At the basis of many important research questions is causality – does X causally impact Y? For behavioural and psychiatric traits, answering such questions can be particularly challenging, as they are highly complex and multifactorial. ‘Triangulation’ refers to prospectively choosing, conducting and integrating several methods to investigate a specific causal question. If different methods, with different sources of bias, all indicate a causal effect, the finding is much less likely to be spurious. While triangulation can be a powerful approach, its interpretation differs across (sub)fields and there are no formal guidelines. Here, we aim to provide clarity and guidance around the process of triangulation for behavioural and psychiatric epidemiology, so that results of existing triangulation studies can be better interpreted, and new triangulation studies better designed.

We first introduce the concept of triangulation and how it is applied in epidemiological investigations of behavioural and psychiatric traits. Next, we put forth a systematic step-by-step guide, that can be used to design a triangulation study (accompanied by a worked example). Finally, we provide important general recommendations for future studies.

While the literature contains varying interpretations, triangulation generally refers to an investigation that assesses the robustness of a potential causal finding by explicitly combining different approaches. This may include multiple types of statistical methods, the same method applied in multiple samples, or multiple different measurements of the variable(s) of interest. In behavioural and psychiatric epidemiology, triangulation commonly includes prospective cohort studies, natural experiments and/or genetically informative designs (including the increasingly popular method of Mendelian randomization). The guide that we propose aids the planning and interpreting of triangulation by prompting crucial considerations. Broadly, its steps are as follows: determine your causal question, draw a directed acyclic graph, identify available resources and samples, identify suitable methodological approaches, further specify the causal question for each method, explicate the effects of potential biases and, pre-specify expected results. We illustrated the guide’s use by considering the question: ‘Does maternal tobacco smoking during pregnancy cause offspring depression?’.

In the current era of big data, and with increasing (public) availability of large-scale datasets, triangulation will become increasingly relevant in identifying robust risk factors for adverse mental health outcomes. Our hope is that this review and guide will provide clarity and direction, as well as stimulate more researchers to apply triangulation to causal questions around behavioural and psychiatric traits.

Depressive disorders are a major public health issue in Western societies, particularly among adolescents, young adults and women. The COVID-19 pandemic has exacerbated mental health challenges, increasing depression and anxiety symptoms, especially in younger people. This study focuses on the hard-hit Emilia-Romagna Region (ERR) in Italy, examining changes in antidepressant (AD) drug use post-COVID-19 to understand the pandemic’s effect on mental health.

A population-based interrupted time series design and a segmented regression analysis was carried out on ERR pharmaceutical data (FED, direct dispensation pharmaceuticals, AFT, territorial pharmaceutical assistance) out to estimate changes in AD use during the three pandemic years (2020, 2021 and 2022) compared to 2017–2019.Analyses were stratified by age, gender, citizenship, population density of the area of residence.

A notable increase in AD consumption compared to what was expected was observed among younger age groups, and especially in females. In the 12–19 age group, a gradual increase was recorded from January 2021 until it reached +48% in 2022 (+58% among women, +30% among men). An even more remarkable growth in AD usage among non-Italian residents in the same age group was recorded compared to expected. A relevant increase, although smaller, was detected among individuals in the 20–34 age group, with a peak of +9% in 2022. These differences persisted up until the end of the observation period.

The study suggests that the COVID-19 pandemic may have had a lasting negative impact on the mental health of younger individuals. The observed increase in AD use may foreshadow a potential long-term need for enhanced mental healthcare and services directed at this subpopulation.

The associations of prior homelessness with current health are unknown. Using nationally representative data collected in private households in England, this study aimed to examine Common Mental Disorders (CMDs), physical health, alcohol/substance dependence, and multimorbidities in people who formerly experienced homelessness compared to people who never experienced homelessness.

This cross-sectional study utilised data from the 2007 and 2014 Adult Psychiatric Morbidity Surveys. Former homelessness and current physical health problems were self-reported. Current CMDs, alcohol dependence and substance dependence were ascertained using structured validated scales. Survey-weighted logistic regression was used to compare multimorbidities (conditions in combination) for participants who formerly experienced homelessness with those who had never experienced homelessness, adjusting for sociodemographic characteristics, smoking status and adverse experiences. Population attributable fractions (PAFs) were calculated.

Of 13,859 people in the sample, 535 formerly experienced homelessness (3.6%, 95% CI 3.2–4.0). 44.8% of people who formerly experienced homelessness had CMDs (95% CI 40.2–49.5), compared to 15.0% (95% CI 14.3–15.7) for those who had never experienced homelessness. There were substantial associations between prior homelessness and physical multimorbidity (adjusted odds ratio [aOR] 1.98, 95% CI 1.53–2.57), CMD–physical multimorbidity (aOR 3.43, 95% CI 2.77–4.25), CMD–alcohol/substance multimorbidity (aOR 3.53, 95% CI 2.49–5.01) and trimorbidity (CMD–alcohol/substance–physical multimorbidity) (aOR 3.26, 95% CI 2.20–4.83), in models adjusting for sociodemographic characteristics and smoking. After further adjustment for adverse experiences, associations attenuated but persisted for physical multimorbidity (aOR 1.40, 95% CI 1.10–1.79) and CMD–physical multimorbidity (aOR 1.55, 95% CI 1.20–2.00). The largest PAFs were observed for CMD–alcohol/substance multimorbidity (17%) and trimorbidity (16%).

Even in people currently rehoused, marked inequities across multimorbidities remained evident, highlighting the need for longer-term integrated support for people who have previously experienced homelessness.

Psychotic-like experiences (PLEs), especially for persistent PLEs, are highly predictive of subsequent mental health problems. Hence, it is crucial to explore the psychopathological associations underlying the occurrence and persistence of PLEs. This study aimed to explore the above issues through a longitudinal dynamic network approach among PLEs and psychological and psychosocial factors.

A total of 3,358 college students completed two waves of online survey (from Oct 2021 to Oct 2022). Socio-demographic information was collected at baseline, and PLEs, depressive and anxiety symptoms, and adverse life events were assessed in both waves. Cross-lagged panel network analyses were used to establish networks among individuals with baseline PLEs as well as those without.

At baseline, 455(13.5%) students were screened positive for PLEs. Distinct dynamic network structures were revealed among participants with baseline PLEs and those without. While ‘psychomotor disturbance’ had the strongest connection with PLEs in participants with baseline PLEs, ‘suicide/self-harm’ was most associated with PLEs in those without. Among all three subtypes of PLEs, bizarre experiences and persecutory ideation were the most affected nodes by other constructs in participants with baseline PLEs and those without, respectively. Additionally, wide interconnections within the PLEs construct existed only among participants without baseline PLEs.

The study provides time-variant associations between PLEs and depressive symptoms, anxiety symptoms, and adverse life events using network structures. These findings help to reveal the crucial markers of the occurrence and persistence of PLEs, and shed high light on future intervention aimed to prevent and relieve PLEs.

The impact of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection on the risk of self-harming behaviours warrants further investigation. Here, we hypothesized that people with a history of hospitalization for self-harm may be particularly at risk of readmission in case of SARS-CoV-2 hospitalization.

We conducted a retrospective analysis based on the French national hospitalization database. We identified all patients hospitalized for deliberate self-harm (10th edition of the International Classification of Diseases codes X60–X84) between March 2020 and March 2021. To study the effect of SARS-CoV-2 hospitalization on the risk of readmission for self-harm at 1-year of the inclusion, we performed a multivariable Fine and Gray model considering hospital death as a competing event.

A total of 61,782 individuals were hospitalized for self-harm. During the 1-year follow-up, 9,403 (15.22%) were readmitted for self-harm. Between inclusion and self-harm readmission or the end of follow-up, 1,214 (1.96% of the study cohort) were hospitalized with SARS-CoV-2 (mean age 60 years, 52.9% women) while 60,568 were not (mean age 45 years, 57% women). Multivariate models revealed that the factors independently associated with self-harm readmission were: hospitalization with SARS-CoV-2 (adjusted hazard ratio (aHR) = 3.04 [2.73–3.37]), psychiatric disorders (aHR = 1.61 [1.53–1.69]), self-harm history (aHR = 2.00 [1.88–2.04]), intensive care and age above 80.

In hospitalized people with a personal history of self-harm, infection with SARS-CoV-2 increased the risk of readmission of self-harm, with an effect that seemed to add to the effect of a history of mental disorders, with an equally significant magnitude. Infection may be a significantly stressful condition that precipitates self-harming acts in vulnerable individuals. Clinicians should pay particular attention to the emergence of suicidal ideation in these patients in the aftermath of SARS-CoV-2 infection.

Epidemiological studies show that despite the episodic nature, the long-term trajectory of depression can be variable. This study evaluated the heterogeneity of 10-year trajectory of major depressive disorder (MDD) related service utilization and associated clinical characteristics among US Veterans with a first diagnosis after 9/11.

Using a cohort design, electronic health record data for 293,265 Operation Enduring Freedom and Iraqi Freedom (OEF/OIF) Veterans were extracted to identify those with MDD between 2001 and 2021 with a full preceding year of clinical data and 10 years following the diagnosis. Latent class growth analysis compared clinical characteristics associated with four depression trajectories. Across all Veterans Affairs (VA)hospitals, 25,307 Veterans met our inclusion criteria. Demographic and clinical information from medical records was extracted and used as predictors of depression 10-year trajectories.

Among the study cohort (N = 25,307), 27.7% were characterized by brief contact, 41.7% were later re-entry, 17.6% were persistent contact and 12.9% were prolonged initial contact for depression related services. Compared to Veterans with trajectories showing brief contact, those with protracted treatment (persistent or prolonged initial contact) were more likely to be diagnosed with comorbid posttraumatic stress disorder (PTSD) and with MDD that was moderate to severe or recurrent.

Depression is associated with a range of treatment trajectories. The persistent and prolonged initial contact trajectories may have distinct characteristics and uniquely high resource utilization and disability income. We can anticipate that patients with comorbid PTSD may need longer-term care which has implications for brief models of care.

Although natural hazards (e.g., tropical cyclones, earthquakes) disproportionately affect developing countries, most research on their mental health impact has been conducted in high-income countries. We aimed to summarize prevalences of mental disorders in Global South populations (classified according to the United Nations Human Development Index) affected by natural hazards.

To identify eligible studies for this meta-analysis, we searched MEDLINE, PsycINFO and Web of Science up to February 13, 2024, for observational studies with a cross-sectional or longitudinal design that reported on at least 100 adult survivors of natural hazards in a Global South population and assessed mental disorders with a validated instrument at least 1 month after onset of the hazard. Main outcomes were the short- and long-term prevalence estimates of mental disorders. The project was registered on the International Prospective Register of Systematic Reviews (CRD42023396622).

We included 77 reports of 75 cross-sectional studies (six included a non-exposed control group) comprising 82,400 individuals. We found high prevalence estimates for post-traumatic stress disorder (PTSD) in the general population (26.0% [95% CI 18.5–36.3]; I2 = 99.0%) and depression (21.7% [95% CI 10.5–39.6]; I2 = 99.2%) during the first year following the event, with similar prevalences observed thereafter (i.e., 26.0% and 23.4%, respectively). Results were similar for regions with vs. without recent armed conflict. In displaced samples, the estimated prevalence for PTSD was 46.5% (95% CI 39.0–54.2; k = 6; I2 = 93.3). We furthermore found higher symptom severity in exposed, versus unexposed, individuals. Data on other disorders were scarce, apart from short-term prevalence estimates of generalised anxiety disorder (15.9% [95% CI 4.7–42.0]; I2 = 99.4).

Global South populations exposed to natural hazards report a substantial burden of mental disease. These findings require further attention and action in terms of implementation of mental health policies and low-threshold interventions in the Global South in the aftermath of natural hazards. However, to accurately quantify the true extent of this public health challenge, we need more rigorous, well-designed epidemiological studies across diverse regions. This will enable informed decision making and resource allocation for those in need.

Most information about the association between childhood maltreatment (CM) and subsequent psychiatric morbidity is based on retrospective self-reports. Findings from longitudinal studies using prospective reports to statutory agencies may be subject to attrition. We therefore compared the prevalence to age 30 of inpatient psychiatric diagnoses in those who experienced agency-reported CM with those of the rest of the cohort using administrative data to minimise loss to follow-up.

We used linked administrative data for two birth cohorts of all individuals born in Queensland, Australia in 1983 and 1984 (N = 83,050) and followed to age 30 years. This was the entire cohort aside from 312 people who died. Information on CM came from statewide child protection data and psychiatric diagnoses from all public and private hospital admissions in Queensland.

On adjusted analyses, the 4,703 participants (5.7%) who had been notified to the statewide child protection authority had three to eight times the odds of being admitted for any of the following psychiatric diagnoses by age 30 years old: schizophrenia-spectrum disorders, bipolar affective disorders, depression, anxiety and post-traumatic stress disorders (PTSD). There were similar findings for all the CM subtypes. Associations were especially strong for PTSD with between a seven – and nine-fold increase in the odds of admission.

This is one of the largest studies of the long-term effects of CM, covering an entire jurisdiction. All types of maltreatment are significantly related to a range of psychiatric disorders requiring hospitalisation. Early identification, intervention and providing appropriate support to individuals who have experienced CM may help mitigate the long-term consequences and reduce the risk of subsequent mental health problems.

Achieving equitable healthcare access is a global challenge. Improving whole-population mental health and reducing the global burden of mental disorders is a key recommendation of the 2018 Lancet Global Mental Health Commission, which proposed monitoring national indicators, including the proportion of people with severe mental disorders who are service-users. This study aims to derive an equity indicator from national datasets integrating need, service utilisation and socioeconomic status, and demonstrate its utility in identifying gaps in mental health service use amongst those with the greatest need, thereby guiding equitable healthcare delivery.

We present a case study of a universal health insurance scheme (Medicare) in Australia. We developed the equity indicator using three national datasets. Geographic areas were linked to an area-based socioeconomic deprivation quintile (Census 2016). Per geographic area, we estimated the number with a mental healthcare need using scores ≥30 on the Kessler-10 (Australian National Health Surveys 2015 and 2018), and obtained the number of services used, defined as mental health-related contacts with general practitioners and mental health professionals (Medicare administrative data 2015–2019). We divided the number of services by the population with an estimated mental healthcare need and averaged these use-rates across each socioeconomic deprivation quintile. The equity indicator is the ratio of the use-rates in the least versus most deprived quintiles.

Those estimated to have the greatest need for mental healthcare in 2019 ranged between 8.2% in the most disadvantaged area quintile (Q1) and 2.4% in the least (Q5), corresponding to a proportional increase of 27.7% in Q1 and 19.5% in Q5 since 2015. Equity-indicator-adjusted service rates of 4.2 (3.8–4.6) and 23.9 (22.4–25.4) showed that individuals with the highest need for care residing in Q1 areas received a stark 6 times fewer services compared to their Q5 counterparts, producing an equity indicator of 6.

As the global prevalence of common mental disorders may be increasing, it is crucial to calculate robust indicators evaluating the equity of mental health service use. In this Australian case study, we developed an equity indicator enabling the direct comparison of geographic areas with different need profiles. The results revealed striking inequities that persisted despite publicly-funded universal healthcare, recent service reforms and being a high-income country. This study demonstrates the importance and feasibility of generating such an indicator to inform and empower communities, healthcare providers and policymakers to pursue equitable service provision.

Although it has been hypothesized that air pollution, particularly PM2.5 and PM10, causes depressed symptoms, their interactions with greenness have not yet been confirmed. This study examined the association between depression symptoms and air pollution, as well as the potential moderating effects of greenness.

A total of 7657 people from all around South Korea were examined using information from the Korean Longitudinal Study of Aging, for the years 2016, 2018 and 2020. Depressive symptoms were assessed using the CES-D 10 score (Center for Epidemiology Studies of Depression scale, Boston form), and annual air pollution levels (PM2.5, PM10) and greenness (NDVI, Landsat Normalized Difference Vegetation Index) at the district level (si-gun-gu) were considered for the association analysis. The investigation was primarily concerned with determining how the CES-D 10 score changed for each 10 ${\mu \text{g/}}{{\text{m}}^{\text{3}}}$ increase in PM2.5 and PM10 according to NDVI quantiles, respectively. The analysis used generalized estimating equation models that were adjusted with both minimal and complete variables. Subgroup analyses were conducted based on age groups (<65, ≥65 years old), sex and exercise status.

The impact of PM10 on depression in the fourth quantile of NDVI was substantially less in the fully adjusted linear mixed model (OR for depression with a 10 ${\mu\text{ g/}}{{\text{m}}^{\text{3}}}$ increment of PM10: 1.29, 95% CI: 1.06, 1.58) than in the first quantile (OR: 1.88, 95% CI: 1.58, 2.25). In a similar vein, the effect of PM2.5 on depression was considerably reduced in the fourth quantile of NDVI (OR for depression with a 10 ${\mu\text{ g/}}{{\text{m}}^{\text{3}}}$ increment of PM2.5: 1.78, 95% CI: 1.30, 2.44) compared to the first (OR: 3.75, 95% CI: 2.75, 5.10). Subgroup analysis results demonstrated beneficial effects of greenness in the relationship between particulate matter and depression.

This longitudinal panel study found that a higher quantile of NDVI was associated with a significantly reduced influence of air pollution (PM10, PM2.5) on depression among older individuals in South Korea.

It is widely recognized that the COVID-19 pandemic exerted an impact on the mental health of the general population, but epidemiological evidence is surprisingly sparse. We aimed to explore the association between serologically confirmed SARS-CoV-2 infection and psychological distress – assessed by symptoms of depression, anxiety and stress – in the general adult population in southern Switzerland, a region widely affected by the pandemic. We also investigated whether this association varied over time and between pandemic waves from late 2020 through 2021.

We used data from 305 adults who participated in the Corona Immunitas Ticino prospective seroprevalence study in southern Switzerland, including results of the serological tests of SARS-CoV-2 infection collected in June 2021, and explored associations with depression, anxiety and stress scores as measured by the 21-item Depression, Anxiety and Stress Scale at three time points between December 2020 and August 2021, accounting for socio-demographic and health characteristics.

In our sample, 84.3% of the participants (mean age of 51.30 years, SD = ±.93) were seronegative at baseline. Seropositive (i.e., infected) participants had a decreasing probability of being depressed and anxious through the COVID-19 pandemic waves compared to the seronegative (non-infected) participants. Further, seropositivity at baseline was also associated with more rapid decline in depressive, anxiety and stress symptomatology, and younger age and the presence of chronic diseases were independently associated with mild anxiety (OR = .97; P = 0.013; 95% CI = 0.95, 0.99; OR = 3.47; P = 0.001; 95% CI = 1.71, 7.04) and stress (OR = .96; P = 0.003; 95% CI = .94, .99; OR = 2.56; P = 0.010; 95% CI = 1.25, 5.22).

Our results suggest that the MH consequences of the pandemic may not be due to the SARS-CoV-2 infection per se, but to fears associated with the risk of infection, and to the pandemic uncertainties.

The enduring impact of the COVID-19 pandemic on mental health and its implications for COVID-19 vaccine uptake necessitate comprehensive investigation. We aimed to characterize the persistence of moderate to severe anxiety and depression symptoms from July 2020 to July 2023, explore demographic associations with symptom persistence, and assess how these symptoms affected COVID-19 vaccination uptake between May 2021 and July 2023.

Participants from the national community-based CHASING COVID Cohort were enrolled between March and June 2020 and completed quarterly follow-ups until December 2023. Scores ≥10 on the Generalized Anxiety Disorder 7-item and the Patient Health Questionnaire 8-item at 14 follow-up assessments indicated moderate to severe anxiety and depression symptoms, respectively. Missing scores were imputed. Persistent anxiety and depression were defined as experiencing moderate to severe anxiety and depression symptoms ≥7 out of 14 follow-up assessments, respectively.

Among 4,851 participants, 15.9% experienced persistent anxiety symptoms and 19.3% persistent depression symptoms from July 2020 to July 2023. Demographic factors associated with symptom persistence included younger age, female or non-binary gender, Hispanic ethnicity, lower education level, household income <$100k, presence of children <18 in the household, greater healthcare barriers and comorbidities. Participants with ongoing moderate to severe anxiety and depression symptoms had 0.95 (95% CI: 0.94, 0.97) and 0.95 (95% CI: 0.93, 0.96) times rates of receiving additional COVID-19 vaccine doses between May 2021 and July 2023, respectively.

Customized support for individuals with mental disorders may mitigate barriers to vaccine uptake. Further investigation is warranted to validate these findings and inform targeted interventions.

Investigate the prevalence of adverse childhood experience (ACE) and intimate partner violence (IPV) using a large representative Chinese sample, explore the association mechanism between ACE and adult exposure to IPV and to examine gender differences.

A total of 21,154 participants were included in this study. The ACE scale was used to assess participants’ exposure to ACE before the age of 18. Participants were evaluated for IPV experienced after the age of 18 using the IPV Scale. Logistic regression model was used to analyse the association between ACE and the risk of IPV exposure in adulthood. Principal component analysis was used to extract the main patterns of ACEs in the Chinese population. Network analyses were employed to identify the most critical types of ACE and IPV, analyse the association mechanisms between ACEs and IPVs, explore gender differences in this association and compare gender differences in the severity of IPVs experienced in adulthood.

Participants with at least one ACE event faced a 215.5% higher risk of IPV compared to those without ACE experiences. In population-wide and gender-specific networks, The ACE and IPV nodes with the highest expected influence are ‘ACE1 (Verbal abuse + physical abuse pattern)’ and ‘IPV5 (Partner compares me to other people and blatantly accuses me, making me feel embarrassed and unsure of myself)’. Positive correlations were found between ‘ACE1 (Verbal abuse + physical abuse pattern)’–‘IPV3 (Partner does not care about me when I am in bad shape [not feeling well or in a bad mood])’, ‘ACE4 (Violent treatment of mother or stepmother + criminal acts in the family pattern)’–‘IPV1 (Partner has ever directly assaulted or hurt me with the help of an instrument)’ and ‘ACE2 (Exposure to sexual assault pattern)’–‘IPV2 (Partner would have physical or sexual contact with me against my will)’, which were the three edges with the highest edge weight values in the ACE pattern and IPV edges. ‘ACE1 (Verbal abuse + physical abuse pattern)’–‘IPV3 (Partner does not care about me when I am in bad shape [not feeling well or in a bad mood])’, ‘ACE2 (Exposure to sexual assault pattern)’–‘IPV2 (Partner would have physical or sexual contact with me against my will)’, ‘ACE4 (Violent treatment of mother or stepmother + criminal acts in the family pattern)’–‘IPV1 (Partner has ever directly assaulted or hurt me with the help of an instrument)’ in the male network and ‘ACE1 (Verbal abuse + physical abuse pattern)’–‘IPV3 (Partner does not care about me when I am in bad shape [not feeling well or in a bad mood])’, ‘ACE4 (Violent treatment of mother or stepmother + criminal acts in the family pattern)’–‘IPV1 (Partner has ever directly assaulted or hurt me with the help of an instrument)’, ‘ACE3 (Substance abuse + mental illness + violent treatment of mother or stepmother pattern)’–‘IPV1 (Partner has ever directly assaulted or hurt me with the help of an instrument)’ in the female network are the three edges with the highest edge weights among the ACE and IPV edges in their networks, respectively, all displaying positive correlations. The strength of ‘IPV3 (Partner does not care about me when I am in bad shape [not feeling well or in a bad mood])’ was higher in the male network than in the female (male = 0.821, female = 0.755, p = 0.002). The edge weight values of ‘ACE3 (Substance abuse + mental illness + violent treatment of mother or stepmother pattern)’–‘IPV1 (Partner has ever directly assaulted or hurt me with the help of an instrument)’ (P = 0.043) and ‘ACE4 (Violent treatment of mother or stepmother + criminal acts in the family pattern)’–‘IPV1 (Partner has ever directly assaulted or hurt me with the help of an instrument)’ (P = 0.032) are greater for females than males.

The most common type of ACE in the Chinese population is verbal violence combined with physical violence, while the predominant type of IPV is verbal violence. Males experience higher levels of emotional neglect from their partners compared to females. The association between witnessing physical violence in childhood and experiencing physical violence from a partner in adulthood is stronger in females than in males. The homotypic continuum between ACE and IPV is a crucial mechanism in understanding intergenerational domestic violence. Enhance economic and educational levels, promote correct parenting concepts, reduce child abuse, establish accurate perceptions of intimate relationships, eliminate shame about violence and further advance gender equality. These efforts are vital for reducing IPV prevalence and breaking the cycle of violence in victims’ lives.

Migraine and post-traumatic stress disorder (PTSD) are both twice as common in women as men. Cross-sectional studies have shown associations between migraine and several psychiatric conditions, including PTSD. PTSD is disproportionally common among patients in headache clinics, and individuals with migraine and PTSD report greater disability from migraines and more frequent medication use. To further clarify the nature of the relationship between PTSD and migraine, we conducted bidirectional analyses of the association between (1) migraine and incident PTSD and (2) PTSD and incident migraine.

We used longitudinal data from 1989–2020 among the 33,327 Nurses’ Health Study II respondents to the 2018 stress questionnaire. We used log-binomial models to estimate the relative risk of developing PTSD among women with migraine and the relative risk of developing migraine among individuals with PTSD, trauma-exposed individuals without PTSD, and individuals unexposed to trauma, adjusting for race, education, marital status, high blood pressure, high cholesterol, alcohol intake, smoking, and body mass index.

Overall, 48% of respondents reported ever experiencing migraine, 82% reported experiencing trauma and 9% met the Diagnostic and Statistical Manual of Mental Disorders-5 criteria for PTSD. Of those reporting migraine and trauma, 67% reported trauma before migraine onset, 2% reported trauma and migraine onset in the same year and 31% reported trauma after migraine onset. We found that migraine was associated with incident PTSD (adjusted relative risk [RR]: 1.26, 95% confidence interval [CI]: 1.14–1.39). PTSD, but not trauma without PTSD, was associated with incident migraine (adjusted RR: 1.20, 95% CI: 1.14–1.27). Findings were consistently stronger in both directions among those experiencing migraine with aura.

Our study provides further evidence that migraine and PTSD are strongly comorbid and found associations of similar magnitude between migraine and incident PTSD and PTSD and incident migraine.

Fifteen years ago, we published an article in Social Science and Medicine seeking to resolve the contentious debate between advocates of two very different frameworks for understanding and addressing the mental health needs of conflict-affected populations. The two approaches, which we labelled trauma-focused and psychosocial, reflect deeply held beliefs about the causes and nature of distress in war-affected communities. Drawing on the burgeoning literature on armed conflict and mental health, the reports of mental health and psychosocial support (MHPSS) staff in the field, and on research on the psychology and psychophysiology of stress, we proposed an integrative model that drew on the strengths of both frameworks and underscored their essential complementarity. Our model includes two primary pathways by which armed conflict impacts mental health: directly, through exposure to war-related violence and loss, and indirectly, through the harsh conditions of everyday life caused or exacerbated by armed conflict. The mediated model we proposed draws attention to the effects of stressors both past (prior exposure to war-related violence and loss) and present (ongoing conflict, daily stressors), at all levels of the social ecology; for that reason, we have termed it an ecological model for understanding the mental health needs of conflict-affected populations.

In the ensuing 15 years, the model has been rigorously tested in diverse populations and has found robust support. In this paper, we first summarize the development and key tenets of the model and briefly review recent empirical support for it. We then discuss the implications of an ecological framework for interventions aimed at strengthening mental health in conflict-affected populations.

We present preliminary evidence suggesting there has been a gradual shift towards more ecological (i.e., multilevel, multimodal) programming in MHPSS interventions, along the lines suggested by our model as well as other conceptually related frameworks, particularly public health.

We reflect on several gaps in the model, most notably the absence of adverse childhood experiences. We suggest the importance of examining early adversity as both a direct influence on mental health and as a potential moderator of the impact of potentially traumatic war-related experiences of violence and loss.

Studies show that people with severe mental illness (SMI) have a greater risk of dying from colorectal cancer (CRC). These studies mostly predate the introduction of national bowel cancer screening programmes (NBCSPs) and it is unknown if these have reduced disparity in CRC-related mortality for people with SMI.

We compared mortality rates following CRC diagnosis at colonoscopy between a nationally representative sample of people with and without SMI who participated in Australia’s NBCSP. Participation was defined as the return of a valid immunochemical faecal occult blood test (iFOBT). We also compared mortality rates between people with SMI who did and did not participate in the NBCSP. SMI was defined as receiving two or more Pharmaceutical Benefits Scheme prescriptions for second-generation antipsychotics or lithium.

Amongst NBCSP participants, the incidence of CRC in the SMI cohort was lower than in the controls (hazard ratio [HR] 0.77, 95% confidence interval [CI] 0.61–0.98). In spite of this, their all-cause mortality rate was 1.84 times higher (95% CI 1.12–3.03), although there was only weak evidence of a difference in CRC-specific mortality (HR 1.82; 95% CI 0.93–3.57). People with SMI who participated in the NBCSP had better all-cause survival than those who were invited to participate but did not return a valid iFOBT (HR 0.67, 95% CI 0.50–0.88). The benefit of participation was strongest for males with SMI and included improved all-cause and CRC-specific survival.

Participation in the NBCSP may be associated with improved survival following a CRC diagnosis for people with SMI, especially males, although they still experienced greater mortality than the general population. Approaches to improving CRC outcomes in people with SMI should include targeted screening, and increased awareness about the benefits or participation.

Australian and New Zealand Clinical Trials Registry (Trial ID: ACTRN12620000781943).

The majority of studies of mental health interventions for young adolescents have only evaluated short-term benefits. This study evaluated the longer-term effectiveness of a non-specialist delivered group-based intervention (Early Adolescent Skills for Emotions; EASE) to improve young adolescents’ mental health.

In this single-blind, parallel, controlled trial, Syrian refugees aged 10-14 years in Jordan who screened positive for psychological distress were randomised to receive either EASE or enhanced usual care (EUC). Primary outcomes were scores on the Paediatric Symptom Checklist (PSC) assessed at Week 0, 8-weeks, 3-months, and 12 months after treatment. Secondary outcomes were disability, posttraumatic stress, school belongingness, wellbeing, and caregivers’ reports of distress, parenting behaviour, and their perceived children’s mental health.

Between June, 2019 and January, 2020, 185 adolescents were assigned to EASE and 286 to EUC, and 149 (80.5%) and 225 (78.7%) were retained at 12 months, respectively. At 12 months there were no significant differences between treatment conditions, except that EASE was associated with less reduction in depression (estimated mean difference -1.6, 95% CI –3.2 to -0.1; p=.03; effect size, -0.3), and a greater sense of school belonging (estimated mean difference -0.3, 95% CI –5.7 to -0.2; p=.03; effect size, 5.0).

Although EASE led to significant reductions in internalising problems, caregiver distress, and harsh disciplinary parenting at 3-months, these improvements were not maintained at 12 months relative to EUC. Scalable psychological interventions for young adolescents need to consider their ongoing mental health needs. Prospectively registered: ACTRN12619000341123.