Matthew Hotopf

Background: People living with long-term physical health conditions who are also experiencing subthreshold depression are at risk of developing major depression. In partnership with lived experience experts, clinicians, and software developers, we co-designed CommonGround, an online peer support and psychoeducation platform. CommonGround’s Theory of Change (ToC) was also conceptualized to outline the intermediate outcomes (and associated indicators of whether such outcomes are achieved) needed to improve the mental health of the target population. Objective: This qualitative study, informed by the ToC, aimed to examine the acceptability, usability, and intervention mechanisms of CommonGround and ultimately refine the ToC and corresponding intervention design ahead of definitive evaluation. Methods: A nested qualitative study was conducted within a remote, 12-week feasibility randomized controlled trial comparing CommonGround (intervention group) to signposting to National Health Service (NHS) web pages (control group). Thirty-nine participants were purposively sampled and interviewed within 6 weeks postintervention about their experiences of CommonGround and trial participation. The interview schedule and thematic framework analysis were rooted in a pretrial ToC while being responsive to emergent themes. Results: Evaluating the assumptions and intermediate outcomes (via their indicators) refined our understanding of how CommonGround may support mental well-being. Overall, participants viewed CommonGround as an appropriate, safe space to access self-management information and support from others with a unique shared understanding of chronic illness. However, many participants felt more time was needed for genuine social relations to form and community culture to evolve. Anonymity helped foster a sense of safety but, for some participants, limited the formation of genuine connections. Other factors contributing to low engagement included personal circumstances (eg, limited time). Key suggestions for future development included creating a mobile app that would be easier and more intuitive to access and navigate. Conclusions: CommonGround was considered an acceptable and appropriate space to access peer support, with the potential to positively impact psychological well-being by fostering shared understanding, support, and knowledge exchange. Beyond informing the future evaluation of CommonGround in a definitive trial, the refined ToC identifies the key preconditions, intermediate outcomes, and associated indicators, and assumptions that may underpin the acceptability and usability of digital peer support interventions in achieving their intended outcomes. We also demonstrate how evaluating and refining components of the ToC after the intervention had been implemented in the trial enabled a more nuanced understanding of how and why the intervention may work, for whom, under what circumstances, and informed changes to improve the co-designed intervention.

Background: The psychosis continuum conceptualises psychotic phenomena as existing on a spectrum from subclinical experiences to psychotic disorders. Despite epidemiological and clinical evidence in support of this model, few studies have examined polygenic liability across the severity continuum: from subclinical experiences to first-episode psychosis (FEP). Aims: To map polygenic liability for schizophrenia (SCZ), bipolar disorder and major depressive disorder (MDD) across the psychosis continuum. We hypothesised that schizophrenia polygenic scores (SCZ-PGS) would track phenotypic severity most closely across this continuum. Method: We harmonised data from 5 south-east London cohorts ( N = 1315; 860 European, 455 African ancestries) across 4 ordered groups indicative of different levels of phenotypic severity: controls ( n = 647), individuals with self-reported psychotic experiences ( n = 89), individuals at clinical high risk (CHR) that did not convert to psychosis ( n = 124) and individuals with FEP ( n = 455). We computed ancestry-specific PGS for SCZ, bipolar disorder and MDD and compared these across groups using ordinal logistic regression. Results: In European ancestry participants, elevated SCZ-PGS (odds ratio 1.72, p = 4.57 × 10 ⁻¹⁵ ), bipolar disorder-PGS (odds ratio 1.56, p = 1.03 × 10 ⁻¹⁰ ) and MDD-PGS (odds ratio 1.44, p = 4.86 × 10 ⁻⁸ ) were individually associated with increasing phenotypic severity. SCZ-PGS explained the largest variance ( R ² = 0.084) across the psychosis continuum and showed pairwise differences between CHR and controls ( β = 0.32, p = 0.027), and between FEP and CHR ( β = 0.38, p = 0.017). Multivariable models, combining three PGS, outperformed univariable models ( R ² = 0.105). Similar trends in SCZ-PGS across the psychosis continuum phenotype were seen in African ancestry participants, but only controls and FEP groups differed significantly ( β = 0.27, p = 0.037). Conclusions: Our findings provide biological evidence supporting the psychosis continuum framework. In European ancestry, SCZ-PGS differences suggest that CHR who do not transition to psychosis have intermediate genetic vulnerability. Multivariable models, utilising all three PGS, explained additional variance compared with univariable ones, supporting transdiagnostic genetic liability. Improving PGS predictive power in African ancestry remains a priority.

Functional neurological disorder (FND) is characterized by involuntary, distressing neurological symptoms that are distinguishable from other neurological diagnoses in presentation, aetiology and mechanisms. Significant disability, impaired quality-of-life and substantial healthcare resource utilization often result from the disorder, yet its underlying mechanisms have not been proven empirically. Importantly, long-term clinical trajectories and predictive markers have been understudied in FND cohorts. Inadequate funding of large-scale mechanistic and longitudinal research, methodological limitations and the absence of standardized outcome measures necessitate development of innovative solutions. Digital remote monitoring technologies (dRMTs), which have been shown to be valuable across a range of other neurological and psychiatric disorders, may have substantial potential to address these research challenges in FND, yet this potential is currently unrealized. We propose that dRMTs could be harnessed to monitor subjective and/or objective functional neurological and associated symptoms, mechanistic variables and important outcome domains, with excellent temporal precision, flexibility and external validity. These methods could allow precise characterization of both short- and long-term fluctuations in FND symptoms, the influences on them and their impact on patients. Such data would strengthen mechanistic models, improve outcome prediction and facilitate personalized FND clinical management in this underserved population. This article is part of the discussion meeting issue ‘Digital healthcare for the management of functional neurological disorders’.

BACKGROUND Remote measurement technologies (RMTs) offer a potential solution for monitoring depression symptoms in situ while reducing reliance on self-report assessments. Sleep and physical activity measures derived from wearable devices may capture behavioural patterns relevant to depressions severity, but their incremental value for predicting current depression severity remains unclear. OBJECTIVE To evaluate the predictive value of wearable-derived sleep and physical activity measures for current depression severity, and to determine whether RMT-derived behavioural subtypes improve prediction compared with individual RMT features and baseline sociodemographic and clinical characteristics. METHODS Longitudinal data from the Remote Assessment of Disease and Relapse in Major Depressive Disorder (RADAR-MDD) cohort were used to predict three-monthly depression severity, at three to 36-months of follow-up, measured using the Inventory of Depressive Symptomatology- Self Report. Elastic net and random forest models were compared across different combinations of baseline individual characteristics, individual sleep and physical activity RMT features, and RMT-derived behavioural subtypes. RMT features were derived from data collected during the seven days preceding each IDS-SR assessment. Behavioural subtypes were previously derived from these longitudinal RMT data using a Hidden Markov model, which estimated participants’ subtype membership probabilities at each timepoint. Prediction models were evaluated using nested group five-fold cross-validation and assessed using root mean square error, R2 and mean absolute error. RESULTS Models including baseline depression severity achieved R2 values of approximately 0.6, with RMSE of approximately 10 and MAE of 8. Adding individual RMT features of RMT-derived subtypes provided little improvement in predictive performance. Excluding baseline depression severity reduced R2 to approximately 0.1, while RMSE increased to approximately 15 and MAE to 12, demonstrating the substantial contribution of baseline depression severity to predictive performance. CONCLUSIONS Wearable-derived sleep and physical activity measures and their behavioural subtypes provided limited additional predictive value for current depression severity beyond baseline individual characteristics, particularly baseline depression severity. Although these measures did not substantially improve prediction in the present models, they may still capture complementary behavioural information relevant to understanding depression.

Twin studies reveal high genetic overlap between anxiety disorders and depression, contributing to the internalising spectrum. Some genetic specificity for fear-based anxiety disorders (fear), distinct from general anxiety and depression (distress), has also emerged. Limited datasets with detailed phenotyping across anxiety disorders have restricted most genome-wide association studies (GWAS) to “any anxiety diagnosis”. Additional genome-wide evidence to discern genetic differences between fear and distress is required. We conducted GWAS meta-analyses of fear (panic, agoraphobia, specific phobia, social anxiety disorder) and generalised anxiety disorder (GAD), measured using brief single-item and detailed symptom-based diagnoses from three datasets. We explored two control group criteria: phenotype-specific (fear/GAD) or broader anxiety/depression screening. We identified one independent genome-wide significant locus and three gene-level associations with fear (up to 35,523 N _cases ; 157,447 N _controls ). Four genome-wide significant loci and three gene-level associations were identified for GAD (up to 60,879 N _cases ; 117,064 N _controls ). The genetic correlation between fear and GAD was significantly different from unity only when excluding a depression-enriched dataset and using phenotype-specific control screening (r _g = 0.87; P = 9.32 × 10 ⁻³ ). Most complex traits had statistically similar genetic correlations with fear and GAD, including depression. Exceptions included general cognitive ability, educational attainment, and coronary artery disease, showing statistically stronger genetic correlations with fear than GAD, while bipolar disorder type I, anorexia nervosa, and neuroticism displayed the opposite pattern. Our findings partially support a distress-fear genetic distinction, but show stronger evidence for an overarching genetic liability to internalising psychopathology driving comorbidity across anxiety disorders and depression.

Background: There is growing evidence that air pollution may impact mental health, however this is the first study to investigate its role in child and adolescent community mental health service use. Methods: This retrospective cohort study used electronic health records from a large South London mental healthcare provider to establish a cohort of young people first in contact with community mental health services between 2008 and 2012, who were then followed for up to 9 years. Baseline (3-month) residential nitrogen dioxide (NO2), particulate matter (PM)10 and PM2.5 exposure was estimated using the KCL urban model. Contacts over 1-, 5-, and 9-year periods were analysed in relation to air pollution using negative binomial regression. The potential effect-modifying role of high vs low greenspace exposure (residential normalised difference vegetation index) and age-, gender-, ethnicity- and disorder-specific differences were investigated using subgroup analyses. Findings: The sample consisted of 11,773 children and adolescents (3-17 years) who had first face-to-face contact with mental health services between 2008 and 2012. Consistent positive associations were observed between air pollution exposure and number of community mental health service (CMHS) contacts. For example, by Year 9 a clear dose-response pattern was evident between NO2 exposure and CMHS contacts, with an adjusted incidence rate ratio of 1.17 (95% CI=1.09 to 1.26) for service users in the highest quartile of exposure compared to the lowest quartile. Population Attributable Fractions (PAFs) suggested that bringing UK air quality in line with WHO recommendations could reduce CMHS contacts by 3-6%. Associations were most pronounced for boys and children (3-12 years) and were substantially attenuated in those living near high levels of greenspace. Interpretation: Residential air pollution exposure was associated with increased community mental health service contacts in children and adolescents, attenuated by high greenspace exposure, suggesting that policy changes acting on air quality and green infrastructure in urban settings could improve population-level mental health and reduce current strain on child and adolescent mental health services.

Background: Efforts to advance our understanding of depression have long been constrained by the disorder’s vast symptom heterogeneity and by the reliance on self-report, which offers only a partial view of phenotypic expression. Digital phenotyping provides an opportunity to address these core challenges by generating real-time, objective data on behavior and physiology, offering new perspectives on understanding depression phenotypes. Yet, prior efforts to identify such objectively derived subtypes have relied on predefined diagnostic labels or supervised models, limiting discovery to existing clinical categories. Objective: This study aimed to identify subtypes of depression based on objective sleep and activity data using an unsupervised learning method and to explore how participants transition between these subtypes over time. Methods: We analyzed longitudinal Fitbit data from 623 participants with recurrent depression enrolled in the Remote Assessment of Disease and Relapse in Major Depressive Disorder study. To identify our subtypes, we applied Gaussian mixture models and hidden Markov models, incorporating a thorough model selection approach that combined grouped cross-validation and seed selection to ensure robustness. Results: Three activity subtypes (high, light, and low activity) and 4 sleep subtypes (efficient early sleepers, efficient late sleepers, disrupted sleepers, and variable late sleepers) were consistently identified. These subtypes align with known associations between depression and behavioral patterns. Transition modeling revealed stability within individuals over follow-up, further suggesting the presence of behavioral phenotypes rather than momentary fluctuations. Conclusions: The results demonstrate that wearable-derived features can identify reproducible and clinically relevant behavioral subtypes of sleep and activity in individuals with major depressive disorder. These subtypes reflect known behavioral correlates of depression and may offer a data-driven framework for reducing phenotypic heterogeneity, improving research stratification, and supporting personalized patient monitoring. Further work is needed to validate these findings in independent cohorts and evaluate their potential use in reducing noise when using sleep or activity data to predict depression outcomes.

Background: There is relatively limited psychological support dedicated to people living with long-term physical health conditions and subthreshold depressive disorder. Online peer support may be an appropriate intervention to help bolster patients’ mental well-being to prevent progression of their symptoms to major depressive disorder. For interventions to be successfully integrated into the self-management routines of people with long-term physical health conditions, they should be co-designed to ensure that they align with the wants and needs of the target audience. Objective: This study aims to coproduce an online peer support intervention with people with lived experience, software experts, clinicians, and academics through an iterative process of co-design and subsequent co-validation through usability testing. Methods: We followed a 4-stage coproduction process: co-assess, co-design, co-validate, and co-deliver. Our research advisory group was actively involved in all stages, consisting of 1 coinvestigator and 6 people with lived experience of long-term physical and/or mental health comorbidities. The co-assess and co-design stages involved our participatory design panel, which included 10 members living with various long-term conditions. The participatory design panel participated in online focus groups to assess their unmet psychosocial needs and then co-designed the intervention prototype through online workshops with software developers. The co-validation stage involved an additional group of participants (n=12) with long-term physical health conditions. During co-validation, the prototype underwent usability testing, including think-aloud exercises and semistructured interviews. Content analysis identified the priorities for the iterative development that formed the basis of further research advisory group co-design workshops. The next stage, co-delivery, involved coproducing the protocol of a feasibility and acceptability randomized controlled trial. Results: Participants highlighted that a platform must feel safe and trustworthy for the space to support the mental well-being of those living with long-term health conditions. The participatory design panel co-designed a platform prototype to meet this need. During the co-validation stage, the think-aloud exercises identified common issues related to navigation challenges and feature glitches. Content analysis of the semistructured interviews confirmed that the community forum, resources, and other platform pages were appropriate and acceptable, but revealed usability concerns. Participants stressed the need for intuitive navigation and suggested new features that would enhance user experience. Facilitators and barriers to engagement were also noted, including the importance of fostering trust in the platform’s ethos and branding to create a safe space. Through iterative development and subsequent usability testing, the final prototype was approved. Conclusions: We have provided a worked example of a comprehensive, coproduction process where we worked alongside people with lived experience to successfully design an online peer support platform with embedded psychoeducation. The platform, called CommonGround, is ready to be evaluated in a feasibility randomized controlled trial.

Background: Online peer support can help people living with long-term physical health conditions to manage their mental well-being. Although the potential negative events that can occur and risks associated with web-based peer communities are well recognized, our understanding of how best to moderate these spaces is relatively limited, particularly with regard to new communities. Previous work has focused on the experiences of either moderators or community members. Objective: This study aims to explore the perspectives of both members and moderators of a new online peer support community to evaluate the moderation procedures and inform recommendations for best practice. Methods: Community members (n=39) who participated in a research trial of a new online peer community, CommonGround, were interviewed. The moderation team (n=5) was invited to a focus group. Community member interviews explored their opinions of moderation policies and the behavior of the moderation team. The moderator focus group explored their experiences of moderating the community, including perceived benefits, common challenges, and areas for improvement. All interviews and the focus group were conducted online, audio-recorded, and transcribed verbatim. An inductive thematic analysis was conducted to sort the data into overarching themes through an iterative process. Results: Effective moderation was considered critical in creating a safe space that members wanted to engage with and for mitigating any risks, particularly around the spread of medical misinformation. Both moderators and community members felt that the moderation policies and practices were appropriate and applicable to the community. Moderators found navigating the moderation threshold, where they balanced safety against free speech, challenging when determining whether to intervene or not. Being part of a team with mixed clinical expertise helped moderators build confidence in navigating this threshold and also presented other benefits of easy access to support and improving the consistency of their moderation practices. It was suggested that in order for a community to flourish, community members would self-moderate. However, moderators and members felt that the strong community culture and high levels of member engagement that are needed to support self-moderation had not yet evolved. Proposed improvements to moderation included new features to support the efficiency of identifying new content for review and reviewing the rule of anonymity. Conclusions: Moderation is critical in making online peer communities feel safe and engaging. Moderation practices should be co-produced with the target audience to ensure that they are aligned with the community’s unique moderation wants and needs, including clear escalation pathways, transparent communication patterns, and plans to review and update policies or procedures as the community evolves. There should be technological features that promote self-moderation, as the community may shift towards self-moderation as it matures. It is also critical to ensure that moderators feel supported so that they are best placed to support the broader community. Trial Registration: ClinicalTrials.gov NCT06222346; https://clinicaltrials.gov/study/NCT06222346 International Registered Report Identifier (IRRID): RR2-10.2196/71513

Twin studies reveal high genetic overlap between anxiety disorders and depression, contributing to the internalising spectrum. Some genetic specificity for fear-based anxiety disorders (fear), distinct from general anxiety and depression (distress), has also emerged. Limited datasets with detailed phenotyping across anxiety disorders have restricted most genome-wide association studies (GWAS) to “any anxiety diagnosis”. Additional genome-wide evidence to discern genetic differences between fear and distress is required. We conducted GWAS meta-analyses of fear (panic, agoraphobia, specific phobia, social anxiety disorder) and generalised anxiety disorder (GAD), measured using brief single-item and detailed symptom-based diagnoses from three datasets. We explored two control group criteria: phenotype-specific (fear/GAD) or broader anxiety/depression screening. We identified one SNP-based independent locus and three gene-level genome-wide significant (GWS) associations with fear (up to 35,523 N _cases ; 157,447 N _controls ). Four GWS SNP-based loci and three gene-level loci were associated with GAD (up to 60,879 N _cases ; 117,064 N _controls ). The genetic correlation between fear and GAD was significantly different from unity only when excluding a depression-enriched dataset and using phenotype-specific control screening (r _g = 0.87; P = 9.32 × 10 ⁻³ ). Most complex traits had statistically similar genetic correlations with fear and GAD, including depression. Exceptions included general cognitive ability, educational attainment, and coronary artery disease, showing statistically stronger genetic correlations with fear than GAD, while bipolar disorder type I, anorexia nervosa, and neuroticism displayed the opposite pattern. Our findings partially support a distress-fear genetic distinction, but show stronger evidence for an overarching genetic liability to internalising psychopathology driving comorbidity across anxiety disorders and depression.