Factors affecting the association between maternal adverse childhood experiences and preschool children’s behavioural problems: a systematic review of mediators and moderators

Background Adverse childhood experiences (ACEs) are pernicious events (e.g., physical abuse) that occur before 18 years of age within the household. Mothers’ ACEs are associated intergenerationally with their preschool children’s (two-to-five-year-old) behavioural problems, impacting lifelong mental health trajectories. Mediators (e.g., mental health) may explain how mothers’ ACEs exert their influence on children's behavioural development whereas moderators (e.g., child sex) may affect the strength and direction of the association, revealing potential points of intervention. Therefore, this study aimed to: (1) identify mediators and moderators of this association; and (2) describe and compare moderation and mediation outcomes. Methods This review (#CRD42022307214) extracted data from peer-reviewed literature formally analyzing at least one moderating or mediating variable of the association between maternal ACEs and preschool children’s behavioural problems using MEDLINE, Embase, APA PsycINFO, Cochrane Central Register of Controlled Trials, CINAHL, SCOPUS, and Web of Science databases (n = 7). Validity was assessed using the Joanna Briggs Institute checklists. Results Thirteen full-text studies with moderate-to-high validity were identified. Most studies examined mediating variables, with unanimous support for mediation through maternal depression. The mediating roles of adult attachment styles and anxiety were conflicting. No studies examined environmental factors (e.g., neurotoxins). Conclusions Mothers’ depression is a significant contributor to children’s behavioural problems in the context of ACEs and should be a target of early intervention to prevent lifelong challenges. Future research should focus on examining more moderating variables and consider the roles of environmental factors. More research is needed on the moderating and mediating roles of genes.


Background
Adverse childhood experiences (ACEs) were first identified by Felitti et al 1 in 1998, an overarching term that reflects stressful life events occurring within the household before 18 years of age including: physical, sexual, and/or emotional abuse or neglect; parental incarceration, loss, divorce, or separation; and parental substance use and mental health concerns.Since then, researchers have identified the pervasive and dose-response impacts of ACEs on developmental trajectories that persist into adulthood and result in poorer quality of life. 1 A report published by the Centers for Disease Control and Prevention determined that ACEs are associated with at least half of the top 10 leading causes of death in the US (e.g., strokes, cancer) and that prevalence of depression among adults could be reduced by 44% if ACEs were prevented. 2Mothers (who are often the primary caregiver due to historically-rooted gender expectations) with mental and physical health problems subsequent to their ACEs are also at risk of providing less optimal care to their children. 1 Consequently, these children are more likely to develop behavioural problems which are associated with adverse developmental trajectories such as mental, physical, and psychosocial problems (e.g., unemployment), reflecting intergenerational transmission of negative impacts via ACEs. 3nderstanding mediators and moderators of these troubling associations could help practitioners to intervene at appropriate timeframes to hinder the transmission of poor health outcomes intergenerationally.
Maternal Adverse Childhood Experiencess are Associated with their Children's Behavioural Problems Many studies provide evidence for the positive association between maternal ACEs and preschool children's internalizing and externalizing behavioural problems. 4Internalizing behaviours (e.g., anxiety, depression, somatic complaints) are psychological in nature, inwardly focused, and related to the emotional state of the child. 5Conversely, externalizing behaviours are manifested in actions directed towards the external environment, and observed in aggressive forms such as yelling, hitting, or hyperactivity. 6cDonald et al 7 reported that three or more maternal ACEs were associated with both internalizing and externalizing behaviours among three-year-old children.In a study on the same cohort, Hetherington et al 8 found that four or more maternal ACEs were associated with internalizing and externalizing behavioural problems at five years of child age.1][12] The substantial evidence for the positive association between maternal ACEs and children's behavioural problems, an increase in the prevalence of behavioural problems globally, and the compounding effects that have resulted from the COVID-19 pandemic calls for more clarity on how other factors (e.g., mediators, moderators) influence this association.Such information is needed to identify intervening strategies and to prevent and mitigate the development of behavioural problems among preschool children. 4,13

Potential Moderators and Mediators and their Importance
The impacts of maternal ACEs on children's behavioural problems may begin as early as the prenatal period.For example, prenatally, ACEs may impact children's socioemotional development, and subsequently, behavioural development, through epigenetic mechanisms involving mothers' hypothalamic-pituitary-adrenal (HPA) axis. 14The HPA axis is a biological system governing the flight-or-fight system through the release of stress hormones and adaptive behavioural functioning. 15Acute stimulation of the HPA axis serves to assist individuals in moments of stress or discomfort; however, chronic stimulation is linked to toxic stress and maldevelopment of several physiological systems including the cardiovascular, metabolic, and immune systems. 15Mothers who have experienced ACEs may be more likely to have an overstimulated HPA axis, 16 which may result in the transmission of cortisol to the developing fetus during pregnancy; 14 studies show that high levels of cortisol during pregnancy may affect fetal development. 14Postnatally, maternal neglect stemming from unresolved ACEs trauma may result in poor caregiving, and in turn, hyperactivate their children's own HPA axis; 17 preschool children's brains are particularly vulnerable to chronic activation of the HPA axis resulting from toxic stress. 18,19When preschool children experience stress for prolonged periods of time due to neglect and abuse, they will likely adopt maladaptive behaviours (i.e., internalizing and externalizing behaviours) as an alternative means of regulating stress or expressing their needs. 202][23][24] Many negative behavioural outcomes for children are associated with maternal ACEs and potential moderators and mediators (defined and depicted in Table 1 and Figure 1, respectively) such as those related to the caregiving environment, 25 genetics, or educational settings, 26 warrant a need for further investigation both pre-and postnatally.
A greater awareness of ACEs' conditional effects and pathways could highlight points of intervention to optimally promote healthy child growth and development (e.g., behaviour).Armed with this knowledge, healthcare professionals and policy/decision makers can be better prepared to intervene and adjust relevant programs and services.Therefore, systematically synthesizing existing knowledge and identifying gaps in the context of moderators and mediators is important to guide researchers in conducting future research studies, clinicians and healthcare providers on the front line, and policy makers focused on programs and services for at-risk children and families at local, regional, and national levels.Further, identifying null findings alongside those that are statistically significant are important to limit reporting bias when synthesizing literature and to ensure that time, energy, and money is funneled toward the most appropriate resources and services. 28

Purpose of the Study
The objectives of this systematic review were to: (1) identify all studies that formally examined at least one potential moderating and/or mediating variable on the association between maternal ACEs and preschool (two-to-five-year-old) children's behavioural problems; and (2) describe and compare the effects of the moderating and mediating variables.

Methods
The systematic review protocol was developed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Systematic Review Protocols (PRISMA-P) guidelines and the review is reported according to the PRISMA guidelines. 29A PRISMA checklist is provided (See Clinical Medicine Insights: Psychiatry n n Supplementary Materials A).This review was registered on the PROSPERO database (ID# CRD42022307214).

Search Strategy and Selection Criteria
The following databases (n = Included studies: (1) were papers published in peer-reviewed journals; (2) were written in English, French, Spanish, Serbian, and Gujarati (languages that represent the fluency of our research team); (3) contained formal analysis related to at least one potential mediating and/or moderating effect on the association between maternal ACEs and preschool children's behavioural problems; (4) were observational studies; and (5) examined mother-preschool child dyads where the mean age for children in the sample was between two-to-five-years of age.To limit reporting bias and overestimation of treatment effects (often more prevalent in meta-analyses), authors: did not use "moderator" and "mediator" as search concepts, broadening the search to capture more studies with potential moderators and mediators; included any studies reporting null findings (when provided) or any studies that used terminology synonymous to moderation and mediation analysis (e.g., interactions); and scanned through CENTRAL.Grey literature, qualitative studies, reviews, commentaries, dissertations, conference abstracts, and opinion papers were excluded.Though the inclusion of grey literature can provide valuable information in reviews, such sources were excluded as they require considerable resources (e.g., time, money) while lacking a peer-review process.Further, there is literature to suggest that the exclusion of grey literature does not usually affect the outcomes of reviews. 30The exposure in this study was maternal ACEs and the outcome was preschool children's behavioural problems (i.e., total, internalizing, externalizing).For this review, the term ACEs reflected any exposure that examined maternal history of childhood adversity to maintain consistency.

Study Selection and Data Extraction
Covidence, an online screening tool, 31 was used to upload all database results and facilitate data management, duplicate removal, and screening of records.To ensure agreement amongst reviewers was met and clarity regarding inclusion and exclusion criteria were well-understood, a calibration exercise with 50 randomly selected studies was completed by two reviewers (SK, LA) through Excel.Once percent agreement was met (>90%), these two reviewers independently screened the titles/abstracts and then full texts using Covidence.After screening through the initial 2695 studies

TERM DEFINITION
Moderator A variable which affects the strength or direction of the relationship between a focal predictor and the outcome of interest; the effect of the focal predictor on the outcome is conditional on the presence, absence, or level of the moderator 27 Mediator A variable quantifying or representing the conduit through, or mechanism by which, the focal predictor transmits its effect of an outcome.A mediator is caused by the focal predictor and causes the outcome and therefore is causal between the focal predictor and outcome.
The focal predictor causally affects the outcome indirectly through the mediator 27

Critical Appraisal
To examine quality and validity of the included studies, critical appraisals were completed using Joanna Briggs Institute (JBI) checklists. 32Although not explicitly assessing risk of bias, the JBI checklists includes questions that assess some degree of risk of bias. 32As only observational studies (analytical cross-sectional and cohort studies) were included, a JBI checklist for analytical crosssectional studies (number of questions = eight) and a checklist for cohort studies (number of questions = 11) were used.For both checklists, each item included the following options as responses: "yes"; "no"; "unclear"; and "not applicable".Critical appraisal results were not used to exclude studies, but rather to allow for description of their rigor and impact.Two individuals critically appraised the studies (SK, SR) independently with at least 90% agreement and then subsequently met to review their results.

Data Analysis and Synthesis
Studies were described by identifying certain characteristics (e.g., demographics, moderators/mediators examined, moderation/ mediation outcomes), providing sample characteristics (through percentages and frequencies), and identifying the presence of moderators and/or mediators to better synthesize results.A summary table was designed to compile and organize all the data (Table 3).

Study Characteristics
General study characteristics and more specific information of each study are provided (Tables 2 and 3).The sample sizes for the three analytical cross-sectional study designs ranged from 121 33 to 152 34 and the sample sizes for the prospective cohort studies ranged from 45 35 to 1994. 36Seven studies only examined mediating variables, [36][37][38][39][40][41][42] two only examined moderating variables, 33,43 and four examined both moderating and mediating variables. 34,35,44,4538,39,44,45 Moderating Variables The moderating effects of certain variables on maternal ACEs and children's behavioural problems acting alone included: 1) child sex, 44 2) maternal perceptions of family social support postnatally, 33 3) frontal EEG alpha asymmetry (greater activity in the right frontal brain region relative to the left) postnatally, 35 4) maternal HPA axis functioning during early or late gestation, 45 and 5) children's serotonin transporter polymorphism (5-HTTLPR) genotype (expression of a gene coding for serotonin) postnatally. 39Overall, maternal ACEs were more likely to: 1) result in internalizing behaviours when the child was female; 44 2) not be associated with children's externalizing behaviours when mothers' perceived moderate to high family social support; 33 3) be associated with total and externalizing behavioural problems when children had higher frontal EEG alpha asymmetry; 35 4) be associated with externalizing behaviours in earlier gestational periods and internalizing behaviours in later gestational periods when maternal diurnal slope was flatter; 45 and 5) be associated with behavioural problems when children had variant alleles for the 5-HTTLPR serotonin-producing genotype. 39

Mediating Variables
42,44 One study examined parenting behaviour (cognitive growth fostering and social-emotional growth fostering behaviours) postnatally, 42 another maternal HPA axis functioning during early and late gestation, 45 and a third examined frontal EEG alpha asymmetry postnatally.35 When attachment was examined as the mediator, maternal secure attachment style (characterized by the ability to develop reliable and secure relationships) 46 during pregnancy demonstrated a protective mediating effect.40 All studies examining maternal depressive symptoms or depression as a mediating variable reported significant effects; 36,38,39,41,44 however, there were minor differences in terms of behavioural outcomes.Except in one study, 36 maternal anxiety mediated the association between maternal ACEs and children's behavioural problems with some reported differences in effect through anxiety in the prenatal and postnatal periods. 42,44 Mother' capacity to foster cognitivegrowth, as opposed to social-emotional growth, 42 as well as maternal cortisol awakening response, 45 mediated the association between maternal ACEs and children's internalizing behaviours.

Complex Model Variables
Some research involved the examination of multiple mediators linked in serial or models that combined mediation and moderation (conditional process models) to examine the impacts of mothers' mind-mindedness, depression, anxiety, sensitivity, and adult attachment styles on the association between maternal ACEs and children's behavioural problems.One study reported that maternal depression mediated the association between maternal ACEs and children's internalizing and externalizing behaviours at low or medium maternal mindmindedness postnatally, which encompasses mothers' abilities to acknowledge that their child has a mind with their own thoughts, wants, and ambitions instead of simply someone requiring fulfillment of their needs. 34Similar mediation outcomes were reported for anxiety but only at low levels of maternal mind-mindedness. 34Maternal ACEs also acted through a sequential mediation mechanism via maternal depression followed by maternal sensitivity. 38,39Boys, as opposed to girls, were more likely to develop both internalizing and externalizing behaviours when exposed to maternal depression and anxiety prenatally and postnatally. 44Further, the influence of maternal Moderator(s) included • Observed a mediated effect of maternal ACEs on both internalizing and externalizing behavioural problems in both the prenatal and postnatal periods when combining the effects of maternal anxiety and maternal depression.Additional moderation through child sex revealed that: the combined effect was significant for boys' externalizing behaviours prenatally; the mediated effect through anxiety was significant for boys' internalizing behaviours prenatally; the combined effect was significant for both boys' and girls' internalizing behaviours prenatally; the mediated effect through depression and combined effect was significant for boys' externalizing behaviours postnatally; and the combined effect was significant for boys' internalizing behaviours postnatally -Child sex assigned at birth cortisol diurnal slope in the earlier gestational period on the association between maternal ACEs and children's behavioural problems was found to differ between boys (externalizing behaviours) and girls (internalizing behaviours). 45Finally, both maternal avoidant (referred to as dismissive in adults and characterized by avoidant tendencies in intimate relationships) 46 and anxious (referred to as preoccupied in adults and characterized by heavy investments in and the need for copious amounts of reassurance from their intimate relationships) 46 attachment styles mediated the association between maternal ACEs and children's behavioural problems, operating sequentially through postnatal maternal depression. 36

Critical Appraisal Results
The quality of the included studies is provided (Table 4).Using the JBI checklists for prospective cohort studies and analytical cross-sectional studies, most studies appeared to be of higher quality.For prospective cohort studies, measuring outcomes in a valid and reliable way and utilizing strategies to address incomplete follow-up were most often scored "no".For analytical cross-sectional studies, measuring outcomes in valid/reliable ways always scored "no".

Discussion
The purpose of this systematic review was to summarize and synthesize existing peer-reviewed observational studies that have examined moderators, mediators, and more complex models of the association between maternal ACEs and preschool children's behavioural problems.Included studies typically examined mediating, rather than moderating variables.Maternal mental health was predominantly examined in the included studies, with consistent support of mediation through maternal depression/ depressive symptoms.Maternal attachment styles, maternal cortisol awakening response, and the ability for mothers to foster cognitive growth among their children were also key mediators.Sex-specific effects were noted, with boys' behaviour more likely to be associated with maternal ACEs.However, the association between maternal ACEs and children's behavioural problems was also conditional on maternal social support, diurnal cortisol slope, children's serotonin genotype, and EEG frontal alpha functioning.Certain variables (e.g., maternal mind-mindedness, maternal sensitivity, maternal attachment styles) demonstrated significant effects when examined via more complex models.A similar review was recently conducted that examined the research question in the context of all children below the age of 18 years, 47 capturing less than half of the studies that this review included and also warranting caution when generalizing findings as development in preschool differs substantially from other developmental periods (i.e., teenage).Both reviews do however identify a pertinent role of maternal depressive symptoms/depression as a mediator of the association between maternal ACEs and children's behavioural problems.

Attachment Styles
Mothers' adult attachment style was found to be a prevalent mediator of the association between maternal ACEs and children's behavioural problems, while children's attachment style was not.However, the results were inconsistent among included studies.Cooke et al 36 observed effects postnatally through maternal anxious attachment style, but not avoidant attachment, for internalizing and externalizing behaviours.However, a more complex model that included depressive symptoms (but not anxiety) revealed that a sequential indirect mediation pathway through maternal anxious and avoidant attachment styles was associated with both externalizing and internalizing behaviours. 36In contrast, maternal secure attachment in the prenatal period was the only significant mediator that Roth et al 40 observed.This suggests an interplay of environmental and temporal factors.
Securely attached mothers are more likely to respond to their children's needs reliably, securely, and appropriately, resulting in better behavioural development trajectories. 48,49However, anxiously attached pregnant mothers (compared to secure) may intergenerationally transmit stress hormones (e.g., cortisol) through placental transfer, 50,51 potentially resulting in behavioural problems during childhood. 52,53In the postnatal period, maternal secure attachment may therefore be more significantly related to appropriate brain development in growing children whereas anxious attachment is more influential in the prenatal period.Lastly, maternal avoidant attachment styles alone may not be predictive of children's behavioural problems and instead operate through mothers' depression, potentially linked to their lack of social connections.While there is evidence that maternal attachment styles are associated with children's behaviours, findings from this review suggest the need to examine the effects of maternal attachment styles temporally and via more complex (sequential) mediation pathways.

Maternal Mental Health
All studies included in this review that examined maternal depressive symptoms or depression as a mediator of the association between maternal ACEs and children's internalizing and externalizing behaviours prenatally and postnatally observed significant mediating effects.Substantial literature provides evidence for a positive association between early childhood adversity and maternal mental health including depression later in life (see meta-analysis by Racine et al 54 ).Consequently, depressive symptoms during motherhood are associated with reduced maternal sensitivity to their children's needs and cues, resulting in mothers' withdrawn behaviours and hostile responses. 55,56As this negative or lack of response continues, children's stress response systems will remain stimulated, potentially manifesting in the display of behavioural problems to self-regulate. 56,57In both cases, maternal neglect and maladaptive parenting skills arising from depressive symptoms,

12
Clinical Medicine Insights: Psychiatry n n whether intentional or unintentional, suggests the potential for intergenerational transmission of ACEs. 1 Therefore, to prevent the development of behavioural problems among children, interventions should target mothers' depression to encourage healthier parenting.As opposed to Cooke et al, 36 Letourneau et al 44 and Shih et al 42 observed significant effects through maternal anxiety but only on internalizing behaviours, suggesting that maternal ACEs act through maternal anxiety to predict inwardly focused, psychological impacts, rather than outwardly focused behaviours.Letourneau et al 44 examined the role of maternal anxiety in both the prenatal and postnatal period, with significant mediating effects only in the prenatal period, whereas Shih et al 42 only examined postpartum anxiety symptoms, with significant results.These discrepancies could be attributable to temporal impacts of confounding factors and methodological differences as studies examining anxiety in the prenatal vs postnatal period often demonstrate conflicting results; 58 however, the co-morbid nature of maternal depressive symptoms and anxiety suggest that anxiety likely plays an important role in preschool children's behavioural development, requiring clarity on how it exerts its influence.
In their meta-analysis, Racine et al 54 observed a significant positive association between maternal ACEs and maternal anxiety, suggesting that like the impacts of ACEs on maternal depressive symptoms, a heightened activation of the HPA axis during childhood may lead to anxiety-related symptoms in adolescence and adulthood without proper intervention. 54onsequently, anxious mothers may not respond optimally to their children's needs, resulting in the internalization of children's emotions to self-regulate. 59Temporally dependent effects of anxiety on children's behaviour may also be present such that anxiety during pregnancy leads to internalizing behaviours in children, again attributable to stress hormone transfer through the placenta but also due to fetal programming, a biological phenomenon suggesting that children may develop internalizing behaviours in preparation for the environment that they are anticipated to grow up in. 50,51,53As the negative effects of maternal anxiety on children's behavioural outcomes may occur prenatally, health professionals should provide resources and interventions to buffer anxiety-inducing stress associated with pregnancy and parenthood.
Dollberg and Hanetz-Gamliel 34 observed interactions with maternal depressive symptoms and anxiety when examining mothers' mind-mindedness as a moderating variable.In their study, the mediating effect of maternal ACEs on children's behaviour through maternal depressive symptoms or anxiety was conditional on mothers' mind-mindedness at low or medium levels. 34This implies that mothers who acknowledge that their infant has desires, feelings, uncertainties, and needs (or in other words, a mind), rather than just needs, are more likely to respond successfully to their infants' cues and engage in more dyadic synchrony. 34,60Similarly, maternal sensitivity refers to the quality of response mothers provide to their children's needs, with increased sensitivity characterized by the ability to respond quickly and adequately. 61In both studies that examined maternal sensitivity, 38,39 effects were only observed sequentially through maternal depression.As discussed, maternal depression was a significant mediating variable on the association between maternal ACEs and children's behavioural problems, but these findings support that depression likely acts by impeding the ability of mothers to respond sensitively to their children.With this knowledge, education, support, and resources are recommended for depressed mothers that promote awareness of their children's needs by highlighting important developmental cues as demonstrated in a review by Letourneau et al 62 .

Maternal Perception of Social Support
Social support is known to be a powerful predictor of mental and physical well-being, buffering the impacts of early childhood adversity. 63In the postpartum period, maternal social support is also known to serve as a protective factor for maternal mental health. 64,65In their study, Hatch, Swerbenski, and Gray 33 observed that the association between maternal ACEs and children's behavioural problems postnatally was conditional on the level of maternal social support, where moderate to high levels buffered the negative impacts of maternal ACEs on children's behaviour.Therefore, not only may social support act as a protective factor for maternal mental health, it can also prevent the intergenerational transmission of poor behavioural outcomes to children. 66

Maternal Parenting
Adults who experienced early adversity from their own parents are more likely to adopt similar parenting behaviours, resulting in the potential intergenerational transmission of ACEs. 67sing the NCAST Parent-Child Interaction Teaching Scale (an assessment that examines parents' sensitivity to their children's cues and the responsiveness of both parents and children to each other), 68 Shih et al 42 observed significant mediating effects through the cognitive-growth fostering subscale.This subscale focuses on parents' use of rich, unambiguous language in their interactions with children, and their support of children's learning, exploratory behaviour, and task performance. 68indings suggest that maternal ACEs may attenuate mothers' ability to foster children's cognitive development, and poor cognitive development is linked to the display of behavioural problems. 69For this reason, when assisting mothers who have ACEs, interventions should focus on engaging caregivers in fostering an environment that promotes interactions which enhance cognition in children. 70e Role of Mothers' HPA Axis Thomas-Argyriou et al 45  Were the two groups similar and recruited from the same population?
Were the exposures measured similarly to assign people to both exposed and unexposed groups?
Was the exposure measured in a valid and reliable way?
Were confounding factors identified?
Were strategies to deal with confounding factors stated?
Were the groups/participants free of the outcome at the start of the study (or at the moment of exposure)?
Were the outcomes measured in a valid and reliable way?
Was the follow up time reported and sufficient to be long enough for outcomes to occur?
Was follow up complete, and if not, were the reasons to loss to follow up described and explored?
Were strategies to address incomplete follow up utilized?
Was appropriate statistical analysis used?Were the criteria for inclusion in the sample clearly defined?
Were the study subjects and the setting described in detail?
Was the exposure measured in a valid and reliable way?
Were objective, standard criteria used for measurement of the condition?
Were confounding factors identified?
Were strategies to deal with confounding factors stated?
Were the outcomes measured in a valid and reliable way?

✖ ✖ ✖
Was appropriate statistical analysis used? pregnancy, with a focus on cortisol awakening response, diurnal slope, and total cortisol output.Of these, cortisol awakening response during pregnancy appeared to be the sole mediating variable of maternal ACEs and children's internalizing problems, suggesting that prenatally, maternal cortisol output in the mornings was more predictive of children's psychological distress at four years of age relative to total daily cortisol secretion or cortisol change during the day. 45McGuinn et al 71 report conflicting results, such that only total cortisol output, and not cortisol awakening response, predicted children's anxiety postnatally.However, that study did not examine the experiences of mothers' early childhood, which is known to influence HPA axis functioning in adulthood including during pregnancy. 72These conflicting findings warrant the need for further examination of maternal HPA axis functioning with consideration of maternal ACEs and temporality to better understand their associations with child behaviour.When examined as moderators, maternal diurnal cortisol slope in the earlier gestational period (six to 26 weeks) appeared to interact with maternal ACEs such that maternal ACEs were negatively associated with children's externalizing behaviours when the diurnal slope was steeper and positively associated when the slope was flatter. 45Further, adding child sex revealed a three-way interaction only in the earlier gestational period such that maternal ACEs interacted with diurnal slope to predict internalizing behaviours among girls and externalizing behaviours among boys. 45he interaction changed in the latter gestational period (27 to 37 weeks), such that diurnal slope was associated with only internalizing behaviours among all children; no sex effects were observed. 45Several studies support that maternal cortisol during pregnancy is more likely to result in internalizing behaviours among girls relative to boys, with some attributing this to the amygdala, a portion of the brain responsible for processing threats and fear-based responses. 73,74Thomas-Argyriou et al.'s 45 study also highlights temporal effects of maternal cortisol such that sexdependent behavioural outcomes were found to occur in the earlier gestational period, when brains and nervous systems of fetuses are beginning to develop. 75Therefore, it is imperative that healthcare professionals provide women considering pregnancy or who are just transitioning into pregnancy with resources and support to reduce chronic stress and persistently high cortisol levels.

Genetic and Neurological Factors
van de Ven et al 35 examined if frontal EEG alpha asymmetry in children moderated and mediated the association between maternal ACEs and children's behavioural problems.Only moderating effects were observed by increased asymmetry (more activity on the right frontal brain and less on the left), suggesting that decreased activation of the left frontal brain is associated with decreased capacity to buffer stress.This is supported in a study conducted by Lopez-Duran et al, 76 which revealed that increased activation of the left frontal brain (greater symmetry) among children at high risk for depressive symptoms residing in families with histories of depression moderated the association between stressful life events and internalizing behaviours.Therefore, for children with frontal EEG alpha asymmetry, the integration of certain tools, such as neurofeedback, 77 can help improve symmetry of the frontal brain region.
With a genetic focus, Bouvette-Turcot et al 43 observed that the association between maternal ACEs and children's behavioural problems was moderated by the 5-HTTLPR genotype such that children with an S or L G variant (rather than homozygous L A /L A functional genotypes) were more likely to develop behavioural problems as maternal ACEs increased due to lower amounts of serotonin. 43Since serotonin is linked to better mood outcomes and psychological functioning, inadequate production of this neurotransmitter may be associated with worsened health outcomes, including poor behaviour. 78More specifically, as this neurotransmitter plays a critical role in brain development, children with lower levels of serotonin may find it difficult to behave appropriately in stressful environments such as school settings. 79Implementing activities (e.g., meditation) into children's lives or pharmacotherapies that boost serotonin levels may prevent the development of behavioural problems among children lacking normal serotonin genotypes. 78mitations and Future Research Some moderating and mediating variables were only identified in a single study, indicating the need for more research prior to suggesting causality.Almost all studies (12/13) were performed in North America and studies differed in terms of ethnicities and vulnerabilities, likely reducing generalizability to other countries or populations.While studies used validated tools to measure maternal ACEs, the tools that were used differed, potentially causing discrepancies in how mothers reported their own ACEs.No studies examined the role of environmental conditions (e.g., nutrition) on the association between maternal ACEs and children's behavioural problems despite evidence for their effects on children's development, suggesting a need for future research.Due to the important role of other caregivers (e.g., fathers, grandparents), studies should consider examining their individual and cumulative effects on preschool children's behavioural problems.
Reporting bias is a prevalent issue in systematic reviews and meta-analyses which is characterized by researchers publishing only statistically significant, or desired, results. 28,80Also termed the file drawer problem, it implies that researchers will be more likely to submit or successfully publish positive or statistically significant results. 28,80In studies of moderators and mediators, this bias may be even more inflated as researchers may prefer to publish their findings only when significant interactions (i.e., moderation) and direct or indirect effects (i.e., mediation) are observed.It is likely that some of the studies captured in this review's search initially sought to examine moderating and mediating effects, with no significant results.However, without explicitly stating this, it is impossible for authors to pinpoint studies that did complete such analyses.Therefore, authors of this review acknowledge the potentiality for reporting bias affecting findings of the review.Nonetheless, the authors of the review utilized a thorough search strategy, comprehensively identifying factors that are relevant to mother-child health and that can prevent intergenerational transmission of poor health outcomes.These factors are also supported by other literature, adding veracity to the findings.

Conclusion
As preschool children are rapidly developing, they are highly susceptible to the negative impacts that may be associated with maternal ACEs.This systematic review aimed to identify all studies that have formally examined at least one moderating and/ or mediating variable on the association between maternal ACEs and preschool children's behavioural problems.Most studies predominantly examined the role of maternal mental health and uniformly observed mediating effects through maternal depressive symptoms/depression.A lack of examination of moderation and of environmental factors such as nutrition, 81 neurotoxicant exposure, 82 and educational settings, 26 highlights the need for more research.Overall, the results of this review reinforce that maternal ACEs are associated with children's behavioural problems and suggest certain factors that could be targeted as points of intervention to improve preschool children's behavioural outcomes and quality of life and prevent the transmission of lifelong poor health outcomes intergenerationally.Clinical Medicine Insights: Psychiatry n n

Figure 1 .
Figure 1.Illustration of moderation (Panel A) and mediation (Panel B).Maternal ACEs represent the focal predictor and children's behavioural problems represents the outcome of interest.

Table 1 .
Definitions of moderators and mediators.

Table 2 .
Study characteristics of all included studies (n = 13).

Table 3 .
Study information for all included studies (n = 13), categorized under analytical cross-sectional and prospective longitudinal cohort study designs.

Table 4 .
examined both the mediating and moderating effects of maternal HPA axis functioning during Critical appraisal of included studies (n = 13) using the Joanna Briggs Institute (JBI) prospective longitudinal cohort and analytical crosssectional study design checklists.