Epidemiology and Psychiatric Sciences最新文献

{"title":"Understanding the patterns and predictors of elevated psychological distress among humanitarian migrants compared to the host population: comparative matched analysis using two national data sources from Australia.","authors":"Demelash Woldeyohannes Handiso, Jacqueline A Boyle, Eldho Paul, Frances Shawyer, Graham Meadows, Joanne C Enticott","doi":"10.1017/S2045796025100139","DOIUrl":"https://doi.org/10.1017/S2045796025100139","url":null,"abstract":"<p><strong>Aims: </strong>Understanding patterns and predictors of elevated psychological distress (EPD) among humanitarian migrants compared to the host population is critical for designing effective mental health interventions. However, existing research presents conflicting findings on the prevalence of EPD. This study examined EPD prevalence and associated factors in humanitarian migrants and Australian-born adults using large population-level datasets.</p><p><strong>Methods: </strong>Kessler 6 scores (range 6-30) were dichotomised, and scores above 19 were defined as EPD and indicative of probable serious mental illness. Comparative 1:2 matched analysis used humanitarian migrant data from the Building a New Life in Australia and Australian-born comparators from the National Health Survey. Each humanitarian migrant was matched by age, sex and location with two Australian-born residents. Modified Poisson regression identified predictors of EPD in both groups.</p><p><strong>Results: </strong>EPD was higher among humanitarian migrants (17.2%, 95% CI: 15.5, 18.9) compared to Australian-born (14.5%, 95% CI: 13.3, 15.6), with an adjusted relative risk (aRR) with 95% confidence intervals (1.16%, 95% CI: 1.11, 1.21) after adjusting for key factors. In both groups, females had a higher aRR than males, with similar effect sizes: 1.06 (95% CI: 1.04, 1.08) among Australian-born and 1.04 (95% CI: 1.02, 1.07) among humanitarian migrants. The impact of age on distress was more pronounced in Australian-born individuals: compared to the 65+ age group, the youngest group (18-24 years) had an aRR of 1.36 (95% CI: 1.28, 1.43) for Australian-born and 1.19 (95% CI: 1.12, 1.27) for humanitarian migrants. Compared to excellent health, poor and fair self-rated health condition had an aRR of 2.13 (95% CI: 2.03, 2.26) and 1.69 (95% CI: 1.61, 1.79), respectively, for humanitarian migrants and 1.94 (95% CI: 1.82, 2.05) and 1.48 (95% CI: 1.43, 1.56), respectively, for Australian born. Australian-born individuals in the lowest-income quintile had higher distress (aRR: 1.11 [95% CI: 1.06-1.15]) compared to the highest-income quintile, with no significant income effect for humanitarian migrants. In both groups, females with poorer self-rated health had higher aRRs than females reporting excellent health.</p><p><strong>Conclusions: </strong>Although distress prevalence was higher in the humanitarian migrants, age and sex differences followed similar patterns in both groups. Income level was a factor in Australian-born adults but not in humanitarian migrants. Clinically, this highlights the need for culturally sensitive and group-specific mental health support. From a policy perspective, the use of matching methodology from large, separate datasets offers a valuable model for generating actionable insights, supporting the development of targeted and equitable mental health programmes.</p>","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e37"},"PeriodicalIF":5.9,"publicationDate":"2025-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144575097","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Experience of financial hardship and depression: a longitudinal population-based multi-state analysis.","authors":"Gustave Maffre Maviel, Alexandra Rouquette, Camille Davisse-Paturet, Arthur Descarpentry, Arnaud Sapin, Nathalie Bajos, Jean-Baptiste Hazo, Anne Pastorello, Josiane Warszawski, M Melchior, Cecile Vuillermoz","doi":"10.1017/S2045796025100115","DOIUrl":"https://doi.org/10.1017/S2045796025100115","url":null,"abstract":"<p><strong>Aims: </strong>Little is known about the effects of both financial hardship and people's perception of it on mental health. This study aimed to evaluate the effect of perceived financial hardship on individuals' depressive symptoms across several strata of objective financial situations.</p><p><strong>Methods: </strong>We used data from a four-wave French national population-based cohort (<i>N</i> = 14,236, 2020-2022) to assess the relationship between depressive symptoms and perceived financial hardship. Multi-state models (MSM) were used on a three-level scale for depressive symptom severity based on the Patient Health Questionnaire (PHQ-9). Analyses were stratified by household income to study the interaction with the objective financial situation.</p><p><strong>Results: </strong>We showed a link between perceived financial hardship and the onset and deterioration of depressive symptoms in subsequent waves, with effect sizes ranging from HR = 1.29 (0.87-1.90) to 2.23 (1.66-2.98). This association was stronger in the high-income population. There was no significant link between perceived financial hardship and the improvement of depressive symptomatology.</p><p><strong>Conclusions: </strong>This study confirms that perceived financial hardship is linked to the onset and deterioration of depressive symptoms. Furthermore, it suggests a stronger effect in high-income households, which could mean that the experience of financial hardship and the objective financial situation interact in their effect on mental health.</p>","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e36"},"PeriodicalIF":5.9,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144539602","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Needs for care of residents with schizophrenia spectrum disorders and association with daily activities and mood monitored with experience sampling method: the DIAPASON study - CORRIGENDUM.","authors":"Alessandra Martinelli, Miriam D'Addazio, Manuel Zamparini, Graham Thornicroft, Gabriele Torino, Cristina Zarbo, Matteo Rocchetti, Fabrizio Starace, Letizia Casiraghi, Mirella Ruggeri, Giovanni de Girolamo","doi":"10.1017/S2045796025100097","DOIUrl":"https://doi.org/10.1017/S2045796025100097","url":null,"abstract":"","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e35"},"PeriodicalIF":5.9,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144526977","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Investigating the association between the number of interpersonal supporters during first-time pregnancy and postpartum depression symptoms.","authors":"Junko Niimura, Syudo Yamasaki, Miharu Nakanishi, Satoshi Yamaguchi, Kaori Baba, Naomi Nakajima, Mitsuhiro Miyashita, Daniel Stanyon, Gemma Knowles, Jordan DeVylder, Mariko Hiraiwa-Hasegawa, Shuntaro Ando, Kiyoto Kasai, Atsushi Nishida","doi":"10.1017/S2045796025000241","DOIUrl":"https://doi.org/10.1017/S2045796025000241","url":null,"abstract":"<p><strong>Aims: </strong>First-year postpartum depression is a common mental health problem among first-time mothers. A younger age of pregnancy often compounds the challenge due to underlying factors such as poverty and limited educational achievement. This study aimed to examine the minimal number of interpersonal supporters during pregnancy associated with lower levels of postpartum depressive symptoms among first-time mothers.</p><p><strong>Methods: </strong>We obtained data from the population-based Mother-Infant/Newborn Tokyo Cohort (MINT cohort) in four municipalities in Tokyo on 429 first-time mothers who responded to two waves of surveys (early pregnancy and one month postpartum). They completed self-report measures of interpersonal support using one item from the Social Support Questionnaire and depressive symptoms using the Edinburgh Postnatal Depression Scale. Segmented regression analyses were conducted to determine the threshold at which the strength of the association changed between the number of interpersonal supporters and postpartum depressive symptoms, with adjustment for depressive symptoms in pregnancy. This analysis was also conducted with the sample stratified into young mothers (≤ 25 years) and older mothers (≥ 26 years).</p><p><strong>Results: </strong>In the overall sample, postpartum depressive symptoms were found to be lower among individuals with more than 3.0 supportive individuals (prepartum). Among young mothers, this threshold was higher, with lower symptom levels observed among those with at least 5.3 supporters. Only 22.9% of young first-time mothers had this level of interpersonal support, compared to 54.8% of all first-time mothers.</p><p><strong>Conclusions: </strong>Our results suggest that having four or more interpersonal supporters in early pregnancy is associated with lower levels of postpartum depressive symptoms among first-time mothers. Additionally, among young mothers, having six or more supporters was associated with lower postpartum depressive symptoms. These findings suggest that tailored strategies to increase supporters around first-time pregnant women might be beneficial depending on their age.</p>","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e34"},"PeriodicalIF":5.9,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144505186","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Developmental trajectories in mental health through adolescence and adulthood: does socio-economic status matter?","authors":"Christine Leonhard Birk Sørensen, Oleguer Plana-Ripoll, Ute Bültmann, Trine Nøhr Winding, Pernille Bach Steen, Karin Biering","doi":"10.1017/S2045796025100073","DOIUrl":"10.1017/S2045796025100073","url":null,"abstract":"<p><strong>Aims: </strong>This study aims to examine the different aspects of socio-economic status (SES) patterns in mental health from adolescence into adulthood by investigating the mean, prevalence, cumulative incidence and trajectories of several mental health measures, including depressive symptoms, mental disorder diagnosis and medication use. The different aspects of SES are investigated through the measures of subjective social status (SSS) in school, SSS in society, income and parental educational level.</p><p><strong>Methods: </strong>Individuals born in 1989 were followed from 2004 to 2021 with surveys at ages 15, 18, 21, 28 and 32 years, supplied with yearly register data. The mean level of depressive symptoms, yearly prevalence of medication use and cumulative incidence of mental disorder diagnosis were calculated for each SES group (low, middle and high) across each measure. Group-Based Trajectory Modelling (GBTM) was used to identify depressive symptom trajectories and logistic regressions were used to analyse the relative odds ratios (ROR) of membership to the different trajectory groups by characteristics.</p><p><strong>Results: </strong>Individuals with low SES at age 15 years across all SES measures showed higher mean depressive symptoms, prevalence of medication use and cumulative incidence of mental disorder diagnosis through adolescence and adulthood (age 15-32 years). Four depressive symptom trajectories were identified: low stable, moderate stable, decreasing and increasing trajectories. Being female, receiving medication or a mental disorder diagnosis in early adulthood and during the study period, having low SSS in school, parents not living together, being bullied, lacking support from teachers or classmates, lower levels of parents' support or higher school pressure resulted in higher RORs of membership to the other trajectory groups compared to the low stable trajectory, while having high SSS in society resulted in a lower ROR.</p><p><strong>Conclusions: </strong>This is the first study to detect the role of social support in relation to depressive symptom trajectories. While individuals with low social status consistently experienced more negative mental health outcomes than those with middle and high social status in the study period (age 15-32 years), low SSS showed the strongest associations. This indicates that SSS may capture vulnerable individuals not identified by traditional SES. Being female, having low SES, low social support, and other mental health outcomes were associated with higher odds of being in trajectories with more depressive symptoms. Preventive initiatives should therefore target individuals with such characteristics. It is worth exploring whether adolescents with increasing depressive symptoms could benefit from increased social support.</p>","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e33"},"PeriodicalIF":5.9,"publicationDate":"2025-06-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12188275/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144332661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Global, regional, and national trends and burden of opioid use disorder in individuals aged 15 years and above: 1990 to 2021 and projections to 2040.","authors":"Shuailei Wang, Yumiao He, Yuguang Huang","doi":"10.1017/S2045796025100085","DOIUrl":"10.1017/S2045796025100085","url":null,"abstract":"<p><strong>Aim: </strong>Opioid use disorder (OUD) is a medical condition associated with problematic opioid use, leading to addiction and severe life impairments. This research delivers an in-depth evaluation of OUD burden and trends at global, regional and national levels.</p><p><strong>Methods: </strong>This study analysed the global burden of OUD from 1990 to 2021 using data from the 2021 Global Burden of Diseases, Injuries, and Risk Factors Study. Key metrics included age-standardized prevalence, incidence, mortality and disability-adjusted life years (DALYs), disaggregated by gender, age, region, country and socio-demographic index (SDI) quintiles. The average annual percentage change described trends, while the age-period-cohort model evaluated age, period and cohort effects. A Bayesian Age-period-cohort model predicted future OUD trends from 2022 to 2040.</p><p><strong>Results: </strong>In 2021, the global burden of OUD remained substantial, with a total of 16,164,876 cases and a prevalence of 154.59 cases per 100,000 population (95% uncertainty interval [UI]: 131.06-181.26). In 2021, the global incidence of OUD reached 1,942,525 cases (95% UI: 1,643,342-2,328,363), and its global mortality reached 99,555 deaths (95% UI: 92,948-108,050), with DALYs amounting to 11,218,519 (95% UI: 9,188,658-13,159,551). Regionally, high SDI regions, particularly in the High-income North America, exhibited the greatest burden. Among countries, the United States faced the most severe burden and increase, with the highest prevalence (2014.62 per 100,000), incidence (151.84 per 100,000), mortality (15.37 per 100,000) and DALYs (1594.63 per 100,000), and all APPC values exceeding 5%. Males aged 20-39 years were the most affected demographic. However, forecasts indicate that the OUD burden among females will significantly increase over the next 20 years, with the prevalence and incidence expected to rise by 39% and 49%, respectively.</p><p><strong>Conclusions: </strong>The global burden of OUD has statistically significantly increased from 1990 to 2021. There are marked disparities across regions, countries and SDI levels. High-SDI regions, particularly High-income North America, bear the heaviest burden, with young males (aged 20-39 years) being the most affected groups. However, caution should be exercised regarding the female population, as the number of affected individuals is rapidly increasing.</p>","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e32"},"PeriodicalIF":5.9,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12175101/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144282967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Lifetime incidence and age of onset of mental disorders, and 12-month service utilization in primary and secondary care: a Finnish nationwide registry study.","authors":"Kimmo Suokas, Ripsa Niemi, Mai Gutvilig, John J McGrath, Kaisla Komulainen, Jaana Suvisaari, Marko Elovainio, Sonja Lumme, Sami Pirkola, Christian Hakulinen","doi":"10.1017/S2045796025100061","DOIUrl":"10.1017/S2045796025100061","url":null,"abstract":"<p><strong>Background: </strong>Previous studies have estimated the lifetime incidence, age of onset and prevalence of mental disorders, but none have used nationwide data covering both primary and secondary care, even though mental disorders are commonly treated in primary care. We aimed to determine lifetime incidence, age-specific incidence, age of onset and service utilization for diagnosed mental disorders.</p><p><strong>Methods: </strong>This register-based cohort study followed the entire population of Finland from 2000 to 2020. We estimated the cumulative incidence of diagnosed mental disorders with the Aalen-Johansen estimator, accounting for competing risks such as death and emigration. We also calculated age-specific incidence and 12-month service utilization as of 31 December 2019, providing diagnosis-, age- and gender-specific estimates.</p><p><strong>Results: </strong>We followed 6.4 million individuals for 98.5 million person-years. By age 100, lifetime incidence of any diagnosed mental disorder was 76.7% (95% CI, 76.6-76.7) in women and 69.7% (69.6-69.8) in men; in psychiatric secondary care, it was 39.7% (39.6-39.8) and 31.5% (31.4-31.6). At age 75, stricter estimates for non-organic disorders (ICD-10: F10-F99) were 65.6% (65.5-65.7) for women and 60.0% (59.9-60.1). Anxiety disorders (F40-F48) had the highest cumulative incidence. Median age of onset of non-organic mental disorders was 24.1 (interquartile range, 14.8-43.3 years) in women and 20.0 (interquartile range, 7.3-42.2 years) in men. Service utilization within 12 months was 9.0% for women and 7.7% for men.</p><p><strong>Conclusions: </strong>Most, though not all, individuals experience at least one type of mental disorder, often during youth. Capturing the overall occurrence of mental disorders requires including both primary and secondary care data.</p>","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e31"},"PeriodicalIF":5.9,"publicationDate":"2025-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12175100/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144233559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Peer-led recovery groups for people with psychosis in South Africa (PRIZE): results of a randomized controlled feasibility trial - CORRIGENDUM.","authors":"Laura Asher, Bongwekazi Rapiya, Julie Repper, Tarylee Reddy, Bronwyn Myers, Gill Faris, Inge Petersen, Charlotte Hanlon, Carrie Brooke-Sumner","doi":"10.1017/S2045796025000253","DOIUrl":"10.1017/S2045796025000253","url":null,"abstract":"","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e30"},"PeriodicalIF":5.9,"publicationDate":"2025-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12090021/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143991648","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Post-acute withdrawal syndrome (PAWS) after stopping antidepressants: a systematic review with meta-narrative synthesis.","authors":"Andri Rennwald, Michael P Hengartner","doi":"10.1017/S204579602500023X","DOIUrl":"10.1017/S204579602500023X","url":null,"abstract":"<p><strong>Aims: </strong>The literature on persistent antidepressant withdrawal symptoms is sparse. This systematic review is the first to examine the prevalence, duration, severity, risk/protective factors and treatment strategies for post-acute withdrawal syndrome (PAWS) following the discontinuation of antidepressant medications.</p><p><strong>Methods: </strong>We searched PubMed, Web of Science and PsycInfo, focusing on newer-generation antidepressants. The electronic database search was complemented with handsearching reference lists of pivotal studies. We included original studies in adults reporting on PAWS and providing data about epidemiology and clinical management of withdrawal symptoms persisting for at least 6 weeks.</p><p><strong>Results: </strong>The literature search yielded 1286 results, with 26 records assessed for eligibility, and seven studies fulfilled our selection criteria. Prevalence data were sparse, with one small cohort study reporting a 15% prevalence rate for PAWS in patients with panic disorder and agoraphobia. The duration of PAWS varied considerably across studies, ranging from 1.5 to 166 months. Long-term paroxetine use emerged as a potential risk factor for the development of PAWS. There was no reliable evidence to support the effectiveness of various treatment strategies, including the reinstatement of antidepressant medication, the use of benzodiazepines and the provision of cognitive-behavioral therapy.</p><p><strong>Conclusions: </strong>The current evidence on PAWS is sparse and predominantly of low certainty. The presence of withdrawal symptoms, lasting several months and possibly even years in some patients, underscores the need for further research with rigorous methodology. Large prospective cohort studies are needed to assess the epidemiology of PAWS, while randomized controlled trials are quired to test the efficacy of clinical interventions to treat PAWS.</p><p><strong>Prospero registration: </strong>CRD42023461793.</p>","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e29"},"PeriodicalIF":5.9,"publicationDate":"2025-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12090023/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143996490","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Identifying most important predictors for suicidal thoughts and behaviours among healthcare workers active during the Spain COVID-19 pandemic: a machine-learning approach.","authors":"Itxaso Alayo, Oriol Pujol, Jordi Alonso, Montse Ferrer, Franco Amigo, Ana Portillo-Van Diest, Enric Aragonès, Andrés Aragon Peña, Ángel Asúnsolo Del Barco, Mireia Campos, Meritxell Espuga, Ana González-Pinto, Josep Maria Haro, Nieves López-Fresneña, Alma D Martínez de Salázar, Juan D Molina, Rafael M Ortí-Lucas, Mara Parellada, José Maria Pelayo-Terán, Maria João Forjaz, Aurora Pérez-Zapata, José Ignacio Pijoan, Nieves Plana, Elena Polentinos-Castro, Maria Teresa Puig, Cristina Rius, Ferran Sanz, Cònsol Serra, Iratxe Urreta-Barallobre, Ronny Bruffaerts, Eduard Vieta, Víctor Pérez-Solá, Philippe Mortier, Gemma Vilagut","doi":"10.1017/S2045796025000198","DOIUrl":"10.1017/S2045796025000198","url":null,"abstract":"<p><strong>Aims: </strong>Studies conducted during the COVID-19 pandemic found high occurrence of suicidal thoughts and behaviours (STBs) among healthcare workers (HCWs). The current study aimed to (1) develop a machine learning-based prediction model for future STBs using data from a large prospective cohort of Spanish HCWs and (2) identify the most important variables in terms of contribution to the model's predictive accuracy.</p><p><strong>Methods: </strong>This is a prospective, multicentre cohort study of Spanish HCWs active during the COVID-19 pandemic. A total of 8,996 HCWs participated in the web-based baseline survey (May-July 2020) and 4,809 in the 4-month follow-up survey. A total of 219 predictor variables were derived from the baseline survey. The outcome variable was any STB at the 4-month follow-up. Variable selection was done using an L1 regularized linear Support Vector Classifier (SVC). A random forest model with 5-fold cross-validation was developed, in which the Synthetic Minority Oversampling Technique (SMOTE) and undersampling of the majority class balancing techniques were tested. The model was evaluated by the area under the Receiver Operating Characteristic (AUROC) curve and the area under the precision-recall curve. Shapley's additive explanatory values (SHAP values) were used to evaluate the overall contribution of each variable to the prediction of future STBs. Results were obtained separately by gender.</p><p><strong>Results: </strong>The prevalence of STBs in HCWs at the 4-month follow-up was 7.9% (women = 7.8%, men = 8.2%). Thirty-four variables were selected by the L1 regularized linear SVC. The best results were obtained without data balancing techniques: AUROC = 0.87 (0.86 for women and 0.87 for men) and area under the precision-recall curve = 0.50 (0.55 for women and 0.45 for men). Based on SHAP values, the most important baseline predictors for any STB at the 4-month follow-up were the presence of passive suicidal ideation, the number of days in the past 30 days with passive or active suicidal ideation, the number of days in the past 30 days with binge eating episodes, the number of panic attacks (women only) and the frequency of intrusive thoughts (men only).</p><p><strong>Conclusions: </strong>Machine learning-based prediction models for STBs in HCWs during the COVID-19 pandemic trained on web-based survey data present high discrimination and classification capacity. Future clinical implementations of this model could enable the early detection of HCWs at the highest risk for developing adverse mental health outcomes.</p><p><strong>Study registration: </strong>NCT04556565.</p>","PeriodicalId":11787,"journal":{"name":"Epidemiology and Psychiatric Sciences","volume":"34 ","pages":"e28"},"PeriodicalIF":5.9,"publicationDate":"2025-05-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12090031/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143979049","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}