Npj mental health research最新文献

Early detection and prevention of eating disorders (EDs) in adolescence are crucial yet challenging. We developed and validated diagnostic and prognostic models to predict EDs using data from 44,357 Danish National Birth Cohort participants. Models were trained to identify ED presence in early and late adolescence (11- and 18-year follow-up), utilizing approximately 100 predictors from self-reported and registry-based data. The machine learning model demonstrated strong discrimination for both tasks (diagnostic Area Under the receiver operating characteristic Curve = 81.3; prognostic AUC = 76.9), while a logistic regression model using the top 10 predictors achieved comparable performance. Sex, emotional symptoms, peer relationship and conduct problems, stress levels, parental BMI values, body dissatisfaction, and BMI at the 7-year follow-up emerged as key predictors. Our models showed potential utility in supporting clinical risk assessment, particularly for low-risk preventive interventions, though further validation studies are needed to evaluate their effectiveness in real-world clinical settings.

Accurate psychiatric diagnosis and assessment are crucial for effective treatment. However, current diagnostic approaches heavily rely on subjective observations constrained by time and clinical resources. This study investigates the potential of using Large Language Models (LLMs) to identify the symptoms in psychiatrist-patient dialogues and use them as intermediate features to predict the diagnostic labels. We collected audio recordings of 1160 outpatients with depressive disorder and anxiety disorder. LLMs were trained and utilized to identify clinical symptoms, rate assessment scales, and an ensemble learning pipeline was designed to classify diagnostic results and symptoms with 10-fold cross-validation. The system achieved 86.9% accuracy for identifying the appearance of clinical annotations and 74.7% (77.2%) accuracy for identifying symptoms of anxiety (depression). In addition, analysis of LLM-generated features shows that depression cases exhibited prominent markers of anhedonia and decreased volition, whereas anxiety disorders were characterized by tension and an inability to relax.

StayWell is a 60-day CBT/DBT-based text messaging intervention which leverages reinforcement learning algorithms to support mental health. Participants were randomly assigned to receiving personalized messaging (adaptive arm), static messaging (random arm) or mood-monitoring only messages (control arm). A diverse sample of 1121 adults participated in a fully remote trial between December 2021 and July 2022. Across study arms, participants showed a 25% reduction in depression symptoms (PHQ-8) and 24% reduction in anxiety symptoms (GAD-7) following the intervention. We did not find statistically significant differences in PHQ-8 and GAD-7 reductions between intervention arms. Participants in the control arm had higher mood-monitoring messages response rates than those in other conditions. Finally, post-hoc exploratory analysis assessing outcomes by condition indicated that patients with minimal to mild depression symptoms (PHQ-8 < 10) benefitted from the reinforcement learning algorithm. The results of this trial suggest that StayWell is a promising text-messaging intervention to achieve reductions in depression and anxiety among diverse populations.

FDA approval is widely regarded as a benchmark of quality for medical devices. However, concerns persist regarding its regulatory framework for digital mental health devices. This perspective article examined FDA-authorized Software as a Medical Device (SaMD) in mental health, tracing the devices' regulatory lineage through the De Novo and 510(k)-clearance pathways while assessing the quality of evidence that led to their authorization. Many 510(k)-cleared devices lacked direct evidence of effectiveness, relying solely on equivalence to predicate devices. Furthermore, we identified four FDA-authorized SaMD whose pivotal randomized controlled trials tested prototypes delivered on different digital platforms than those of the final marketed products. Strengthening regulatory standards requires randomized controlled trials evaluating the final marketed product on its intended platform and the use of context-appropriate control conditions. Sham placebo controls may be considered feasible; however, evidence supporting fully inert and fully blinding sham controls for digital interventions remains limited at present. This should occur alongside consistent application of the FDA's discretionary authority to require new 510(k) submissions when substantial product changes occur.

In their recent paper¹, Maples et al. surveyed users of the Replika app². Among their results, they reported that participants were relatively lonely and used Replika for diverse purposes, and emphasized that "3% reported that Replika halted their suicidal ideation"¹. However, important context about how Replika has been marketed and used was missing. We provide context about Replika's sexual component, and discuss the threat of industry interests to scientific integrity.

This reply addresses concerns raised in the Matters Arising letter, emphasizing the rigor of our empirical study on student well-being outcomes with ISAs. We clarify methodological decisions, address speculative claims regarding Replika's marketing and usage, and highlight our study's focus on peer-reviewed, evidence-based findings. Ethical considerations and potential conflicts of interest are transparently discussed, reinforcing our commitment to scientific integrity and advancing knowledge in the field of AI and mental health.

Negative experiences with psychotherapy are common, affecting 3-25% of patients. However, their causes remain underexplored despite their substantial impact on therapy outcomes. Online forums provide unique insights into patients' concerns due to their anonymity. We collected and anonymized forum posts and used a large language model to identify psychotherapy dissatisfaction. Human raters validated the outputs. To identify and analyze themes, we applied clustering, topic modeling, sentiment analysis, and classification based on an existing meta-analytic framework. In total, we extracted 28,079 text passages reflecting dissatisfaction. Clustering yielded 55 subthemes, covering therapist misbehavior, negative treatment effects, poor alliance, treatment mismatch, and healthcare-related frustrations, extending existing taxonomies. Our NLP-based, mixed-methods approach highlights dissatisfaction as both frequent and multifaceted, surfacing themes often overlooked in traditional research, such as structural barriers and lasting psychological consequences. These findings expand previous frameworks and underscore the need for better recognition of negative therapy experiences.

Prosocial behavior, the act of helping others, is essential to social life, yet spontaneous environmental triggers for such behavior remain underexplored. This study tested whether an unexpected event, such as the presence of a person dressed as Batman, could increase prosocial behavior by disrupting routine and enhancing attention to the present moment. We conducted a quasi-experimental field study on the Milan metro, observing 138 rides. In the control condition, a female experimenter, appearing pregnant, boarded the train with an observer. In the experimental condition, an additional experimenter dressed as Batman entered from another door. Passengers were significantly more likely to offer their seat when Batman was present (67.21% vs. 37.66%, OR = 3.393, p < 0.001). Notably, 44% of those who offered their seat in the experimental condition reported not seeing Batman. These findings suggest that unexpected events can promote prosociality, even without conscious awareness, with implications for encouraging kindness in public settings. Trial registration: ClinicalTrials.gov n° NCT06481748; registered on July 1, 2024.

Depression, anxiety, and stress are major mental health concerns globally, especially in India. This study examines the prevalence of mental health symptoms in overweight and normal BMI individuals living near thermal power plants and develops a depression prediction model using binary logistic regression using the DASS-21 score. A community-based cross-sectional study was conducted from October 2018 to March 2019, with data collected through face-to-face interviews. Socio-demographic factors like age, gender, cooking fuel type, and income were analyzed. Significant associations were found between stress and household air pollution (p = 0.011, OR = 17.408, 95% CI) and between anxiety and income below 1 lakh in normal BMI individuals (p = 0.045, OR = 0.303, 95% CI). Depression, anxiety, and stress were more prevalent in females. The depression prediction model demonstrated high performance with an ROC-AUC of 0.8754. These findings highlight the need to address environmental and socio-demographic factors to protect mental health in populations living near thermal power plants.