A systematic review and meta‐analysis investigating the impact of childhood adversities on the mental health of LGBT+ youth

Abstract Background The presence of trauma as a backdrop to the lives of LGBT+ youth has been recognised in recent literature. LGBT+ youth report a higher frequency, severity and pervasiveness of adverse childhood experiences when compared to their heterosexual and cisgender counterparts. This exposure has been directly related to an increased risk of mental health problems. Method A systematic literature search of Medline, Embase, PsycINFO, PubMed and Web of Science was conducted from the date of their inception until the 1st September 2021. The study protocol was registered in PROSPERO (CRD42021240472). Results A total of 27 studies satisfied the inclusion criteria and were used in the systematic review, representing 199,285 participants, 26,505 of whom identified as LGBT+ (mean age 16.54). Female participants (ranging from 11% to 74%) and white participants (7.7%–96%) made up the largest percentage of most samples. Depressive symptoms were the most commonly described psychiatric outcome (n = 17, 63%), followed by anxiety symptoms (n = 6, 31.5%). 18 studies provided meta‐analysable data, compromising 21,781 LGBT+ young people. LGBT+ youth reported a higher prevalence of adverse experiences in comparison to their heterosexual or cisgender counterparts (p < .001), with sexual abuse representing the most commonly reported experience (29.7%), followed by verbal abuse (28.7%), physical abuse (26.5%) and cyberbullying (19.1%). LGBT+ youth were also at a heightened risk of mental health disorders (p < .001), with 36.9% and 31.5% of sample meeting the clinical criteria for depression and anxiety, respectively. Conclusions Continued advocacy is needed from communities and Allies to support and empower LGBT+ youth in the face of adversity. Longitudinal and longer‐term studies are required to further understand the relationship between adverse experiences in LGBT+ youth and the impact on mental health.


INTRODUCTION
The first study to assess adverse childhood experiences (ACEs) captured a series of ongoing stressful or traumatic events experienced by an individual before the age of 18 (Felitti et al., 1998). The most commonly reported included physical, psychological and sexual abuse, physical and emotional neglect, and five aspects of household dysfunction. Previous research suggests that the risk of exposure to a traumatic event escalates for individuals with minority identities related to sexual orientation, gender expression, or gender identity (the LGBT+ community; Clements-Nolle et al., 2018). They are the target of unwanted victimisation across a wide range of contexts, for example, parental physical assault, harassment at school, and on social media platforms (Williams et al., 2021).
Despite burgeoning global activism and statutory advances in legislative reform, research continues to document how sexual and gender minority groups are at greater risk of mental health difficulties (Semlyen et al., 2016). This is illustrated by increased prevalence of depressive symptoms, anxiety, substance abuse disorders, eating disorders and behavioural disorders (Austin et al., 2013;Day et al., 2017;Russel & Fish, 2016;Scannapieco et al., 2018). The most frequently invoked explanation for this stems in part from a disproportionate exposure to traumatic events (Blosnich & Andersen, 2015).
It is suggested that adversity exposure in childhood can impact healthy emotional growth (Spinazzola et al., 2005). For example, in LGBT+ adults, adversity has been linked with depressive symptoms, anxiety, isolation, rejection, and eating and behaviour disorders (Blosnich & Andersen, 2015).

Importance
We know thus far that LGBT+ individuals face a significant and poorly understood set of ACEs and bear a higher burden of mental health disorders than the general cohort. However, previous research has focused more on outcomes in adulthood rather than adolescence (Blosnich & Andersen, 2015).

Study objectives
The current review aims to be the first to present findings of studies measuring associations between adverse experiences and mental health outcomes across the full spectrum of LGBT+ youth. We aim to fill the gaps in existing research by (1) examining the prevalence rates of trauma in the lives of LGBT youth, (2) synthesising the mental health implications of ACEs, and (3) comparing the presence of mental health disorders in individuals exposed to ACEs compared to those not exposed.

This study was conducted in accordance with the Preferred
Reporting Items for Systematic Reviews and Meta-analyses Reporting Guideline (Page et al., 2021) and the Meta-analysis of Observational Studies in Epidemiology (MOOSE) reporting guidelines (Stroup et al., 2000; Tables S1 and S2). The study protocol was registered in PROSPERO (CRD42021240472).

Information sources
A multistep search of the literature was undertaken by two independent researchers (LJ and JV)

Eligibility criteria
Included studies were those that (i) studied young people with a mean age of ≤18 identifying as LGBT+; (ii) examined the relationship between childhood adversity and mental health outcomes in LBGT+ individuals; (ii) included exposure to trauma subtypes; (iii) reported outcomes compatible with a diagnosis from the Diagnostic and Statistical Manual of Mental Disorders (DSM) or International Classification of Diseases (ICD); (iv) provided relevant data on the relationship between any form of adversity and mental health; (v) were empirical, longitudinal or transversal, with original data, and (vi) had reports written in English. Consideration of the amount of available research led to the inclusion of all countries, settings, and developmental ages.
Studies with insufficient information were included in the review but excluded from the meta-analysis (Lipsey & Wilson, 2001). Additional inclusion criteria for the meta-analysis were those with (i) meta-analysable data and (ii) non-overlapping samples. Overlap between the included studies was actively searched by evaluating the country, recruitment period, and setting from which the study sample was obtained.

Key points
� LGBT+ youth report a higher prevalence of adverse experiences in comparison to their heterosexual or cisgender counterparts (p < .001), with sexual abuse the most commonly reported � LGBT+ youth were at a heightened risk of mental health disorders (p < .001), evidenced by elevated rates of depression and anxiety in LGBT+ youth (36.9% and 31.5%, respectively) � Continued advocacy is needed from communities and allies to support and empower LGBT+ youth in the face of adversity 2 of 12 -JONAS ET AL.

Data extraction and outcome measures
Independent data extraction was performed by two researchers (LJ and MS), and discrepancies were resolved through discussions with the senior author (JV). If available, the variables extracted included: study and sample-level descriptors (first author and year; study name; country; recruitment; design; sample size; age: mean [SD], % females; population category; sexuality measure); ACE information (exposure category, instrument to assess the traumatic experiences and the proportion in the ACE group); diagnoses/outcomes information (mental health disorder and diagnostic instrument) and all relationships of interest were reported along within the main and quantitative findings.

Quality assessment
Study quality was evaluated in all the included studies. The Newcastle-Ottawa scale (NOS; Wells et al., 2000) was performed to avoid incorporating any methodologically deficient or biased literature. Variations of the scale (Alameda et al., 2020;Modesti et al., 2016) were needed to assess different study designs (Table S4).
Decisions specific to the paper were made by the authors about representative cohorts, important factors to control for and adequate response rates.

Data synthesis and meta-analysis
As the studies in this meta-analysis were expected to be heterogeneous, the random-effects model was used (DerSimonian & Laird, 1986). Heterogeneity between studies was measured with the Q statistic and its magnitude was evaluated with the I-squared index (Lipsey & Wilson, 2001). A sensitivity analysis was performed stratified by group, that evaluated the design of the study to determine whether there were differences between cross-sectional and longitudinal studies. Publication bias was assessed by visually inspecting funnel plots (Sterne et al., 2001) and applying the regression intercept of Egger (Egger et al., 1997). Finally, meta-analytical regressions were conducted to evaluate the association between our outcomes and the factors that might be associated with the presence of any ACE or the presence of a mental health problem in LBGT+ youth. The fixed meta-regressor factors were age, gender (female) and ethnicity (white). All p-values reported in the meta-analysis were two-sided and the level of significance was set at a p-value of less than .05.

Screening and selection
The search strategy was initially broad with 9831 articles retrieved through the initial database search. Following the removal of duplicates (n = 2504), 7323 titles and abstracts were screened. 7078 studies were inconsistent to the inclusion criteria and these efforts produced a list of 243 unique citations. Of this set of articles, a further 214 were removed, leaving 27 studies (Table S5). Three independent reviewers randomly selected and reviewed the texts to ensure eligibility. After removing studies that did not provide metaanalysable data, 18 studies were included in the meta-analysis ( Figure 1; PRISMA 2020 flow diagram).

Study characteristics
This review includes 199,285 participants, of whom 26,505 iden-  Table S6).

Sociodemographic data
The majority of studies (n = 22) reported on the mean age of participants. All mean ages were 18 or below (mean age = 16.45), with a small proportion of participants exceeding 18 but not skewing the mean. The ethnic makeup of the samples varied greatly from study to study. White participants made up the largest range across studies (7.7%-96%). In 6 studies, racial/ethnic participants represented at least 50% of the sample, with a range between 4% and 72%. Female participants made up the largest percentage of most studies (n = 16), ranging between 11% and 74% (referenced in Table S7).

Quality assessment
The quality of the studies ranged from 5 to 9 using the Newcastle-Ottawa scale (NOS). The full results are detailed in Tables S8 and S9. INVESTIGATING THE IMPACT OF CHILDHOOD ADVERSITIES ON THE MENTAL HEALTH OF LGBT+ YOUTH -3 of 12

Exposures of interests
A range of adverse experiences were examined (Table 1) Three studies focused on cyberbullying and the power imbalance between perpetrator and victim on a virtual platform (Byrd, 2015;Garaigordobil & Larrain, 2020;Duarte et al., 2018), ranging from 11.76% (Zhao et al., 2020) to 47% (Byrd, 2015). Three studies examined unique LGBT+ variables. One study (Pollit et al., 2017) examined the impact of disclosure stress to friends and family, another measured the degree of comfort regarding a transgender person's identity and appearance (appearance congruence; Chozden et al., 2019). Lastly, one study (Veale et al., 2017) used a stigma index based on reports of discriminatory behaviours.

Outcomes measures
Various mental health outcomes were studied, and diverse measures were used to assess the same constructs across studies (Table 1). The F I G U R E 1 PRISMA 2020 flow diagram for new systematic reviews (Page et al., 2021)  6 of 12prevalence of LGBT+ youth meeting criteria for an identifiable mental health problem ranged from 8.3% (Donahue et al., 2017) to 82% (Battalen et al., 2020) and were most common in LGBT+ youth who had experienced ACEs.

T A B L E 1 Exposures of interests and outcomes measures
From the available data, the most commonly described psychiatric outcome was depressive symptoms (n = 17, 63%), which ranged between 12% (Peters et al., 2020) to 82% (Byrd, 2015)  victimisation and eating disorder symptoms, ADHD and autism spectrum disorder (Donahue et al., 2017).
Proportion of LGBT+ youth exposed to any adverse childhood Experience Overall, the meta-analytical results revealed that 55.2% (95% included was statistically significant (I2: 00.582-99.984, all p < .001) in all the primary analyses ( Figure 2, Figure 1).

Proportion of LGBT+ youth with any mental disorder
The meta-analysis on the prevalence of mental health disorders in LGBT+ youth showed that 23.2% had any mental disorder (95%  (Figure 3, Figure 2).
Furthermore, all our outcomes in terms of mental health conditions were more prevalent and statistically significant in LGBT+ individuals compared to heterosexual youth (all p < .001). The sensitivity analyses comparing the studies by their design (cross-sectional vs. longitudinal studies) for both sets of analyses (proportion of LGBT+ youth exposed to any ACEs and proportion of LGBT+ youth with any mental health problem) did not reveal any statistically significant differences. Lastly, meta-regression analyses, controlled for age, gender and ethnicity, showed that an older age (β = 0.227, p = 0.0003) was associated with a higher prevalence of any ACEs in LGBT+ youth.
There was no significant association between gender and ethnicity and the presence of any ACEs (all p > .05) nor between age, gender or ethnicity, and the presence of mental health conditions (all p > .05).

Further information and figures produced by the meta-analysis
are shown in the supporting information.

DISCUSSION
This is the first meta-analysis that comprehensively addresses the presence of ACEs in LGBT+ youth. A thorough search of relevant research yielded no existing reviews that investigated a range of adverse experiences across the full spectrum of LGBT+ youth and mental health outcomes. We systematically reviewed 27 studies, focusing on the distinctive features of adversity in childhood and mental health outcomes in 22,309 LGBT+ adolescents. We further addressed via the meta-analysis the presence of four forms of adverse childhood experiences and two mental health outcomes experienced by 21,781 LGBT+ youth from 18 articles. In most studies, a higher prevalence of female participants was reported compared to males. Possible explanations include an account of sexual prejudice that posits how negative attitudes towards sexual minority individuals are more severe for males than females (Bettinsoli et al., 2019) and that disclosure stress is more prevalent among men than women (Pollit et al., 2017).
After reviewing the literature, we cast a broad, atheoretical net for multiple types of adversity to overcome limitations of previous ACE reviews, recommending more thorough measures (Craig et al., 2020). The results of the systematic review build on knowledge regarding the different patterns of adverse experiences reported among LGBT+ youth. Our meta-analysis showed that more than 50% of LGBT+ youth suffered any ACE, which is in keeping with prior literature in which ACEs were reported by 50% of young people (Craig et al., 2020). The most commonly cited forms of victimisation were physical, sexual and verbal abuse, where sexual and gender minority youth have increased odds of exposure to sexual abuse, verbal abuse and physical abuse. Such findings are similar to findings in LGBT+ adults (Andersen & Blosnich, 2013).
The results from our meta-analysis showed that experiences of sexual abuse were the most prevalent ACE in LGBT youth (29.7% of the sample), with one study finding the risk of experiencing sexual abuse for LGBT youth was tenfold compared to heterosexual youth (Atteberry Ash et al., 2020). This prevalence is in keeping with a national LGBT+ survey reporting a prevalence rate of 33% (Bradford et al., 1997). 26.5% of LGBT+ youth in this sample also reported physical abuse. This result was empirically similar to previous literature reporting that between 22.2% and 33% of LGBT+ students were physically harassed due to their F I G U R E 3 Proportion of LGBT+ youth with any mental disorder. Forest plot 8 of 12sexual orientation or gender expression (Craig et al., 2020;Kann et al., 2018;Kosciw et al., 2016).
There was a noticeably smaller proportion of the sample reporting verbal harassment (28.7%), compared to previous metaanalysis reviews (55% in Katz-Wise & Hyde, 2012). A possible explanation could be the relative recency of most of studies included in the meta-analysis. In the past decade, we have been witnessing to a social climate becoming more supporting of LGBT+ young people and stricter anti-bullying policies, which may in turn reduce verbal harassment for example, around school institutions. In this metaanalysis, 19.1% of LGBT+ youth had undergone cyberbullying.
Recent national data from the Youth Risk Behaviour Survey (Kann et al., 2018) found a slightly higher prevalence (27%), but this was still higher than heterosexual counterparts.
The psychological sequelae from victimisation have been shown to increase the risk of mental health problems. We assessed outcomes defined by diagnostic symptoms or the criteria of any mental health disorder present in any edition of the DSM or ICD, to prevent overestimating the prevalence of mental health problems among this population.
A meta-analysis examining the worldwide-pooled prevalence of mental disorders affecting children and adolescents was 13.4% (Polanczyk et al., 2015). The meta-analysis was also subject to limitation. We were unable to conduct a meta-analysis using the subpopulations of LGBT+ as moderating factors since the data collected were not homogeneous.
The sample sizes were not concordant, and the populations were broadly categorised, grouping together the small samples of those who did not identify as heterosexual. As a result, we note that our ability to speak to differences within sexual-and gender minority populations is limited. Further to this, we were unable to distinguish between different gender identities (e.g., transgender man, transgender woman).
Most studies were based in the United States of America (US) or in Europe. Consequently, a generalisation of the data is limited to Western nations and can only be applicable to the rest of the world with caution. In such areas, an overwhelming majority hold the view that homosexuality or diverse gender identities is highly problematic and the consequences for expressing such an identity can be fatal.
Although there are various reasons why it is difficult to obtain data from participants in these areas, future research should find creative ways to have a better understanding of ACEs and mental health outcomes in participants living within cultures under severe constraints and conflict zones to combat state-sponsored repression and social stigma.

Implications
How should schools prepare for working with LGBT+ youth?
The strong evidence that school-based victimisation is causative of mental illness as well as suboptimal academic outcomes, less motivation and less engagement, highlights the need for establishing inclusive programming to support the mental health of LGBT+ youth.
It is well documented that the impact of childhood adversity on LGBT+ youth can be lessened through relationships with supportive peers. Theory supports that other peer-based support groups would be helpful, and these have been piloted (e.g., Ramsey et al., 2015).
Further, key adults must be vigilant in recognising victimisation efforts, strive to establish rapport with vulnerable students and offer a safe space where minority youth feel valued in their choices. For example, by always ensuring they are using a transgender youth's preferred pronouns.
How should clinicians prepare for working with LGBT+ youth?
Outside education institutions, continued advocacy is required to modify the longstanding history of poor patient care towards LGBT+ people (Riggs & Sion, 2017). Service professionals should aspire to foster inclusive and safe health care environments in which youth can INVESTIGATING THE IMPACT OF CHILDHOOD ADVERSITIES ON THE MENTAL HEALTH OF LGBT+ YOUTH -9 of 12 disclose sensitive histories is recommended to avoid poor engagement and poor treatment adherence.
Clinicians assessing young people may wish to pay particular attention to patients endorsing a minority identity status and consider the aetiological role of trauma in their practice and adopt a trauma-informed approach. This places a strong emphasis on enhancing support systems, mitigating trauma reminders and fostering safety. Alternatively, cognitive-behavioural interventions may serve to alter negative self-narratives attributed to minority stress in both sexual and gender minorities. Best practice may call for matching LGBT+ youth with clinicians where possible to provide positive role modelling or educate clinicians using evidencebased care.

Directions for future research
To enhance future study comparability and precision, a standardised Despite clear evidence that LGBT+ populations are among the most vulnerable groups for mental health problems, stigma has driven a long-standing shortage of funding. It is important that LGBT + specific work is continuously reviewed and updated in light of rapid developments in public discourse and academic knowledge.

CONCLUSIONS
Existing literature has produced compelling evidence of the pervasive nature of the trauma that LGBT+ young people endure, but in the absence of systematic reviews, we offer an up-to-date understanding of the various sources of adversity which can contribute to poor mental health outcomes among LGBT+ youth. Information demonstrating higher prevalence of mental health disorders among LGBT+ individuals has been published previously, but the present results suggest that they emerge in adolescence. In addition to confirming and extending findings related to ACEs in LGBT+ individuals, we also documented an association between ACEs and poor mental health in transgender respondents. Continued advocacy is needed from communities and Allies to support and empower LGBT+ youth in the face of adversity. Longitudinal and longer-term studies are required to further understand the relationship between adverse experiences in LGBT+ youth and the impact on mental health.

ACKNOWLEDGMENTS
The open access publication charge was covered by the MRC (grant ref MR/S026460/1). The authors have no further funding from any agency in the public, commercial, or not-for-profit sectors, to report.