Hostname: page-component-797576ffbb-58z7q Total loading time: 0 Render date: 2023-12-06T05:57:54.580Z Has data issue: false Feature Flags: { "corePageComponentGetUserInfoFromSharedSession": true, "coreDisableEcommerce": false, "useRatesEcommerce": true } hasContentIssue false

Child maltreatment, migration and risk of first-episode psychosis: results from the multinational EU-GEI study

Published online by Cambridge University Press:  28 October 2022

Giuseppe D'Andrea
Affiliation:
Department of Medical and Surgical Sciences, Bologna Transcultural Psychosomatic Team (BoTPT), University of Bologna, Bologna, Italy Community Mental Health Center of Sassuolo, Department of Mental Health and Drug Abuse, AUSL Modena, Modena, Italy
Jatin Lal
Affiliation:
Department of Medical and Surgical Sciences, Bologna Transcultural Psychosomatic Team (BoTPT), University of Bologna, Bologna, Italy
Sarah Tosato
Affiliation:
Section of Psychiatry, Department of Neuroscience, Biomedicine and Movement, University of Verona, Piazzale L.A. Scuro 10, 37134, Verona, Italy
Charlotte Gayer-Anderson
Affiliation:
ESRC Center for Society and Mental Health, Institute of Psychiatry, Psychology, and Neuroscience, King's College London, London, UK
Hannah E. Jongsma
Affiliation:
Centre for Transcultural Psychiatry ‘Veldzicht’, Balkbrug, The Netherlands University Centre for Psychiatry, University Medical Centre Groningen, Groningen, The Netherlands
Simona A. Stilo
Affiliation:
Department of Mental Health and Addiction Services, ASP Crotone, Crotone, Italy Department of Psychosis Studies, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, UK
Els van der Ven
Affiliation:
Department of Clinical, Neuro- and Developmental Psychology, Vrije Universiteit Amsterdam, Amsterdam, Netherlands
Diego Quattrone
Affiliation:
Social, Genetic and Developmental Psychiatry Centre, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, SE5 8AF, UK
Eva Velthorst
Affiliation:
Department of Psychiatry, Early Psychosis Section, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands Department of Psychiatry, Icahn School of Medicine at Mount Sinai, New York, USA
Domenico Berardi
Affiliation:
Department of Biomedical and Neuromotor Sciences (DIBINEM), Section of Psychiatry, University of Bologna, Bologna, Italy
Paulo Rossi Menezes
Affiliation:
University Hospital, Section of Epidemiology, University of São Paulo, São Paulo, Brazil
Celso Arango
Affiliation:
Department of Child and Adolescent Psychiatry, Institute of Psychiatry and Mental Health, Hospital General Universitario Gregorio Marañón, IiSGM, School of Medicine, Universidad Complutense, CIBERSAM, Centro de Investigación Biomédica en Red de Salud Mental, Madrid, Spain
Mara Parellada
Affiliation:
Department of Child and Adolescent Psychiatry, Institute of Psychiatry and Mental Health, Hospital General Universitario Gregorio Marañón, IiSGM, School of Medicine, Universidad Complutense, CIBERSAM, Centro de Investigación Biomédica en Red de Salud Mental, Madrid, Spain
Antonio Lasalvia
Affiliation:
Section of Psychiatry, Department of Neuroscience, Biomedicine and Movement, University of Verona, Piazzale L.A. Scuro 10, 37134, Verona, Italy
Caterina La Cascia
Affiliation:
Department of Biomedicine, Neuroscience and Advanced Diagnostics (BiND), Psychiatry Section, University of Palermo, Palermo, Italy
Laura Ferraro
Affiliation:
Department of Biomedicine, Neuroscience and Advanced Diagnostics (BiND), Psychiatry Section, University of Palermo, Palermo, Italy
Daniele La Barbera
Affiliation:
Department of Biomedicine, Neuroscience and Advanced Diagnostics (BiND), Psychiatry Section, University of Palermo, Palermo, Italy
Lucia Sideli
Affiliation:
Department of Human Science, LUMSA University, Rome, Italy
Julio Bobes
Affiliation:
Department of Medicine, Psychiatry Area, School of Medicine, Universidad de Oviedo, ISPA, Ineuropa, CIBERSAM, Oviedo, Spain
Miguel Bernardo
Affiliation:
Barcelona Clinic Schizophrenia Unit, Department of Medicine, Neuroscience Institute, Hospital Clinic, University of Barcelona, Institut d'Investigacions Biomèdiques August Pi i Sunyer, Centro de Investigación Biomédica en Red de Salud Mental, Barcelona, Spain
Julio Sanjuán
Affiliation:
Department of Psychiatry, School of Medicine, Universidad de Valencia, Centro de Investigación Biomédica en Red de Salud Mental, Valencia, Spain
Jose Luis Santos
Affiliation:
Department of Psychiatry, Servicio de Psiquiatría Hospital “Virgen de la Luz”, Cuenca, Spain
Manuel Arrojo
Affiliation:
Department of Psychiatry, Psychiatric Genetic Group, Instituto de Investigación Sanitaria de Santiago de Compostela, Complejo Hospitalario Universitario de Santiago de Compostela, Santiago, Spain
Cristina Marta Del-Ben
Affiliation:
Neuroscience and Behavior Department, Ribeirão Preto Medical School, University of São Paulo, São Paulo, Brazil
Giada Tripoli
Affiliation:
Department of Psychosis Studies, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, UK Department of Biomedicine, Neuroscience and Advanced Diagnostics (BiND), Psychiatry Section, University of Palermo, Palermo, Italy Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties, University of Palermo, Piazza delle Cliniche, 290127 Palermo, Italy
Pierre-Michel Llorca
Affiliation:
Université Clermont Auvergne, EA 7280 Npsydo, Clermont-Ferrand, France
Lieuwe de Haan
Affiliation:
Department of Psychiatry, Amsterdam UMC, Amsterdam, The Netherlands
Jean-Paul Selten
Affiliation:
School for Mental Health and Neuroscience, University of Maastricht, Maastricht, The Netherlands
Andrea Tortelli
Affiliation:
Etablissement Public de Santé Maison Blanche, Paris, France
Andrei Szöke
Affiliation:
Univ Paris Est Creteil, INSERM, IMRB, AP-HP, Hôpitaux Universitaires « H. Mondor », DMU IMPACT, Fondation Fondamental, F-94010 Créteil, France
Roberto Muratori
Affiliation:
Department of Mental Health and Pathological Addiction, Local Health Authority, Bologna, Italy
Bart P. Rutten
Affiliation:
Department of Psychiatry and Neuropsychology, School for Mental Health and Neuroscience, Maastricht University Medical Centre, Maastricht, The Netherlands
Jim van Os
Affiliation:
Department of Psychosis Studies, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, UK Department of Psychiatry and Neuropsychology, School for Mental Health and Neuroscience, Maastricht University Medical Centre, Maastricht, The Netherlands
Peter B. Jones
Affiliation:
Department of Psychiatry, University of Cambridge, Cambridge, UK
James B. Kirkbride
Affiliation:
PsyLife Group, Division of Psychiatry, UCL, London, UK
Robin M. Murray
Affiliation:
Department of Psychosis Studies, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, UK
Marta di Forti
Affiliation:
Social, Genetic and Developmental Psychiatry Centre, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, SE5 8AF, UK
Ilaria Tarricone*
Affiliation:
Department of Medical and Surgical Sciences, Bologna Transcultural Psychosomatic Team (BoTPT), University of Bologna, Bologna, Italy Department of Mental Health and Pathological Addiction, Local Health Authority, Bologna, Italy
Craig Morgan
Affiliation:
ESRC Center for Society and Mental Health, Institute of Psychiatry, Psychology, and Neuroscience, King's College London, London, UK
*
Author for correspondence: Ilaria Tarricone, E-mail: ilaria.tarricone@unibo.it
Rights & Permissions [Opens in a new window]

Abstract

Background

Child maltreatment (CM) and migrant status are independently associated with psychosis. We examined prevalence of CM by migrant status and tested whether migrant status moderated the association between CM and first-episode psychosis (FEP). We further explored whether differences in CM exposure contributed to variations in the incidence rates of FEP by migrant status.

Methods

We included FEP patients aged 18–64 years in 14 European sites and recruited controls representative of the local populations. Migrant status was operationalized according to generation (first/further) and region of origin (Western/non-Western countries). The reference population was composed by individuals of host country's ethnicity. CM was assessed with Childhood Trauma Questionnaire. Prevalence ratios of CM were estimated using Poisson regression. We examined the moderation effect of migrant status on the odds of FEP by CM fitting adjusted logistic regressions with interaction terms. Finally, we calculated the population attributable fractions (PAFs) for CM by migrant status.

Results

We examined 849 FEP cases and 1142 controls. CM prevalence was higher among migrants, their descendants and migrants of non-Western heritage. Migrant status, classified by generation (likelihood test ratio:χ2 = 11.3, p = 0.004) or by region of origin (likelihood test ratio:χ2 = 11.4, p = 0.003), attenuated the association between CM and FEP. PAFs for CM were higher among all migrant groups compared with the reference populations.

Conclusions

The higher exposure to CM, despite a smaller effect on the odds of FEP, accounted for a greater proportion of incident FEP cases among migrants. Policies aimed at reducing CM should consider the increased vulnerability of specific subpopulations.

Type
Original Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
Copyright © The Author(s), 2022. Published by Cambridge University Press

Introduction

The World Health Organization (WHO) defines child maltreatment (CM) as ‘all forms of physical and/or emotional ill-treatment, sexual abuse, neglect or negligent treatment or commercial or other exploitation, resulting in actual or potential harm to child's health, survival, development or dignity in the context of a relationship of responsibility, trust or power’ (World Health Organization (WHO), 2006). Estimates of lifetime prevalence of maltreatment vary by sex and region (Moody, Cannings-John, Hood, Kemp, & Robling, Reference Moody, Cannings-John, Hood, Kemp and Robling2018). Most studies have been conducted in the United States (US), where pooled estimates range between 8.0% and 36.5% according to sex and type of abuse (Stoltenborgh, Bakermans-Kranenburg, Alink, & van Ijzendoorn, Reference Stoltenborgh, Bakermans-Kranenburg, Alink and van Ijzendoorn2015). There is some evidence that the prevalence of CM varies by social class (or income levels) and ethnicity (Pelton, Reference Pelton2015; Sedlak et al., Reference Sedlak, Mettenburg, Basena, Peta, McPherson and Greene2010). The fourth National Incidence Study of Child Abuse and Neglect (NIS-4) in the US reported that risk of maltreatment was 5.8-times higher for children who grew up in poverty and nearly 2 times higher for those of black ethnicity (Sedlak et al., Reference Sedlak, Mettenburg, Basena, Peta, McPherson and Greene2010). However, this report did not consider socioeconomic status (SES), introducing a potential bias in the analyses, given the high degree of economic stratification of the US and its implications (Adler & Rehkopf, Reference Adler and Rehkopf2008; Braveman, Cubbin, Egerter, Williams, & Pamuk, Reference Braveman, Cubbin, Egerter, Williams and Pamuk2010; Kawachi, Daniels, & Robinson, Reference Kawachi, Daniels and Robinson2005). In fact, more recent studies have shown that, when accounting for SES, differences in the estimates of CM between ethnic groups disappear, whereas the effect of socioeconomic disadvantage remains (Cancian, Yang, & Slack, Reference Cancian, Yang and Slack2013; Dworsky, Courtney, & Zinn, Reference Dworsky, Courtney and Zinn2007; Kim & Drake, Reference Kim and Drake2018; Putnam-Hornstein, Needell, King, & Johnson-Motoyama, Reference Putnam-Hornstein, Needell, King and Johnson-Motoyama2013; Slack, Lee, & Berger, Reference Slack, Lee and Berger2007). That is, ethnic disparities in risk of socioeconomic disadvantage may drive the observed differences in CM. In turn, the association with socioeconomic disadvantage undoubtedly reflects the consequence of structural inequalities in the access to basic material resources. European studies, such as the Netherlands' Prevalence study of Maltreatment (NPM) (Alink, Euser, van IJzendoorn, & Bakermans-Kranenburg, Reference Alink, Euser, van IJzendoorn and Bakermans-Kranenburg2013; Euser, van IJzendoorn, Prinzie, & Bakermans-Kranenburg, Reference Euser, van IJzendoorn, Prinzie and Bakermans-Kranenburg2011; van Berkel, Prevoo, Linting, Pannebakker, & Alink, Reference van Berkel, Prevoo, Linting, Pannebakker and Alink2020), yielded similar results, showing that migrants and their descendants from different ethnic backgrounds had from around a 2- to 3- fold increased likelihood of maltreatment compared with Dutch (van Berkel et al., Reference van Berkel, Prevoo, Linting, Pannebakker and Alink2020). These differences were partly explained by measures of social disadvantage (Alink et al., Reference Alink, Euser, van IJzendoorn and Bakermans-Kranenburg2013).

CM is linked to subsequent development of physical and mental disorders (World Health Organization (WHO), 2006). Several meta-analyses (Arango et al., Reference Arango, Dragioti, Solmi, Cortese, Domschke, Murray and Fusar-Poli2021; Varese et al., Reference Varese, Smeets, Drukker, Lieverse, Lataster, Viechtbauer and Bentall2012) have shown that the odds of developing psychosis among those exposed to maltreatment is around 2–3 times the odds among those not exposed. The effect of CM on odds of psychosis is cumulative and affected by type, age of exposure, and severity (Morgan et al., Reference Morgan, Gayer-Anderson, Beards, Hubbard, Mondelli, Di Forti and Fisher2020).

Psychosis incidence is higher among migrants and their descendants (Morgan, Knowles, & Hutchinson, Reference Morgan, Knowles and Hutchinson2019; Selten, Van Der Ven, & Termorshuizen, Reference Selten, Van Der Ven and Termorshuizen2019) and vary by region of origin, being higher among those from non-Western countries, and by region of resettlement, peaking in sites where incidence of psychosis is higher also for local populations (Termorshuizen et al., Reference Termorshuizen, Van Der Ven, Tarricone, Jongsma, Gayer-Anderson, Lasalvia and Selten2020). Morgan et al. ( Reference Morgan, Knowles and Hutchinson2019) have proposed a ‘socio-developmental pathway’ to psychoses, according to which greater exposure to social adversities across the life course accounts for the elevated rates of psychosis in migrants and their descendants. In two recent studies (Gatt et al., Reference Gatt, Alexander, Emond, Foster, Hadfield, Mason-Jones and Wu2020; Solà-Sales et al., Reference Solà-Sales, Pérez-González, Van Hoey, Iborra-Marmolejo, Beneyto-Arrojo and Moret-Tatay2021) the effect of traumatic events on mental health was lower among migrants compared to non-migrants, hindering at possibly greater resilience. However, none of these studies specifically looked at the risk of psychosis.

In this context, we aimed to: (1) estimate the prevalence of specific CM subtypes among migrants considering generational status and regions of origin compared with individuals of host country's ethnicity; (2) test whether migrant status moderated the association between CM and first-episode psychosis (FEP); (3) examine whether CM exposure accounted for a more sizeable proportion of incident FEP cases among migrants compared with the reference populations.

Methods

Study design and participants

This study is part of the European network of national schizophrenia networks studying Gene–Environment Interactions (EU-GEI) project (http://www.eu-gei.eu) (Gayer-Anderson et al., Reference Gayer-Anderson, Jongsma, Di Forti, Quattrone, Velthorst, de Haan and Morgan2020). Participants were recruited between 1 May 2010, and 1 April 2015 from sixteen centers in England, the Netherlands, Spain, France, Italy and Brazil. All persons aged 18–64 years who made contact with mental health services for a probable FEP were invited to participate (Gayer-Anderson et al., Reference Gayer-Anderson, Jongsma, Di Forti, Quattrone, Velthorst, de Haan and Morgan2020). Cases were included if they met International Classification of Disease (ICD)-10 criteria for psychotic disorders (F20–33), ascertained using the Operational Criteria Checklist (OPCRIT) algorithm (Quattrone et al., Reference Quattrone, Di Forti, Gayer-Anderson, Ferraro, Jongsma, Tripoli and Cristofalo2019; Williams et al., Reference Williams, Farmer, Ackenheil, Kaufmann, McGuffin and Group1996). Assessment with OPCRIT relied on semistructured clinical interviews or review of case notes and other relevant information.

In each center, we recruited population-based controls aged 18–64 years using a mixture of random and quota-sampling strategies, to maximize representativeness to the population-at-risk by age, sex and ethnicity. Quotas were derived from the most accurate local demographic data. Strategies for control recruitment included random sampling from lists of all postal addresses; stratified sampling via General Practitioners lists from randomly selected surgeries; and advertisement via multiple channels (internet, newspapers, leaflets). Individuals with a history of psychotic disorder, or taking anti-psychotic medication, were not eligible (Di Forti et al., Reference Di Forti, Quattrone, Freeman, Tripoli, Gayer-Anderson, Quigley and van der Ven2019; Gayer-Anderson et al., Reference Gayer-Anderson, Jongsma, Di Forti, Quattrone, Velthorst, de Haan and Morgan2020). As shown by a previous EU-GEI publication, controls were broadly representative of the population-at-risk on sex and ethnicity, but in some sites they were younger (Jongsma et al., Reference Jongsma, Gayer-Anderson, Tarricone, Velthorst, van der Ven, Quattrone and Kirkbride2020).

Ethical approval was granted in each center. All participants gave written informed consent.

Measures

Outcome

Our primary outcome was case–control status, with cases defined as receiving an OPCRIT-confirmed ICD-10 diagnosis of any psychotic disorder (ICD-10 codes F20–F33).

Socio-demographic and migrant status

We collected data on age at first diagnosis/control recruitment, sex, ethnicity, education (years), parental social class (long-term unemployed, working class, intermediate, salariat), and personal and parental country of birth, using an amended version of the Medical Research Council Socioeconomic Schedule (Mallet, Reference Mallet1997). First-generation migrants were identified based on their birth country. Information on parental country of birth and site-specific ethnicity were used to identify migrants' descendants. Consistent with a previous EU-GEI publication (Termorshuizen et al., Reference Termorshuizen, Van Der Ven, Tarricone, Jongsma, Gayer-Anderson, Lasalvia and Selten2020), migrants and their descendants were also grouped according to personal or parental regions of origin into individuals from Western and non-Western countries. Western countries included Europe, US, Canada, Australia, New Zealand, and countries of the former Soviet Union with a predominantly Christian religion. Non-Western countries included the following areas: Middle East & The Maghreb; sub-Saharan Africa; Asia; Latin America; The Caribbean Islands and the French overseas departments. We refer to the prior paper for more detailed information (Termorshuizen et al., Reference Termorshuizen, Van Der Ven, Tarricone, Jongsma, Gayer-Anderson, Lasalvia and Selten2020). In each site, the reference population was composed of individuals of host country's ethnicity, born in the country of recruitment and with both parents born in that same country. We excluded all individuals who could not be ascribed with certainty to any of the aforementioned groups.

Child maltreatment

CM was assessed using the Childhood Trauma Questionnaire (CTQ) (Bernstein et al., Reference Bernstein, Stein, Newcomb, Walker, Pogge, Ahluvalia and Zule2003). It consists of five items for each subtype of trauma (emotional abuse, emotional neglect, physical abuse, physical neglect, and sexual abuse). Every item is rated on a 5-point Likert-type scale (1 = never true, 5 = very often true) and allows for computation of both mean scores for each type of abuse (range 5–25) and categorical scores, with fixed threshold values for each subdomain (none to minimal, slight to moderate, moderate to severe, severe to extreme) (Bernstein et al., Reference Bernstein, Stein, Newcomb, Walker, Pogge, Ahluvalia and Zule2003). Walker et al. (Walker et al. Reference Walker, Gelfand, Katon, Koss, Von Korff, Bernstein and Russo1999) have proposed another procedure for the CTQ rating and, based on a comparison between self-reported CTQ scores and expert ratings of structured clinical interviews, determined threshold scores with both sensitivity and specificity ⩾0.85 (online Supplementary Materials). Walker et al.'s (Walker et al., Reference Walker, Gelfand, Katon, Koss, Von Korff, Bernstein and Russo1999) method provides dichotomized measures of emotional abuse, emotional neglect, physical abuse, physical neglect, and sexual abuse. Conform previous analyses on the EU-GEI dataset (Sideli et al., Reference Sideli, Schimmenti, La Barbera, La Cascia, Ferraro, Aas and van der Ven2022), CM exposure was operationalized as a dichotomous variable using the 80th percentile of the mean CTQ score in the control group as cut-off value. For the moderation analyses, we used the mean of the CTQ subscales scores (range: 5–25) as an overall maltreatment score.

Other exposures

Parental history of psychosis and mental illness was recorded using the Family Interview for Genetic Studies questionnaire (Maxwell, Reference Maxwell1992). Current cannabis use (no/yes) was derived from a modified version of the Cannabis Experience Questionnaire (Di Forti et al., Reference Di Forti, Morgan, Dazzan, Pariante, Mondelli, Marques and Murray2009).

Missing data

The proportion of participants with missing data ranged from none on sex and country of recruitment to 269 (13.5%) on parental psychosis. Complete data were available for 1502 participants (75.4%). Missing data were handled by multiple imputation (details in online Supplementary Materials).

Statistical analyses

Differences in exposures and covariates distribution by case–control status on the whole sample and by migrant status were conducted using χ2 tests for categorical variables and student's t tests for continuous variables.

The analyses on prevalence of CM by migrant status were conducted only on controls, who were broadly representative of the population at risk within each catchment area (Gayer-Anderson et al., Reference Gayer-Anderson, Jongsma, Di Forti, Quattrone, Velthorst, de Haan and Morgan2020). We estimated the prevalence ratios (PR) of each type of CM [dichotomized yes/no (Walker et al., Reference Walker, Gelfand, Katon, Koss, Von Korff, Bernstein and Russo1999)] by migrant generational status and by migrants' region of origin (with white majority as reference population) using Poisson regression with robust variance (Barros & Hirakata, Reference Barros and Hirakata2003; Santos et al., Reference Santos, Fiaccone, Oliveira, Cunha, Barreto, Do Carmo and Amorim2008). PRs were adjusted for age, sex, parental social class, educational attainment, parental mental illness, and site of recruitment. Models' goodness-of-fit was tested with χ2d tests.

To test our second hypothesis, we used a mixed-effects logistic regression with random intercepts at the site level to account for clustering by recruitment site (n = 14) while estimating odds ratios (OR) for case–control status from the following independent variables: CM (CTQ mean score), migrant status, age, sex, education, parental psychosis, cannabis use, and parental social class at birth. We used polychoric correlations to describe associations between confounders and exposures and estimated Variance Inflation Factor to further check for multicollinearity (online Supplementary Materials). We finally introduced a multiplicative interaction term to the model between migrant status and CTQ mean score. The CTQ score was standardized for the purpose of interaction analysis. We performed a likelihood ratio test to assess whether the addition of the interaction term improved the model (Aiken, West, & Reno, Reference Aiken, West and Reno1991). Then, we estimated the stratified OR for each migrant group from the model with the interaction term by combining the coefficients. Finally, we used the command ‘margins’ in Stata 17 to estimate the adjusted predicted probabilities with 95% CI, by migrant status, of being a case at fixed values of standardized mean CTQ using the marginal standardization method (Muller & MacLehose, Reference Muller and MacLehose2014). Predicted probabilities at the selected values of CTQ along with standard errors and p values were reported in online Supplementary Materials and were plotted for better reporting the interaction effect. Residual diagnostics of the mixed models were performed using the R package ‘DHARMa’ (Hartig, Reference Hartig2022) (online Supplementary Materials). We controlled the false discovery rate using the Benjamini-Hochberg procedure (Benjamini & Hochberg, Reference Benjamini and Hochberg1995), tolerating a 10% false discovery rate. Both unadjusted and adjusted p values are reported.

Lastly, we refitted the same mixed-effect logistic regression models substituting the total CTQ score for a dichotomous variable obtained using the 80th percentile of CTQ total score of controls as cut-off value (Sideli et al., Reference Sideli, Schimmenti, La Barbera, La Cascia, Ferraro, Aas and van der Ven2022). We used the command ‘punafcc’ in Stata 17 to calculate the population attributable fraction (PAF) for CM exposure by migrant status. Assuming causality, PAF represents the estimated fraction of FEP cases that would not have occurred if there had been no exposure to CM.

Sensitivity analyses were conducted on complete cases only for the main outcomes (online Supplementary Materials).

Analyses were performed using Rstudio R version 3.6.3 (Rstudio Team (2020). Rstudio: Integrated Development for R. Rstudio, PBC, Boston, MA URL http://www.rstudio.com/) and Stata 17 (StataCorp. 2021. Stata Statistical Software: Release 17. College Station, TX: StataCorp LLC). All analyses were conducted following imputation of the missing values.

Results

Between 2010 and 2015, 1130 cases and 1497 controls were recruited and assessed across 17 sites in 6 countries (UK, The Netherlands, France, Spain, Italy, and Brazil). We excluded Brazil (n = 494) due to differences in the context and in the migration patterns with respect to European countries. We also excluded the French site of Maison-Blanche (n = 36) as no control was recruited there, and the Spanish site of Santiago (n = 64) due to the underrepresentation of migrants (n = 2, 3.2%). For 40 participants, it was not possible to determine their region of origin or migrant generational status due to missing data on either personal/parental country of birth or ethnicity. This resulted in an analysis sample of 1991 individuals (849 cases and 1142 controls).

Characteristics of the case–control sample

The distribution of exposures and covariates in the case–control sample is described in Table 1. Cases were more likely than controls to be men (χ2 = 46.7; p < 0.001) and younger (t = 10.2; p < 0.001). They also received less education (t = 9.4; p < 0.001) and more often reported lower parental SES (χ2 = 13.3; p = 0.004). Parental mental illness (χ2 = 24.3; p < 0.001) and psychosis (χ2 = 33.8; p < 0.001) were more frequent among cases. The latter were more likely to report use of cannabis (χ2 = 41.4; p < 0.001) and scored higher than controls on the total CTQ score (t = −14.5; p < 0.001). Finally, emotional abuse2 = 101.2; p < 0.001), emotional neglect2 = 71.8; p < 0.001), physical abuse2 = 95.4; p < 0.001), physical neglect2 = 129.2; p < 0.001), and sexual abuse2 = 47.1; p < 0.001), were more frequent among cases than among controls. Distribution of exposures and covariates among cases and controls by migrant generational status and region of origin is detailed in online Supplementary Materials (Tables S3, S4).

Table 1. Distribution of exposures and covariates in the case–control sample

s.d., standard deviation.

a Pearson's χ2 test.

b t student's test

Defined as mean CTQ > 80th percentile of the control group.

Prevalence of the different forms of CM among migrants and descendants compared with reference population (controls only analyses)

In controls, the distribution of CM exposure differed significantly among the study groups, with both first-generation migrants (N = 56, 26.2%) and their descendants (N = 36, 22.6%) being more likely to report CM compared with the reference populations (N = 114, 14.8%) (χ2 = 17.2, p < 0.001). Similarly, a considerably greater proportion of migrants from both non-Western (N = 63, 26.0%) and Western countries (N = 29, 22.1%) was exposed to CM compared with the reference populations (N = 114, 14.8%) (χ2 = 17.2, p < 0.001).

Unadjusted and adjusted PR of the measured subtypes of CM by migrant generational status and region of origin are presented in Tables 2 and 3 (Tables 2, 3). The prevalence of emotional abuse (aPR = 1.47, 95% CI 1.06–2.04), physical abuse (aPR = 3.28, 95% CI 2.18–4.94), physical neglect (aPR = 1.96, 95% CI 1.52–2.53), and sexual abuse (aPR = 1.97, 95% CI 1.17–3.33) were higher among first-generation migrants compared with the reference populations. Children of migrants reported greater PRs of physical abuse (aPR = 2.50, 95% CI 1.61–3.88) and sexual abuse (aPR = 1.97, 95% CI 1.17–3.33) than the reference populations. Considering region of origin, migrants and their descendants from non-Western countries showed higher prevalence of emotional abuse (aPR = 1.50, 95% CI 1.09–2.06), physical abuse (aPR = 3.57, 95% CI 2.38–5.36), physical neglect (aPR = 1.83, 95% CI 1.40–2.40), and sexual abuse (aPR = 2.44, 95% CI 1.50–3.97) compared with the reference populations. People with a Western migrant heritage only had a higher prevalence of physical abuse (aPR = 2.08, 95% CI 1.24–3.48) in relation to the reference group.

Table 2. Unadjusted and adjusted PRs of CM subtypes by migrant generational status

PR, prevalence ratio; 95% CI, 95% confidence interval; aPR, adjusted prevalence ratio; EA, emotional abuse; EN, emotional neglect; PA, physical abuse; PN, physical neglect; SA, sexual abuse.

PRs were adjusted for age, gender, educational attainment, parental social class, parental mental illness, and site of recruitment. PRs in bold are significant at 0.05 level.

Table 3. Unadjusted and adjusted PRs of CM subtypes by migrants' area of origin

PR, prevalence ratio; 95% CI, 95% confidence interval; aPR, adjusted prevalence ratio; EA, emotional abuse; EN, emotional neglect; PA, physical abuse; PN, physical neglect; SA, sexual abuse.

PRs were adjusted for age, gender, educational attainment, parental social class, parental mental illness, and site of recruitment. PRs in bold are significant at 0.05 level.

Associations of CM with FEP in reference populations and migrant populations

Both first-generation migrants (OR 1.88, 95% CI 1.49–2.38; p < 0.001, p adj < 0.001) and their descendants (OR 1.47, 95% CI 1.11–1.95; p = 0.008, p adj = 0.014) had greater odds of FEP compared with the reference populations in the unadjusted mixed-effects model. Migrants and their descendants from non-Western countries also had increased odds of FEP compared with the reference population (OR 2.03, 95% CI 1.60–2.58; p < 0.001, p adj < 0.001); odds were not statistically different for migrants from Western countries compared with the reference population (OR 1.24, 95% CI 0.91–1.68; p = 0.171, p adj = 0.243).

In the model with migrants classified by generational status, every 1-s.d. increase in the CTQ score was associated with almost 2-fold increased odds of FEP (OR 1.98, 95% CI 1.75–2.23; p < 0.001, p adj < 0.001). The addition of the interaction term ‘migrant × maltreatment’ improved the model (likelihood test ratio: χ2 = 11.4, p = 0.003, p adj = 0.007). The stratified OR from the model with the interaction term suggested that the association between CTQ and FEP was stronger among individuals from the reference population (OR 2.42, 95% CI 2.03–2.89; p < 0.001, p adj < 0.001) than among first-generation migrants (OR 1.54, 95% CI 1.26–1.89; p < 0.001, p adj < 0.001) or their descendants (OR 1.83, 95% CI 1.43–2.33; p < 0.001, p adj < 0.001) (Table 4). Figure 1a shows the estimated predicted probabilities of FEP among the different groups for varying scores of CTQ. For increasing values of CTQ, predicted probabilities of FEP were higher among the reference population compared with first-generation migrants or their children (Fig. 1a).

Fig. 1. Predicted probabilities of FEP by CM and migrant status. (a) represents predicted probabilities of psychosis by CM and migrant generational status. (b) represent predicted probabilities of psychosis by CM and Western/non-Western migrant status.

Table 4. Odds ratios of FEP by migrant status and migrant-CM interaction

OR, odds ratio; 95% CI, 95% confidence interval; PAF, proportion of attributable fraction.

a Odds ratios were adjusted for age, gender, educational attainment, cannabis use, parental social class, and parental psychosis.

b Odds ratios were estimated by linear combination of the coefficients from the interaction model.

c PAFs were estimated from the fully adjusted models with the interaction terms.

All models were mixed effect models accounting for clustering by site of recruitment (n = 14).

Similarly, in the model with migrants and their descendants classified by region of origin, the addition of the interaction term ‘migrant × maltreatment’ improved the model (likelihood test ratio: χ2 = 11.4, p = 0.003, p adj = 0.005). The association between CM and FEP was more robust among individuals from the reference population (OR 2.42, 95% CI 2.03–2.89; p < 0.001, p adj < 0.001) compared with migrants from Western (OR 1.95, 95% CI 1.43–2.65; p < 0.001, p adj < 0.001) or non-Western countries (OR 1.57, 95% CI 1.31–1.87; p < 0.001, p adj < 0.001) (Table 4). Predicted probabilities of FEP by migrant status and CM are plotted in Fig. 1b. The effect of higher CTQ scores on the probability of FEP was weaker among migrants and their descendants from Western and non-Western countries compared with the reference population (Fig. 1b).

Population attributable fractions

The PAF for CM exposure in the total sample was 31.9% (95% CI 28.9–34.7%). PAF was higher for among first-generation migrants (37.3%, 95% CI 33.8–40.5%) and descendants (36.7%, 95% CI 33.4–39.9%) as well as among Western (34.7%, 95% CI 31.5–37.7%) and non-Western migrants (37.9%, 95% CI 34.5–41.2%) compared with the reference populations (27.9%, 95% CI 25.4–30.3%).

Sensitivity analyses

Results of the analyses conducted in the complete-case sample did not differ from imputed analyses (online Supplementary Materials).

Discussion

Main findings

We found that migrants and their descendants and participants from non-Western countries reported greater exposure to CM specific subtypes compared with the reference population. In adjusted models, first-generation migrants and those of non-Western heritage also had about 1.5-times higher odds of FEP compared with the reference group. Migrant status moderated the association between CM and FEP by attenuating the effect of trauma on the predicted probabilities of psychosis. Finally, according to our results, CM accounted for about one third of incident FEP cases (PAF = 31.9%, 95% CI 28.9–34.7%). Estimated PAFs were higher among all migrant groups compared with the reference populations.

Strengths and limitations

To the best of our knowledge, this is the first study to investigate the association between CM and FEP between migrants and majority populations. We used data from a large case–control study. All relevant information was collected using validated instruments and a standardized data-entry was ensured to minimize collection biases across sites. The broad range of confounders that have been considered provides more certainty that the associations we found are independent of potentially confounding effects. The control recruitment strategy was designed to obtain a sample representative of the population-at-risk, allowing us to calculate reliable estimates of the prevalence of CM in local populations. CM measures for the estimation of prevalence were operationalized using highly sensitive and specific cut-off scores which were proposed in 1999 by Walker et al. (Walker et al. Reference Walker, Gelfand, Katon, Koss, Von Korff, Bernstein and Russo1999) and have consistently been utilized over the past years (Brühl, Kley, Grocholewski, Neuner, & Heinrichs, Reference Brühl, Kley, Grocholewski, Neuner and Heinrichs2019; Khosravani, Messman-Moore, Mohammadzadeh, Ghorbani, & Amirinezhad, Reference Khosravani, Messman-Moore, Mohammadzadeh, Ghorbani and Amirinezhad2019; Renna et al., Reference Renna, Peng, Shrout, Madison, Andridge, Alfano and Kiecolt-Glaser2021; Vaskinn, Melle, Aas, & Berg, Reference Vaskinn, Melle, Aas and Berg2021). We also dichotomized the CTQ total score to calculate PAFs for each study group conform previous analyses in this sample (Sideli et al., Reference Sideli, Schimmenti, La Barbera, La Cascia, Ferraro, Aas and van der Ven2022). Nevertheless, we acknowledge that dichotomization of variables is associated with increased risk of both type I and II error.

This study, however, has several limitations. First, the total EU-GEI case–control comprised over 1000 cases and 1000 controls. A sample of this size has over 90% power to detect OR of 1.5 or greater at p < 0.05 when the prevalence of the exposure is 15% or greater. Our analyses were conducted on a slightly smaller sample (N = 1991) with less power to detect relevant associations. We examined the interplay between CM and migrant status on psychosis risk. This was not a primary objective of the EU-GEI study. Nonetheless, our hypotheses, though secondary, were built on previous EU-GEI publications (Aas et al., Reference Aas, Alameda, Di Forti, Quattrone, Dazzan, Trotta and Morgan2021; Tarricone et al., Reference Tarricone, D'Andrea, Jongsma, Tosato, Gayer-Anderson, Stilo and Morgan2021; Termorshuizen et al., Reference Termorshuizen, Van Der Ven, Tarricone, Jongsma, Gayer-Anderson, Lasalvia and Selten2020). The retrospective assessment of the exposures may lead to recall bias. Specifically, the CTQ relies on the subjects' recollection, and we therefore cannot exclude the possibility of recall or desirability bias entering the study; we do not know whether these effects would have been differential by migrant status. Nonetheless, evidence suggests that this limitation may not be as critical as commonly thought (Brewin, Andrews, & Gotlib, Reference Brewin, Andrews and Gotlib1993). In our case, individuals with FEP were more likely to have missing data on CTQ than controls. This may have led to either underestimate or overestimate the true association between CM and FEP. However, imputation of missing values was implemented to minimize loss of precision and selection biases potentially deriving from complete-case analysis (Stekhoven & Bühlmann, Reference Stekhoven and Bühlmann2012; Waljee et al., Reference Waljee, Mukherjee, Singal, Zhang, Warren, Balis and Higgins2013). The reliability of retrospective reports of CM by individuals with psychosis has been questioned as recall of events could be potentially biased by features of the disease such as cognitive impairment or delusional beliefs (Susser & Widom, Reference Susser and Widom2012). However, previous research showed that memories of CM seem reasonably accurate and stable in individuals with psychosis (Fisher et al., Reference Fisher, Craig, Fearon, Morgan, Dazzan, Lappin and Morgan2011) and equally biased independently of psychiatric status (Fergusson, Horwood, & Woodward, Reference Fergusson, Horwood and Woodward2000). Importantly, the CTQ does not provide any information on the perpetrators, nor about the context where the maltreatment took place or its duration over time. Finally, it is possible that the experiences captured among migrants differ from those captured among the reference population (e.g. sexual violence in the context of war and ‘neglect’ in the context of social deprivation). Another limitation is that this study did not consider the age at exposure to the adversity. Evidence from neuroimaging suggest that abuse could have a more ‘detrimental’ effect if perpetrated within a specific age range (Pechtel, Lyons-Ruth, Anderson, & Teicher, Reference Pechtel, Lyons-Ruth, Anderson and Teicher2014). In conclusion, study results need replication on prospective studies with larger sample size.

Comparison with previous evidence: prevalence of maltreatment by migrant status

Migrants reported greater exposure to cumulative trauma than the reference population, independently of generational status. Regarding the specific maltreatment subtypes, rates were 1.5 to 3.0 times higher for emotional abuse, physical abuse, physical neglect, and sexual abuse among first-generation migrants, and 2.0 to 2.5-times higher for sexual abuse and physical abuse respectively among second-generation migrants in the fully adjusted models. We also found higher PRs of almost all maltreatment subtypes, except for emotional neglect, for migrants and their descendants from non-Western countries, while Western migrants only showed a 2-times higher exposure to physical abuse compared with the reference population. Previous studies comparing the prevalence of CM in migrants with non-migrants have reported mixed, and partially contradictory findings. For example, in a study conducted in the US (Vaughn et al., Reference Vaughn, Salas-Wright, Huang, Qian, Terzis and Helton2017), both the reference population and second-generation migrants were more likely than first-generation migrants to report any form of abuse, but less likely to report any form of neglect. Other research from the same country, found that children from ethnic minority backgrounds reported more maltreatment (Sedlak et al., Reference Sedlak, Mettenburg, Basena, Peta, McPherson and Greene2010). Among European studies, papers from the NPM (Alink et al., Reference Alink, Euser, van IJzendoorn and Bakermans-Kranenburg2013; Euser et al., Reference Euser, van IJzendoorn, Prinzie and Bakermans-Kranenburg2011; van Berkel et al., Reference van Berkel, Prevoo, Linting, Pannebakker and Alink2020) of youth consistently replicated the finding of an increased risk of maltreatment among migrants compared with Dutch majority, with the last study reporting a RR of 3.41 (95% CI 2.47–4.70) for first- and of 1.93 (95% CI 1.52–2.45) for second-generation migrants (van Berkel et al., Reference van Berkel, Prevoo, Linting, Pannebakker and Alink2020). In the group of migrants which the authors defined as ‘traditional’ based on a relatively longer migration history in the Netherlands (i.e. Turkish, Moroccan, Surinamese, and Antillean), the higher risk of maltreatment compared with the Dutch majority reference group was partially explained by socioeconomic and family factors (i.e. SES, parental education, and single parenthood) (Alink et al., Reference Alink, Euser, van IJzendoorn and Bakermans-Kranenburg2013). In our study, the addition of measures of socioeconomic and family factors (i.e. parental SES and parental mental illness) to the models attenuated the PRs of some CM subtypes, such as emotional abuse and physical abuse. Measures of social deprivation that we have not captured (e.g. household-related variables, income, or home ownership data) may explain the higher levels of maltreatment. We also did not assess specific factors within the family, including single/step-parenthood, family separation, parental stress, or traumatization of migrant families in their country of origin. The latter is particularly important, as post-traumatic stress disorder and traumatic experiences are predictors of CM (Mbagaya, Oburu, & Bakermans-Kranenburg, Reference Mbagaya, Oburu and Bakermans-Kranenburg2013; Montgomery, Just-Østergaard, & Jervelund, Reference Montgomery, Just-Østergaard and Jervelund2019), and can be up to 10 times more common in migrants compared with the general population (Fazel, Wheeler, & Danesh, Reference Fazel, Wheeler and Danesh2005). Together, this research suggests we need to do more to identify the upstream causes of the causes; that is what factors explain why migrant and their descendants report higher levels of maltreatment. It is these proximal causes that may be susceptible to interventions.

Comparison with previous evidence: FEP risk associated with maltreatment by migrant status

Consistent with the existing literature (Morgan & Gayer-Anderson, Reference Morgan and Gayer-Anderson2016; Morgan et al., Reference Morgan, Gayer-Anderson, Beards, Hubbard, Mondelli, Di Forti and Fisher2020; Varese et al., Reference Varese, Smeets, Drukker, Lieverse, Lataster, Viechtbauer and Bentall2012), we found that experiences of CM increased the odds of adulthood FEP in all groups. Migrant status attenuated this association. This is in line with two recent studies which found that, despite a greater exposure to traumatic events, migrants showed greater resilience and a smaller detrimental effect of trauma on their mental health, compared with the reference populations (Gatt et al., Reference Gatt, Alexander, Emond, Foster, Hadfield, Mason-Jones and Wu2020; Solà-Sales et al., Reference Solà-Sales, Pérez-González, Van Hoey, Iborra-Marmolejo, Beneyto-Arrojo and Moret-Tatay2021).

Finally, to our knowledge no prior study has examined PAFs for CM according to migrant status. Our results on the overall sample (PAF = 32.8%, 95% CI 29.4–35.6%), though, are consistent with the evidence provided by previous meta-analyses (Dragioti et al., Reference Dragioti, Radua, Solmi, Arango, Oliver, Cortese and Fusar-Poli2022; Varese et al., Reference Varese, Smeets, Drukker, Lieverse, Lataster, Viechtbauer and Bentall2012). The finding that CM accounts for a greater PAF in migrants and their descendants compared with non-migrants aligns with the hypothesis of a ‘socio-developmental pathway’ to psychosis (Morgan et al., Reference Morgan, Knowles and Hutchinson2019).

Relevance and implications

CM is a widespread public health issue which needs global and effective interventions. Our results confirm the well-established association between CM and psychosis. Migrants and their descendants in our study had greater prevalence of most maltreatment subtypes. Despite greater exposure to CM, the association between maltreatment and psychosis was stronger for those from the reference population compared with migrants and their descendants. Migrants experience 2 to 3-times higher rates of psychosis compared with the non-migrant population (Selten et al., Reference Selten, Van Der Ven and Termorshuizen2019). Reasons for the increased rates are yet to be completely understood. Our results, however, suggest that, despite a marginally smaller effect on the odds of FEP, the considerably greater exposure to CM among migrants and their children accounted for a more sizable proportion of incident FEP cases. Therefore, public health policies aimed at addressing the issue of CM by the means of adequate interventions for each level of prevention should consider the greater vulnerability of some groups, such as migrants and their descendants, and ensure that such programs are accessible to these population sections. Exposure to CM is not the sole explanation for the higher psychosis risk among migrants and their descendants. The ‘social defeat’ hypothesis (Selten & Cantor-Graae, Reference Selten and Cantor-Graae2005) posits that the prolonged experience of exclusion from the majority groups increases the risk of psychosis by sensitizing the dopamine pathway. Further evidence on the contribution of cumulative exposure to psycho-social adversities derives from previous reports on this issue from the EU-GEI study (Jongsma et al., Reference Jongsma, Gayer-Anderson, Tarricone, Velthorst, van der Ven, Quattrone and Kirkbride2020; Misra et al., Reference Misra, Gelaye, Williams, Koenen, Borba, Quattrone and Morgan2021; Tarricone et al., Reference Tarricone, D'Andrea, Jongsma, Tosato, Gayer-Anderson, Stilo and Morgan2021). More studies are needed to better understand the association between migrant status and psychosis focusing on putative causal factors.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S003329172200335X

Acknowledgments

Special acknowledgment to all the participants and the EU-GEI team.

Financial support

The EU-GEI Project was funded by grant agreement Health-F2-2010-241909 from the European Community's Seventh Framework program. The Brazilian study was funded by grant 2012-0417-0 from the São Paulo Research Foundation.

Conflict of interest

None.

References

Aas, M., Alameda, L., Di Forti, M., Quattrone, D., Dazzan, P., Trotta, A., … Morgan, C. (2021). Synergistic effects of childhood adversity and polygenic risk in first-episode psychosis: The EU-GEI study. Psychological Medicine, 19. https://doi.org/10.1017/S0033291721003664.Google ScholarPubMed
Adler, N. E., & Rehkopf, D. H. (2008). U.S. Disparities in health: Descriptions, causes, and mechanisms. Annual Review of Public Health, 29, 235252. https://doi.org/10.1146/ANNUREV.PUBLHEALTH.29.020907.090852.CrossRefGoogle ScholarPubMed
Aiken, L., West, S., & Reno, R. (1991). Multiple regression: Testing and interpreting interactions.Google Scholar
Alink, L. R. A., Euser, S., van IJzendoorn, M. H., & Bakermans-Kranenburg, M. J. (2013). Is elevated risk of child maltreatment in immigrant families associated with socioeconomic status? Evidence from three sources. International Journal of Psychology, 48(2), 117127. https://doi.org/10.1080/00207594.2012.734622.CrossRefGoogle ScholarPubMed
Arango, C., Dragioti, E., Solmi, M., Cortese, S., Domschke, K., Murray, R. M., … Fusar-Poli, P. (2021). Risk and protective factors for mental disorders beyond genetics: An evidence-based atlas. World Psychiatry: Official Journal of the World Psychiatric Association (WPA), 20(3), 417436. https://doi.org/10.1002/WPS.20894.CrossRefGoogle ScholarPubMed
Barros, A. J. D., & Hirakata, V. N. (2003). Alternatives for logistic regression in cross-sectional studies: An empirical comparison of models that directly estimate the prevalence ratio. BMC Medical Research Methodology, 3, 113. https://doi.org/10.1186/1471-2288-3-21.CrossRefGoogle ScholarPubMed
Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society. Series B (Methodological), 57(1), 289300.CrossRefGoogle Scholar
Bernstein, D. P., Stein, J. A., Newcomb, M. D., Walker, E., Pogge, D., Ahluvalia, T., … Zule, W. (2003). Development and validation of a brief screening version of the childhood trauma questionnaire. Child Abuse & Neglect, 27(2), 169190.CrossRefGoogle ScholarPubMed
Braveman, P. A., Cubbin, C., Egerter, S., Williams, D. R., & Pamuk, E. (2010). Socioeconomic disparities in health in the United States: What the patterns tell us. American Journal of Public Health, 100(Suppl 1), 8696. https://doi.org/10.2105/AJPH.2009.166082.CrossRefGoogle ScholarPubMed
Brewin, C. R., Andrews, B., & Gotlib, I. H. (1993). Psychopathology and early experience: A reappraisal of retrospective reports. Psychological Bulletin, 113(1), 8298. https://doi.org/10.1037/0033-2909.113.1.82.CrossRefGoogle ScholarPubMed
Brühl, A., Kley, H., Grocholewski, A., Neuner, F., & Heinrichs, N. (2019). Child maltreatment, peer victimization, and social anxiety in adulthood: A cross-sectional study in a treatment-seeking sample. BMC Psychiatry, 19(1), 418. https://doi.org/10.1186/s12888-019-2400-4.CrossRefGoogle Scholar
Cancian, M., Yang, M.-Y., & Slack, K. S. (2013). The effect of additional child support income on the risk of child maltreatment. Social Service Review, 87(3), 417437.CrossRefGoogle Scholar
Di Forti, M., Morgan, C., Dazzan, P., Pariante, C., Mondelli, V., Marques, T. R., … Murray, R. M. (2009). High-potency cannabis and the risk of psychosis. British Journal of Psychiatry, 195, 488491. https://doi.org/10.1192/bjp.bp.109.064220.CrossRefGoogle ScholarPubMed
Di Forti, M., Quattrone, D., Freeman, T. P., Tripoli, G., Gayer-Anderson, C., Quigley, H., … van der Ven, E. (2019). The contribution of cannabis use to variation in the incidence of psychotic disorder across Europe (EU-GEI): A multicentre case–control study. The Lancet Psychiatry, 6(5), 427436. https://doi.org/10.1016/S2215-0366(19)30048-3.CrossRefGoogle ScholarPubMed
Dragioti, E., Radua, J., Solmi, M., Arango, C., Oliver, D., Cortese, S., … Fusar-Poli, P. (2022). Global population attributable fraction of potentially modifiable risk factors for mental disorders: A meta-umbrella systematic review. Molecular Psychiatry, 25(2), 110. https://doi.org/10.1038/s41380-022-01586-8.Google Scholar
Dworsky, A., Courtney, M. E., & Zinn, A. (2007). Child, parent, and family predictors of child welfare services involvement among TANF applicant families. Children and Youth Services Review, 29(6), 802820.CrossRefGoogle Scholar
Euser, E. M., van IJzendoorn, M. H., Prinzie, P., & Bakermans-Kranenburg, M. J. (2011). Elevated child maltreatment rates in immigrant families and the role of socioeconomic differences. Child Maltreatment, 16(1), 6373. https://doi.org/10.1177/1077559510385842.CrossRefGoogle ScholarPubMed
Fazel, M., Wheeler, J., & Danesh, J. (2005). Prevalence of serious mental disorder in 7000 refugees resettled in western countries: A systematic review. The Lancet, 365(9467), 13091314. https://doi.org/10.1016/S0140-6736(05)61027-6.CrossRefGoogle ScholarPubMed
Fergusson, D. M., Horwood, L. J., & Woodward, L. J. (2000). The stability of child abuse reports: A longitudinal study of the reporting behaviour of young adults. Psychological Medicine, 30(3), 529544. https://doi.org/10.1017/S0033291799002111.CrossRefGoogle ScholarPubMed
Fisher, H. L., Craig, T. K., Fearon, P., Morgan, K., Dazzan, P., Lappin, J., … Morgan, C. (2011). Reliability and comparability of psychosis patients’ retrospective reports of childhood abuse. Schizophrenia Bulletin, 37(3), 546553. https://doi.org/10.1093/SCHBUL/SBP103.CrossRefGoogle ScholarPubMed
Gatt, J. M., Alexander, R., Emond, A., Foster, K., Hadfield, K., Mason-Jones, A., … Wu, Q. (2020). Trauma, resilience, and mental health in migrant and non-migrant youth: An international cross-sectional study across six countries. Frontiers in Psychiatry, 10(997). https://doi.org/10.3389/fpsyt.2019.00997.CrossRefGoogle ScholarPubMed
Gayer-Anderson, C., Jongsma, H. E., Di Forti, M., Quattrone, D., Velthorst, E., de Haan, L., … Morgan, C. (2020). The European network of national schizophrenia networks studying gene–environment interactions (EU-GEI): Incidence and first-episode case–control programme. Social Psychiatry and Psychiatric Epidemiology, 51(5), 645657. https://doi.org/10.1007/s00127-020-01831-x.CrossRefGoogle Scholar
Hartig, F. (2022). Residual Diagnostics for Hierarchical (Multi-Level / Mixed) Regression Models [R package DHARMa version 0.4.5].Google Scholar
Jongsma, H. E., Gayer-Anderson, C., Tarricone, I., Velthorst, E., van der Ven, E., Quattrone, D., … Kirkbride, J. B. (2020). Social disadvantage, linguistic distance, ethnic minority status and first-episode psychosis: Results from the EU-GEI case–control study. Psychological Medicine, 51(9), 15361548. https://doi.org/10.1017/S003329172000029X.CrossRefGoogle ScholarPubMed
Kawachi, I., Daniels, N., & Robinson, D. E. (2005). Health disparities by race and class: Why both matter. Health Affairs, 24(2), 343352.CrossRefGoogle ScholarPubMed
Khosravani, V., Messman-Moore, T. L., Mohammadzadeh, A., Ghorbani, F., & Amirinezhad, A. (2019). Effects of childhood emotional maltreatment on depressive symptoms through emotion dysregulation in treatment-seeking patients with heroin-dependence. Journal of Affective Disorders, 256, 448457. https://doi.org/10.1016/j.jad.2019.06.021.CrossRefGoogle ScholarPubMed
Kim, H., & Drake, B. (2018). Child maltreatment risk as a function of poverty and race/ethnicity in the USA. International Journal of Epidemiology, 47(3), 780787. https://doi.org/10.1093/IJE/DYX280.CrossRefGoogle Scholar
Mallet, R. (1997). Sociodemographic schedule. London, UK: Section of Social Psychiatry, Institute of Psychiatry.Google Scholar
Maxwell, M. E. (1992). Family interview for genetic studies (FIGS): A manual for FIGS. Bethesda, MD: Clinical Neurogenetics Branch, Intramural Research Program, National Institute of Mental Health.Google Scholar
Mbagaya, C., Oburu, P., & Bakermans-Kranenburg, M. J. (2013). Child physical abuse and neglect in Kenya, Zambia and the Netherlands: A cross-cultural comparison of prevalence, psychopathological sequelae and mediation by PTSS. International Journal of Psychology: Journal International de Psychologie, 48(2), 95107. https://doi.org/10.1080/00207594.2012.691975.CrossRefGoogle ScholarPubMed
Misra, S., Gelaye, B., Williams, D. R., Koenen, K. C., Borba, C. P. C., Quattrone, D., … Morgan, C. (2021). Perceived major experiences of discrimination, ethnic group, and risk of psychosis in a six-country case–control study. Psychological Medicine, 19. https://doi.org/10.1017/S0033291721000453.Google Scholar
Montgomery, E., Just-Østergaard, E., & Jervelund, S. S. (2019). Transmitting trauma: A systematic review of the risk of child abuse perpetrated by parents exposed to traumatic events. International Journal of Public Health, 64(2), 241251. https://doi.org/10.1007/S00038-018-1185-4.CrossRefGoogle ScholarPubMed
Moody, G., Cannings-John, R., Hood, K., Kemp, A., & Robling, M. (2018). Establishing the international prevalence of self-reported child maltreatment: A systematic review by maltreatment type and gender. BMC Public Health, 18(1), 115. https://doi.org/10.1186/S12889-018-6044-Y/FIGURES/5.CrossRefGoogle ScholarPubMed
Morgan, C., & Gayer-Anderson, C. (2016). Childhood adversities and psychosis: Evidence, challenges, implications. World Psychiatry, 15(2), 93102. https://doi.org/10.1002/wps.20330.CrossRefGoogle ScholarPubMed
Morgan, C., Gayer-Anderson, C., Beards, S., Hubbard, K., Mondelli, V., Di Forti, M., … Fisher, H. L. (2020). Threat, hostility and violence in childhood and later psychotic disorder: Population-based case–control study. British Journal of Psychiatry, 217(4), 575582. https://doi.org/10.1192/bjp.2020.133.CrossRefGoogle ScholarPubMed
Morgan, C., Knowles, G., & Hutchinson, G. (2019). Migration, ethnicity and psychoses: Evidence, models and future directions. World Psychiatry, 18(3), 247258. https://doi.org/10.1002/wps.20655.CrossRefGoogle ScholarPubMed
Muller, C. J., & MacLehose, R. F. (2014). Estimating predicted probabilities from logistic regression: Different methods correspond to different target populations. International Journal of Epidemiology, 43(3), 962. https://doi.org/10.1093/IJE/DYU029.CrossRefGoogle ScholarPubMed
Pechtel, P., Lyons-Ruth, K., Anderson, C. M., & Teicher, M. H. (2014). Sensitive periods of amygdala development: The role of maltreatment in preadolescence. NeuroImage, 97, 236244. https://doi.org/10.1016/j.neuroimage.2014.04.025.CrossRefGoogle ScholarPubMed
Pelton, L. H. (2015). The continuing role of material factors in child maltreatment and placement. Child Abuse & Neglect, 41, 3039. https://doi.org/10.1016/J.CHIABU.2014.08.001.CrossRefGoogle ScholarPubMed
Putnam-Hornstein, E., Needell, B., King, B., & Johnson-Motoyama, M. (2013). Racial and ethnic disparities: A population-based examination of risk factors for involvement with child protective services. Child Abuse & Neglect, 37(1), 3346.CrossRefGoogle ScholarPubMed
Quattrone, D., Di Forti, M., Gayer-Anderson, C., Ferraro, L., Jongsma, H. E., Tripoli, G., … Cristofalo, D. (2019). Transdiagnostic dimensions of psychopathology at first episode psychosis: Findings from the multinational EU-GEI study. Psychological Medicine, 49(8), 13781391. https://doi.org/10.1017/S0033291718002131.CrossRefGoogle ScholarPubMed
Renna, M. E., Peng, J., Shrout, M. R., Madison, A. A., Andridge, R., Alfano, C. M., … Kiecolt-Glaser, J. K. (2021). Childhood abuse histories predict steeper inflammatory trajectories across time. Brain, Behavior, and Immunity, 91, 541545. https://doi.org/10.1016/J.BBI.2020.11.012.CrossRefGoogle ScholarPubMed
Santos, C. A. S. T., Fiaccone, R. L., Oliveira, N. F., Cunha, S., Barreto, M. L., Do Carmo, M. B. B., … Amorim, L. D. (2008). Estimating adjusted prevalence ratio in clustered cross-sectional epidemiological data. BMC Medical Research Methodology, 8(1), 80. https://doi.org/10.1186/1471-2288-8-80.CrossRefGoogle ScholarPubMed
Sedlak, A. J., Mettenburg, J., Basena, M., Peta, I., McPherson, K., & Greene, A. (2010). Fourth national incidence study of child abuse and neglect (NIS-4), Vol. 9. Washington, DC: US Department of Health and Human Services.Google Scholar
Selten, J. P., & Cantor-Graae, E. (2005). Social defeat: Risk factor for schizophrenia? British Journal of Psychiatry, 187(2), 101102. https://doi.org/https://doi.org/10.1192/bjp.187.2.101.CrossRefGoogle ScholarPubMed
Selten, J. P., Van Der Ven, E., & Termorshuizen, F. (2019). Migration and psychosis: A meta-analysis of incidence studies. Psychological Medicine, 50(2), 303313. https://doi.org/10.1017/S0033291719000035.CrossRefGoogle ScholarPubMed
Sideli, L., Schimmenti, A., La Barbera, D., La Cascia, C., Ferraro, L., Aas, M., … van der Ven, E. (2022). Childhood maltreatment, educational attainment, and IQ: Findings from a multicentric case–control study of first-episode psychosis (EU-GEI). Schizophrenia Bulletin, 48(3), 575589. https://doi.org/10.1093/SCHBUL/SBAC004.CrossRefGoogle ScholarPubMed
Slack, K. S., Lee, B. J., & Berger, L. M. (2007). Do welfare sanctions increase child protection system involvement? A cautious answer. Social Service Review, 81(2), 207228.CrossRefGoogle Scholar
Solà-Sales, S., Pérez-González, N., Van Hoey, J., Iborra-Marmolejo, I., Beneyto-Arrojo, M. J., & Moret-Tatay, C. (2021). The role of resilience for migrants and refugees’ mental health in times of COVID-19. Healthcare, 9(9), 1131. https://doi.org/10.3390/HEALTHCARE9091131.CrossRefGoogle ScholarPubMed
Stekhoven, D. J., & Bühlmann, P. (2012). Missforest-non-parametric missing value imputation for mixed-type data. Bioinformatics (Oxford, England), 28(1), 112118. https://doi.org/10.1093/bioinformatics/btr597.Google ScholarPubMed
Stoltenborgh, M., Bakermans-Kranenburg, M. J., Alink, L. R. A., & van Ijzendoorn, M. H. (2015). The prevalence of child maltreatment across the globe: Review of a series of meta-analyses. Child Abuse Review, 24(1), 3750. https://doi.org/10.1002/CAR.2353/FORMAT/PDF.CrossRefGoogle Scholar
Susser, E., & Widom, C. S. (2012). Still searching for lost truths about the bitter sorrows of childhood. Schizophrenia Bulletin, 38(4), 672675. https://doi.org/10.1093/SCHBUL/SBS074.CrossRefGoogle ScholarPubMed
Tarricone, I., D'Andrea, G., Jongsma, H. E., Tosato, S., Gayer-Anderson, C., Stilo, S. A., … Morgan, C. (2021). Migration history and risk of psychosis: Results from the multinational EU-GEI study. Psychological Medicine, 113. https://doi.org/10.1017/s003329172000495x.Google ScholarPubMed
Termorshuizen, F., Van Der Ven, E., Tarricone, I., Jongsma, H. E., Gayer-Anderson, C., Lasalvia, A., … Selten, J. P. (2020). The incidence of psychotic disorders among migrants and minority ethnic groups in Europe: Findings from the multinational EU-GEI study. Psychological Medicine, 52(7), 13761385. https://doi.org/10.1017/S0033291720003219.CrossRefGoogle ScholarPubMed
van Berkel, S. R., Prevoo, M. J. L., Linting, M., Pannebakker, F. D., & Alink, L. R. A. (2020). Prevalence of child maltreatment in the Netherlands: An update and cross-time comparison. Child Abuse and Neglect, 103, 104439. https://doi.org/10.1016/j.chiabu.2020.104439.CrossRefGoogle ScholarPubMed
Varese, F., Smeets, F., Drukker, M., Lieverse, R., Lataster, T., Viechtbauer, W., … Bentall, R. P. (2012). Childhood adversities increase the risk of psychosis: A meta-analysis of patient-control, prospective-and cross-sectional cohort studies. Schizophrenia Bulletin, 38(4), 661671. https://doi.org/10.1093/schbul/sbs050.CrossRefGoogle ScholarPubMed
Vaskinn, A., Melle, I., Aas, M., & Berg, A. O. (2021). Sexual abuse and physical neglect in childhood are associated with affective theory of mind in adults with schizophrenia. Schizophrenia Research: Cognition, 23, 100189. https://doi.org/10.1016/j.scog.2020.100189.Google ScholarPubMed
Vaughn, M. G., Salas-Wright, C. P., Huang, J., Qian, Z., Terzis, L. D., & Helton, J. J. (2017). Adverse childhood experiences among immigrants to the United States. Journal of Interpersonal Violence, 32(10), 15431564. https://doi.org/10.1177/0886260515589568.CrossRefGoogle ScholarPubMed
Waljee, A. K., Mukherjee, A., Singal, A. G., Zhang, Y., Warren, J., Balis, U., … Higgins, P. D. R. (2013). Comparison of imputation methods for missing laboratory data in medicine. BMJ Open, 3(8), e002847. https://doi.org/10.1136/bmjopen-2013-002847.CrossRefGoogle ScholarPubMed
Walker, E. A., Gelfand, A., Katon, W. J., Koss, M. P., Von Korff, M., Bernstein, D., & Russo, J. (1999). Adult health status of women with histories of childhood abuse and neglect. American Journal of Medicine, 107(4), 332339. https://doi.org/10.1016/S0002-9343(99)00235-1.CrossRefGoogle ScholarPubMed
Williams, J., Farmer, A. E., Ackenheil, M., Kaufmann, C. A., McGuffin, P., & Group, T. O. R. R. (1996). A multicentre inter-rater reliability study using the OPCRIT computerized diagnostic system. Psychological Medicine, 26(04), 775. https://doi.org/10.1017/S003329170003779X.CrossRefGoogle ScholarPubMed
World Health Organization (WHO). (2006). Preventing child maltreatment: A guide to taking action and generating evidence / World Health Organization and International Society for Prevention of Child Abuse and Neglect. Geneva PP, Geneva: World Health Organization.Google Scholar
Figure 0

Table 1. Distribution of exposures and covariates in the case–control sample

Figure 1

Table 2. Unadjusted and adjusted PRs of CM subtypes by migrant generational status

Figure 2

Table 3. Unadjusted and adjusted PRs of CM subtypes by migrants' area of origin

Figure 3

Fig. 1. Predicted probabilities of FEP by CM and migrant status. (a) represents predicted probabilities of psychosis by CM and migrant generational status. (b) represent predicted probabilities of psychosis by CM and Western/non-Western migrant status.

Figure 4

Table 4. Odds ratios of FEP by migrant status and migrant-CM interaction

Supplementary material: File

D'Andrea et al. supplementary material

D'Andrea et al. supplementary material

Download D'Andrea et al. supplementary material(File)
File 142 KB