Adverse childhood experiences, mental health, and social functioning: A scoping review of the literature

Background: Adverse childhood experiences (ACEs) negatively impact people's physical and mental health and social functioning. Research literature focuses on the impact of ACEs on physical and mental health, yet to our knowledge, no study has examined the literature on ACEs, mental health, and social functioning outcomes. Objective: To map how ACEs, mental health, and social functioning outcomes have been defined, assessed, and studied in the empirical literature and identify gaps in the current research which need further investigation. Methods: A scoping review methodology following a five-step framework was implemented. Four databases were searched CINAHL, Ovid (Medline, Embase) and PsycInfo. The analysis involved both numerical and a narrative synthesis in line with the framework. Results: Fifty-eight studies were included in the analysis, and three key issues were identified a) the limitations of research samples to date, b) the choice of outcome measures for ACEs, social and mental health outcomes, and c) the limitations of current study designs. Conclusion: The review demonstrates variability in the documentation of participant characteristics and inconsistencies in the definitions and applications of ACEs, social and mental health and related measurements. There is also a lack of longitudinal and experimental study designs, studies on severe mental illness, and studies including minority groups, adolescents, and older adults with mental health problems. Existing research is highly variable methodologically and limits our broader understanding of the relationships between ACEs, mental health, and social functioning outcomes. Future research should implement robust methodologies to provide evidence that could be used for developing evidence-based interventions. * Corresponding author. E-mail address: vasiliki.tzouvara@kcl.ac.uk (V. Tzouvara).


Introduction
Adverse childhood experiences (ACEs) are traumatic events that children and adolescents under 18 years of age have experienced (Crouch et al., 2019). ACEs cover a broad range of traumatic events, including physical and emotional neglect, physical, sexual, and emotional abuse, exposure to domestic violence, mental health problems, family incarceration, separation and substance misuse (Felitti et al., 1998). Various individual, family, and community factors can affect a child's likelihood of experiencing ACEs, such as living in unstable housing, having parents who have experienced ACEs, and growing up in communities with high levels of social and environmental dysfunction (CDC, 2021;Fagrell Trygg et al., 2019;Hargreaves et al., 2019;Walsh et al., 2019). Epidemiological research reveals that millions of children are affected by ACEs each year globally (Asmussen et al., 2020). A World Health Organization (WHO) study of 51,945 adults revealed that more than half of the respondents surveyed had experienced multiple ACEs and that ACEs were significantly associated with the risk of DSM-IV disorders across all countries (Kessler et al., 2010).
Research has long established the link between poor mental health outcomes and ACEs (Beilharz et al., 2020;McLafferty et al., 2019). People with a history of ACEs are at greater risk of experiencing a range of mental health problems, such as depression, bipolar disorder, suicide, and substance misuse (Fuller-Thomson et al., 2016;Leza et al., 2021;Merrick et al., 2017). Research has found an association between ACEs and alterations in adverse childhood experiences are associated with changes in biological systems. Children exposed to maltreatment showed smaller volume of the prefrontal cortex, greater activation of the hypothalamic-pituitary-adrenal (HPA) axis, and elevation in inflammation levels, while adults with a history of maltreatment showed smaller volume of the prefrontal cortex and hippocampus, greater activation of the HPA axis, and elevation in inflammation levels compared to non-maltreated individuals (Danese & McEwen, 2012).
Social functioning has also been identified as key in the relationship between ACEs and poor mental health outcomes Hyland et al., 2019;McCrory et al., 2019). A meta-analysis of social measures (Valtorta et al., 2016) has established two dimensions of social relationships: objective (i.e., the structure and function of relationships) and subjective (i.e., involvement in relationships, perceived availability, perceived adequacy, feelings/emotions). People who are subjected to ACEs are more likely than their peers to experience difficulty developing healthy relationships due to lack of trust, poor emotional regulation skills, and maladaptive coping strategies (Poole et al., 2018). Consequently, ACEs are associated with increased social isolation (an objective deficiency in high-quality social relationships) and loneliness (a subjectively perceived gap between desired social contact and actual contact) in later life (Forster et al., 2020;Sheikh, 2018aSheikh, , 2018bWeber Ku et al., 2021). These two social functioning factors have been identified as mediators in the development of adult psychiatric morbidity for people with ACEs (Hyland et al., 2019;Shevlin et al., 2015). As research has established that loneliness and social isolation can hinder recovery from mental illness , these concepts have particular clinical relevance for mental health practitioners. A trauma-informed approach to clinical care takes into consideration the need to assess for ACEs and recognises the importance of the psychosocial aspects of recovery to ensure effective mental health care delivery (Oral et al., 2016;Ranjbar & Erb, 2019). Towards that aim, mental health policies (e.g., UK NHS Mental Health Implementation Plan 2019/20-2023/24) now recognise the development and implementation of trauma-informed care as a key priority for future mental health care services.
Recently published reviews have focussed on investigating the links between ACEs and health and measures and methods of ACEs in a broader context. For example, Hughes et al. (2017) systematically reviewed the effects of multiple ACEs on health and found associations between ACEs and various health outcomes, including mental ill health and substance abuse. Liu et al.'s (2021) systematic review and meta-analysis of the lifetime prevalence of ACEs in homeless people found an association between ACEs exposure, functional health and mental health problems. Karatekin et al. (2022) recently conducted a scoping review of the ACEs literature to determine the direction of current research and found that studies had predominantly focused on the effects of ACEs rather than on the causes of ACEs or how to prevent them from occurring (Karatekin et al., 2022). None of these reviews, however, focused solely on ACEs in people with mental health problems across the lifespan, nor the contribution of social functioning outcomes, such as loneliness and social isolation, in this population. There is, therefore, currently limited understanding of how ACEs are defined in the mental health literature, and a lack of clarity as to the types of mental health problems and social functioning outcomes most often examined in ACEs research. Whilst the literature has yet to be comprehensively reviewed, the breadth and potential heterogeneity of ACEs research may make it challenging to conduct a meta-analytic review in this area. Given this, an initial scoping review was considered an appropriate way to map research in mental health, social functioning outcomes and ACEs, as well as to also identify gaps and limitations to date, and provide guidance on what research is needed to advance the field (Levac et al., 2010).

Scoping review procedure
A scoping approach provides a preliminary assessment of the research area with the aim of identifying and determining the nature and extent of research literature in a particular area (Grant & Booth, 2009). This scoping review follows the five-step framework set out by Arksey and O'Malley (2005) (adapted by Levac et al., 2010): namely, stage 1) Identify the research question, stage 2) Identify relevant studies (search strategy), stage 3) Select studies, stage 4) Chart the data (data extraction) and stage 5) Collate, summarise, and report results.

Stage1
Targeted research questions being developed to guide the scoping review: 1. How have the concepts of ACEs, mental health and social functioning outcomes been operationalised and assessed in the empirical literature? 2. What are the gaps and limitations in current research, and which areas require further investigation? Girls had significantly higher median Children's Depression Inventory (CDI) and Child Posttraumatic Stress Reaction Index (CPTS-RI) scores than boys, while no significant difference was determined between boys and girls in terms of Perceived Social Support Scale-Revised (PSSS-R) scores. In addition, a statistically significant negative correlation was determined between CDI and PSSS-R scores, CPTS-RI scores and PSSS-R scores in girls, while no significant correlation was identified in male victims. Perceived social support in girls was observed to bestow a greater psychological benefit compared to boys.
Further studies are needed to determine the sub-factors involved in the differences exhibited by perceived social support in reducing psychological symptoms for the male and female gender, that social support needs to be increased and expanded for both genders, and that priority must be attached to protective approaches.
2. Baiden et al. (2017), Canada To examine the effect of social support and disclosure of child abuse to Child Protection Service on lifetime suicidal ideation among Canadian adults who were abused when they were children.      (self-report) • Social Support Questionnaire 6 (SSQ-6) [α = N/R] Childhood sexual abuse was found to be an independent predictor of suicidal ideation and behaviour. Both problem-solving confidence and social support moderated the relationship between childhood abuse and suicidal ideation. The results of this study underscore the importance of both problem-solving appraisal and social support for suicidality The results of this study are generalizable only to delinquent adolescents. It is possible that adolescents from the general population would respond differently than those with conduct disordered behaviour. These are questions that deserve study in future research.
(continued on next page) • Six questions adapted from the original adverse childhood experiences study (Felitti et al., 1998)  To advance our understanding of school-based sources of social support future work should consider using more robust tools to assess these constructs.
12. Gallus et al. (2015), USA     Respondents with a first episode of schizophrenia Respondents with a high risk for psychosis In univariate analysis, high risk for psychosis (HR) individuals had more childhood trauma, more recent life events and less social support than the healthy control (HC) group, and these findings were also supported by ANCOVA analysis except for the results related to social support after taking age, education, marital and employment status as covariates. HR Calls for further exploration to develop optimal psychosocial interventions which may be beneficial in improving symptoms of high-risk individuals and may therefore help to delay and reduce conversion to psychosis.
(continued on next page)   Future research could make further contributions by examining the interaction between childhood protective factors and later life physical health quality. It would also be worthwhile to expand the scope and specificity of protective experiences and explore aspects of these experiences, such as duration of involvement and degree of satisfaction derived from particular activities.
Mixed methods research could elucidate processes involved in the development of ACEinformed BH service and begin to study health outcomes. There is a need to identify and examine (continued on next page)           The experience of child abuse, rape, and a lack of perceived support, enormously increases women's likelihood of probable clinical depression and, especially, probable posttraumatic stress disorder (PTSD). Child abuse was a significant risk factor for adult rape, such that women who were abused as children were twice as likely to experience adult rape, compared to those not abused. Women who reported both child abuse and adult rape reported more severe depressive moods and were more likely to be at risk for major depression than those without trauma histories. Women with trauma in both developmental periods were over three times as likely to be at high risk for probable depression compared to those with no trauma. Finally, women who reported child abuse or adult rape exclusively reported higher depressive mood than women not reporting child abuse or adult rape, but less Future research should attempt to examine the impact of social support for women with cumulative traumatic experiences in a longitudinal design.
(continued on next page)      Childhood maltreatment was found to be negatively associated with social support, positive coping skills, and positive mental health but positively associated with psychological distress.
Social support and positive coping skills predicted higher rates of positive mental health but lower rates of psychological distress. Social support and positive coping skills partially mediated the negative consequences of Future longitudinal studies following children exposed to maltreatment are required to definitively establish causal relationships and test how mental health issues and resilience interact over time. Childhood maltreatment needs to be more comprehensively assessed to capture more dimensions of such experiences.
(continued on next page)       Additional longitudinal research is needed to assess its stability.
(continued on next page) studies, the majority of participants were female (n = 49), while nine studies reported a higher percentage of males in the sample. Twenty-two studies did not provide information on participants' ethnicity/race. Papers reporting ethnicity/race information stated the highest percentages of participants from White ethnic backgrounds (n = 21), followed by two studies with the highest percentage of participants from African American origin.

ACEs
There were 302 discrete variables measured under the category of ACEs across the studies included in this review. While many studies measured physical abuse (n = 47), sexual abuse (n = 41), and emotional abuse (n = 35), the operational definition and measurement tool for each type of abuse varied across studies. For example, 39 out of 41 studies examining sexual abuse defined the concept simply as sexual abuse, but a further four studies adopted different definitions: inappropriate touching, sexual abuse by a parent or adult in the home, sexual abuse by anyone, and sexual assault.
Most articles (87.93 %) used variants of the original ACES measure (Felitti et al., 1998) as their operational definition of ACES, while 2 % used the original framework. In addition, 22.4 % of articles used additional sub-categories to ACEs, such as bullying or witnessing mugging. The most frequently used tool for assessing ACES was the Childhood Trauma Questionnaire (CTQ). However, it was common that modifications were made to its format and analysis approach. For example, Yearwood et al. (2019) used the short form of the questionnaire (CTQ-SF) to create a latent variable called Complex Trauma, which was then used in the analysis. Additionally, Schumm et al. (2006) also used an "abbreviated version" of CTQ-SF. It is unclear whether this is a reference to the accepted Short Form version of the tool or if they made extra changes. In addition, only 18 papers reported any reliability and/or validity scores for ACEs measures. A comprehensive list of ACEs discrete variables and measurements can be found in Table 3.

Social functioning outcomes
Studies measured both objective and subjective aspects of social functioning outcomes (loneliness: n = 8; social isolation/social support: n = 63). Loneliness included a range of subjective variables such as emotional loneliness, family and friend harmony, and perception of others' fondness for self. Social isolation also included a range of objective variables such as family resources, financial assistance, and neighbourhood cohesion. A further four studies also considered variables measuring social relationships, such as annual income (Haahr-Pedersen et al., 2020) and transitions and changes . These variables were included in bespoke tools created for the purpose of the study, and no validity or reliability figures were reported.
Of the measures that included social functioning outcome measurements, 35 % reported any reliability or validity scores for social relationship measures. Like other measures in this review, many of the social isolation and loneliness scales used were modified or ad hoc instruments (e.g., Larkin et al., 2018;Steine et al., 2020;Wang et al., 2018). Of the 58 papers included in this review, over 20 relied on non-validated measures of social isolation and loneliness, including large-scale population surveys (n = 3) and other bespoke questionnaires (n = 15). A further two studies assessed aspects of social relationships qualitatively through semi-structured interviews (n = 2). Social support emerged as a distinct, with 46 studies examining some facets of social support such as perceived and actual emotional support, peer support, and family support. However, consideration of the complexity of social support was lacking, and the distinctions between the objective and subjective nature of social support in the research were not regularly addressed in the choice of research measures. A comprehensive list of social functioning outcome measurements can be viewed in Table 4.

Mental health outcomes
Categories of mental health examined in the 58 studies covered mood disorders (n = 44), anxiety disorders (n = 19), PTSD (n = 11), psychotic disorders (n = 12), personality disorders (n = 1). Additionally, there were 37 miscellaneous aspects of mental health examined, such as sleep disorders (n = 3), suicide/self-harm (n = 14), and substance misuse (n = 7). The most frequently examined mental illness was depression (n = 36). A broad range of tools was used to assess mental health, but only 52/58 studies used established and validated measures, and only 33 studies reported reliability or validity scores. The most frequently used tool was the Patient Health Questionnaire (PHQ) or a PHQ variant version (n = 7). However, other measures were frequently present in modified and shortform versions. For example, five studies used the Beck Depression Inventory (BDI) scale, four studies used the Epidemiologic Studies -Depression Scale (CES -D), three studies used the Generalized Anxiety Disorder 7-item Scale (GAD-7) and the Hopkins Symptom Checklist (HSCL-10), and two studies used the PTSD Symptom Scale-Self Reported (PSS-SR). A comprehensive list of Mental health categories and outcome measures can be found in Table 5.

Discussion
The aims of this review were to examine how the concepts of ACEs, mental health, and social functioning outcomes have been defined, assessed, and studied in the empirical literature to date and to identify gaps or weaknesses in current research to guide further investigation. This review identified 58 studies examining ACEs, social functioning outcomes, and mental health outcomes in a wide range of populations. Whilst many studies employed a robust design and methodology, there was high variability and substantial discrepancies in the definition and operationalisation of key concepts and the outcome measures used. The following issues emerged from the narrative synthesis: a) the limitations of research samples to date, b) the choice of outcome measures for ACEs, social functioning and mental health outcomes, and c) the limitations of current study designs.

Table 3
ACEs, measurements and discrete variable in included studies.    x One question on the perceived size of participants' peer network: x One question on perceived social isolation x Two items on social support in adulthood x One question on instrumental/tangible support x One question on emotional support x One item on loneliness x Five items on relational difficulties x Social Supports Provision Scale (SSPS) x Relationship Scales Questionnaire (RSQ) x Perceived Social Support Questionnaire (F-SozU) x Social Provisions Scale (SPS) x Premorbid Adjustment Scale (PAS) Two questions on perceived social support from the Childhood Experience of Care and Abuse Questionnaire (CECA.Q) x Adolescent Social Support Scale x Social Support Rating Scale (SSRS) x Questions on mentoring x x (continued on next page)

Diversity in research samples
The documentation of sample characteristics was highly variable across the included studies, which makes results difficult to replicate, validate, or generalize. This review found that within the studies which provided information, the majority of the populations were women and binary gender populations. This disparity highlights the need to examine ACEs, and social and mental health outcomes in populations with different gender identities, including LGBTQ+ and non-binary populations, and men to advance our understanding of these subgroups, who have varying risks for poor social, emotional, and mental health (Almuneef et al., 2017;Haahr-Pedersen et al., 2020;Jones et al., 2022). Similarly, only twenty-six studies documented participants' ethnicity information, while twenty of them examined White ethnic populations (n = 20). Previous studies have focused on ethnic/racial disparities in ACEs , while others examined these differences predominantly in relation to mental health (Lee & Chen, 2017;Zhang & Monnat, 2022); however, there is very little research that examines the role of social outcomes in these populations. Research suggests that there are racial differences in both the number and types of ACEs experienced by racial groups (Maguire-Jack et al., 2020). Racial/ethnic differences in ACEs, social outcomes, and mental health have important implications for intervention development and clinical practice if they are to effectively address poly victimization, racism, racial stigma, stereotypes, and discrimination which increase poor mental health and negative social outcomes (Zhang & Monnat, 2022). The majority of studies were conducted in the US, UK and Canada, and thus it is desirable that ACES research from other countries, with different cultural environments and potentially different attitudes to mental health and social support, is adequately represented in the published literature to avoid bias.
This review found that children, adolescents, and older adults remain under-researched in this area. Cross-sectional and longitudinal studies have established the negative relationships between ACEs and poor mental health outcomes in children and adolescents, yet the impacts on social functioning outcomes have been largely overlooked. Teenagers with one ACE are also at greater risk of experiencing Attention Deficit Hyperactivity Disorder, behavioural/conduct problems, substance abuse disorders or a mental health diagnosis compared to youth without ACEs (Bomysoad & Francis, 2020). Traditionally, most of the published literature has focused on adult samples and retrospective reports on adversities experienced during childhood due to the definition of ACEs as experiences that occur before the age of 16 years (Struck et al., 2021). Although an increasing trend of studying the effects of ACEs on younger samples is noted, there is still an urgent need for more research on this population.
The wider literature also suggests that ACEs exposure is a key risk factor for serious mental illness (SMI), including substance use disorders (Bryant et al., 2020) and schizophrenia disorders (Prokopez et al., 2018). However, a limited number examined people with SMI, psychiatric comorbidities and/or inpatients in mental health care facilities and only one study in this review examined social functioning outcomes and ACEs in people with personality disorders. Certainly, more research is needed to understand the breadth and extent of these relationships in people experiencing SMI and/or receiving care in mental health services. The latter will help us to understand the mechanisms of ACEs and will support trauma-informed care in mental health care services (NHS England, 2019).

Inconsistent outcome measures for ACEs, social functioning and mental health
Conceptualisation and operationalisation discrepancies can negatively impact the interpretation of the literature, while discrepancies in conceptual models explaining the relationships and predictive power of ACEs on outcomes impose a risk for policy and intervention development (Bentall et al., 2012). This review shows that the majority of studies have focused on physical abuse, sexual abuse and emotional abuse, with little research on neglect, and family dysfunction variables. Felitti's (1998) typology of ACES was used to inform the scope of ACES literature in this study. Indeed, it is a widely accepted framework and is currently used by the Centres for Disease Control and Prevention and the Scottish Government. Although this approach to quantifying exposure to ACEs, which is based on 10 adverse experiences, has "predictive validity" (Lewer et al., 2020, p. 493), it has been the focus of criticism in recent years. Indeed, it has been argued that it does not adequately capture a sufficient range of adverse events, particularly in vulnerable populations (Zhang & Monnat, 2022) and that other indicators such as socioeconomic adversities (e.g., economic hardship) are more important in predicting health outcomes. Recent studies argue that rather than counting the number of ACEs as a cumulative, linear dose, research should reflect the interplay between different ACEs that occur. It is suggested that a mixture of modelling approaches to identifying clusters of adversities might explain the impact of adverse experiences more effectively (Zhang & Monnat, 2022). Notably there was a lack of consistency and clarity to the measurement of social functioning outcomes, with variability between studies and with the distinction between objective and subjective social support not regularly addressed. Whilst the concepts are directly related and often used interchangeably, they are distinct in that one can occur with or without the other. Therefore it is argued that social isolation and loneliness should be researched as separate but related entities (Newall & Menec, 2019).
The poor operationalisation of certain outcomes was also reflected by the range of measures used to capture ACEs, social    (2015) functioning outcomes, and mental health outcomes. A significant number of studies used non-validated or ad hoc measures, particularly in the assessment of social functioning outcomes, which have less established conceptual definitions. In ACEs measurement, although the use of validated instruments was more common, it was also noted that tools were modified or adapted for use, often without justification or reliability data for the version used. Similarly, in the assessment of mental health outcomes, there was still a failure to provide reliability and validity data for population use on the tools or measures used. This wide variability of the types of instruments used across the literature means that comparisons, replicability, and accurate conclusions on the relationships between the constructs is challenging. The lack of psychometric data also raises significant questions about the validity, generalisability, and quality of each study (Bryman, 2012). Finally, the use of self-reported measures also raises concerns due to social desirability bias, response bias, the risk for discrepancies in the interpretations of questions and lack of response flexibility (Demetriou et al., 2015).

Limitations of current research designs
This review found that cross-sectional designs were the most widely implemented research design, followed by cohort studies, while only a limited number of longitudinal studies were identified. In addition, power calculations were only performed in two studies, highlighting the limitations of the current literature in relation to a precise and accurate conclusion in the absence of an appropriate sample size (Nayak, 2010). Even adequately powered cross-sectional retrospective designs are vulnerable to attrition and recall biases due to the nature of their design (Hardt & Rutter, 2004;Hartas, 2019;Reuben et al., 2016). They require retrospective recall of childhood adversities which could be affected by a) inconsistencies in early memories, b) limited capacity to remember the time of event sequences, c) the frequency of recall and consolidation of traumatic memory, d) childhood amnesia, and e) individual's personality (Hartas, 2019). Limitations related to retrospective cross-sectional designs highlight the need for more longitudinal studies since they allow both within-and between-group comparisons through the collection of multiple data at different points in time. They are also able to differentiate between environmental effects on personal outcomes and personal effects on the environment, as well as assess attrition bias which increases the validity of the study findings (Hardt & Rutter, 2004). Finally, increasing reports of domestic violence, and therefore ACEs during and after the Covid-19 pandemic, along with poor social and mental health outcomes (World Health Organisation (WHO), 2020, Dawson et al., 2021), signifies the need to study the impact of ACEs longitudinally. However, it is perhaps not surprising that cross-sectional studies continue to dominate, given that longitudinal studies require large sample sizes and are time-consuming and expensive compared to cross-sectional designs (Caruana et al., 2015).

Limitations
Whilst a scoping review framework was implemented in this review, ensuring trustworthiness and allowing replicability, only four databases were searched, and grey literature and non-peer reviewed studies were not included. In addition, papers were limited to those published in the English language, and thus the review may have missed key papers, leading to a failure to capture culturally specific issues regarding ACES research and limiting the generalisability of the review's findings. Another limitation relates to the use of Felitti et al.'s (1998) framework for conceptualising ACEs which fails to capture more recent potential constructs of ACEs such as socio-economic dimensions. However, the framework is still valid and is so widely used in the ACEs literature to date that a pragmatic decision was made to employ this approach for the review.

Conclusion
This study mapped the evidence in relation to the definitions and operationalisation of ACEs and mental health and social functioning outcomes in current research. It highlighted the limited evidence on populations from diverse backgrounds and various gender identities, as well as the lack of evidence on minority groups. The shift towards examining ACEs clusters and attributing relationships at the cluster level is crucial since there is evidence that specific types of mental health disorders are associated with specific types of childhood adversities (Bruni et al., 2018). Similarly, the mechanisms of social predictors in mental health need to be established for comparisons between outcomes. The validity and reliability and the development of established measurements for assessing ACEs and social outcomes are important. By doing so, we will be in a better future position to effectively examine the relationships between the concepts and identify significant mechanisms and pathways that will allow comparisons between studies and inform future research and interventions.

Funding statement
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.

Declaration of competing interest
None.

Data availability
Data will be made available on request.