European Psychiatry Impact of adverse childhood experiences on www.cambridge.org/epa educational achievements in young people at clinical high risk of developing psychosis Stefania Tognin1,2 , Ana Catalan1,3* , Matthew J. Kempton1,4,5 , Research Article 6,7 8,9 10 Barnaby Nelson , Patrick McGorry , Anita Riecher-Rössler , Cite this article: Tognin S, Catalan A, 11 12 13,14 Kempton MJ, Nelson B, McGorry P, Riecher- Rodrigo Bressan , Neus Barrantes-Vidal , Marie-Odile Krebs , Rössler A, Bressan R, Barrantes-Vidal N, 15 16 17 Krebs M-O, Nordentoft M, Ruhrmann S, Merete Nordentoft , Stephan Ruhrmann , Gabriele Sachs , Sachs G, Rutten BPF, van Os J, de Haan L, 18 1,8,19 20 van der Gaag M, EU-GEI High-Risk Study, Bart P. F. Rutten , Jim van Os , Lieuwe de Haan , 21,22 McGuire P, Valmaggia LR (2023). Impact of Mark van der Gaag , EU-GEI High-Risk Study, Philip McGuire1,2,5 and adverse childhood experiences on educational 23 achievements in young people at clinical high Lucia R. Valmaggia risk of developing psychosis. European Psychiatry, 66(1), e16, 1–8 1 Department of Psychosis Studies, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, London, https://doi.org/10.1192/j.eurpsy.2022.2351 United Kingdom; 2Outreach and Support in South London (OASIS) Service, South London and Maudsley NHS Foundation Trust, London, United Kingdom; 3Biobizkaia Health Research Institute, Basurto University Hospital, OSI Received: 15 May 2022 Bilbao-Basurto, University of the Basque Country UPV/EHU, Centro de Investigación en Red de Salud Mental Revised: 25 October 2022 Accepted: 06 November 2022 (CIBERSAM), Instituto deSalud Carlos III, Barakaldo, Spain; 4Department of Neuroimaging, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, London, United Kingdom; 5National Institute for Health Research Keywords: (NIHR), Biomedical Research Centre (BRC), London, United Kingdom; 6The National Centre of Excellence in Youth Adverse childhood experiences; clinical high Mental Health, Orygen, Parkville, Victoria 3052, Australia; 7Centre for Youth Mental Health, The University of Melbourne, risk for psychosis; education Melbourne, Parkville, Victoria 3052, Australia; 8Center for Neuropsychiatric Schizophrenia Research (CNSR) and Center for Clinical Intervention and Neuropsychiatric Schizophrenia Research (CINS), University of Copenhagen, Mental Health Author for correspondence: Centre Glostrup, Copenhagen, Denmark; 9Faculty of Health and Medical Sciences, Department of Clinical Medicine, *Ana Catalan, University of Copenhagen, Copenhagen, Denmark; 10Faculty of Medicine, University of Basel, Basel, Switzerland; 11LiNC - E-mail: ana.catalan@kcl.ac.uk Lab Integrative Neuroscience, Depto Psiquiatria, Escola Paulista de Medicina, Universidade Federal de São Paulo See the Supplementary material for EU-GEI – UNIFESP, São Paulo, Brazil; 12Departament de Psicologia Clínica i de la Salut (Universitat Autònoma de Barcelona), High-Risk Study Group Author list. Fundació Sanitària Sant Pere Claver (Spain), Centre for Biomedical Research in Mental Health (CIBERSAM), Madrid, Spain; 13INSERM, IPNP UMR S1266, Laboratoire de Physiopathologie des Maladies Psychiatriques, Université Paris S.T. and A.C. contributed equally to this work. Descartes, Université de Paris, CNRS, GDR3557-Institut de Psychiatrie, Paris, France; 14Faculté de Médecine Paris Descartes, GHU Paris - Sainte-Anne, Service Hospitalo-Universitaire, Paris, France; 15Mental Health Center Copenhagen and Center for Clinical Intervention and Neuropsychiatric Schizophrenia Research, CINS, Mental Health Center Glostrup, Mental Health Services in the Capital Region of Copenhagen, University of Copenhagen, Copenhagen, Denmark; 16Department of Psychiatry and Psychotherapy, Faculty of Medicine and University Hospital, University of Cologne, Cologne, Germany; 17Department of Psychiatry and Psychotherapy, Medical University of Vienna, Vienna, Austria; 18Department of Psychiatry and Neuropsychology, Faculty of Health, Medicine and Life Sciences, School for Mental Health and Neuroscience (MHeNS), Maastricht University, Maastricht, The Netherlands; 19Department of Psychiatry, UMC Utrecht Brain Center, Utrecht University Medical Centre, Utrecht, The Netherlands; 20Amsterdam UMC, Early Psychosis Department, Amsterdam, The Netherlands; 21VU University, Faculty of Behavioural and Movement Sciences, Department of Clinical Psychology and Amsterdam Public Mental Health Research Institute, Amsterdam, The Netherlands; 22Department of Psychosis Research, Parnassia Psychiatric Institute, The Hague, The Netherlands and 23Department of Psychology, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, London, United Kingdom Abstract Background. Adverse childhood experiences (ACE) can affect educational attainments, but little is known about their impact on educational achievements in people at clinical high risk of psychosis (CHR). Methods. In total, 344 CHR individuals and 67 healthy controls (HC) were recruited as part of © The Author(s), 2023. Published by Cambridge the European Community’s Seventh Framework Programme-funded multicenter study the University Press on behalf of the European European Network of National Schizophrenia Networks Studying Gene–Environment Inter- Psychiatric Association. This is an Open Access article, distributed under the terms of the actions (EU-GEI). The brief version of the Child Trauma Questionnaire was used to measure Creative Commons Attribution licence (http:// ACE, while educational attainments were assessed using a semi-structured interview. creativecommons.org/licenses/by/4.0), which Results. At baseline, compared with HC, the CHR group spent less time in education and had permits unrestricted re-use, distribution and higher rates of ACE, lower rates of employment, and lower estimated intelligence quotient (IQ). reproduction, provided the original article is Across both groups, the total number of ACE was associated with fewer days in education and properly cited. lower level of education. Emotional abuse was associated with fewer days in education in HC. Emotional neglect was associated with a lower level of education in CHR, while sexual abuse was associated with a lower level of education in HC. In the CHR group, the total number of ACE, physical abuse, and neglect was significantly associated with unemployment, while emotional neglect was associated with employment. 2 Stefania Tognin et al. Conclusions. ACE are strongly associated with developmental outcomes such as educational achievement. Early intervention for psychosis programs should aim at integrating specific interventions to support young CHR people in their educational and vocational recovery. More generally, public health and social interventions focused on the prevention of ACE (or reduce their impact if ACE occur) are recommended. Introduction associations between ACE and employment. We hypothesized that (1) compared with HC, the CHR population will have completed Adverse childhood experiences (ACE), including physical and fewer days in education and will have achieved lower levels of emotional abuse and neglect and sexual abuse, are associated with academic achievement and lower rates of employment. Moreover, the emergence of psychosis [1, 2] as well as with several cognitive (2) CHR and HC participants who experienced ACE will present deficits, including impaired memory, attention, emotional discrim- worse educational outcomes (less days in education, lower level of ination [3–5], and a decline in social and occupational functioning education) and lower employment rates than those who did not. [6, 7]. Not surprisingly, ACE per se have been associated with worse Finally, (3) the associations between ACE and educational out- educational attainments in both clinical and nonclinical popula- comes will be more pronounced in CHR than in HC. tions [8–10]. Poor educational outcomes have been consistently associated with psychosis [11], as well as with several other mental health conditions. For example, autism spectrum disorder [12], Methods attention-deficit/ hyperactivity disorder [13], and substance-use disorder [14] have been associated with low educational attain- Sample ments. An exception is anorexia nervosa, which has been associated In total, 344 CHR participants and 67 HC were recruited as part of with high educational attainments [8]. the European Network of National Schizophrenia Networks Study- Poor educational outcomes have been identified as one of the ing Gene–Environment Interactions (EU-GEI) study [28] from most significant long-term consequences of traumatic experiences 11 centers (London, Amsterdam, Den Haag, Vienna, Basel, in childhood [15]. One of the reasons is that poor educational Cologne, Melbourne, Copenhagen, Paris, Barcelona, and Sao attainment is linked to higher rates of unemployment later in life Paolo) from July 2010 to August 2015. The overall design of the [10]. Possible mechanisms linking ACE to poor academic achieve- study was naturalistic, longitudinal, and prospective, consisting of a ments are the neuropsychological changes that result from chronic baseline and three follow-up time points when extensive clinical, childhood stress. For example, previous studies have shown how cognitive, and biological measures were collected. CHR partici- ACE can adversely impact the emotional stress response [16], social pants were recruited from local early detection for psychosis ser- functioning [17], and cognitive functions [18], such as memory, vices, while HC participants with similar age, gender-ratio, attention and verbal learning, and, more generally, how ACE can ethnicity, and socioeconomic status were recruited from the same interfere with early development [19]. geographical areas. The inclusion criteria for all participants were as A recent review and meta-analysis highlighted that all forms of follows: being aged 15–35, being able to consent, and having ACE, including witnessing domestic violence, doubled the risk of adequate language skills local to each center. In addition, CHR school dropout [20]. Similarly, other studies reported that when participants had to meet clinical high risk for psychosis criteria individuals with ACE attend school, they are more likely to suffer defined according to the Comprehensive Assessment of At-Risk bullying [21] and do less well in areas such as overall school per- Mental States scale (CAARMS) [29]. The exclusion criteria for CHR formance and completion of tasks such as homework [22]. Poor participants were having had prior experience of a psychotic epi- educational attainments can have long-term repercussions on several sode for more than 1 week as determined by the CAARMS and the areas, including mental health, financial security, and, more gener- Structural Clinical Interview for DSM Disorders (SCID); previous ally, needs to be met. For example, poor educational attainment is a treatment with an antipsychotic for a psychotic episode; and an risk factor for adolescent and adult difficulties, including conduct estimated intelligence quotient (IQ) < 60. Exclusion criteria for HC disorders [23] and substance abuse [24]. More generally, ACE might included meeting criteria for CHR or psychosis, reporting personal impact the achievement of formal qualifications, which in turn or (first-degree) family history of psychosis, and an estimated affects the likelihood of securing future employment [25, 26]. IQ < 60. The study received ethical approval at each site. All At their first contact with Early Intervention in Psychosis ser- procedures contributing to this work comply with the ethical vices (EIP), between 46 and 78% of the young people are not in standards of the relevant national and institutional committees employment, education, or training [27]. Adolescence and early on human experimentation and with the Helsinki Declaration of adulthood are critical times to complete studies and enter the work 1975, as revised in 2008. environment and it is therefore important to investigate the impact that psychosis risk factors, such as ACE, might have on the voca- Measures tional, including educational, pathway. To date, little is known about the impact of ACE on educational achievements in people Sociodemographics and clinical data at clinical high risk for developing psychosis (CHR). Despite ACE Detailed sociodemographic characteristics were assessed using the being highly prevalent in early psychosis, to the best of our know- modified Medical Research Council Sociodemographic Schedule ledge, no prior studies directly examined the implications of ACE [28, 30]. The CAARMS [29] was used to measure CHR status and to on educational achievements in the CHR population. In this study, determine the transition to psychosis. The SCID [31] was used to we investigated the baseline associations between some of the ACE establish the presence of other psychiatric disorders and to exclude subtypes (i.e., emotional abuse and neglect, physical abuse and the presence of current or past psychotic disorders. neglect, and sexual abuse) and education in a sample of CHR participants and a sample of healthy controls (HC) with similar Adverse childhood experiences measures sociodemographic characteristics. While the main focus of this The brief version of the Child Trauma Questionnaire (CTQ-B) [32] work is education, for completeness we also analyzed at the baseline was used to measure ACE up to the age of 17. The CTQ is a 25-item European Psychiatry 3 self-report questionnaire that assesses five domains: emotional variables of interest and ACE (as measured by the CTQ scale). We abuse, emotional neglect, sexual abuse, physical abuse, and physical also performed binary regression models to examine the relation- neglect. A higher total score is related to a higher amount of adverse ship between employment status (employed/not employed) and experiences. ACE. The following confounding variables were used to adjust the Responses were used to calculate an individual’s total ACE model results: gender, age, estimated IQ, social class, recruitment score, which was then categorized into ACE counts (0, 1, 2–3, site, and ethnicity. and ≥4), consistent with methodologies applied elsewhere [33]. All tests were two-sided, and significance was set at p < 0.05. Educational attainment was assessed using a semi-structured Analyses were carried out with the Statistical Package for the Social interview on years of education and level of education. The highest Sciences (SPSS) software program, version 24.0 [34]. level of education was determined using six categories: (1) school, no qualifications—depicting those who had completed compulsory education but passed no exams, tests, and so forth to gain formal Results qualifications; (2) school, with qualifications—those who had com- Descriptive analyses pleted compulsory education and passed one or more exams, tests, and so forth to gain formal qualifications; (3) tertiary, further— In total, 411 participants were included in the analyses, 67 HC and those with the first level of non-compulsory education, for example, 344 CHR. Sociodemographic characteristics of the samples are A-levels or the International Baccalaureate; (4) vocational—those detailed in Table 1. At baseline, HC participants presented with with job-related education, for example, teacher training, plumber, higher estimated IQ, more years of education, lower rates of ACE, electrician, etc.; (5) higher (undergraduate)—those with a univer- and higher rates of employment compared with CHR individuals. sity degree; and (6) higher (postgraduate)—those with qualifica- Approximately 52.6% of CHR compared with 18.2% of HC tions higher than first-degree level, for example, Masters or PhD. reported at least one ACE before the age of 17 (Table 2). This These categories were clustered into three variables: higher educa- information was missing for 26 participants (1 HC and 25 CHR; tion (5–6), first-level non-compulsory education (3–4), and part-all 6.6%), who were therefore excluded from subsequent analyses. compulsory education (1–2). Relationship between time in education and ACE Statistical analyses Higher CTQ total scores were significantly associated with fewer Sociodemographic data were analyzed using means and standard days in education in both the CHR sample (R2 = 0.22; F deviation for continuous data and frequencies for categorical data. [4, 272] = 20.725, p < 0.001; B = 9.3, SE = 4.2; p = 0.028) and Analysis of the variance (ANOVA) was used to examine group the HC sample (R2 = 0.49; F [3, 59] = 19.42, p < 0.001; B = –22.5, differences in continuous variables. Chi-squares were used to assess SE = 9; p = 0.015). differences in adverse experiences variables across groups and other When different types of adverse experiences were analyzed categorical sociodemographic variables. separately, we found that emotional abuse was significantly asso- Linear (time of study measured as days) and multinomial (edu- ciated with fewer days of education in the HC sample (R2 = 0.50, cational level-part-all compulsory education [reference category]) adjusted = 0.48; F [3, 59] = 20.030, p < 0.001; B = 81.77, regression models were used to examine the relationship between SE = 30.27; p = 0.009) but not in the CHR sample. Table 1. Sociodemographic and clinical and characteristics of the sample. CHR (N = 344) HC (N = 67) Statistics Age in years mean (SD) 22.4 (4.9) 22.8 (4.1) t = 0.8, df = 106.5, p = 0.4 Female N (%) 159 (46.2%) 33 (49.3%) X2 = 0.21, df = 1, p = 0.37 Estimate of IQ mean (SD) 98.7 (16.2) 112.3 (17.9) t = 5.8, df = 88.3, p < 0.001 Ethnicity white N (%) 246 (71.7%) 42 (66.7%) X2 = 18.7, df = 5, p = 0.02 Childhood trauma total mean (SD) 47.5 (14.9) 34.8 (10.6) t= 8.3, df = 112.7, p < 0.001 Emotional neglect 13.2 (4.8) 9.4 (4.7) t= 5.9, df = 94.9, p < 0.001 Emotional abuse 11.9 (5.2) 7.4 (3.1) t= 6.8, df = 383, p < 0.001 Physical neglect 8.2 (3.1) 6.4 (2.5) t= 4.3, df = 383, p < 0.001 Physical abuse 7.2 (3.6) 5.9 (2.1) t= 2.9, df = 383, p < 0.001 Sexual abuse 7.1 (4.2) 5.5 (1.9) t= 3, df = 383, p = 0.003 Years in education mean (SD) 14.4 (3.1) 16.0 (2.8) t = 102.3, df = 102.3, p < 0.001 Levels of education N (%) Part/all compulsory education 126 (41.7%) 19 (28.8) X2 = 15.9, df = 2, p < 0.001 First-level non-compulsory education 133 (44%) 24 (36.4%) Higher education 43 (14.2%) 23 (34.8%) Employed N (%) 180 (59%) 60 (90.9%) X2 = 24.2, df = 1, p < 0.001 4 Stefania Tognin et al. Table 2. Number of ACE reported. significantly associated with employment. No significant associ- ations were found in the HC sample. > = 4 ACE 2–3 ACE 1 ACE 0 ACE HC 1 (1.5%) 7 (10.6%) 4 (6.1%) 54 (81.8%) Discussion CHR 29 (9.1%) 78 (20.3%) 74 (23.2%) 138 (43.3%) To the best of our knowledge, this is the first study to analyze the Abbreviations: CHR, clinical high risk of psychosis; HC, healthy controls. relationship between ACE and educational and vocational achieve- ments in a large sample of CHR individuals and a sample of HC with similar sociodemographic characteristics. Relationship between educational level and ACE Our first hypothesis was confirmed as our results showed that The level of education (higher education/first level noncompulsory CHR individuals present significantly worse educational attain- education/part-all compulsory education-reference category) was ments, including fewer years in education than HC. Additionally, significantly associated with CTQ total scores in the CHR group CHR individuals present significantly lower employment rates than (odds ratio [OR] = 0.97, 95% confidence interval [CI] [0.95, 0.99]; HC. This is in line with previous studies reporting an association p = 0.016, and OR = 0.94, 95%CI [0.9, 0.97]; p = 0.001) but not in between most mental health conditions and poor educational the HC sample. achievement [8, 11–14]. For example, Cotter and colleagues [35] When different types of adverse experiences were analyzed reported high rates of unemployment in a CHR sample (23%). At separately, we found that emotional neglect was significantly asso- baseline, those who were unemployed at follow-up had a longer ciated with lower rates of first-level noncompulsory education in duration of untreated illness, more severe negative symptoms, the CHR group (OR = 0.92, 95%CI [0.90, 0.98]; p = 0.006), while in lower IQ, poorer social and occupational functioning, and reported the HC group sexual abuse was significantly associated with lower more childhood trauma than the employed group [35]. In our rates of first-level noncompulsory education (OR = 0.52, 95%CI sample, 41% of CHR were unemployed during the baseline assess- [0.27, 0.99]; p = 0.045) (Table 3 and Figure 1). ment. This high percentage is also consistent with previous reports [27, 36]. Taken together, these findings strengthen the idea that people in the early stages of psychosis generally experience severe Relationship between employment and ACE difficulties with both education and employment. In the CHR group, CTQ total score (OR = 0.98, 95%CI [0.96, 0.99]; Our second hypothesis was confirmed. Our results showed that p = 0.03), physical abuse (OR = 0.9, 95%CI [0.82, 0.98]; p = 0.025), high CTQ scores were associated with significantly fewer years of and psychical neglect (OR = 0.82, 95%CI [0.73, 0.92]; p = 0.001) were education in both the HC and CHR samples. In addition, emotional significantly associated with unemployment. In the CHR group, abuse was negatively associated with days of education in the HC emotional neglect (OR = 1.1, 95%CI [1.003, 1.15]; p = 0.04) was but not in the CHR sample. These results carry significant clinical, Table 3. Relationship between types of ACE and level of education. CHR group Level of education B SE Wald df p Exp (B) 95%CI First-level non-compulsory education Intercept 4.19 1.56 7.22 1 0.007 Age 0.23 0.04 27.01 1 < 0.001 1.25 [1.10, 1.5] Estimate of total IQ 0.02 0.01 5.41 1 0.020 1.02 [1.00, 1.03] CTQ emotional neglect score 0.08 .030 7.55 1 0.006 0.92 [0.90, 0.98] Higher education Intercept 15.03 2.61 33.22 1 < 0.001 Age 0.47 0.06 57.25 1 < 0.001 1.61 [1.42, 1.81] Estimate of total IQ 0.07 0.014 26.03 1 < 0.001 1.07 [1.04, 1.10] CTQ emotional neglect score 0.20 0.05 18.12 1 < 0.001 0.81 [0.74. 0.9] HC group Level of education B SE Wald df p Exp (B) 95%CI First-level non-compulsory education Intercept 26.12 13.02 4.02 1 0.04 Age 0.93 0.41 5.05 1 0.02 2.53 [1.13, 5.68] Estimate of total IQ 0.07 0.044 2.68 1 0.102 1.07 [0.99, 1.72] CTQ sexual abuse score 0.65 0.326 4.00 1 0.045 0.52 [0.27, 0.99] Higher education Intercept 63.71 13.910 20.98 1 < 0.001 Age 1.29 0.427 9.13 1 0.003 3.64 [1.57, 8.41] Estimate of total IQ 0.125 0.050 6.33 1 0.012 1.13 [1.03, 1.25] CTQ sexual abuse score 0.34 0.238 2.01 1 0.156 0.71 [0.45, 1.14] Note: Reference category: part-all compulsory education. Abbreviations: ACE, adverse childhood experiences; CHR, clinical high risk of psychosis; CTQ, Child Trauma Questionnaire; HC, healthy controls. European Psychiatry 5 Figure 1. Adjusted odds ratios (adjOR) and 95% confidence intervals (CI) between CTQ scores and levels of education. Reference group: Part-all compulsory education. public health, and societal implications as ACE are a modifiable and scores in the CHR sample compared to HC. Higher ACE rates in mitigable risk factor. When they come to the attention of Early CHR individuals have been consistently described [40, 41]. ACE Intervention Services, many young CHR individuals are already have also been associated with lower school attainment in clinical experiencing difficulties with education and employment and non-clinical populations [10, 42, 43]. Chronic stress can nega- [36]. While ACE seem to be affecting educational attainments in tively impact cognition, school connectedness, and school attend- both HC and CHR individuals, CHR individuals are reporting ance [44], all of which may mediate the relationship between ACE overall significantly worse educational attainments than HC. The and lack of educational achievement. In line with our results, fact that ACE are similarly associated with educational attainments Hardcastle and colleagues reported that positive history of child- in both HC and CHR is important from a clinical point of view. HC hood adversity doubled the risk of having no qualifications in a with ACE might not come to the attention of clinical services, and general population sample [10], suggesting a dose–response rela- yet they might require support to complete their studies. In this tionship as a higher number of ACE was related to poorer educa- case, ACE screening and related interventions delivered within a tional outcomes. school setting would be most appropriate. Schools are in a unique Finally, our study also showed some inconsistent findings with position to offer a non-stigmatizing intervention to support young physical abuse and neglect associated with unemployment, in line people to complete their studies successfully [37, 38]. If the young with previous studies in the general population [10, 25]; emotional person is already in contact with mental health services, interven- neglect associated with employment, rather than unemployment; tions such as Individual Placement and Support with a dual focus and with the HC not being affected. Future studies should investi- on employment and education might be most appropriate [36]. gate additional factors such as ACE duration and family socio- Higher ACE scores were significantly associated with lower economic status to further clarify these relationships. educational levels achieved in the CHR sample, but not in HC. In Our third hypothesis was partially confirmed. CHR individuals the CHR group, emotional neglect was significantly associated with generally have lower educational attainments than HC. However, lower educational achievements, while in the HC group, sexual both CHR and HC are affected by ACE and our results indicate that abuse was associated with lower educational attainments. Interest- HC are particularly affected in terms of missed days of education. ingly, IQ did not significantly contribute to the prediction model up Our findings suggest that HC who experienced ACE might miss a to the first level of compulsory education, but it contributed sig- significant number of days in education (i.e., 22.5, and 81 when nificantly to progressing to higher education. This supports the specifically affected by emotional abuse) while CHR still miss days, results of a previous study [10] where in those who achieved but less so (i.e., 9.3). This, however, needs to be put into context: secondary level qualifications, ACE did not impact the progression while HC are missing more days compared to CHR individuals, to higher qualifications raising the possibility that completing the they also spend overall more time in education (i.e., 16 vs 14.4 years) compulsory secondary level qualifications could mitigate some of and achieve higher levels of education (14.2% CHR vs 34.8% HC the negative effects of trauma. Exposure to ACE in this sample has complete higher education). Nonetheless, both HC and CHR are been previously described elsewhere [39], with higher rates of CTQ affected by ACE and should be provided with the appropriate 6 Stefania Tognin et al. Social Determination of Health Adverse childhood Mental Health Care experiences Intervention: Disparities: School-based health Access Mental health centers with mental Utilization disorder health services Quality specialized in ACE Funding Poor academic treatment achievement Decreased Social Capital Chronic physical and mental illness Figure 2. Conceptual framework. Adapted from Larson et al. [54]. Conceptual model created by Larson and colleagues that combined Link and Phelan’s 1995 social determinants of health, Braveman’s 2006 measurements of health disparities, and Felitti et al’s 1998 study of exposure to childhood adverse events and negative adult health outcomes [57–59]. support to overcome their difficulties, perform at their best and be Finally, although factors such as parental illness [52], parental able to complete their studies. socioeconomic status [53], duration of ACE [54], and migration are This work examined the relationship between ACE and educa- known to be linked with lower educational attainments, we could tional achievements. Most previous research on CHR individuals not explore these associations in this study due to a high number of focused on the rate of transition to psychosis [45–47]. However, the missing data or to the information not being collected (i.e., duration majority of those at CHR of developing psychosis do not develop a of ACE). psychotic disorder even after 10 years post-identification [48] while Although a previous review highlighted that there is still a need many remain symptomatic and functionally impaired [49]. Based for stronger and broader evidence base in the field of mental health on our findings, we propose that ACE influence academic achieve- promotion in young people within a school setting [55], a multidi- ment directly (Figure 2), and that related difficulties, such as mensional and integrated approach to mental health promotion in symptoms of posttraumatic stress disorder, poor stress- this population is recommended [56]. In our opinion, this approach management, low self-esteem, impairments in mentalization abil- should include the appropriate and timely assessment of ACE and ities, and interpersonal skills should be considered as a treatment their potential impact on the young person’s life, including school target priority for children and adolescents [50], especially those attainments. This would ensure a holistic approach to mental more likely to present poor educational attainments, such as the health. CHR group. As the vast majority of ACE happen either within or close to the family, early family-centered therapeutic and educa- tional interventions should be put in place to support the family as a Conclusions whole as this is expected to both reduce ACE instances and to The effects of ACE are associated with developmental outcomes improve children’s and parents’ mental health [51]. These measures such as educational achievement. From earlier research, we know should also aim at targeting socioeconomic inequalities. that ACE are associated with an increased risk of psychosis [2]. This This work has several strengths. First, the sample is relatively may generate two distinct sources of societal disadvantage and, in large and very well characterized as all participants went through some cases, stigma. Public health and social interventions focused extensive and in-depth clinical interviews. Second, we were able to on prevention and early intervention around ACE are key include an estimate of the IQ in our models, and we could therefore [50]. Moreover, interventions to reduce the impact of ACE and exclude that IQ uniquely explains educational attainments. This to support vocational and educational achievements should be has important clinical implications in terms of the type and mode of routinely implemented within early detection for psychosis services intervention that can be offered to support pupils with mental and in school settings. health difficulties during school age. This work has also several limitations. First, the HC sample is Supplementary Materials. To view supplementary material for this article, relatively small compared to the CHR one which might have limited please visit http://doi.org/10.1192/j.eurpsy.2022.2351. the statistical power to detect differences. Secondly, even though the overall sample was large, when considering the different levels of Data Availability Statement. The data that support the findings of this study education, we had to merge categories to allow sufficient represen- are available on request from the corresponding author. The data are not tation. This might have affected the specificity of our results. Third, publicly available due to privacy or ethical restrictions. we were unable to investigate the relationship between educational Author Contributions. Conceptualization: S.T., A.C., M.J.K., P.McG., A. R., attainments and transition to psychosis. This was due to the rela- J.V., P.M., L. R.V.; Data curation: S.T., A.C., B.N., R.B., N.B, M.K., M.N., S.R., tively low number of transitions to psychosis and the fact that not G.S., B.P.F.R., L.H., M.G; Formal analysis: S.T., A.C., M.K., L. V.; Investigation: enough observations for all variables of interest (in particular when S.T., A.C., M.J.K., B.N., P.McG., A.R., R.B., N.B, M.K., M.N., S.R., G.S., B.P.F.R., analyzing levels of education) were available. J.V., L.H., M.G, P.M., L.V.; Methodology: S.T., A.C., M.J.K., B.N., P.McG., A.R., European Psychiatry 7 R.B., N.B, M.K., M.N., S.R., G.S., B.P.F.R., J.V., L.H., M.G, P.M., L.V.; Investi- [15] Pan J, Zaff JF, Porche M. Social support, childhood adversities, and gation: S.T., A.C., M.J.K., B.N., P.McG., A.R., R.B., N.B, M.K., M.N., S.R., G.S., academic outcomes: a latent class analysis. J Educ Stud Placed Risk. B.P.F.R., J.V., L.H., M.G, P.M., L.V.; Supervision: P.McG., A.R., G.S., B.P.F.R., 2020;25:251–71. J.V., L.H., M.G, P.M., L.V.; Validation: P.M., L.V.; Writing–original draft: S.T., [16] Bunea IM, Szentagotai-Tatar A, Miu AC. Early-life adversity and cortisol A.C., M.J.K., P.M., L.V.; Writing–review and editing: S.T., A.C., M.J.K., B.N., response to social stress: a meta-analysis. Transl Psychiatry. 2017;7:1274. P.McG., A.R., R.B., N.B, M.K., M.N., S.R., G.S., B.P.F.R., J.V., L.H., M.G, P.M., L.V. [17] Vaskinn A, Melle I, Aas M, Berg AO. Sexual abuse and physical neglect in childhood are associated with affective theory of mind in adults with Funding Statement. The European Network of National Schizophrenia Net- schizophrenia. Schizophr Res Cogn. 2021;23:100189. works Studying Gene–Environment Interactions (EU-GEI) Project is funded by [18] Majer M, Nater UM, Lin JM, Capuron L, Reeves WC. Association of grant agreement HEALTH-F2–2010–241909 (Project EU-GEI) from the childhood trauma with cognitive function in healthy adults: a pilot study. European Community’s Seventh Framework Programme. Additional support BMC Neurol. 2010;10:61. was provided by a Medical Research Council Fellowship to M. Kempton (grant [19] Pechtel P, Pizzagalli DA. Effects of early life stress on cognitive and MR/J008915/1). S. Tognin is supported by a Maudsley Charity Grant (1510). affective function: an integrated review of human literature. Psycho- B. Nelson was supported by an NHMRC Senior Research Fellowship (1137687). pharmacology. 2011;214:55–70. Conflicts of Interest. The authors declare none. [20] Fry D, Fang X, Elliott S, Casey T, Zheng X, Li J, et al. The relationships between violence in childhood and educational outcomes: a global sys- tematic review and meta-analysis. Child Abuse Negl. 2018;75:6–28. [21] Forster M, Gower AL, McMorris BJ, Borowsky IW. Adverse childhood References experiences and school-based victimization and perpetration. J Interpers Violence. 2020;35:662–81. [1] Baudin G, Szoke A, Richard JR, Pelissolo A, Leboyer M, Schurhoff F. [22] McDowell N. The association between adverse childhoodvexperiences and Childhood trauma and psychosis: beyond the association. Child Abuse educational outcomes among children ages 6–17. Atlanta, GA: Georgia Negl. 2017;72:227–35. State University; 2017. [2] Varese F, Smeets F, Drukker M, Lieverse R, Lataster T, Viechtbauer W, [23] Murray J, Farrington DP. Risk factors for conduct disorder and delin- et al. Childhood adversities increase the risk of psychosis: a meta-analysis quency: key findings from longitudinal studies. Can J Psychiatr. 2010;55: of patient-control, prospective- and cross-sectional cohort studies. Schi- 633–42. zophr Bull. 2012;38:661–71. [24] Stone AL, Becker LG, Huber AM, Catalano RF. Review of risk and [3] Morkved N, Johnsen E, Kroken RA, Gjestad R, Winje D, Thimm J, et al. protective factors of substance use and problem use in emerging adult- Does childhood trauma influence cognitive functioning in schizophrenia? hood. Addict Behav. 2012;37:747–75. The association of childhood trauma and cognition in schizophrenia [25] Topitzes J, Pate DJ, Berman ND, Medina-Kirchner C. Adverse childhood spectrum disorders. Schizophr Res Cogn. 2020;21:100179. experiences, health, and employment: a study of men seeking job services. [4] Bilgi MM, Taspinar S, Aksoy B, Oguz K, Coburn K, Gonul AS. The Child Abuse Negl. 2016;61:23–34. relationship between childhood trauma, emotion recognition, and irrit- [26] Houtepen LC, Heron J, Suderman MJ, Fraser A, Chittleborough CR, Howe ability in schizophrenia patients. Psychiatry Res. 2017;251:90–6. LD. Associations of adverse childhood experiences with educational attain- [5] Tognin S, Catalan A, Modinos G, Kempton MJ, Bilbao A, Nelson B, et al. ment and adolescent health and the role of family and socioeconomic Emotion recognition and adverse childhood experiences in individuals at factors: a prospective cohort study in the UK. PLoS Med. 2020;17:e1003031. clinical high risk of psychosis. Schizophr Bull. 2020;46:823–33. [27] Fusar-Poli P, Byrne M, Badger S, Valmaggia LR, McGuire PK. Outreach [6] Boyda D, McFeeters D. Childhood maltreatment and social functioning in and support in South London (OASIS), 2001-2011: ten years of early adults with sub-clinical psychosis. Psychiatry Res. 2015;226:376–82. diagnosis and treatment for young individuals at high clinical risk for [7] Alameda L, Ferrari C, Baumann PS, Gholam-Rezaee M, Do KQ, Conus P. psychosis. Eur Psychiatry. 2013;28:315–26. Childhood sexual and physical abuse: age at exposure modulates impact [28] (EU-GEI) EnonnsG-EiiS, van Os J, Rutten BP, Myin-Germeys I, Delespaul on functional outcome in early psychosis patients. Psychol Med. 2015;45: P, Viechtbauer W, et al. Identifying gene–environment interactions in 2727–36. schizophrenia: contemporary challenges for integrated, large-scale inves- [8] Dalsgaard S, McGrath J, Ostergaard SD, Wray NR, Pedersen CB, Mor- tigations. Schizophr Bull. 2014;40:729–36. tensen PB, et al. Association of mental disorder in childhood and adoles- cence with subsequent educational achievement. JAMA Psychiatry. 2020; [29] Yung AR, Yuen HP, McGorry PD, Phillips LJ, Kelly D, Dell’Olio M, et al. 77:797–805. Mapping the onset of psychosis: the comprehensive assessment of at-risk [9] Brannlund A, Strandh M, Nilsson K. Mental-health and educational mental states. Aust N Z J Psychiatry. 2005;39:964–71. achievement: the link between poor mental-health and upper secondary [30] Mallett R. Sociodemographic schedule. In: Section of social psychiatry. school completion and grades. J Ment Health. 2017;26:318–25. London: Institute of Psychiatry; 1997. [10] Hardcastle K, Bellis MA, Ford K, Hughes K, Garner J, Ramos Rodriguez G. [31] First MB, Spitzer R, Gibbon M, Williams J. Structured clinical interview for Measuring the relationships between adverse childhood experiences and Axis I Disorders-Patient Edition. New York: Biometrics Research, New educational and employment success in England and Wales: findings from York State Psychiatric Institute; 1996. a retrospective study. Public Health. 2018;165:106–16. [32] Bernstein DP, Stein JA, Newcomb MD, Walker E, David Pogge D, Ahlu- [11] Hakulinen C, McGrath JJ, Timmerman A, Skipper N, Mortensen PB, valia T, et al. Development and validation of a brief screening version of the Pedersen CB, et al. The association between early-onset schizophrenia childhood trauma questionnaire. Child Abuse Negl. 2003;27:169–90. with employment, income, education, and cohabitation status: nationwide [33] Hughes K, Bellis MA, Hardcastle KA, Sethi D, Butchart A, Mikton C, et al. study with 35 years of follow-up. Soc Psychiatry Psychiatr Epidemiol. The impact of multiple adverse childhoodvexperiences on health: a sys- 2019;54:1343–51. tematic review and metaanalysis lancet public health 2017;2:e356e66. [12] Keen D, Webster A, Ridley G. How well are children with autism spectrum [34] IBM Corporation. IBM SPSS statistics for windows, version 27.0. Armonk, disorder doing academically at school? An overview of the literature. NY: IBM Corp; 2020. Autism. 2016;20:276–94. [35] Cotter J, Lin A, Drake RJ, Thompson A, Nelson B, McGorry P, et al. Long- [13] Tosto MG, Momi SK, Asherson P, Malki K. A systematic review of term employment among people at ultra-high risk for psychosis. Schi- attention deficit hyperactivity disorder (ADHD) and mathematical ability: zophr Res. 2017;184:26–31. current findings and future implications. BMC Med. 2015;13:204. [36] Tognin S, Grady L, Ventura S, Valmaggia L, Sear V, McGuire P, et al. The [14] Hooper SR, Woolley D, De Bellis MD. Intellectual, neurocognitive, and provision of education and employment support at the outreach and academic achievement in abstinent adolescents with cannabis use dis- support in South London (OASIS) service for people at clinical high risk order. Psychopharmacology. 2014;231:1467–77. for psychosis. Front Psych. 2019;10:799. 8 Stefania Tognin et al. [37] Dray J, Bowman J, Campbell E, Freund M, Hodder R, Wolfenden L, et al. [48] Nelson B, Yuen HP, Wood SJ, Lin A, Spiliotacopoulos D, Bruxner Effectiveness of a pragmatic school-based universal intervention targeting A, et al. Long-term follow-up of a group at ultra high risk student resilience protective factors in reducing mental health problems in (“prodromal”) for psychosis: the PACE 400 study. JAMA Psychiatry. adolescents. J Adolesc. 2017;57:74–89. 2013;70:793–802. [38] Salerno JP. Effectiveness of universal school-based mental health aware- [49] Luo X-j, Huang L, van den Oord EJ, Aberg KA, Gan L, Zhao Z, et al. ness programs among youth in the United States: a systematic review. J Sch Common variants in the MKL1 gene confer risk of schizophrenia. Schi- Health. 2016;86:922–31. zophr Bull. 2015;41:715–27. [39] Kraan T, van Dam DS, Velthorst E, de Ruigh EL, Nieman DH, Durston S, [50] Lorenc T, Lester S, Sutcliffe K, Stansfield C, Thomas J. Interventions to et al. Childhood trauma and clinical outcome in patients at ultra-high risk support people exposed to adverse childhood experiences: systematic of transition to psychosis. Schizophr Res. 2015;169:193–8. review of systematic reviews. BMC Public Health. 2020;20:657. [40] Addington J, Stowkowy J, Cadenhead KS, Cornblatt BA, McGlashan TH, [51] Salokangas RKR. Childhood adversities and mental ill health. Finland: Perkins DO, et al. Early traumatic experiences in those at clinical high risk University of Turku; 2020. for psychosis. Early Interv Psychiatry. 2013;7:300–5. [52] Bortes C, Strandh M, Nilsson K. Parental illness and young people’s [41] Loewy RL, Corey S, Amirfathi F, Dabit S, Fulford D, Pearson R, et al. education. Child Indic Res. 2020;13:2069–91. Childhood trauma and clinical high risk for psychosis. Schizophr Res. [53] Rodriguez-Hernandez CF, Cascallar E, Kyndt E. Socio-economic status 2019;205:10–4. and academic performance in higher education: a systematic review. Educ [42] Romano E, Babchishin L, Marquis R, Frechette S. Childhood maltreatment Res Rev. 2020;29:100305. and educational outcomes. Trauma Violence Abuse. 2015;16:418–37. [54] Larson S, Chapman S, Spetz J, Brindis CD. Chronic childhood trauma, [43] Porche MV, Fortuna LR, Lin J, Alegria M. Childhood trauma and psychi- mental health, academic achievement, and school-based health center atric disorders as correlates of school dropout in a national sample of mental health services. J Sch Health. 2017;87:675–86. young adults. Child Dev. 2011;82:982–98. [55] O’Reilly M, Svirydzenka N, Adams S, Dogra N. Review of mental health [44] Basch CE. Aggression and violence and the achievement gap among urban promotion interventions in schools. Soc Psychiatry Psychiatr Epidemiol. minority youth. J Sch Health. 2011;81:619e25. 2018;53:647–62. [45] Ruhrmann S, Schultze-Lutter F, Salokangas RK, Heinimaa M, Linszen D, [56] Weare K, Murray M. Building a sustainable approach to mental health Dingemans P, et al. Prediction of psychosis in adolescents and young work in schools. Int J Ment Health Promot. 2004;6:53–9. adults at high risk: results from the prospective European prediction of [57] Link B, Phelan J. Social conditions as fundamental causes of disease. J psychosis study. Arch Gen Psychiatry. 2010;67:241–51. Health Soc Behav. 1995:80–94. PMID: 7560851. [46] Kraan TC, Velthorst E, Themmen M, Valmaggia L, Kempton MJ, McGuire P, [58] Braveman P. Health disparities and health equity: concepts and measure- et al. Child maltreatment and clinical outcome in individuals at ultra-high risk ment. Annu Rev Public Health. 2006;27:167–94. for psychosis in the EU-GEI high risk study. Schizophr Bull. 2018;44:584–92. [59] Felitti VJ, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards V, [47] Fusar-Poli P, Bonoldi I, Yung AR, Borgwardt S, Kempton MJ, Valmaggia et al. Relationship of childhood abuse and household dysfunction to many L, et al. Predicting psychosis meta-analysis of transition outcomes of the leading causes of death in adults. The adverse childhood experiences in individuals at high clinical risk. Arch Gen Psychiatry. 2012;69 220–9. (ACE) study. Am J Prev Med. 1998;14:245–58.