Early prediction of depressive and anxiety disorders is challenging due to substantial heterogeneity in risk pathways. Conventional machine-learning models trained on aggregated populations may obscure subgroup-specific mechanisms and limit interpretability for prevention. We evaluated whether a hybrid unsupervised–supervised framework can identify meaningful subgroups and yield more interpretable risk prediction.

We analyzed cohort data of 15,897 Japanese adults who completed baseline (August–September 2020) and 6-month follow-up (February–March 2021) surveys and did not screen positive for depressive and anxiety disorders at baseline (K6 score < 13). Using 169 baseline demographic, psychosocial, lifestyle, and behavioral variables, we performed hierarchical clustering to derive data-driven subgroups. Within each cluster, we trained Random Forest models to predict incident screened depressive and anxiety disorders at follow-up (K6 ≥ 13) and interpreted predictors using SHapley Additive exPlanations (SHAP).

The overall 6-month incidence was 6.23%. A five-cluster solution revealed two high-risk subgroups: an older-adult profile with poor quality of life (12.9%) and a working-parent profile characterized by work–family overload (29.8%). Compared with a global model trained on the full sample, the cluster-then-predict framework showed broadly similar overall performance but performed better in the highest-risk subgroup and revealed more differentiated predictor profiles. Loneliness, health-related quality of life, happiness, and personality traits predominated in clusters with moderate adversity, whereas lifestyle disruption (sleep, diet, and irregular routines) characterized the high-risk late-life subgroup and alcohol dependence and work–family burden characterized the high-risk working-parent subgroup.

Addressing risk-factor heterogeneity before prediction may enable more interpretable, context-tailored prevention strategies.

The COVID-19 pandemic raised concerns about the mental health of an already burdened healthcare workforce. This study examined mental health trajectories among healthcare workers (HCWs) across the pandemic and identified personal and employment factors associated with different symptom patterns.

Longitudinal data were drawn from the NHS CHECK cohort, including clinical and non-clinical staff from 18 NHS Trusts in England (April 2020–April 2023). Growth curve and growth mixture models identified latent classes of HCWs characterized by distinct trajectories of probable common mental disorders. Secondary outcomes included anxiety, depression, alcohol misuse, and post-traumatic stress symptoms. Logistic regression examined associations between baseline personal and employment characteristics and class membership.

The analytical sample included 22,764 participants. For each outcome, growth mixture models identified two latent classes. Approximately 31% of HCWs experienced persistently high symptoms of probable common mental disorders, while 69% experienced persistently low symptoms. Similar patterns were observed for secondary outcomes, with small subgroups demonstrating worsening symptoms followed by improvement. Logistic regression analyses showed that being female, younger, single, working as a nurse, or having a pre-existing mental health diagnosis increased the odds of belonging to a high symptom class. Perceived support from colleagues and managers was protective.

While many HCWs reported consistently low mental health symptom levels, almost a third belonged to a latent class characterized by persistently high symptoms across all time points. These findings underscore the need for mental health support for vulnerable HCW groups, embedded within routine NHS practice rather than limited to crisis periods.

Global initiatives to reduce restrictive practices in mental health settings have gained increasing attention. However, discrepancies in restrictive practice rates create uncertainties about whether these variations reflect true differences in clinical practices or arise from inconsistent classification and reporting methods.

This study investigated how healthcare professionals classify and report potential restrictive practice scenarios, and examined variations in classification and documentation across diverse facilities.

This was an international survey conducted using an online questionnaire via the Qualtrics platform. Healthcare professionals working in adult mental health in-patient settings were recruited through multiple media platforms and snowball sampling. The questionnaire included 44 potential restrictive practice case scenarios. Participants rated each scenario as follows: (a) whether it should be classified as a restrictive practice; (b) whether it should be recorded as such; (c) whether it would be classified as a restrictive practice within their facility; and (d) whether it would be reported as a restrictive practice in their facility. Survey development was guided by systematic reviews and co-design work with stakeholders. Data were analysed using ordered regression models, with clustering by participant identity and country. Robust standard errors were applied to ensure accurate estimation of variability.

A total of 491 healthcare professionals from 41 countries participated. Results indicated substantial inconsistencies in clinicians’ perspectives regarding what constitutes restrictive practices and whether a given action should be reported as a restrictive practice. Although participants frequently identified scenarios as restrictive practices, their intention to report them was considerably lower. Additional discrepancies were observed between clinicians’ individual perspectives and their expectations of how these practices were actually being classified and reported as restrictive practices within the in-patient facilities where they work.

Discrepancies between healthcare professionals’ classification of restrictive practices and their reporting intentions, as well as between their perspectives and actual institutional practices, highlight potential errors in current reporting systems. These findings underscore the need for standardised definitions, enhanced reporting frameworks and structured training programmes and monitoring mechanisms to improve consistency in the management of restrictive practices across mental health settings.

Psychotic experiences (PEs) in are associated with elevated risk for mental health difficulties. This study examined predictors of PEs, inclusive of the role of gender, ethnicity, and protective factors.

Data were drawn from a 2021 Planet Youth survey of adolescents (n = 4,005). PEs were measured using the adolescent psychotic symptom screener. Effects of psychosocial predictors on PEs were measured by fitting multivariable logistic regression main effect and joint exposure models.

29.8% reported PEs. Black/Asian/Other minorities had elevated odds (aOR = 1.59, 95% CI 1.26–2.02, p < .001). Increased odds in males, females, undisclosed gender and non-binary/transgender with elevated emotional/behavioural difficulties (aOR = 4.47, 95% CI 3.53–5.67, p < .001; aOR = 3.25, 95% CI 2.59–4.08, p < .001; aOR = 4.83, 95% CI 2.58–9.02, p < .001; aOR = 4.33, 95% CI 2.69–6.97, p < .001 respectively). High odds in undisclosed gender with low emotional/behavioural difficulties (aOR = 4.36, 95% CI 1.50–12.66, p = .007). Lower odds from perceived school/home safety (aOR = 0.69, 95% CI 0.58–0.83, p < .001 and (aOR = 0.81, 95% CI 0.66–0.99, p = .038, respectively). Elevated odds from recent adversities (aOR = 1.91, 95% CI 1.47–2.49, p = .011) attenuated by parental support (aOR = 1.76, 95% CI 1.17–2.65, p < .001). Each additional adversity (>12 months) increased odds (aOR = 1.12, 95% CI 1.07–1.17, p < .001).

Findings highlight the interplay of risk and protective factors in adolescent PEs, with increased vulnerability among minoritized youth. Results support targeted interventions to reduce mental health disparities.

The QualityRights initiative has shown benefits in decreasing coercive practices and enhancing the recognition of human rights in healthcare.

To translate the World Health Organization’s (WHO) QualityRights Practices Questionnaire into Spanish and assess the relationships between coercive practices in healthcare settings, perceptions of mental disorders, and attitudes and commitment to human rights among primary healthcare professionals from multiple disciplines. Also, we sought to compare these variables between trained and untrained professionals, and evaluate longitudinal outcomes of integrating QualityRights training into undergraduate medical education.

A quasi-experimental study with a non-equivalent control group was conducted. Instruments included the Community Attitudes Toward the Mentally III, Human Rights Exposure in Social Work and Human Rights Engagement in Social Work scales, and the WHO QualityRights Practices Questionnaire. A total of 260 professionals from 12 Colombian regions participated in the study. Multiple linear regression analyses were conducted to examine the effects of the QualityRights initiative on various dimensions related to human rights.

Translation and validation of the WHO QualityRights Practices Questionnaire yielded excellent psychometric properties (Kaiser–Meyer–Olkin value: 0.945; Cronbach’s α = 0.891–0.923; single component explaining 73.4% of variance). The QualityRights initiative was significantly associated with greater human rights knowledge and lower endorsement of coercive practices and authoritarian beliefs. In relation to coercive practices, the initiative was significantly associated with lower scores on the QualityRights Practices Questionnaire (B = −3.118, p < 0.001).

In Colombia, incorporating the QualityRights initiative into medical education appears to be a promising strategy for reducing stigma, enhancing knowledge and commitment to human rights, and minimising coercive practices in primary mental healthcare.

As Alzheimer’s disease and related dementias (ADRD) progress, family caregivers may experience grief before the death of the person living with ADRD. This type of grief is commonly referred to as anticipatory grief, which can contribute to increased psychological distress (i.e., depressive and anxiety symptoms) and potentially affect caregivers’ long-term mental health. This scoping review aimed to synthesize evidence on risk and resilience factors associated with anticipatory grief, its relationship with caregiver mental health, and psychosocial interventions targeting anticipatory grief among caregivers of people living with ADRD.

Guided by the Stress Process Model and the Grief-Stress Model, a literature search was conducted in PubMed, CINAHL, Embase, Scopus, Web of Science, and PsycINFO in May 2025.

Thirty articles met the inclusion criteria. Caregiver characteristics, such as being a spousal caregiver and greater caregiving involvement, were associated with higher levels of anticipatory grief. Caregiving-related stressors and relationship changes across the ADRD trajectory were consistently linked to anticipatory grief across studies, while psychosocial resources, such as adaptive coping and social support, were generally associated with lower levels of anticipatory grief. Anticipatory grief was consistently associated with depressive symptoms. Intervention studies were limited, but those focused on acceptance and preparedness showed potential for reducing anticipatory grief.

Anticipatory grief represents an important dimension of caregiver mental health that reflects ongoing loss. Conceptualizing anticipatory grief within caregiving stress frameworks highlights how vulnerability to distress may emerge from the interplay between caregiving stressors, relationship changes, and psychosocial resources. This conceptual framing may inform future research and palliative care interventions to support the well-being of family caregivers across the dementia trajectory.

Attention-deficit hyperactivity disorder (ADHD) is increasingly recognised as a social identity as well as a medical diagnosis. Social identity theory suggests that group identification can benefit self-esteem, well-being and mental health, but little is known about ADHD social identification or preferred terminology in English.

We aimed to measure ADHD social identification and preferred terminology in a sample of adults with ADHD in the UK and to understand whether ADHD social identification is related to improved self-esteem, well-being and mental health.

Three hundred and nineteen adults with ADHD in the UK participated. They were aged between 18 and 73 years and 59% were female. Participants completed self-report measures of ADHD social identification, self-esteem, well-being, anxiety, depression, terminology preferences, medication use and sources of learning about ADHD. Descriptive statistics were used to identify the percentage of participants who preferred ADHD-first versus person-first terminology. Pre-registered serial mediation models tested hypothesised pathways from ADHD identification to mental health via self-esteem and well-being. Further analyses examined associations between terminology preferences, medication use and sources of learning about ADHD.

ADHD identification was not significantly correlated with self-esteem, anxiety or depression. Most participants (77%) preferred person-first terminology (‘person with ADHD’). Higher ADHD identification was associated with identity-first language preference and medication use. Social media was the only source of learning about ADHD related to higher ADHD identification. In mediation models, ADHD identification was not associated with self-esteem or well-being; however, a subcomponent of ADHD social identification – satisfaction – was indirectly related to better mental health via self-esteem and well-being.

These cross-sectional findings indicated that ADHD identification did not show the hypothesised protective associations with mental health. Preferences for person-first terminology suggest ADHD is not always central to identity. Longitudinal and qualitative studies are needed to clarify causal relationships and clinical implications.

Homelessness is increasing and associated with poor mental health (MH). Few studies have examined how experiences of homelessness and sexual identity intersect to effect MH. We used an intersectional approach to examine MH inequalities related to sexual identity and past homelessness in a nationally representative private household sample, and whether associations were explained by discrimination.

Analysis of the 2007 and 2014 Adult Psychiatric Morbidity Surveys included 10,428 individuals aged 16–64 (58% female/3.8% non-heterosexual). The Clinical Interview Schedule-Revised (CIS-R) identified common mental disorders (CMDs). Self-harm, attempted suicide, alcohol dependence, substance use, sexual identity, discrimination/bullying, past homelessness and health behaviours were self-reported. Associations between sexual identity and homelessness were examined using multivariable Poisson regression. Prevalence ratios (PRs) for MH and health behaviours by intersectional sexual identity-past homelessness were examined using Poisson regression and adjusted for age, sex, area-level deprivation and further for discrimination/bullying.

Bisexual (adjusted PR [aPR]: 2.52, 95% CI: 1.48–4.29) and gay/lesbian (aPR: 1.76, 0.97–3.19) individuals were more likely to report past homelessness than heterosexual peers. Sexual minority (SM) and heterosexual individuals with past homelessness had higher prevalence of all MH outcomes compared to heterosexual peers without homelessness, with associations strongest in the SM-homelessness group (e.g., CMD: aPR: 2.67 [2.37–3.01] for heterosexual-homeless, aPR: 4.11 [3.00–5.63] for SM-homeless, aPR: 1.82 [1.45–2.28] for SM-not homeless groups), and similarly for depression/self-harm/attempted suicide. Likewise, the SM-homeless group had highest prevalence for drug dependence (aPR, 7.38 [3.15–17.29]) compared to the heterosexual-homeless (aPR, 4.03 [3.00–5.42]) and SM-not homeless (aPR, 2.19 [1.27–3.79]) groups. Adjustment for discrimination and bullying substantially attenuated point estimates, with the greatest attenuation (30–50%) in the SM-homeless compared to the heterosexual-homeless groups.

Individuals with past experiences of homelessness have significantly worse MH than heterosexuals without homelessness, with associations highest in the SM-homeless group. Considering experiencing homelessness and SM identity together identifies a group facing particular adversity, which is often lost when examined separately. Discrimination and bullying explained much of the worse MH in SM- and heterosexual-homeless groups, but especially the former. Investigation into the mechanisms leading to MH inequalities is needed, alongside policies and services to support this group.