Mental health triage-level intent classification in linguistically diverse and resource-constrained clinical environments requires models that simultaneously meet strict patient data privacy regulations, cross-lingual generalization to low-resource languages, and real-time inference on edge hardware. No existing federated NLP system has jointly demonstrated this combination. This work proposes Fed-XLM-R, a unified federated learning framework designed to close that gap.

Fed-XLM-R integrates XLM-RoBERTa with lightweight bottleneck adapter layers, adapter-scoped differentially private stochastic gradient descent (DP-SGD), proximal regularization for straggler-robust aggregation (FedProx), and INT8 post-training quantization for edge deployment. Gradient perturbation is confined exclusively to the 0.87% trainable adapter subspace rather than the full 270-million-parameter model. The framework was evaluated in a simulated 10-client federation on a public benchmark, with intent classification trained and evaluated on English data. A cross-lingual diagnostic was additionally run on the XNLI benchmark across five languages (English, Spanish, Hindi, Swahili, Arabic) against an mBERT baseline, and canary phrase memorization analysis was conducted across all 50 federated rounds to assess data extraction risk.

At the clinically meaningful privacy budget of ε = 1.0 (δ = 10^-5, Fed-XLM-R achieves 96% accuracy retention, yielding an overall accuracy of 0.923, F1-score of 0.918, and AUC-ROC of 0.956, within 0.5% of a centralized baseline while providing formal differential privacy guarantees. Canary phrase analysis confirms the DP-SGD mechanism prevents training data extraction. The XLM-R backbone retains strong discriminative capacity in low-resource languages relative to mBERT on the XNLI diagnostic. Adapter-only parameter transmission reduces cumulative communication cost by 40.7-fold at the 85% accuracy threshold, at a per-round volume compatible with standard mobile data networks. INT8 quantization reduces inference latency from 1,974 ms to 523 ms on low-end CPU hardware, a 3.8-fold speedup.

These findings characterize the mechanism's behavior under controlled conditions in a simulated federation rather than demonstrating field deployment. The cross-lingual results support the feasibility of future multilingual extension but do not themselves constitute multilingual intent classification. The three-category label scheme (Anxiety/Depression/Normal) supports triage-level intent routing but has not been validated against standardized clinical instruments; the system is not proposed as a clinical diagnostic or screening tool. Overall, Fed-XLM-R demonstrates that privacy-preserving, communication-efficient, and edge-deployable federated NLP is achievable without substantial accuracy loss, offering a practical pathway toward privacy-compliant mental health triage systems pending real-world and multilingual validation.

Moral injury is at a critical juncture in its conceptualization. Current dominant approaches to conceptualizing and assessing moral injury are rooted in syndromal frameworks with implicit medical model assumptions. In these approaches, painful moral emotions and cognitions are often conceptualized as pathological symptoms to be reduced or corrected. Such approaches risk pathologizing deeply human moral responses and may inadvertently position clinicians as arbiters of which moral appraisals or emotional reactions are appropriate, proportionate, or adaptive. In this paper, we present Contextual Behavioral Science (CBS), the philosophical foundation of Acceptance and Commitment Therapy for Moral Injury (ACT-MI), as an alternative framework for conceptualizing and treating moral injury without pathologizing moral pain or requiring clinicians to adjudicate the validity of clients' moral judgments. Drawing on CBS, Relational Frame Theory (RFT), and evolutionary science, we examine the social functions of moral emotions, the human capacity to construct verbal moral narratives, and the implications of these processes for intervention. Building on these foundations, ACT-MI is described as an approach aimed at disrupting the behavioral patterns that maintain moral injury at both individual and group levels. Rather than targeting moral pain to reduce symptoms, using ACT-MI can facilitate transforming the individual's relationship to it. Through functional assessment, clients' efforts to avoid, control, or resolve moral pain are identified and addressed with core ACT processes, including acceptance, defusion, contact with the present moment, self-as-context, values, and committed action. Within this framework, individuals learn to hold moral pain while continuing to engage in values-based living and meaningful social functioning at individual and group levels of treatment. Assessment targets, methods, and hypotheses are identified to support testing the mechanisms and impacts of ACT-MI in future research.

Brain and body metabolic disruptions have been suggested in several studies as a mechanism underlying posttraumatic stress disorder (PTSD). Therapies targeting metabolic pathways might provide new treatment options. This review explores the potential of ketogenic therapies, such as the ketogenic diet and ketone supplementation, as novel metabolic treatments for PTSD.

An integrative review was conducted. We searched PubMed, Scopus, PsycINFO, PsycArticles, PSYNDEX, MEDLINE, ClinicalTrials.gov, International Clinical Trials Registry Platform, and EU Clinical Trials Register for clinical, preclinical, observational, mechanistic, and review studies that reported outcomes or mechanisms on the intersection of ketogenic therapy and PTSD. Methodological quality of included studies was assessed independently by two reviewers based on predefined criteria.

Sixteen articles (9 human, 4 animal studies, 3 reviews) were eligible and included in the present review. Several metabolic and signaling pathways through which ketogenic therapies could impact PTSD pathophysiology were identified. Preclinical data suggestive of potential efficacy were identified, while only minimal reliable information is available from human studies.

Importantly, the absence of larger clinical trials and the reliance on low-level evidence substantially limit current interpretations. High-quality randomized and adequately powered clinical trials are required to examine the potential role of ketogenic therapies in treating PTSD.

Spanning multisensory integration, ageing, attention, and motor control, this study is, to our knowledge, the first to examine how audiovisual integration differentially influences the initiation and execution of goal-directed motor actions across the adult lifespan.

Young adults (<30 years) and older adults (between 50 and 69 years and ≥70 years) completed a simple detection task and a visual discrimination task on a portable touchscreen device. In the detection task, participants tapped to indicate the perceived location of auditory, visual, or audiovisual stimuli. In the discrimination task, they responded to a visual target presented either with or without distractors, accompanied by auditory stimulation that was either spatially aligned or misaligned with the target.

Across both tasks, multisensory stimulation significantly reduced reaction times in younger and older adults, whereas movement speed slowed modestly, most reliably in the detection task and under audiovisual distraction during visual discrimination. Distractors disproportionately impaired performance in older adults, but spatially aligned auditory information partly alleviated this effect, especially in the oldest group. Importantly, these modest movement-speed costs did not outweigh the overall facilitation of goal-directed actions.

The results suggest that multisensory signals enhance rapid motor planning and initiation, followed by a later calibration phase that improves precision during movement execution. They also indicate that the spatial window of multisensory integration may narrow in the presence of distractors, particularly in older adults. Taken together, these findings demonstrate that multisensory facilitation remains functionally advantageous even under distracting conditions, albeit at a relatively small task-dependent cost to movement speed.

Racial-ethnic minoritized individuals face persistent disparities in mental health service utilization, which are shaped by structural and interpersonal racism. This study examined whether racial microaggressions in counseling were indirectly associated with substance use through their impact on help-seeking intentions and psychological distress. A serial mediation model was tested using self-report data from a community-based sample of racial-ethnic minoritized adults (N = 223). Results indicated that prior experiences of racial microaggressions in counseling were associated with lower help-seeking intentions, which in turn were linked to greater psychological distress and increased substance use. Although the total effect of racial microaggressions on substance use was not significant, the indirect pathway was significant, underscoring the importance of examining mediating mechanisms that impact substance use outcomes. These findings highlight how interpersonal racism in therapeutic settings may contribute to adverse mental health outcomes and emphasize the need for culturally humble, affirming mental health care. Implications for clinical practice and equity are discussed using frameworks from Critical Race Theory and the Social Determinants of Health framework.

Menstrual distress - the physiological and emotional stress response to menstrual symptoms that are perceived as threatening and difficult to manage - is common among menstruating people and can have a profound negative impact on mental health and well-being. Romantic partners can support dealing with emotional and physical stress, yet their role in dealing with menstrual distress specifically remains underexplored. Adopting a biopsychosocial perspective, this qualitative study aims to shed light on the experience of menstrual distress within romantic relationships. Semi-structured interviews were conducted with 27 individuals (aged 19-59 years, living in Austria), who were in romantic relationships (n = 23 heterosexual, n = 3 homosexual/queer, n = 1 unlabelled) and who self-reported to have experienced menstrual distress. Indeed, romantic partners play a crucial role in shaping both the impact of and coping mechanisms for menstrual distress. Due to the intimate nature of romantic relationships, partners influence experiences of menstrual distress, whether actively engaging with the topic or not. Supportive relationships fostering authenticity, communication and teamwork enhance coping, whereas negative partner responses can perpetuate negative stereotypes, stigma and symptom trivialisation. To enhance individuals' and couples' well-being, educational campaigns should raise menstrual health awareness early, and couples may benefit from guidance in managing the challenges effectively.

We aimed to use Latent Profile Analysis (LPA) to identify distinct profiles of psychosocial health among adults aged 50-70 years in Victoria, Australia, to identify predictors of profile membership and determine if profile membership is associated with differences in hazardous alcohol consumption.

A cross-sectional exploratory study using observational data from the Beyond 50 longitudinal cohort collected between September 2023 and July 2024.

Victoria, Australia.

Eligible participants were aged 50-70 years and residing in the Frankston or Mornington Peninsula Local Government Areas. Their recent alcohol use was comparable to the wider Australian population (past 12-month use 85.2% cohort vs. 84.4% age-matched national estimate). The cohort consisted of 1059 adults (47% female and 53% males) aged 50-70 years [mean 61.0, standard deviation (SD) 6.0], representing 1.2% of the local target population.

University of California Loneliness Scale 4-item (UCLA-LS-4), Duke Social Support Index (DSSI), General Anxiety Disorder-7 (GAD-7), Patient Health Questionnaire-9 (PHQ-9) and Alcohol Use Disorder Identification Test (AUDIT-C). Low social connection was defined as limited social interaction and support alongside high loneliness. Poor mental health was reflected in elevated anxiety and depressive symptoms.

Four psychosocial profiles were identified, reflecting differing levels of social connection and mental health: (1) low social connection with good mental health, (2) socially connected with good mental health, (3) low social connection with poor mental health and (4) socially connected with poor mental health. People who were younger [odds ratio (OR) = 0.90, 95% confidence interval (CI) = 0.85-0.94] and had lower educational attainment (OR = 0.44, 95% CI = 0.34-0.58) were more likely to belong to the profile 'low social connection, poor mental health' than 'socially connected, good mental health'. The profile 'low social connection, good mental health' showed the lowest probability (0.04) of receiving a positive screen for hazardous alcohol consumption (an AUDIT-C score of 3/4 or higher for men/women), with the profile 'socially connected, poor mental health' having the highest (0.20). The overall chi-squared (χ²) test indicated statistically significant differences in the distribution of the risk of hazardous alcohol consumption across profiles [χ² = 21.15 (6), P < 0.05].

Latent profiles of psychosocial health appear to be statistically significantly associated with differences in alcohol consumption among adults aged 50-70 years in Victoria, Australia. The profile 'socially connected, poor mental health' has the highest probability of hazardous alcohol consumption, while 'low social connection and good mental health' has the lowest.

This study aimed to compare clinical outcomes among children and adolescents with autism spectrum disorder (ASD) who did or did not have documented telehealth (TH) utilization.

Data were aggregated from electronic health records of 112 health care organizations (queried in March 2026). Patients ages 5-18 years who had received a diagnosis of autistic disorder (ICD-10-CM code F84.0) were assigned to a TH or non-TH cohort. After 1:1 propensity score matching (PSM) across 14 comorbid condition categories, the analysis included 12,380 patients per cohort at 6 months and 15,689 patients per cohort at 12 months. Eight outcomes were assessed: medical noncompliance, antipsychotic initiation, emergency department (ED) visit, inpatient hospitalization, antidepressant initiation, attention-deficit hyperactivity disorder (ADHD) medication initiation, irritability-related encounters, and adaptive behavior treatment. Risk ratios (RRs), hazard ratios, and encounter frequencies were derived.

PSM cohorts were well balanced. In the 6-month analysis, TH use was associated with lower antipsychotic initiation (RR=0.81) and hospitalization (RR=0.69) but more ED visits (RR=1.39), irritability-related encounters (RR=2.49), and adaptive behavior treatment (RR=5.17). In the 12-month analysis, higher irritability (RR=2.29) and adaptive behavior treatment (RR=3.40) persisted in the TH cohort, which also had increased antidepressant (RR=1.17) and ADHD medication initiation (RR=1.12) and greater medical noncompliance (RR=1.84).

TH use by young people with ASD was associated with lower rates of hospitalization and antipsychotic initiation at 6 months, although not at 12 months. TH use was also associated with increased irritability-related encounters and adaptive behavior treatment. These findings highlight the need for tailored hybrid care strategies.

This pilot study examined the effectiveness of high and low measurement-based care (MBC) feedback strategies for depression treatment.

Twenty providers at a U.S. Department of Veterans Affairs facility were randomly assigned to a no-feedback, low-feedback (automated collection and scoring of a depression self-report measure), or high-feedback (automated collection and scoring plus automated decision support) condition. Patients (N=336) completed the electronic version of the Patient Health Questionnaire-9 biweekly for 16 weeks. Outcomes included number of medication changes and depression symptom severity.

Patients completed 75% of the assessments. No statistically significant differences in medication changes or depression symptom severity were observed by condition. Approximately 46% of the patients received a medication change. The MBC groups had lower variability in management compared with the control group. Depression symptoms improved among all patients.

MBC may reduce variability in medication management; however, automated collection was insufficient. Further research is needed to support implementation of all components of MBC.