To understand the influence of adjuvant ketogenic diet on behavioural dysregulation in individuals with bipolar disorder through a review of the available literature.

A systematic review was conducted across the databases of Medline Ovid, PubMed, PsycINFO, the Cochrane Library, and the University of Sheffield Star Plus library, following the PRISMA guidelines. Relevant articles were identified using Inclusion and Exclusion criteria. Critical appraisal tools (the Critical Appraisal Skills Programme (CASP) and the Joanna Briggs Institute (JBI)) were used to assess the quality of the selected publications.

Twenty-two articles were identified as suitable for the review. Out of this, 10 were systematic reviews, and 6 were case reports. Eighteen publications have deduced that ketogenic diets aid in stabilising mood.

The review identified that ketogenic diets have the potential to reduce behavioural dysregulation in bipolar patients. The descriptive analysis suggests the need for additional clinical research, as the available studies were mainly uncontrolled and included a limited number of participants.

This systematic review aims to synthesize the current literature on Artificial Intelligence (AI)-based interventions for improving sleep outcomes in individuals with neuropsychiatric conditions, excluding those with a primary diagnosis of insomnia.

This paper is a systematic review conducted in accordance with the PRISMA 2020 guidelines, with two independent reviewers conducting all stages using a double-blind method and conflicts being resolved after each section. A comprehensive electronic search was performed across EMBASE, Medline and PubMed using MeSH terms for Artificial Intelligence, Sleep and Neurological/Neuropsychiatric Conditions. Studies included were English-language, peer-reviewed primary research published between 2022 and November 2025, focusing on AI-based interventions for sleep outcomes in individuals with neuropsychiatric conditions (excluding primary insomnia). Data extraction used the Joanna Briggs Institute (JBI) systematic review framework.

Five studies met the inclusion criteria with wide variations in study type. Sample sizes ranged from 65 to 540 across studies conducted in the UK, USA and Australia. The interventions ranged from AI-based chatbots, mobile apps, wearable devices and sociallyassistive robots. Positive and statistically significant increases in participant engagement were observed, with two papers reporting interactive AI groups attending up to two additional sessions on average (b=1.65, p<0.001) and using the intervention up to 3.8 times longer (p<0.01) compared to the control group. However, overall sleep-related outcomes were mixed; the intention-to-treat analysis in the US randomised control trial (Paper 4) found no statistically significant difference, but post hoc analysis showed significant improvements in sleep quality in the high-device adherence subgroup. Conversely, a cluster randomised control trial (Paper 5) showed no evidence of improved sleep patterns but did note a significant reduction in daytime and night-time motor activity.

Positive and statistically significant increases in participant engagement were observed with the use of artificial intelligence interventions. However, the increased engagement did not result in statistically significant improvements in overall sleep outcomes. The effectiveness of AI interventions is limited by the lack of robust research and universally low sample sizes. Whilst AI interventions have shown to improve engagement, there is currently insufficient evidence to suggest AI interventions improve sleep outcomes in the population group studied.

Severe psychiatric illness (SPI), defined as schizophrenia and related psychoses, bipolar disorder (BPAD), and depression with psychosis, is associated with a reduced life expectancy of 15–20 years, largely attributable to physical illness.

SPI, and the sometimes-compulsory treatments, are associated with adverse effects on physical health, while other illness features like impaired decision-making ability, abnormal mood, and negative symptoms can impair engagement with health services. Consequently, regular monitoring and proactive management of physical illnesses are especially important. For these reasons, General Practice (GP) patient registers and outreach were previously considered necessary; however, the Quality and Outcomes Framework (QOF), which paid GPs to perform these roles, was withdrawn in Scotland in 2016 and cut back in England in 2025.

Therefore, we undertook a service evaluation of physical health monitoring in CMHT SPI outpatients. A constituent survey of contact with primary care services is reported here.

CMHT keyworkers recorded BMI and asked their SPI outpatients when they last saw their GP and when physical health was last discussed in primary care.

Schizophrenia (n=212): 41% obese; schizoaffective disorder (n=11): 70% obese; BPAD (n=23): 56% obese; depression with psychosis (n=6): 80% overweight or obese; and delusional disorder (n=2). Among SPI patients with BMI ≥30, 47.1% had not seen their GP in over a year; for BMI ≥35, the figure was 46.5%; and for BMI ≥40, 45%. Overall, 51%(n=134) SPI patients thought they had seen their GP in the previous year, and only 36% (n=94) reported discussing physical health with a primary care provider during that time.

32% of Scottish adults are obese, and 92% report seeing their GP at least annually, with contact rates higher among those with limiting long-term conditions. This contrasts with our SPI patient data, in which most were obese and only around half could recall seeing their GP within the past year, with only 36% recalling any discussion of physical health.

In 2012, SPI patients saw their GPs approximately three times annually. Our data suggest this level of engagement has declined markedly since then. As noted, the QOF incentives for GP management of the physical comorbidities associated with reduced life expectancy in SPI have been withdrawn in Scotland and cut back in England; this may have compounded an existing healthcare inequality and geographical disadvantage.

These data will be expanded to advocate for better community management of the physical health risks faced by people with SPI.

Voyeurism represents a significant public safety concern in Singapore, ranking third amongst crimes of concern with a 9% increase in cases from 2023 to 2024. Despite its prevalence, limited research exists on the psychological profiles and risk factors of individuals engaging in voyeuristic behaviour in local contexts.

This study seeks to develop a foundational understanding of potential patterns and risk factors for voyeuristic acts in Singapore’s local context. It postulates that there are predictable, recurring patterns for how individuals engage in voyeuristic acts, with risk factors such as social isolation, dysfunctional interpersonal relationships, the lack of a meaningful romantic relationship, antisocial personality traits, and concurrent mental illness.

A retrospective medical records review was conducted of cases known to the Department of Forensic Psychiatry at the Institute of Mental Health, Singapore.

The study was approved by the National Healthcare Group Domain Specific Review Board, which granted waiver of informed consent (reference number 2024-4055). The study uses de-identified data in a secure manner that would not pose risk to study subjects.

Sample Selection: Records were reviewed for individuals who had engaged in voyeuristic acts, either allegedly or confirmed through legal conviction, between January 2023 and December 2024. No age restrictions were applied. These subjects presented to the department through one or more of the following routes: remand assessments, suitability for Mandatory Treatment Order, or follow-up after their release from prison.

Data Collection: Data was extracted from medical records covering their demographics, offending patterns, relationship history, social factors, mental health diagnoses, substance use, criminal history, and treatment history. Data was de-identified by an independent party before analysis to ensure confidentiality.

All 30 subjects were male (median age 28.5 years), with 73.3% having tertiary education. Social isolation was prevalent (68.8% of those with available data), alongside significant family dysfunction. The median number of voyeuristic acts was 3.5 per individual. Offending patterns were quite evenly divided between upskirting and toilet/shower voyeurism. Mental health diagnoses were present in 73.3% of participants, with Voyeuristic Disorder being most common (23.3%). Antisocial traits were rare, while substance use was virtually absent.

Voyeuristic behaviour in this clinical sample was characterised by interpersonal dysfunction rather than traditional antisocial risk factors. The high educational attainment and low antisocial traits suggest this represents a distinct subgroup of help-seeking individuals. Treatment interventions should prioritise addressing underlying interpersonal difficulties and attachment dysfunction rather than focusing solely on traditional risk factors.

Rapid tranquillisation (RT) is widely used to manage acute agitation in psychiatric inpatient settings when non-pharmacological strategies fail. Older adults represent a particularly vulnerable population due to frailty, multimorbidity, polypharmacy, and age-related pharmacokinetic and pharmacodynamic changes, which increase the risk of adverse drug events.

Despite routine clinical use, most evidence regarding RT practice is derived from mixed-age adult populations, with limited age-specific guidance for elderly psychiatric inpatients. This systematic review aims to evaluate the evidence available and assess its applicability to older adults.

Aim was to systematically review and evaluate global evidence on the efficacy and safety of pharmacological rapid tranquillisation strategies in older psychiatric inpatient settings.

1. To compare the effectiveness of different RT pharmacological agents and combinations.

2. To evaluate reported adverse outcomes associated with different RT pharmacological agents.

3. To identify gaps in age-specific evidence, targeting older persons and international practice variations.

A systematic literature search of PubMed, EMBASE and related databases, Cochrane Library, grey literature and reference lists was conducted following PRISMA 2020 guidelines. Randomised control trials evaluating pharmacological RT agents for acute agitation in psychiatric or emergency settings were included.

Outcomes of interest included sedation efficacy, time to tranquillisation and adverse events. Risk of bias was assessed using the Cochrane RoB 2 tool. Due to inconsistent reporting of quantitative outcomes and a lack of elderly-specific data, a formal meta-analysis was not done.

Fifteen high-quality RCTs were identified, all involving mixed-age adult populations, none exclusively enrolled older psychiatric inpatients (>65 years).

Midazolam consistently demonstrated a faster onset of sedation compared with antipsychotic or benzodiazepine monotherapy. Combination regimens, particularly haloperidol plus promethazine or lorazepam, were more effective than monotherapy in achieving rapid tranquillisation and reducing the need for additional medication. Adverse events were generally infrequent; however, haloperidol monotherapy was associated with a higher rate of extrapyramidal symptoms, and midazolam with transient respiratory depression in some studies.

The absence of RCTs that are age-specific, targeting older persons, limits the direct applicability of existing evidence to older psychiatric inpatients. Current RT practicesfor older adults are largely extrapolated from general adult data, despite the heightened safety risks associated with this population. Inconsistent outcome reporting further restricts quantitative analysis.

There is a critical lack of elderly-specific evidence to guide rapid tranquillisation in psychiatric inpatient settings. Future research should prioritise age-stratified trials with standardised outcome reporting to inform safer, evidence-based practice.

To assess the association between access to supervised therapeutic leave during mandated inpatient substance use disorder rehabilitation and post-discharge relapse outcomes in a national cohort in Qatar. Specifically, the study examines time to relapse following discharge and explores early recovery stability during the transition from inpatient care to the community.

A retrospective cohort study was conducted using medical records of 72 patientsdischarged from a mandated residential rehabilitation service in 2023. Exposure was defined as receipt of any supervised therapeutic leave during admission. Relapse was defined as a positive urine drug screen during the 26-week follow-up period. Time-to-relapse was analysed using Kaplan–Meier methods, with between-group comparisons conducted using log-rank testing. Multivariable Cox proportional hazards modelling adjusted for demographic and clinical covariates, including comorbid personality disorder and forensic history.

Twenty-eight patients (38.9%) received supervised therapeutic leave, while 44 (61.1%) did not. Relapse occurred in 40.3% of the cohort, with a median time-to-relapse of 28 days. Patients granted therapeutic leave demonstrated significantly longer relapse-free survival than those without leave (log-rank p=0.034). During follow-up, relapse was observed in 25% of patients who received therapeutic leave compared with 50% among those without leave. After adjustment for covariates, supervised therapeutic leave was independently associated with a marked reduction in relapse hazard (hazard ratio 0.27; p=0.0088). Greater frequency of leave exposure was associated with lower relapse risk, suggesting a graded protective effect.

As one of the first evaluations of supervised therapeutic leave within a compulsory addiction treatment model in the region, these findings indicate that structured community leave may meaningfully enhance early recovery stability and delay relapse following discharge. Integrating carefully governed therapeutic leave into mandated treatment pathways may strengthen reintegration while maintaining clinical and public safety safeguards during the critical post-discharge period.

The NHS Long Term Plan advocates for ambient voice technology (AVT) tools to alleviate administrative burden and improve patient access. However, psychiatric assessment relies heavily on linguistic nuance in culturally diverse presentations. Moreover, accent diversity may affect AVT speech recognition. This raises concerns about inaccurate records and the widening of health inequalities.

This study aims to evaluate the accuracy of three NHS-approved AVT tools–Company A, B, and C–in summarising psychiatric consultations across diverse cultural presentations and accents.

We evaluated three AVTs using test scripts (n=3) simulating depression, dementia, and psychosis. The study included:

1. Cultural Test: Scripts adjusted for British, Nigerian, Pakistani, and Polish presentations using standardised controlled speakers.

2. Accent Test: Participants (n=9) with diverse UK and international accents reading identical scripts.

3. Output summaries were evaluated against a gold-standard checklist. Errors were classified as ‘missing/incomplete’, ‘incorrect/misleading’, or ‘hallucination’ (information not stated in the script).

Data are presented as mean accuracy [95% CI].

Cultural test: Company B demonstrated the highest stability: British 85% [83–87], Polish 85% [78–91], Nigerian 81% [74–88], and Pakistani 79% [72–86]. Company A displayed some variability, performing better on Nigerian 76% [69–82] and Pakistani 75% [63–86] scripts than on British 65% [43–87] and Polish 61% [34–89]. Company C consistently underperformed, ranging from British 58% [41–75] to Nigerian 56% [40–72].

Accent test: Performance was notably more stable. Company B remained stable across local (South England 84% [79–89]) and international (Spanish 80% [76–84]) speakers. Company A peaked in South England 82% [78–86] but degraded for Nigerian 69% [60–78]. Company C generally mirrored this stability (e.g. Romanian 78% [68–88]) but suffered a technical speech recognition failure for the South England accent (19% [14–23]).

Error distribution per tool: Missing/incomplete errors (>92%) dominated across all tools; 98.1% of Company A’s errors were missing/incomplete, with minimal hallucinations (1%). Company B and Company C showed higher rates of active error: Company C recorded the highest percentage of both hallucinations (2.3%) and incorrect/misleading data (5.4%), followed by Company B (1.7% and 4.7%, respectively).

No statistically significant difference was found in overall accuracy acrosscultures or accents; however, Company B produced the most consistent summaries, while Company C produced the least. Limitations include small sample size, limited representation, and simulation design. Although seemingly robust across cultures/accents, incomplete summaries highlight the need for human oversight in clinical use. Future work must assess these tools in larger, live clinical settings to ensure safety for diverse populations.

Growing research indicates that attention-deficit/hyperactivity disorder (ADHD) may occur at higher rates in individuals with functional neurological disorder (FND). However, reported prevalence differs substantially across studies and FND subtypes. This systematic review and meta-analysis sought to estimate the frequency of ADHD comorbidity within FND populations.

A systematic literature search of PubMed, Embase, Web of Science, PsycINFO, and Google Scholar was conducted through October 2025. Seventy-five studies fulfilled inclusion criteria. Information was extracted regarding study design, sample characteristics, diagnostic methods, and reported ADHD prevalence. Pooled prevalence estimates were calculated using random-effects meta-analytic models.

Across 63 studies involving 115,331 participants, the pooled prevalence of ADHD among individuals with FND was 15% (95% CI: 12–19; I²=92%). Subgroup analyses indicated ADHD rates of 14% (95% CI: 10–18) in functional seizures (22 studies, n=11,117), 6% (95% CI: 1–16) in functional movement disorder (4 studies, n=166), and 27% (95% CI: 19–35) in functional tic-like behaviours (21 studies, n=1057). In studies limited to paediatric populations (33 studies, n=2141), the pooled prevalence was 16% (95% CI: 11–22).

Approximately one in six individuals with FND also meet criteria for ADHD, with particularly elevated rates observed in functional tic-like presentations and in children and adolescents. Considerable heterogeneity was present across studies. These findings underscore the importance of evaluating for ADHD during FND assessment, especially when difficulties with attention regulation or impulse control are prominent.

ADHD drivers have up to four times higher risk of being involved in road traffic accidents than other drivers. Symptom severity has some influence on risk but comorbidity, especially for antisocial behaviours and alcohol misuse are strong associations with risk. It seems that ADHD drivers engage in more distracting tasks while driving and there is also an increased risk of sleep-related accidents. ADHD drivers tend to drive faster and it has been suggested that this is a strategy to escape the boredom.

This study tests the hypothesis that those with ADHD choose faster cars and that this may be another factor explaining this group’s higher accident risk. In addition, since those diagnosed with ADHD in childhood are more likely to take up motor sport, which is associated with increased risk of road collisions, this study also looked for an association of ADHD with engagement with motorsport in adult life.

Participants, all employees at an NHS mental health trust, were invited to fill out an electronic questionnaire. A link was included within the weekly Trust newsletter. Questions covered demographic details, ADHD screening (ASRS), depression and anxiety screening (PHQ-4), two questions asking preference, if buying a car, between luxury and sports variants and three questions about motorsport interest. Research department advice, to use the NHS Health Research Authority’s “Is my study research?” decision tool, determined that the survey was not classified as research as results may not be generalisable. Power calculations suggested minimum n of 121.

In total 114 replies were received, 91 were female with a wide spread of age groups. Ninety-eight scored less than 6 on PHQ-4. Of these 27 scored on 4 or more items on ASRS, who were classified as ADHD group. More of the non-ADHD group chose the faster models, in fact for one of the two 31% chose the sportier version compared with only 7.4% of the ADHD group, p<0.05 (Chi squared). There was little difference in motorsport interest between groups. Of the 16 with higher PHQ-4 scores, 62% scored above ADHD cut-off.

Contrary to the hypothesis, reasons why the ADHD group seemed to choose the more sedate cars are discussed in terms of the intriguing possibility that strategies adopted enabling management of responsible employment may influence other aspects of living. The importance of removing those with higher anxiety and depression scores from analysis is emphasised.

Legal herbal supplements are increasingly used as “natural” alternatives for mental health issues. Nymphaea caerulea (Blue Lotus) is promoted online as a relaxant, though its neuropsychiatric effects remain poorly understood. This short review aims to examine the available literature onNymphaea caerulea(Blue Lotus), with a focus on its pharmacology, psychoactive and neuropsychiatric properties, and to describe illustrative examples from clinical practice. The review also seeks to raise awareness of potential associations between new-onset psychotic presentations and legal herbal supplements, and identify gaps in the evidence base requiring further research and regulation.

A narrative literature search was conducted in PubMed, Google Scholar, and grey literature up to 2025 using the terms “Nymphaea caerulea”, “Blue Lotus”, “herbal psychoactive”, and “psychiatric symptoms”. Pharmacological sources and regulatory reports were also reviewed. Two brief anonymised clinical vignettes from clinical practice were also included to illustrate real-world psychiatric presentations temporally associated with Blue Lotus use.

The results of the search revealed fewer than ten publications, with only one peer-reviewed case series, which described five individuals presenting with agitation, confusion, derealisation, and anxiety following ingestion or vaping. Routine toxicology screening was negative in all cases. Two brief anonymised clinical vignettes from recent clinical practice were also included, illustrating hypomanic and psychotic presentations following Blue Lotus use.

Regulatory alerts from the UK Office for Product Safety and Standards have raised concerns around the safety of Nymphaea caerulea (Blue Lotus), particularly the lack of information on dosages, variability in psychoactive content, and concerns regarding purity standards. Pharmacological reports have described the main active compounds, including nuciferine and aporphine, which act on dopaminergic and serotonergic receptors and may induce both calming and euphoric effects. The concentration of these alkaloids may vary depending on the plant, extraction method, and storage, resulting in unpredictable clinical effects. Furthermore, nuciferine has been identified in Blue Lotus resin used in electronic-cigarette devices, confirming the presence of psychoactive compounds in products marketed for inhalation.

Nymphaea caerulea is a legally obtainable, yet psychoactive herbal supplementthat may provoke neuropsychiatric symptoms. Clinicians should inquire about herbal/“natural” supplement use in new-onset psychiatric presentations. Regulation and further research are needed.

A number of subtypes of obsessive-compulsive disorder (OCD) have been described, one of which is religious OCD (ROCD), also referred to as Scrupulosity Disorder. Whilst there is an expanding literature which investigates the link between OCD and religion, there is very little research which examines the impact of clergy attitudes and intervention upon a sufferer of ROCD. The study aims to begin the task of developing an understanding of how those in Christian pastoral ministry in the UK typically view ROCD, how they are likely to deal with it in someone for whom they have pastoral care, and whether such interventionsare compatible with current models of clinical care.

The study involved conducting in-depth interviews with six respondents working professionally in pastoral ministry. The respondents were interviewed using a questionnaire comprising a series of open questions. The initial phase of the interview was designed toencourage discussion as to how a respondent might react to someone with typical ROCD symptoms, having had such explained to them. The nature of ROCD as a psychiatric diagnosis was then explained, including some cognitive theory and a description of typical treatment. The remainder of the interview concerned itself with exploring respondents’ views on these matters.

There was general consistency of thought between correspondents; all were willing to give credence to a psychopathological attribution of ROCD and the implication that a psychological approach is useful in its management. All were clear that an empathetic and sensitive approach is necessary and that a response which conveys a sense of judgement is to be avoided. Recommendations were generated which would broadly be viewed as psycho education from a clinical standpoint: reflection on the nature of God; an explanation that obsessions do not confer culpability nor are they indicative of an underlying spiritual or moral deficit; an exploration of the dissonance created by the sufferer’s unsuccessful attempts to resist compulsive behaviour.

The principal source of conflict between pastoral care and treatment occurs when a minister seeks to offer reassurance to combat the doubts and anxieties of a sufferer. Conversely, respondents’ recommendations are clearly well suited to pastoral intervention, being of a theological nature. It is here, with the ‘theological examination’ of a sufferer’s negative appraisal of obsessions, that the minister might best use the authority invested in him/her, rather than by the giving of permission to follow the edicts of the therapist.

Patients discharged from inpatient mental health services can experience multiple adverse outcomes, including a 191-fold higher suicide risk within three months of discharge from inpatient mental health services. The SAFER-MH intervention was co-designed to improve quality and safety in this vulnerable period. Preliminary research suggests it is feasible, acceptable, and offers vital support during inpatient and pre-discharge phases. However, its use is limited to secondary care. This study aimed to strengthen SAFER-MH by co-producing a community follow-up element of the intervention to enhance continuity of care post-discharge.

A three-phase sequential design was used. Phase 1 involved focus groups with patients, carers, and professionals from primary and community care to explore current practices. Phase 2 used a Nominal Group Technique (NGT) workshop to generate and prioritise feasible solutions. Phase 3 consisted of a co-design workshop with the stakeholder groups described above to collaboratively refine and assess the feasibility of proposed interventions.

Seventeen participants took part across three workshops: Phase 1 included fourindividuals from three abovementioned stakeholder groups; Workshops 2 and 3 added three inpatient professionals and two replacements for those who withdrew. Six themes emerged in Phase 1, including fragmented communication and coordination between services, inadequate discharge preparedness, service limitations and inequalities, patient/carer related barriers, quality and safety concerns and pharmacological care challenges. The NGT workshop generated 33 potential solutions, with 17 receiving votes. Top priorities included a one-stop hub and peer support, assuming no resource or implementation constraints. In Phase 3, participants focused on which elements would be feasible as part of SAFER-MH and the top three interventions included adopting a true multidisciplinary team approach to care, improving medication follow-up, and devising a community peer support element.

This study highlighted shared problems faced by all stakeholder groups post-discharge centring around communication breakdowns, medication challenges, and limited patient involvement. Integrating multidisciplinary teamwork, medication follow-up, and peer support within SAFER-MH is feasible and could enhance continuity and safety during transitions from inpatient to community care.

This study was co-designed and co-produced with patient and public contributors and key stakeholders throughout all stages, resulting in a strengthened SAFER-MH intervention that is now ready for evaluation in a national randomised controlled trial.

Due to their higher acquisition costs, long-acting injectable (LAI) antipsychotics are often seen as more expensive substitutes for oral antipsychotics (OAs). However, Schizophrenia is a chronic and relapsing- remitting illness; non-compliance to oral therapy frequently results in hospitalisation and emergency care, which are significant contributors to overall healthcare expenses. Therefore, rather than concentrating solely on pharmacies, evaluating treatment value necessitates a total cost of care (TCOC) perspective.

Aims were to determine, using economic data, whether the higher costs of LAI antipsychotics result in an overall increased healthcare cost when compared to oral antipsychotics in adults with Schizophrenia.

Our data sources for the study included administrative claims databases (Medicaid and commercial), cost-effectiveness models, mirror-image (pre-post) studies, national health system cohorts, systematic reviews and meta-analyses. We conducted a structured narrative evidence synthesis from the above data.

We focused on real-world economic and pharmacoeconomic studies comparing LAI antipsychotics with OAs. The total healthcare costs and/or component costs (pharmacy, inpatient, emergency department, and outpatient) were reported in the studies. Results were analysed directionally rather than pooling them statistically, due to heterogeneity in settings,currencies, and time horizons.

It was noted that across US, European, and Asian studies, LAIs consistently demonstrated lower inpatient and emergency department costs but higher pharmacy costs in comparison to OA’s.

Several pre-post analyses revealed a reduction in hospitalisation costs after LAI initiation. However, few claims-based and meta-analytic studies reported no statistically significant difference in total healthcare costs between LAIs and oral therapy.

Cost-effectiveness models indicated that LAIs may be cost-effective over multi-year horizons, despite similar or slightly higher total costs, driven by improved compliance and reduced relapse rates.

The findings demonstrate a robust “cost-offset effect”, in which the higher upfront medication costs of LAIs are balanced by the downstream savings from avoided hospitalisations.

The generalisability of this pattern for stakeholders and policy/decision-makers was supported by the fact that it was consistent across healthcare systems, study designs, and LAI molecules.

LAI antipsychotics do not significantly raise overall healthcare costs and may be economically beneficial by lowering expensive relapse-related care, though they may appear to be raising pharmacy expenditures initially. Treatment evaluations for schizophrenia should focus on the overall cost of care rather than just the cost of the medication.

Recently the therapies that utilize psychedelics have emerged as promising treatments for mood disorders, trauma-related disorders, and substance use disorders. While there is a growing body of research comparing various interventions directly, the majority of available research has been conducted in an indirect manner; therefore, Network Meta-Analysis (NMA), a statistical method for comparing the efficacy of multiple interventions at once, is well-suited to evaluate the available evidence and create a comparison of relative efficacy among the various psychedelic assisted interventions. This study aims to compare the efficacy of psychedelic-assisted interventions and provide a comparative efficacy ranking of each treatment modality using a Network Meta Analytic approach.

A contrast-based NMA was used to synthesize randomized and controlled trials of psychedelic-assisted interventions. Each treatment node was created based upon the type of drug being utilized (psilocybin, ketamine, MDMA) and the specific therapeutic context (psychotherapy) along with comparator treatments (escitalopram, active placebo, and waitlist). Means and standard deviations (SD) of continuous outcomes were obtained, standardized (as Hedges g) and incorporated into a fixed-effect inverse-variance weightedmodel. Trials with crossover designs contributed their data prior to crossover. Trials that reported binary or time-to-event outcomes were incorporated into the geometric structure of the network. Surface Under the Cumulative Ranking Curve (SUCRA) was used to estimate rankings of each treatment.

A total of six studies comprised the network and all but one study had psilocybin-assisted psychotherapy as the focus of the study. All studies demonstrated that psychedelic-assisted treatments were more effective than control groups (pharmacological, active placebo, and waitlist). Study-level estimates of treatment effect favoured psilocybin (range=−0.6 to −1.4) when compared to all other treatments. Psilocybin at high doses was found to be more effective than low dose psilocybin in reducing symptoms. A large and statistically significant pooled summary effect favouring psychedelic-assisted therapy was observed (−0.89, 95% CI: −1.07 to −0.72). Using SUCRA, psilocybin-assisted psychotherapy ranked highest among all treatments examined and both MDMA-assisted psychotherapy and ketamine-assisted therapy ranked higher than all control condition treatments.

In this network meta-analysis we present encouraging quantitative evidence supporting the clinical efficacy of psychedelic-assisted therapies, particularly psilocybin-assisted psychotherapy. These findings support the potential therapeutic utility of these treatments and suggest the need for additional head-to-head trials to provide more precise comparative effectiveness estimates and develop future clinical practice guidelines.

There is increasing evidence that suggests that different subtypes of major depressive disorders, particularly treatment-resistant depression and anhedonia-predominant forms, have dysregulation in reward processing and chronic neuroimmune activation. Low-dose naltrexone (LDN), which is administered at low doses (1–5 mg/day), modifies both the signalling from the body’s endogenous system and the inflammatory pathways of microglia and thus can be used as a potential therapeutic for depression. This mechanistic systematic review aims to evaluate the dual mechanism of Low Dose Naltrexone in Treatment-Resistant Depression. This study aims to provide an analysis of the human clinical and biological effects of low dose naltrexone (LDN) in depression based upon a mechanistic convergence framework.

Human studies examining the effects of LDN on outcomes such as mood, fatigue, pain, quality of life, inflammatory markers, or opioid-related variables, published in PubMed, EMBASE, Scopus, Web of Science, the Cochrane Library, or Google Scholar, wereidentified. The quality of randomized controlled trials was evaluated according to the Risk of Bias 2 tool, and the quality of non-randomized studies was evaluated according to the Risk Of Bias In Non-randomized Studies–Instrument (ROBINS-I). Results from studies of human subjects were synthesized narratively using a mechanism-based approach.

Improvements in quality of life, fatigue, and mood-related symptoms in disorders characterized by immune activation and central sensitization (e.g. fibromyalgia, multiple sclerosis) were observed across randomized trials, observational cohorts, and case reports after administration of LDN. Biomarkers of inflammation (pro-inflammatory cytokine levels) were lower in human studies after administration of LDN. Indirect indicators of enhancement of endogenous opioid function (e.g. improved pain tolerance, decreased consumption of analgesics and psychotropics) were also reported in human studies after administration of LDN. The mechanisms by which LDN exerts these beneficial effects occur in the same patient population and therefore suggest a biological convergence of immune suppression and opioid-mediated reward modulation.

The results from the literature support a “Dual Mechanistic Model” where LDN suppresses inflammation by activated microglia and at the same time increases signalling of endogenous opioids, both of which have been implicated in the pathophysiology of depression and anhedonia. Thus there is a strong mechanistic basis for testing LDN as a treatment for patients who have failed to respond to current treatments for depression.

There is growing evidence that certain subtypes of major depressive disorder, especially treatment-resistant depression and anhedonia-predominant phenotypes, are marked by both chronic neuroimmune activation and disruptions in reward processing circuitry. These results cast doubt on purely monoaminergic models of depression and highlight the need to comprehend how psychotherapeutic interventions may work through biological systems controlling immune function, stress regulation, and plasticity rather than just suppressing symptoms. Some therapies like Cognitive behavioural therapy, Mindfulness based and Trauma focused therapies have been shown to bring positive neuro immune, autonomic and epigenetic changes in brain.With an emphasis on finding common neurobiological pathways and potential biomarkers of treatment response across psychotherapy modalities, this review attempts to outline the biological correlates and mechanisms of action of psychotherapy.

To find systematic reviews, meta-analyses, and important longitudinal studies looking at biological changes linked to psychotherapy, a targeted systematic search ofMEDLINE/PubMed and major online publishing platforms was carried out. Studies that used biological measurements at two time points (pre- and post-intervention) were eligible. Four predetermined domains were used to classify the extracted biological outcomes:

1. Neuroimaging techniques, such as task-based and resting-state functional neuroimaging;

2. Physiological and autonomic measures of stress management, especially heart rate variability;

3. Indicators of inflammatory and immune function;

4. DNA methylation alterations linked to treatment response are the main focus of epigenetic markers.

People receiving Cognitive behavioural therapy (CBT) for depression consistently show changes in the fronto-cingulate–limbic circuitry, specifically in the medial prefrontal cortex and ventral anterior cingulate cortex (vACC), according to longitudinal functional magnetic resonance imaging (fMRI) studies. Standardized measures like the Beck Depression Inventory (BDI) show a strong correlation between these brain alterations and clinical improvement. Additionally, convergent reductions in limbic reactivity–most notably decreased right amygdala activation–across a variety of successful depression treatments are suggested by coordinate-based meta-analytic evidence, supporting the existence of shared neural pathways of symptom improvement.

Autonomic biomarkers have also emerged as promising predictors of the effectiveness of psychotherapy, particularly in the treatment of post-traumatic stress disorder (PTSD). Higher levels of high-frequency heart rate variability (HF-HRV) under basal conditions, associated with parasympathetic tone, have been related to the magnitudes of improvements following psychological therapy including CBT.

Available evidence for immune-related consequences is inconsistent. On the one hand, trials of mindfulness-based interventions show consistently that mindfulness may have limited and variable effects on inflammation with extensive methodology variations between studies that use similar populations, biological markers used for immune evaluation, and interventions. There is some promising evidence that trauma-focused psychotherapy may show differences in longitudinal changes in peripheral DNA methylation between responders and non-responders to trauma-focused psychotherapy while targeting immune function, stress pathways, and neuroplasticity pathways.

Within psychotherapeutic approaches, the strongest consistently accumulated biological evidence points to two related psychological constructs:

(i) the fronto-cingulate–limbic neural networks that are involved in emotion regulation and self-reference;

(ii) regulation of stress and immune system functions.

However, the existing evidence base is limited by small sample size, heterogeneous experimental designs and biomarker measurements, and the timing of the biological measurements. Next generation randomized controlled trials of psychotherapy should employ harmonized multimodal protocols of biomarkers to examine the longitudinal patterns of the causal mechanisms and to elucidate the clinically relevant biomarkers of the response to the treatment

Large language models (LLMs) are increasingly proposed to support healthcare delivery, yet their ability to address the distinctive challenges of dementia care remains unclear. Dementia is characterised by progressive cognitive impairment, communication barriers, high caregiver burden, and substantial demands on memory services and old age psychiatry. We systematically reviewed and classified empirical applications of LLMs in dementia assessment and care, and critically appraised the maturity of evidence, intended users, and reporting of safety, fairness, and governance relevant to clinical implementation.

We searched EMBASE, MEDLINE, PsycINFO, PubMed, the ACL Anthology, the ACM Digital Library, arXiv, medRxiv, and bioRxiv (2017–June 2025). Eligible studies were empirical evaluations (simulation-based or experimental) of LLMs used for dementia-relevant clinical, caregiver-facing, or research tasks; non-evaluative commentaries were excluded. Titles/abstracts were screened by multiple reviewers, with full-text screeningconducted in accordance with PRISMA; disagreements were resolved by discussion. Using a standardised extraction form, we recorded study design, model attributes, input modalities, dementia-related use cases, evaluation setting (benchmark/vignette vs clinical data), outcomes, and reporting of ethics approval, bias assessment, and hallucination-related issues. Findings were synthesised narratively and grouped using a predefined taxonomy of clinical, patient-facing, and research-oriented applications.

We identified 35 dementia-focused studies, several covering multiple applications. The most frequently reported functions were screening/early detection (26/35) and diagnostic support (28/35), typically framed as classification tasks. Other common applications included research facilitation (19/35), evaluation/outcome measurement (19/35), information extraction (14/35), and prediction/risk stratification (13/35). Multimodal approaches were reported in (14/35). Intended users were most often researchers (22/35) and clinicians (20/35), with fewer studies targeting people living with dementia (10/35) or caregivers. Inputs included speech/audio (20/35) and free text (18/35); outputs most commonly included classification labels (26/35) and risk scores (11/35). Performance metrics were reported in 18/35. Reporting of safeguards was inconsistent: ethics approval (10/35, 29%), bias assessment (10/35, 29%), and explicit hallucination reporting (5/35, 14%).

LLMs show promising experimental performance for dementia-related classification and early detection tasks, suggesting potential roles in augmenting assessment workflows. However, evidence is largely preclinical, with limited real-world validation and inconsistent reporting of fairness, safety, and governance. Future research should prioritise prospective evaluation in memory services, robust bias and hallucination assessment, and co-designed implementation with people living with dementia and caregivers to support safe, equitable integration in psychiatric practice.

This study evaluated the efficacy of an immersive Acceptance and Commitment Therapy (ACT)-based intervention in reducing subclinical depressive, anxiety, and obsessive–compulsive symptoms in medical students at Cardiff University.

A quantitative within-subjects repeated-measures design was employed using a single-group pretest–post-test approach. Participants completed baseline measures of obsessive–compulsive (OCI-R), anxiety (GAD-7), and depressive (PHQ-9) symptoms prior to engaging with the immersive ACT intervention, and the same measures were repeated post-intervention. Participants were recruited via the medical school’s student support unit and through advertising across all years of the undergraduate medical programme.

Seventeen participants completed all study components. No significant pre–postdifferences were observed on the total OCI-R, GAD-7, or PHQ-9 scores. However, subscale analyses revealed a significant reduction in the neutralising dimension of the OCI-R following the intervention (z=−1.941, p <0.05). Post-intervention differences were also identified between participants with high versus low depressive symptoms (PHQ-9), specifically in obsessional symptoms (z=−2.298, p <0.05). In addition, participants with high versus low anxiety scores (GAD-7) demonstrated a significant reduction in ordering symptoms post-intervention (z=−2.179, p=0.029).

Although no significant changes were observed on overall measures of obsessive–compulsive, anxiety, or depressive symptoms, targeted reductions were evident in specific symptom dimensions following the immersive ACT-based intervention. These findings suggest potential symptom-specific benefits for medical students with elevated obsessive–compulsive, depressive, and anxiety traits. While limited by a small, subclinical sample, the results support further investigation of immersive ACT interventions in larger and clinical populations.

The representativeness of study populations relative to the clinical populations they aim to reflect remains contested. Evidence suggests that some research cohort populations present with less severe symptoms compared to clinical populations. This phenomenon may also occur in studies investigating motivational impairment within schizophrenia, whereby recruited patients have less severe motivational impairment than typical patients. Thus, findings of studies investigating motivational impairments in schizophrenia may not fully generalise to the wider patient population.

We aim to assess the representativeness of participants diagnosed with treatment-resistant schizophrenia who volunteered for a study on motivation compared with patients attending the Cambridge Clozapine Clinic who did not volunteer for that study, across demographics and clinical outcomes, with a specific focus on motivation and pleasure.

The CHANSS (Characterising negative symptoms in schizophrenia) is a multi-centre international study aiming to characterise the cognitive mechanisms underlying motivational impairment within the negative symptoms of schizophrenia. Data used were collected from the CHANSS study (n=69 clozapine-treated) and from the Cambridge Clozapine Clinic, included in the ethically approved database (n=244 post duplicate removal). Statistical analysis was conducted via R to assess the difference in population demographics (Age, sex, family history, smoking, dose of clozapine) and outcomes (Outcomes were measured using the Positive and Negative Scale (PANSS), Brief Negative Symptom Scale (BNSS), Calgary Depression Scale (CDS) and Global Assessment of Functioning (GAF) Scale). BNSS-MAP, a measure of apathy, was the primary variable studied.

Clinical outcomes followed a non-normal distribution with unequal variance and high levels of heterogeneity. BNSS overall motivation and pleasure score (MAP), as well as its subcomponents of anhedonia, asociality and avolition, showed no significant difference. CHANSS participants taking clozapine had a significantly younger age. The CHANSS cohort showed significantly higher scores across several other symptom measures. CHANSS participants were more likely to have a higher positive score for positive symptom severity, expression, anhedonia, blunting and CDS total score. Power calculations for all significant results returned a high power above (all >0.9).

For the primary outcome of differences in motivation and pleasure, no significant differences occurred between the clozapine-treated CHANSS participants and the Cambridge clozapine clinic cohort. Thus, for studying motivational impairment, the CHANSS Study was successful in recruiting a study population representative of the wider patient population. However, demographic differences and higher symptom scores in domains beyond motivation and pleasure should be considered when interpreting any findings beyond motivational impairment.

South Asia bears a substantial burden of developmental and psychosocial adversity among children and adolescents, yet specialist mental health resources remain scarce. Scalable, non-specialist delivered interventions have been proposed as a solution, but the breadth, characteristics, and evidentiary foundations of such interventions across developmental stages remain poorly synthesised.

This review aimed to systematically examine the existing evidence on child-centred interventions delivered by non-specialists in South Asia, including whether such interventions demonstrate successful outcomes, are predominantly preventive or promotive in focus, are developmentally targeted, and are implemented with acceptable participant retention across diverse settings.

A scoping review was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidance (PRISMA). Web of Science and PubMed were searched for peer-reviewed intervention studies published from 2010 onwards. Eligible studies were clinical trials with a comparison group, conducted in South Asia, targeting infants, children, or adolescents up to 19 years, and aiming to improve developmental or wellbeing outcomes through prevention, promotion, or treatment approaches. Pharmacological studies, qualitative designs, and samples with high baseline psychiatric morbidity were excluded. Screening, full-text review, and data extraction were undertaken independently by multiple reviewers with consensus resolution.

Forty-seven studies met inclusion criteria, encompassing 57,951 participants at baseline and 49,321 at endline, with an overall attrition rate of approximately 15%. Nearly half of studies were conducted in India, followed by Bangladesh and Pakistan. Most interventions were randomised or cluster randomised trials and were delivered in schools, homes, or community settings. Non-specialists delivered 87% of interventions, including health workers, lay counsellors, peers, and teachers. Interventions predominantly focused on promotion (68%) or prevention, with only two treatment-focused trials.

Perinatal and early childhood interventions predominantly emphasised developmental stimulation (n=21), parenting practices (n=18), psychoeducation related to child health, development, nutrition, or maternal wellbeing (n=19), and responsive caregiving or maternal responsiveness (n=11). In contrast, adolescent interventions primarily targeted psychosocial wellbeing (n=8), positive youth development approaches (n=5), life skills, stress management, or resilience-focused outcomes (n=multiple studies; exact counts varied by domain).

Cognitive behavioural components were present in a minority of interventions (n=7 overall), most commonly embedded within multi-component programmes rather than delivered as stand-alone therapies. The use of standardised developmental and psychosocial outcome measures was common, with the Bayley Scales of Infant and Toddler Development (Second edition: n=6; Third edition: n=11) most frequently used for early childhood outcomes, and the Strengths and Difficulties Questionnaire (n=7) most commonly employed for child and adolescent psychosocial wellbeing.

Over the past decade, South Asia has generated a substantial and methodologically robust body of evidence demonstrating the feasibility and acceptability ofnon-specialist delivered child and adolescent developmental interventions across settings and developmental stages. However, the literature remains heavily weighted toward promotion and prevention, with notable gaps in middle childhood and treatment-focused interventions. Future research should prioritise theoretically explicit, mechanism-informed designs and address under-represented age groups to inform scalable, equitable child mental health strategies in low-resource settings.