Suicide is a leading cause of death among young adults, yet few studies have examined longitudinal trajectories of suicidal ideation and attempts across the academic year in UK university students.

The study aimed to identify demographic and psychosocial factors associated with lifetime suicide risk, examine factors associated with changes in risk trajectories, and evaluate the relationship between suicidal ideation and attempts over an academic year.

Participants comprised 3295 students from 5 UK universities who completed longitudinal surveys in autumn and spring terms 2021–2024 in the Nurture-U study. Suicidal ideation and attempts were assessed with Columbia Suicide Severity Rating Scale items. Demographic, psychosocial and mental health measures were examined as potential risk and protective factors. Regression models with multiple imputation were used, with case–control analyses for smaller subgroups.

Approximately 45.5% of students reported baseline lifetime suicidal ideation and 9.5% reported a past attempt. Factors associated with lifetime suicidal ideation included minoritised sexual orientation or ethnic background, male or gender-diverse identity, previous mental health treatment and adverse childhood experiences. Among students with lifetime ideation at baseline, 61.1% did not report ideation in the spring term, whereas new-onset ideation occurred in 3.5% of participants. Among students with prior ideation only, 1.3% made a suicide attempt during the academic year, whereas 10.5% with prior attempt history made a suicide attempt. Increases in depressive symptoms and rumination were associated with new attempts at follow-up.

Most students experiencing suicidal ideation during university have risk factors before university entry. Suicidal ideation alone showed limited specificity for identifying those most at risk of future attempts, highlighting the importance of targeted support for high-risk groups and ongoing monitoring of symptoms throughout the academic year.

Healthcare can be a demanding career, and there is a rising interest in fostering the well-being of healthcare workers (HCWs).

To examine self-reported use of mental health and well-being support available to HCWs and explore predictors of use.

Data were collected by a longitudinal cohort study. The baseline survey was distributed from April 2020, followed by three surveys at 6 months (time 1; n = 10 514), 12 months (time 2; n = 10 429) and 32 months (time 3; n = 6382) post baseline. Use of mental health support (e.g. therapy) was assessed at each follow-up, and 96 supports were measured at time 1 only. Multilevel regression modelling was used to explore baseline predictors.

At time 1, we developed seven support categories. Practical supports (e.g. free or discounted food or parking) were the most commonly used (58%), whereas employee assistance programmes were least commonly used (5%). Furthermore, 21% of HCWs reported never using any support. Mental health support access varied: 23% of participants had used support before August 2021 (time 1), 17% between March 2020 and January 2022 (time 2), and 21% between February 2022 and May 2023 (time 3). HCWs who were male, older and/or of Asian ethnicity had lower odds of accessing support, whereas those who reported symptoms of common mental disorders at baseline had higher odds.

Most HCWs (79%) reported having accessed some support, whereas 21% had not. Practical and informal supports were most used, whereas uptake of employee assistance programmes was low. Future research should explore and address barriers faced by male, older, Asian and non-permanent staff to improve access and, in turn, the well-being, performance and patient care of these individuals.

Emerging evidence suggests that attention-deficit hyperactivity disorder (ADHD) may be associated with increased likelihood of experiencing interpersonal harm, including from intimate partners.

This study aims to test the long-term associations between ADHD traits in childhood and experiences of intimate partner violence (IPV) up to midlife (age 51 years).

Data were from the 1970 British Cohort Study, a nationally representative longitudinal study of people born during 1970. ADHD traits were assessed at age 10 years. Lifetime experience of five types of IPV was measured at age 51 years (harassment, economic abuse, physical violence, threats/coercive control, sexual violence). Regression models were used to test associations between childhood ADHD traits and midlife IPV, adjusting for gender, ethnicity and childhood social class.

Adjusted models showed that ADHD traits in childhood were associated with higher risk of experiencing any form of IPV, a greater number of different IPV experiences and all individual subtypes of IPV by age 51 years. Post-estimation showed that around one in seven men and one in four women without high ADHD traits experience some form of IPV by midlife. In people with high ADHD traits (4.40%), these estimates increase to approximately one in five men and one in three women.

Higher levels of childhood ADHD traits are associated with an increased risk of experiencing IPV by middle age. These findings highlight that ADHD may be associated with increased vulnerability to interpersonal harm in relationships. Identifying people with ADHD at greatest risk, as well as addressing the needs of people with ADHD within IPV prevention and response systems is important for reducing harm and addressing inequalities faced by people with ADHD over the life course.

Receipt of violence (victimisation) is associated with a higher risk of severe mental illness and with worse illness severity; however, associations with routine clinical outcomes are less clear.

To investigate associations between victimisation and subsequent clinical outcomes in people presenting with severe mental illness, using a south London health records data resource.

Data were extracted for 16 372 patients presenting between January 2007 and October 2022, aged 18+ years and receiving diagnoses of schizophrenia-related disorders, mania or bipolar disorder. Recorded victimisation and subtype (any, physical, domestic or sexual) were ascertained via natural language processing from the first 3 months of the record following presentation. Cox regression models were deployed to investigate subsequent risk of mental health emergency care assessment, crisis intervention, in-patient care, Mental Health Act (MHA) detention and mortality. Poisson regression models investigated the numbers of attended healthcare events and antipsychotic agents received. Covariates included sociodemographic characteristics and clinical status (service receipt, medication, symptom profile).

All victimisation exposures were associated with a higher risk of all outcomes, apart from mortality and number of antipsychotics, following adjustment for sociodemographic factors, and with emergency assessment (hazard ratio for any victimisation 1.17, 95% CI 1.09–1.27), in-patient/MHA detention (hazard ratio 1.32, 95% CI 1.06–1.65/1.82, 95% CI 1.31–2.55) and higher numbers of attended events (incidence rate ratio 1.17, 95% CI 1.16–1.19) following full adjustment. Associations were similar for subtypes of victimisation and exposures, and between men and women following sociodemographic adjustment, but the associations were slightly weaker for women following full adjustment.

Experiences of violence, to the extent to which patients report their experiences and these are recorded, are risk factors for worse outcomes in severe mental illness, only partly accounted for by clinical status around the time of presentation. More systematic ascertainment and recording of victimisation needs to be considered if interventions are to be appropriately targeted.

Adults with intellectual disabilities who display behaviours that challenge (BtC) are more prone to poor health.

This study seeks to evidence the long-term health outcomes for those who display BtC.

We conducted a longitudinal cohort study of adults with intellectual disabilities aged ≥18 years in England, using data from the Clinical Practice Research Datalink Aurum (January 2003 to December 2023) linked to Hospital Episode Statistics and Office for National Statistics. Main outcome measures were annual health checks, general practitioner referrals, emergency visits, out-patient attendance, in-patient admissions and mortality.

Among 83 166 adults with intellectual disabilities (mean age 38.6 years), 18.5% had a record of BtC, with similar sociodemographic distributions to those without BtC, but higher rates of physical and mental health comorbidities and uptake of annual health checks. A total of 72.5% of participants with BtC were receiving psychotropic medication(s). Adults with BtC had higher rates of mental health out-patient attendance (odds ratio: 1.42, 95% CI 1.33–1.52) and in-patient admissions (incidence rate ratio (IRR): 1.19, 95% CI 1.09–1.29), but consistently lower rates of physical health out-patient attendance (IRR = 0.81, 95% CI 0.78–0.84) and in-patient admissions (IRR = 0.77, 95% CI 0.74–0.79), after adjusting for demographic and clinical characteristics. BtC were not associated with mortality after adjustment for comorbidities (hazard ratio: 0.97, 95% CI 0.93–1.00).

This longitudinal study not only corroborated the markedly elevated burden of physical and mental health comorbidities among individuals displaying BtC, but also indicated that repeated efforts to improve health outcomes have yielded minimal measurable benefit over time. The apparent absence of progress is likely underpinned by a combination of insufficiently effective or poorly tailored interventions and wider systemic constraints that limit the capacity of services to respond to the complex needs of this population.

Late-life depression (LLD) is associated with cognitive impairment and an elevated risk of dementia, yet the influence of age at depression onset on cognitive prognosis remains unclear. Emerging evidence suggests that late-onset depression, defined as first depressive episode in later adulthood, may reflect distinct neuropathological mechanisms and predict more severe cognitive decline and greater dementia risk than early-onset depression.

This study aimed to investigate whether late-onset depression is linked to domain-specific cognitive impairment and higher risk of incident dementia among older adults with major depressive disorder.

We analysed UK Biobank data from older adults (aged ≥60 years) with primary care linkage, classifying participants into depression-free controls, early-life depression, late-life depression with early onset (LLD-EO) and late-life depression with late onset (LLD-LO). Cognitive performance across these five domains was assessed cross-sectionally at baseline using touchscreen tasks. Incident dementia was evaluated prospectively using clinical records up to 2022. Multi-level models with inverse-probability weighting and survey-adjusted mixed modelling were applied to assess group differences in cognitive function, controlling for demographic covariates, lifestyle factors and physical and mental health conditions. A Cox regression model was employed to estimate dementia risk among groups.

Among 75 064 participants aged ≥60 years, the LLD-LO group (n = 4858) showed significantly worse cognitive performance than healthy controls, particularly on fluid intelligence and visuospatial memory. The LLD-LO group performed worse than LLD-EO on fluid intelligence. During follow-up, LLD-LO was associated with a higher risk of incident dementia (hazard ratio 1.42–1.52) across all adjusted models. Deficits in fluid intelligence and visuospatial memory partially mediated the link between LLD-LO and subsequent dementia.

Late-onset depression showed more severe impairment in fluid intelligence compared with LLD-EO. Late-onset depression was associated with increased incident dementia compared with depression-free individuals.

Social connections might be protective against depressive and anxious symptoms and dementia in later life. The extent to which social connections are heritable versus modifiable in older age remains unknown.

We aimed to investigate the heritability of social connections and their influence on mental and cognitive health over time among older adults in a longitudinal cohort.

We analysed data from the Older Australian Twins Study (333 monozygotic, 266 dizygotic twins; 65+ years) at three time-points over 6 years. We examined the factor structure and heritability of baseline social connections and their associations with mental and cognitive health longitudinally.

We found three weakly heritable social connections factors: (a) interacting with friends/neighbours/community (h2 = 0.09, 95% CI: 0.00, 0.44); (b) family interactions/childcare (h2 = 0.13, 95% CI: 0.00, 0.43); (c) involvement in religious groups/caregiving (h2 = 0.00, 95% CI: 0.00, 0.19). Strong genetic correlations were observed between depressive symptoms and factors a (r = −0.96) and b (r = −0.60). More frequent baseline interactions with friends/neighbours/community were associated with fewer depressive symptoms cross-sectionally (B = −0.14, p = .004) and longitudinally (B = −0.09, p = 0.006), but the associations between social connections and cognitive health were not significant.

Social connections were weakly heritable, suggesting large environmental determination. Connections with friends/neighbours/community were associated with better mental health cross-sectionally and over time.

Anorexia nervosa is associated with high personal and financial costs for sufferers, carers and society in general, but little is known about the long-term health economic burden.

To examine healthcare utilisation, social assistance, sick leave and disability pension in individuals with anorexia nervosa over a period of 30 years.

Fifty-one individuals with adolescent-onset anorexia nervosa and 51 matched comparison cases (COMP) were recruited in the community and followed prospectively from 1985. All individuals were examined on five occasions. At the 30-year follow-up, mean age 44, data on in- and out-patient care, prescribed medications, social assistance, sick leave and disability pension were collected from Swedish national registers.

The anorexia nervosa group had more days of in-patient care (p < 0.001) and out-patient visits to psychiatry (p < 0.001), more days of sick leave (p = 0.006), more days of disability pension (p = 0.002) and were prescribed more psychotropic medication (p = 0.045) compared with the COMP group. Of the anorexia nervosa group, 22% had ever received a disability pension compared with 2% in the COMP group (p = 0.004) and less than half the anorexia nervosa group worked full-time at the 30-year follow-up. In the anorexia nervosa group, 45% had received social assistance at some point, compared with 22% in the COMP group (p = 0.02). Age at onset of anorexia nervosa emerged as a predictor of healthcare utilisation with significant odds ratios for psychiatric in-patient (odds ratio 0.61, 95% CI: 0.39, 0.94; p = 0.027) and out-patient care (odds ratio 0.63, 95% CI: 0.40, 0.98; p = 0.042), i.e. individuals with a later onset of anorexia nervosa were less likely to require psychiatric care.

The long-term burden of adolescent-onset anorexia nervosa comprises increased utilisation of healthcare and dependence on society for a significant minority. A later onset of anorexia nervosa predicted a lower healthcare utilisation.

The Resilience Hub was established to support people in need of psychological/psychosocial support following the 2017 Manchester Arena terrorist attack.

To use mental health screening measures over 3 years following the Arena event to examine the variation in symptoms reported by adults registered with the Hub, and whether this was associated with treatment access characteristics.

Adults engaging with Hub services were separated into eight cohorts depending on when they registered post-incident. Participants completed screening measures for symptoms of trauma, depression, generalised anxiety and work/social functioning. Baseline and follow-up scores over 3 years were compared among the eight admission groups. All types of appointment were recorded in terms of the number of minutes of clinical ‘contact time’ involved, to explore associations with time taken to register.

Overall, baseline screening scores increased as time to register post-event increased. Over the 3 years of follow-up, a decrease in scores was observed for all 4 screening measures, indicating improvement in mental well-being. Those taking longer to register had higher follow-up scores. However, they showed a slightly stronger decrease in average change of score per follow-up month. Mean contact time per month was greater (apart from the 18-months admission group) in individuals delaying registration. Increased contact time was associated with decreased follow-up screening scores for depression and anxiety.

People who registered earlier were less symptomatic, suggesting there may be a potential beneficial impact of early engagement with support services following traumatic events. All who registered showed improvement in symptoms, including those delaying registration, with increased contact time being beneficial. This reinforces the benefits of encouraging early and sustained engagement with services as soon as possible post-incident.

Adolescence is a key developmental period associated with an increased risk of experiencing cannabis-related problems. Identifying modifiable risk factors prior to the onset of cannabis use could help inform preventative interventions.

Analysis nested within a UK prospective birth cohort study, the Avon Longitudinal Study of Parents and Children. Participants (n = 6,049) provided data on cannabis use and symptoms of cannabis problems using the Cannabis Abuse Screening Test at two or more time points between the ages of 15–24 years. Risk factors included internalizing and externalizing disorders assessed at age 10 years, and cognitive function assessed at age 8 years via short-term memory, emotion recognition, divided attention, and listening comprehension.

Participants were mostly female (59.1%) and white (95.73%). Five patterns of adolescent cannabis use problems were identified using longitudinal latent class analysis: stable-no problems (n = 5,157, 85%), early-onset high (n = 104, 2%), late-onset high (n = 153, 3%), early onset low (n = 348, 6%), and late-onset low (n = 287, 5%). In adjusted models, externalizing disorders were associated with early-onset high [RR, 95% CI: 2.82 (1.72, 4.63)], late-onset high [RR, 95% CI: 1.62 (1.02, 2.57)], and early-onset low [RR, 95% CI: 1.82 (1.30, 2.55)] compared to the stable-no problems class. Internalizing disorders were associated with late-onset low only [RR, 95% CI: .50 (.26, .96)], and short-term memory with late-onset high only [RR, 95% CI: 1.09 (1.01, 1.18) compared to the stable-no problems class.

Childhood externalizing disorders were consistently associated with increased risk of problematic patterns of cannabis use over adolescence, particularly early-onset and high levels of problems.

The Kahramanmaraş earthquakes in February 2023 represented a disaster within a disaster, as northwest Syria had been affected by years of war. Literature on the immediate psychological impact of such natural disasters in high-adversity populations is lacking.

To examine prevalences, longitudinal trajectories and cognitive predictors of post-traumatic stress disorder (PTSD), depression and generalised anxiety disorder (GAD) in survivors of armed conflict in northwest Syria exposed to the Kahramanmaraş earthquakes.

We assessed self-reported PTSD, depression and GAD symptoms, as well as self-efficacy and repetitive negative thinking (RNT), at 4, 11 and 18 weeks post-earthquake (T1, T2 and T3, respectively) in 204 war survivors exposed to recent earthquakes. Retention rates for T2 and T3 were 84.4 and 75.8%, respectively. To determine trajectories of PTSD, depression and GAD, we conducted latent class growth analyses with time, self-efficacy and RNT as predictors, and trauma history, education and gender as covariates.

Prevalences of probable PTSD, depression and GAD according to questionnaire cut-offs were 80.4, 79.9 and 70.1% at T1; 62.2, 57.2 and 54.2% at T2; and 62.1, 55.2 and 51.1% at T3. Across all disorders, three developmental trajectories emerged, with most participants following a recovery or low-symptom trajectory. RNT was associated with protracted recovery.

Natural disasters are associated with poor mental health in individuals in war-torn regions. Although latent class growth analyses indicated prevailing recovery trajectories, prevalence remained alarmingly high across time. RNT emerged as a potential transdiagnostic factor across disorders. Research and interventions should prioritise northwest Syrians’ unprecedented mental health needs.

People with an intellectual disability are vulnerable to additional disorders such as dementia. Psychometrically sound and specific instruments are needed for assessment of cognitive functioning in cases of suspected dementia.

To evaluate the construct and item validity, internal consistency and test–retest reliability of a new neuropsychological test battery, the Dementia Test for People with Intellectual Disability (DTIM).

The DTIM was applied to 107 individuals with intellectual disability with (n = 16) and without (n = 91) dementia. The psychometric properties of the DTIM were assessed in a prospective study. The assessors were blinded to the diagnostic assignment.

Confirmatory factor analysis at the scale level showed that a one-factor model fitted the data well (root mean square error of approximation < 0.06, standardised root mean square residual < 0.08, comparative fit index > 0.9). At the domain level, one-factor models showed reasonable-to-good fit index for five of seven domains. Internal consistency indicated excellent reliability of the overall scale (Cronbach’s α: 0.94 for dementia and 0.95 for controls). Item analysis revealed a wide range of difficulties (0.19–0.75 for dementia, 0.31–0.87 for controls), with minimal floor and ceiling effects. Eleven items (26%) had discrimination value ≤ 0.50. Test–retest reliability (n = 82) was high, with intraclass correlations of 0.95 (total score) and 0.69–0.96 (domains).

The DTIM fits a one-factor model and demonstrates internal and test–retest reliability; thus, it is suitable for use in cases of suspected dementia in people with various intellectual disabilities.