Epidemiology and Psychiatric Sciences最新文献

Aims: The epidemiology and age-specific patterns of lifetime suicide attempts (LSA) in China remain unclear. We aimed to examine age-specific prevalence and predictors of LSA among Chinese adults using machine learning (ML).

Methods: We analyzed 25,047 adults in the 2024 Psychology and Behavior Investigation of Chinese Residents (PBICR-2024), stratified into three age groups (18-24, 25-44, ≥ 45 years). Thirty-seven candidate predictors across six domains-sociodemographic, physical health, mental health, lifestyle, social environment, and self-injury/suicide history-were assessed. Five ML models-random forest, logistic regression, support vector machine (SVM), Extreme Gradient Boosting (XGBoost), and Naive Bayes-were compared. SHapley Additive exPlanations (SHAP) were used to quantify feature importance.

Results: The overall prevalence of LSA was 4.57% (1,145/25,047), with significant age differences: 8.10% in young adults (18-24), 4.67% in adults aged 25-44, and 2.67% in older adults (≥45). SVM achieved the best test-set performance across all ages [area under the curve (AUC) 0.88-0.94, sensitivity 0.79-0.87, specificity 0.81-0.88], showing superior calibration and net clinical benefit. SHAP analysis identified both shared and age-specific predictors. Suicidal ideation, adverse childhood experiences, and suicide disclosure were consistent top predictors across all ages. Sleep disturbances and anxiety symptoms stood out in young adults; marital status, living alone, and perceived stress in mid-life; and functional limitations, poor sleep, and depressive symptoms in older adults.

Conclusions: LSA prevalence in Chinese adults is relatively high, with a clear age gradient peaking in young adulthood. Risk profiles revealed both shared and age-specific predictors, reflecting distinct life-stage vulnerabilities. These findings support age-tailored suicide prevention strategies in China.

Aims: The COVID-19 pandemic exacerbated psychological distress, but limited information is available on the shifts in mental health symptoms and their associated factors across different stages. This study was conducted to more reliably estimate shifts in mental health impacts and to identify factors associated with symptoms at different pandemic stages.

Methods: We performed a national repeated cross-sectional study at stable (2021), recurrence (2022), and end-of-emergency (2023) stages based on representative general national population with extensive geographic coverage. Anxiety, depression, post-traumatic stress disorder (PTSD) and insomnia symptoms were evaluated by GAD-7, PHQ-9, IES-R and ISI scales, respectively, and their associated factors were identified via multivariable linear regression.

Results: Generally, 42,000 individuals were recruited, and 36,218, 36,097 and 36,306 eligible participants were included at each stage. The prevalence of anxiety, depression and insomnia symptoms increased from 13.7-16.4% at stable to 17.3-22.2% at recurrence and decreased to 14.5-18.6% at end of emergency, while PTSD symptom continuously increased from 5.1% to 7.6% and 9.2%, respectively (all significant, P < 0.001). Common factors associated with mental health symptoms across all stages included centralized quarantine, frontline work and residence in initially widely infected areas. Centralized quarantine was linked to anxiety, depression, PTSD and insomnia during the stable, recurrence and end-of-emergency stages. Frontline workers exhibited higher risks of anxiety, depression and insomnia throughout these stages. Individuals in initially widely infected areas were more likely to experience depression and PTSD, particularly during the stable and recurrence stages. Stage-specific risk factors were also identified. Lack of outdoor activity was associated with anxiety, depression and insomnia during the stable and recurrence stages. Residents in high-risk areas during the recurrence stage correlated with increased anxiety and insomnia. Suspected infection was tied to anxiety and insomnia in the recurrence and end-of-emergency stages, while the death of family or friends was linked to PTSD during recurrence and to depression, PTSD and insomnia at the end-of-emergency stage.

Conclusions: Mental health symptoms increased when pandemic recurred, and could remain after end-of-emergency, requiring prolonged interventions. Several key factors associated with mental symptoms and their variations were identified at different pandemic stages, suggesting different at-risk populations.

Aims: Although the United States incarcerates nearly two million people, the epidemiology of psychiatric disorders in correctional populations is not well understood, and no study has examined temporal trends in psychiatric disorder prevalences within a single correctional system. This study assessed how psychiatric disorder prevalences have changed in the Texas Department of Criminal Justice (TDCJ), the largest American state prison system housing post-conviction, sentenced individuals.

Methods: This retrospective cohort study of TDCJ electronic medical record data from 1 January 2016 through 31 December 2023 included all persons incarcerated for any duration during that period. Diagnoses were based on International Classification of Disease (ICD-10) diagnostic codes. Outcomes were annual prevalences of depressive, bipolar and schizophrenia spectrum disorders stratified by age, race and sex. Cochran-Armitage Tests were used to assess temporal trends within each stratum. Two-way interactions were assessed by fitting Generalized Estimating Equations models using autoregressive correlation with repeated subjects.

Results: The overall population ranged from 170,269 to 222,798 individuals. Approximately, one-third were White (34.5-35.4%), one-third Black (31.0-32.3%), and one-third Hispanic (32.7-33.5%). Most were aged 30-49 (52.8-57.3%), and men (88.9-90.7%) outnumbered women (9.3-11.1%). The prevalences (per 100 [95% CI]) of psychiatric disorders generally increased when comparing 2016 to 2023. Depressive disorders increased the most among those aged 30-49 (5.23 [5.10-5.35] to 6.71 [6.56-6.86]), Hispanic individuals (3.86 [3.72-4.00] to 5.72 [5.53-5.90]), and men (4.72 [4.63-4.82] to 6.53 [6.42-6.65]). Bipolar disorders increased the most among those aged ≥50 (2.57 [2.42-2.72] to 3.46 [3.29-3.63]), Hispanic individuals (1.31 [1.23-1.40] to 2.23 [2.11-2.35]), and men (2.26 [2.20-2.33] to 3.12 [3.04-3.20]). Schizophrenia spectrum disorders increased the most among those aged ≤29 (1.33 [1.24-1.42] to 2.52 [2.35-2.68]), Hispanic individuals (1.53 [1.44-1.62] to 3.21 [3.35-4.40]), and women (1.27 [1.56-1.89] to 4.24 [3.95-4.53]). When stratified by demographic variables, trend tests were significant for nearly all comparisons (P < 0.0001), and all two-way interactions were significant (P < 0.0001).

Conclusions: The prevalences of major psychiatric disorders in the Texas prison system increased when comparing 2016 to 2023, with certain disorders rising more rapidly than others within specific subgroups. These findings emphasize the need for expanded mental health treatment options and resources within correctional settings.

Aims: Evidence on the effects of parental Adverse Childhood Experiences (ACEs) on adolescent mental health remains limited. This study investigates the associations between parental ACEs, children's exposure to threat- and deprivation-related ACEs, and adolescent depression and anxiety using data from the Longitudinal Study of Australian Children.

Methods: We conducted a secondary analysis of the Longitudinal Study of Australian Children (LSAC), a population-based longitudinal cohort study. Parental ACEs were retrospectively reported by caregivers. Children's exposure to ACEs was assessed from ages 4-17 years and categorised as threat-related ACEs (e.g., bullying, hostile parenting, unsafe neighbourhoods, family violence) or deprivation-related ACEs (e.g., financial hardship, parental substance abuse, parental psychological distress, death of a family member, parental separation, parental legal problems). Depressive and anxiety symptoms were self-reported by adolescents at ages between 12 and 17 years. Modified Poisson regression models were used to examine the independent and combined associations of parental ACEs and children's threat- and deprivation-related ACEs (assessed before ages 12, 14, and 16 years) with depression and anxiety outcomes, including tests for interaction effects.

Results: The analysis included 3,956 children aged 12-13 years, 3,357 children aged 14-15 years, and 3,089 children aged 16-17 years. Males comprised 50.8-59.8% and females 40.2-49.2% across all ages. By the age of 17, 30.4% and 9.4% of the adolescents had depression and anxiety, respectively. Parental ACEs (≥2) were associated with increased depression risk at ages 12 to 13 years (RR = 1.42; 95% CI: 1.10-1.84) and at 16-17 years (RR = 1.19; 95% CI: 1.02-1.39). Exposure to ≥ 2 deprivation-related ACEs significantly increased the risk of depression across all ages, with relative risks ranging from 1.31 to 2.18. High threat-related ACEs (≥2) were associated with increased depression risk only at 12 to 13 years (RR = 2.01; 95% CI: 1.28-3.17). No significant interactions were observed.

Conclusions: The findings reinforce the ACEs model by showing that, at the population level, early identification of children exposed to early life deprivations rooted in financial crisis or familial adversities, combined with targeted interventions for both children and parents and supportive social policies, can reduce long-term mental health risks.

Aims: Loneliness is a common public health concern, particularly among mid- to later-life adults. However, its impact on early mortality (deaths occurring before reaching the oldest old age of 85 years) remains underexplored. This study examined the predictive role of loneliness on early mortality across different age groups using data from the Health and Retirement Study (HRS).

Methods: A retrospective cohort study was conducted using data from the 2010-2020 waves of the HRS, restricted to participants aged 50-84 years at baseline. Loneliness was measured using the 11-item UCLA Loneliness Scale, categorized into four levels: low/no loneliness (scores 11-13), mild loneliness (14-16), moderate loneliness (17-20) and severe loneliness (21-33). Cox proportional hazards models and time-varying Cox regression models with age as the time scale were created to evaluate the relationship between loneliness and early mortality, adjusting for sociodemographic, lifestyle, and physical and mental health factors.

Results: Among 6,392 participants, the overall mortality rate before the age of 85 years was 19.1 per 1,000 person-years. A dose-response relationship was observed, with moderate and severe loneliness associated with 23% (adjusted hazard ratio [aHR]: 1.23, 95% confidence interval [CI] = 1.02-1.48) and 36% (aHR: 1.36, 95% CI = 1.13-1.65) higher mortality risk, respectively. Significant associations existed for the 65-74-year-old (aHR = 1.37, 95% CI = 1.03-1.83) and 75-84-year-old (aHR = 1.77, 95% CI = 1.23-2.56) age groups in the fully-adjusted models, but not for the 50-64-year-old age group. Time-varying Cox models showed a stronger association for severe loneliness (aHR = 1.65, 95% CI = 1.37-1.99).

Conclusions: Loneliness is a significant predictor of mortality among older adults. Preventive and interventional programs targeting loneliness may promote healthy ageing.

Aims: There is a substantial body of literature on environmental risk associated with schizophrenia. Most research has largely been conducted in Europe and North America, with little representation of the rest of the world; hence generalisability of findings is questionable. For this reason, we performed a mapping review of studies on environmental risk for schizophrenia spectrum disorders, recording the country where they were conducted, and we linked our findings with publicly available data to identify correlates with the uneven global distribution. Our aim was to evaluate how universal is the 'common knowledge' of environmental risk for psychosis collating the availability of evidence across different countries and to generate suggestions for future research identifying gaps in evidence.

Methods: We performed a systematic search and mapping of studies in the PubMed and PsycINFO electronic databases reporting on exposure to environmental risk for schizophrenia including obstetric complications, paternal age, migration, urbanicity, childhood trauma, and cannabis use and subsequent onset of schizophrenia spectrum disorders. This search focused on articles published from the date of the first available publication until 31 May 2023. We recorded the country where they were conducted. We downloaded publicly available data on population size, measures of wealth, medical provisions, research investment, and of quality research outputs per country and performed regression analyses of each predictor with the number of studies and recruited cases in each country.

Results: We identified 308 publications that included a sample size of 445,000 patients with schizophrenia spectrum disorders. The majority were conducted in northern Europe and North America, with large parts of the world totally unrepresented. In the associations between the number of environmental risk studies for schizophrenia with potential predictors, we found that neither population nor wealth or research investment were strong predictors of research outputs in the field. Interestingly, the stronger correlations were found for number of researchers per population and for indicators of top-end scientific achievements, such as number of Nobel laureates per country.

Conclusions: Our results demonstrate a gap of knowledge due to the underrepresentation of studies on environmental risk of schizophrenia spectrum disorders in large parts of the world. This has implications not only in the generalisability of any findings from research conducted in the Northern hemisphere but also in our ability to progress in efforts to make causal inferences about biological pathways to schizophrenia. These findings reinforce the need to focus research on populations that are underrepresented in research and underserved in health care.

Introduction: Suicide rates in the United States have been increasing, necessitating an understanding of demographic variations by ethnicity, age, sex and method to inform effective prevention strategies.

Objective: To dissect suicide rates in the US population from 2001 to 2023 by age, sex, ethnicity, and method.

Methods: This retrospective observational study utilized suicide data and population statistics from the CDC's WISQARS database for the years 2001 (n = 30,418), 2018 (n = 48,132), 2020 (n = 45,721) and 2023 (n = 49,014). Cases were stratified by age, sex, ethnicity, and suicide method to assess trends and demographic differences.

Results: From 2001 to 2023, the overall US suicide rate rose from 10.7 to 14.6 per 100,000, with a temporary decrease in 2019 and 2020 (14.4 and 13.8, respectively). The primary driver of the increase was firearm-related suicides among White males, contributing 25.8% of the rise from 2001 to 2018 and 51.6% from 2020 to 2023. Decline between 2018 and 2020 was mainly due to reductions in firearm and drug-related suicides among White males, but firearm suicides surged again from 2020 to 2023. Additionally, firearm suicides among ethnic minorities, especially Black/African-American males, accounted for 14.0% of the increase during 2020-2023. Drug-related suicides also increased by 8.6% among White females aged 45 and older in the same period.

Conclusions: Firearm suicides are the leading factor in the changing suicide rates in the United States from 2001 to 2023, alongside rising drug-related suicides among White females. These trends highlight the necessity for targeted prevention efforts that consider demographic-specific factors and method accessibility.

Aim: The environment shapes the risk of psychosis. In particular, urbanicity, deprivation or inequality, migrant density and cannabis availability may not only influence psychosis incidence, but also the characteristics of individuals who arrive at clinical services. This study examined how socioeconomic factors influence the incidence and characteristics of cases of First-Episode Psychosis (FEP).

Methods: We analysed prospective data collected from the FEP early detection programme of Emilia-Romagna, a high-income Italian region. Participants were 1240 individuals aged 18-35 years, who presented at the public healthcare services for a FEP. Exposures were derived from area-level data of 331 municipalities. We used population density, socioeconomic deprivation, educational deprivation, economic inequality, migrant density (proportion of migrants), frequent cannabis use (proportion of people aged 15-19 years old who reported frequent cannabis use). Outcome measures were FEP incidence (cases/100 000 inhabitants at risk per year) and characteristics (age of onset, migrant status, unemployment, substance use, treatment lag [DUP], family and resource problems). We reviewed pertinent literature, and formulated a Directed Acyclic Graph to present causal assumptions and provide adjustment sets for Bayesian spatial and multilevel models of social causation. To compare the effects of different exposures, we computed Average Marginal Effects and report the outcome changes that correspond to one standard deviation change of the exposure, incidence rate ratios (IRR) or odds ratios (OR).

Results: The exposures and incidence of FEP displayed heterogeneous spatial distribution, with no spatially organized pattern. Accordingly, incidence and characteristics were best modelled as non-spatial, three-level hierarchical models. The incidence of FEP was influenced by population density (IRR, 1.14; 95% CrI, 1.03; 1.29), educational deprivation (IRR, 1.15; 95% CrI, 1.02; 1.31) and frequent cannabis use (IRR, 1.31; 95% CrI, 0.98; 1.82), more than socioeconomic deprivation. Higher migrant density in an area shortened the DUP on average by 3.4 months (95% CrI, -1.122; 0.76), while an increase of cannabis use of one standard deviation increased the DUP of 12.9 months (95% CrI, -2.86; 6229). Socioeconomic deprivation increased the likelihood of FEP cases being substance users (OR, 1.12; 95% CrI, 1.01; 1.26), while population density decreased it (OR, 0.91; 95% CrI, 0.83; 1.00).

Conclusions: Area-level socioeconomic features affect both the incidence and the characteristics of FEP, including the probability of individual being migrants, substance users or having a different DUP. Educational deprivation may function as a proxy for culture- or cognitive-related factors. Area-level socioeconomic data may inform public healthcare strategies for early identification and availability of tertiary

Aims: The Diagnostic and Statistical Manual of Mental Disorders - 5th Edition (DSM-5) and International Classification of Diseases - 11th Revision (ICD-11) employ different post-traumatic stress disorder (PTSD) criteria, necessitating updated prevalence estimates. Most of the existing evidence is still based on ICD-Tenth Revision and DSM-Fourth Edition criteria, leading to varied estimates across populations. This study provides current PTSD prevalence rates in the German general population, comparing DSM-5 and ICD-11 criteria and examines variations by age and gender.

Methods: In a 2016 cross-sectional survey of 2404 adults (18-94 years) representative of the German general population, participants completed the Life-Events-Checklist for DSM-5 (LEC-5) for trauma exposure and the PTSD Checklist for DSM-5 (PCL-5) for PTSD symptoms. Probable PTSD diagnoses were based on DSM-5-, ICD-11-algorithms and suggested cut-off scores. Chi-square and McNemar's tests were used to test differences in prevalence rates by diagnostic framework, age and gender.

Results: Of the total sample, 47.2% (n = 1135) reported experiencing at least one lifetime traumatic event (TE), with transportation accidents (7.3%) and life-threatening injuries (4.9%) being most common. Probable PTSD prevalence was 4.7% under both DSM-5 and ICD-11 criteria, and 2.6% based on a conservative cut-off normed for prevalence estimation. Gender and age were not significantly associated with TE exposure or PTSD prevalence, though trauma types varied: female participants more often reported sexual violence and severe suffering, while more male participants reported physical assaults and various types of accidents. DSM-5 and ICD-11 diagnostic algorithms had substantial yet not perfect agreement (κ = 0.62). Particularly within the re-experiencing symptoms, cluster agreement was only moderate (κ = 0.57). The cut-off method aligned more closely with DSM-5 (κ = 0.60) than ICD-11 algorithm (κ = 0.42).

Conclusions: This study provides updated PTSD prevalence estimates for the German general population and underscores differences between DSM-5 and ICD-11 in identifying cases, particularly with respect to re-experiencing symptoms. These findings emphasize that while overall PTSD prevalence rates under DSM-5 and ICD-11 criteria are similar, the diagnostic frameworks identify partially distinct cases, reflecting differences in symptom definitions. This highlights the need to carefully consider the impact of evolving diagnostic criteria when interpreting prevalence estimates and comparing results across studies.

Aims: Social determinants of health (SDHs) exert a significant influence on various health outcomes and disparities. This study aimed to explore the associations between combined SDHs and mortality, as well as adverse health outcomes among adults with depression.

Methods: The research included 48,897 participants with depression from the UK Biobank and 7,771 from the US National Health and Nutrition Examination Survey (NHANES). By calculating combined SDH scores based on 14 SDHs in the UK Biobank and 9 in the US NHANES, participants were categorized into favourable, medium and unfavourable SDH groups through tertiles. Cox regression models were used to evaluate the impact of combined SDHs on mortality (all-cause, cardiovascular disease [CVD] and cancer) in both cohorts, as well as incidences of CVD, cancer and dementia in the UK Biobank.

Results: In the fully adjusted models, compared to the favourable SDH group, the hazard ratios for all-cause mortality were 1.81 (95% CI: 1.60-2.04) in the unfavourable SDH group in the UK Biobank cohort; 1.61 (95% CI: 1.31-1.98) in the medium SDH group and 2.19 (95% CI: 1.78-2.68) in the unfavourable SDH group in the US NHANES cohort. Moreover, higher levels of unfavourable SDHs were associated with increased mortality risk from CVD and cancer. Regarding disease incidence, they were significantly linked to higher incidences of CVD and dementia but not cancer in the UK Biobank.

Conclusions: Combined unfavourable SDHs were associated with elevated risks of mortality and adverse health outcomes among adults with depression, which suggested that assessing the combined impact of SDHs could serve as a key strategy in preventing and managing depression, ultimately helping to reduce the burden of disease.