and digital tools to accelerate discovery and improve youth mental health worldwide.
For press questions, contact cmiscience@ssmandl.com or mediaoffice@childmind.org.
The post Child Mind Institute Launches Research Initiative To Inform Safer AI Systems for Youth appeared first on Child Mind Institute.
The association between worry, rumination, and metacognitive beliefs in female adolescents with Anorexia Nervosa
BackgroundA growing body of literature has emphasized the role of worry and rumination in eating disorders (EDs). According to the metacognitive model, worry and rumination are maintained by dysfunctional metacognitive beliefs. Although metacognitive beliefs have been investigated in adults with EDs, their role in adolescents with EDs remains underexplored. The present exploratory study aimed to examine the specific association among worry, rumination, and metacognitive beliefs and eating disorder symptomatology in adolescents with AN.MethodsThirty Italian female adolescents (mean age ± SD: 15.387 ± 1.35 years, mean Body Mass Index (BMI): 16.34 ± 2.5 kg/m2) completed a self-report survey composed of the Penn State Worry Questionnaire (PSWQ), the Ruminative Response Scale (RRS), the Metacognitions Questionnaire – 30 (MCQ-30), and the Eating Disorder Inventory – 3 (EDI-3).ResultsThus, Model 1, including RRS and PSWQ, was significant (F = 17.5, p<0.001) and explained 57% of the variance (R² = 0.574). Within this model, RRS was the only significant predictor (p = 0.019) of the Eating Disorder Risk Composite (EDRC) of EDI-3, indicating that higher rumination scores were associated with greater eating disorder symptomatology. Once MCQ was entered as a predictor of EDRC, there was no significant incremental variance compared with Model 2 (ΔR² = 0.0034, p = 0.658), even though the model remained significant overall (F = 11.4, p < 0.001). The final model explained 58% of variance (R² = 0.578; Cohen’s f2 = 1.37).ConclusionRumination was significantly associated with eating disorder symptomatology. Although worry showed strong positive associations with eating disorder symptomatology at the correlational level, it did not maintain an independent role once rumination was included in the regression model. The final model showed a very large effect size (Cohen’s f² = 1.37), suggesting strong associations between predictors and eating disorder symptoms in the present sample. However, this estimate should be interpreted cautiously, given the limited sample size and the exploratory nature of the analysis. Contrary to our expectations, metacognitive beliefs did not significantly explain the variance of eating disorder symptomatology. This finding may reflect the more proximal role of RNT in explaining the variance in eating disorder symptomatology, or alternatively, the possibility that only specific metacognitive domains—rather than overall metacognitive functioning—are relevant in this population.
Clinical Study of Non-invasive 40-Hz Audiovisual Stimulation Guided by Gamma EEG Monitoring for Promoting Perioperative Brain Function Recovery in Children
Conditions: ADHD – Attention Deficit Disorder With Hyperactivity
Interventions: Device: 40-Hz non-invasive synchronized audiovisual stimulation
Sponsors: General Hospital of Ningxia Medical University
Not yet recruiting
Interventions: Device: 40-Hz non-invasive synchronized audiovisual stimulation
Sponsors: General Hospital of Ningxia Medical University
Not yet recruiting
Diagnostic Discrimination of BOKE STARS, a Bimodal Continuous Performance Test, for Attention-Deficit/Hyperactivity Disorder Assessment in Chinese Children: Single-Center Case-Control Study
Background: Attention-deficit/hyperactivity disorder (ADHD) affects 6.3% of Chinese children, but only 10% are diagnosed, as diagnosis is hindered by low awareness and lack of culturally adapted, objective tools. Existing continuous performance tests are unimodal and lack validity. Objective: This study aimed to conduct an initial evaluation of the diagnostic discrimination of BOKE STARS (Sustained Task and Attention Response Screening), a culturally adapted bimodal continuous performance test, for distinguishing children with ADHD from typically developing children in a Chinese clinical setting. Methods: In this prospective, single-center diagnostic accuracy study with a case-control design, 100 children aged 6 to 12 years (n=50 with ADHD and n=50 controls) were recruited at Xinhua Hospital between January and May 2024. Parents completed the Swanson, Nolan, and Pelham Rating Scale–fourth version (SNAP-IV), and children completed the BOKE STARS assessment on a tablet device under standardized conditions. Group comparisons were conducted using independent-sample 2-tailed tests or Mann-Whitney tests, as appropriate. Receiver operating characteristic (ROC) curve analysis was performed in the full case-control sample to assess diagnostic discrimination. Sensitivity and specificity were reported descriptively across ROC-derived cutoffs; these cutoffs were not interpreted as clinically validated diagnostic thresholds. Secondary exploratory analyses included comparisons among ADHD subtypes and correlations between BOKE STARS indices and parent-reported SNAP-IV symptom severity scores. Results: Compared with controls, children with ADHD performed significantly worse on all major BOKE STARS indices, including errors of omission, errors of commission, reaction time, reaction time variability (RTV), discrimination prime, and total score. In the full case-control sample, the total score showed the strongest diagnostic discrimination (area under the ROC curve [AUC] 0.962, 95% CI 0.931-0.992), followed by RTV (AUC 0.919, 95% CI 0.867-0.971), errors of omission (AUC 0.884, 95% CI 0.818-0.950), and discrimination prime (AUC 0.819, 95% CI 0.737-0.900). Errors of commission (AUC 0.689, 95% CI 0.585-0.792) and reaction time (AUC 0.634, 95% CI 0.524-0.743) showed comparatively weaker discrimination. No significant differences were observed among ADHD subtypes. Several BOKE STARS indices were modestly correlated with SNAP-IV inattention and hyperactivity or impulsivity scores. Conclusions: BOKE STARS showed promising preliminary diagnostic discrimination for identifying Chinese children with ADHD in this case-control sample, with the total score and RTV showing the strongest discriminatory performance. However, because the case-control design artificially fixed the ratio of ADHD cases to controls, diagnostic performance estimates and exploratory cutoffs should be interpreted cautiously and should not be considered representative of real-world clinical diagnostic performance. BOKE STARS may serve as an adjunctive assessment tool to complement clinical interviews and caregiver-reported rating scales, but further external validation in larger and clinically heterogeneous populations is required before broader clinical implementation.
<img src="https://jmir-production.s3.us-east-2.amazonaws.com/thumbs/a352c65e649b80c4b66c3ed01598115e" />
Deprexis for Veteran Depression: Open-Label Pilot Trial Examining Feasibility, Acceptability, and Preliminary Efficacy
Background: Depression carries the highest burden of mental health–related disability in the United States. Approximately 13% of military veterans report elevated rates of depression. Despite the availability of evidence-based treatments for depression, nearly 50% of veterans in need of mental health care remain untreated. Internet-based interventions show promise in reducing this gap; however, there are currently no standard self-guided internet-based interventions for depressive symptoms in veterans. Deprexis is one such intervention that leverages cognitive behavioral therapy to target depressive symptoms. Objective: This pilot study evaluated the feasibility, acceptability, and preliminary effectiveness of Deprexis, a fully self-guided internet-based intervention for depression, in US military veterans with mild to severe depressive symptoms. Methods: This open-label pilot trial recruited 19 veterans with mild to severe depression (mean age 55.5, SD 8.2 y; baseline Quick Inventory of Depressive Symptomatology—Self-Report [QIDS-SR]: mean 16.2, SD 4.1) for an 8-week course of Deprexis, with self-report assessments at baseline, posttreatment (8 wk), and follow-up (16 wk). Primary outcomes included depressive symptoms (QIDS-SR), functional disability (World Health Organization Disability Assessment Schedule 2.0), and symptom-related disability (Sheehan Disability Scale). Feasibility was assessed through recruitment and retention rates, and acceptability was measured using validated questionnaires (Credibility and Expectancy Questionnaire and Client Satisfaction Questionnaire). Multilevel models examined change over time, with effect sizes calculated using pooled SDs from unconditional models. Results: Recruitment and retention targets were met, with 15 out of 19 (79%) participants meeting the adherence criteria (ie, ≥60 min of active program use). Of these, 14 participants completed posttreatment questionnaires and were included in the completer analyses. The program received a positive acceptability rating: of the 18 participants who completed follow-up assessments, 78% (n=14) rated services as good or excellent and 72% (n=13) were satisfied with the amount of help received. No safety concerns were reported. Among completers (n=14), QIDS-SR scores decreased from baseline to posttreatment (estimate −2.22, SE 1.44; =.14; =−0.54, 95% CI −1.07 to 0.13) and follow-up (estimate −2.85, SE 1.19; =.02; =−0.70, 95% CI −1.21 to −0.08) with moderate-to-large effect sizes. Effect sizes were similar in the total sample. Functioning (World Health Organization Disability Assessment Schedule 2.0) improved among completers at follow-up (estimate −8.09, SE 3.80; =.045; =−0.41, 95% CI −0.96 to −0.05). Disability (Sheehan Disability Scale) did not significantly improve from baseline to posttreatment or follow-up. Conclusions: This pilot trial demonstrates that Deprexis is feasible and acceptable for veterans with mild to severe depression, with preliminary evidence of effectiveness for depressive symptoms. The delayed emergence of functional improvements and sustained gains at follow-up support the potential of this scalable intervention. The results provide a strong foundation for the ongoing randomized controlled trial. Trial Registration: ClinicalTrials.gov NCT06217198; https://clinicaltrials.gov/study/NCT06217198 International Registered Report Identifier (IRRID): RR2-10.2196/59119
<img src="https://jmir-production.s3.us-east-2.amazonaws.com/thumbs/70826d4d70665042dcb7552593cb44c1" />
Developing a Text Messaging Intervention to Increase Uptake of the Screening and Treatment for Anxiety and Depression Program Among Community College Students: Formative Study Using a Human-Centered Design Approach
Background: Community college (CC) students face significant mental health concerns but are unlikely to receive treatment. Barriers to mental health service uptake among CC students have been delineated, but few studies have identified strategies to improve uptake. Text messaging has been used to address engagement barriers to mental health services among adolescents and adults, but little research has explored this strategy for CC students. Objective: The goal of this study was to partner with CC students to co-design and conduct pilot usability testing of a text messaging intervention to address barriers and increase uptake of a mental health screening and treatment program, called Screening and Treatment for Anxiety and Depression (STAND), offered to CC students. Methods: We conducted 2 parallel sets of 4 co-design focus groups with CC students who had varying levels of engagement with STAND. We used rapid qualitative analysis to extract key themes, create text message prototypes and refine them, and present updated prototypes to gather feedback across workshops. We also assessed six usability factors on a 5-point Likert scale: satisfaction, helpfulness, attractiveness, readability, comprehension, and likelihood of getting started with STAND after receiving texts. Results: Key themes emerged about perceptions of texting, barriers to STAND, a basic framework for the text message intervention, feedback about the format of messages, and feedback about the content of messages. Students expressed positive regard for text messaging and general agreement on key barriers to STAND. Students codeveloped a framework for the intervention, including (1) delivering introductory texts to engage students in the text messages, (2) providing a personalized approach for students to select barriers most salient for them, and (3) delivering tailored content designed by students to address each barrier. Across workshops, several themes emerged with regard to how messages should be formatted and delivered, including the following: use short messages; use not too many messages; use relevant language; use images, memes, and short videos; and make messages “human-like.” Themes related to the content of messages included the following: reminders that you are not alone, knowledge that STAND has worked for other students, expressing understanding of student context and stressors, and providing an option to speak to a team member. Mean ratings on usability factors ranged from 3.88 (SD 0.64) to 4.25 (SD 0.46). Conclusions: This study describes a process for co-designing a text messaging mental health engagement intervention with CC students that is grounded in a human-centered design approach. Further research is needed to rigorously test this intervention and make iterative refinements to improve response and effectiveness.
<img src="https://jmir-production.s3.us-east-2.amazonaws.com/thumbs/8205377cceae781ec5169b5fbc0a42cc" />
Optimizing Treatment Strategies in the Bipolar Disorder Spectrum With Classical AI Approaches: Systematic Review of Performance, Bias, and Clinical Applicability
Background: Bipolar disorder (BD) is a complex and heterogeneous psychiatric condition, characterized by fluctuating clinical courses that affect approximately 1%‐2% of the global population in their lifetime. Despite pharmacological advances, treatment response varies significantly among patients, making the identification of individualized treatment strategies a major challenge. Artificial Intelligence (AI), through its classical approaches, has emerged as a powerful tool in precision psychiatry to identify subtle patterns in complex data and inform personalized clinical decisions. Objective: The present systematic review aimed to examine the current evidence on classical AI-supported treatment optimization in the BD spectrum. Methods: The review was conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines. Four databases (PubMed, Web of Science, Scopus, and Embase) were searched for original studies published after 2015 on the application of classical AI in the treatment of BD in adult patients. Publication bias was evaluated by visual inspection of a funnel plot. The methodological quality, risk of bias, and clinical applicability of the predictive models were assessed using the Prediction Model Risk Of Bias Assessment Tool for prediction models using regression or AI methods (PROBAST+AI; PROBAST+AI Working Group) tool. Results: A total of 35 studies were included and classified into 5 outcome-based categories, including acute symptomatic response, long-term maintenance response, relapse and readmission risk, safety and dose optimization, and brain aging and phenotyping. Acute symptomatic response models performed modestly (pooled area under the curve [AUC] 0.68), while imaging improved accuracy (74%‐77%). Long-term maintenance response models showed moderate-to-high performance (pooled AUC 0.80), with biomarker- and cellular-based models reaching 96%‐99% accuracy. Relapse and readmission prediction achieved a pooled AUC of 0.71, with digital phenotyping and rule-based methods performing best (AUC 0.85‐0.88). Safety and dose optimization models achieved 85%‐97% accuracy. Brain aging and phenotyping studies highlighted accelerated brain aging in BD, partially mitigated by lithium, and revealed novel data-driven subgroups. However, 3 studies were considered at high risk of bias due to small sample sizes associated with disproportionately high-performance estimates. An additional study was identified as potentially biased because it lay markedly distant from the funnel plot’s confidence line. Finally, the PROBAST+AI assessment revealed a high risk of bias in most studies, primarily due to data analysis limitations, small sample sizes, and lack of external validation. Conclusions: The adoption of classical AI tools in BD serves as a driver for therapeutic optimization, although current AI tools in BD should still be considered exploratory rather than ready for clinical use. Effective implementation in real-world clinical scenarios requires more robust, transparent, and externally validated models to ensure reliability and generalizability.
<img src="https://jmir-production.s3.us-east-2.amazonaws.com/thumbs/df358c951a2c44fbd20077fdda306baf" />