Parenting after a history of childhood maltreatment: A scoping review and map of evidence in the perinatal period

Background and aims Child maltreatment is a global health priority affecting up to half of all children worldwide, with profound and ongoing impacts on physical, social and emotional wellbeing. The perinatal period (pregnancy to two years postpartum) is critical for parents with a history of childhood maltreatment. Parents may experience ‘triggering’ of trauma responses during perinatal care or caring for their distressed infant. The long-lasting relational effects may impede the capacity of parents to nurture their children and lead to intergenerational cycles of trauma. Conversely, the perinatal period offers a unique life-course opportunity for parental healing and prevention of child maltreatment. This scoping review aims to map perinatal evidence regarding theories, intergenerational pathways, parents’ views, interventions and measurement tools involving parents with a history of maltreatment in their own childhoods. Methods and results We searched Medline, Psychinfo, Cinahl and Embase to 30/11/2016. We screened 6701 articles and included 55 studies (74 articles) involving more than 20,000 parents. Most studies were conducted in the United States (42/55) and involved mothers only (43/55). Theoretical constructs include: attachment, social learning, relational-developmental systems, family-systems and anger theories; ‘hidden trauma’, resilience, post-traumatic growth; and ‘Child Sexual Assault Healing’ and socioecological models. Observational studies illustrate sociodemographic and mental health protective and risk factors that mediate/moderate intergenerational pathways to parental and child wellbeing. Qualitative studies provide rich descriptions of parental experiences and views about healing strategies and support. We found no specific perinatal interventions for parents with childhood maltreatment histories. However, several parenting interventions included elements which address parental history, and these reported positive effects on parent wellbeing. We found twenty-two assessment tools for identifying parental childhood maltreatment history or impact. Conclusions Perinatal evidence is available to inform development of strategies to support parents with a history of child maltreatment. However, there is a paucity of applied evidence and evidence involving fathers and Indigenous parents.


Introduction
Child maltreatment is a global health priority affecting 25 to 50% of children worldwide [1] and can have profound and ongoing impacts on physical and social and emotional wellbeing and development [2,3]. Conflicting infant attachment and defence (fear or 'flight, fright and freeze') systems can be activated in response to child maltreatment which can lead to internal confusion and behavioural responses that are an attempt to manage distress and promote selfregulation, but may also result in increased confusion and harm [4]. These responses can be maintained into adulthood as part of a cluster of symptoms associated with recently proposed criteria for diagnosis of complex post-traumatic stress disorder (complex PTSD) [5]. Complex PTSD, caused by cumulative exposure to traumatic experiences that often involve interpersonal violation within a child's care giving system (sometimes referred to as 'developmental' or 'relational' trauma), can occur in families, within the context of social institutions [6], and be exaberbated by cumulative traumatic experiences as an adult.
Long-term associations with childhood maltreatment include smoking, eating disorders, adolescent [7] and unplanned pregnancies [8], adverse birth outcomes [9], and a range of physical and psychological morbidities [10]. Critically, these long-lasting relational effects can impede the capacity of parents to nurture and care for children, leading to 'intergenerational cycles' of trauma [4]. Parental fear responses can be triggered by their child's distress, and are often re-experienced as conflicting sensations and emotions, rather than as a thought-out narrative [11]. This in turn can give rise to hostile or helpless responses to the growing child's needs [12]. In addition, the intimate nature of some perinatal experiences associated with pregnancy, birth and breastfeeding pose a high risk of triggering childhood trauma responses.
Conversely, the transition to parenthood during the perinatal period (pregnancy to two years postpartum) offers a unique life-course opportunity for improving health [13], for healing [3] and for relational growth [14]. Growing research into the 'neurobiology of attachment' demonstrates that healing can occur despite severe experiences of maltreatment, by restoring a sense of safety and well-being through nurturing, supportive relationships with others-a transition sometimes referred to as 'earned security'. A positive strengths-based focus during the often-optimistic perinatal period has the potential to disrupt the 'vicious cycle' of intergenerational trauma into a 'virtuous cycle' that contains positively reinforcing elements that promote healing [14].
Despite the critical importance of the perinatal period for parents who have a history of maltreatment in their own childhoods; and frequent scheduled contacts with service providers during pregnancy, birth and early parenthood; there is limited specific guidance for care in the perinatal period for parents who have experienced maltreatment in their own childhoods [15]. In Australia, National Trauma Guidelines emphasize the need for trauma-informed care and trauma-specific support for the general population [6], and Perinatal Mental Health Guidelines recommend that perinatal care providers conduct a psychosocial assessment for mothers, including childhood maltreatment history [16]. However, there is limited guidance on trauma-specific care and support interventions in the perinatal period for parents with a history of childhood maltreatment, which is required for any population-based screening program [17].
Child maltreatment is not randomly distributed. The World Health Organization (WHO) use a socioecological framework [18] to explain why some people are at higher risk of experiencing interpersonal violence [19]. Parents with a history of childhood maltreatment are also more likely to have multiple socio-economic challenges, including unintended pregnancies [8], antenatal and postnatal depression [20], contact with the justice system and low employment [21]. Within Australia, there have been harrowing reports documenting high rates of child maltreatment and violence among Aboriginal and Torres Strait Islander (Aboriginal) communities in Australia [22]. The reasons for this are complex and lie beyond the scope of this review but are influenced by a legacy of past governmental policies of forced removal of Aboriginal children from their Aboriginal families and communities, intergenerational effects of these previous separations from family and culture, ongoing oppressive policies, social exclusion, marginalisation and poverty [23,24]. Thus, Australian National Trauma Guidelines emphasize the need for understandings of complex trauma to be contextualised within socioecological environments [6].
The primary aim of this scoping review is to map perinatal evidence involving parents with a history of childhood maltreatment. Our specific purpose is to identify relevant evidence to support the co-design of strategies for perinatal trauma-informed care (awareness), recognition, assessment and support for Aboriginal parents with a history of maltreatment in their own childhoods. We briefly illustrate how this scoping review is being incorporated into the co-design process in the discussion, with a project led by our team [25]. While this review is designed to address a direct practical purpose, we expect that it will have broader applicability to those working with parents who have experienced maltreatment in their own childhoods and help to prioritise future research in this critical area. A secondary aim is to use this scoping review to refine the search strategy and develop detailed protocols for further in-depth systematic reviews (see S1 Appendix for overview of planned reviews).
The specific questions for this 'phase 1' scoping review are: 1. What theories are used during the perinatal period to understand and frame the impact of a parental history of childhood maltreatment?
2. What risk and protective factors are identified in epidemiological evidence of life-course and intergenerational pathways (mediators/moderators) between a parental history of childhood maltreatment and behavioural/health outcomes for parents and their infants?
3. What are the perinatal experiences of parents with a history of childhood maltreatment? And what perinatal strategies do these parents report using to heal and prevent intergenerational transmission of trauma to their child? 4. What perinatal interventions are described to support parents with a history of childhood maltreatment to improve parental and child wellbeing? 5. What tools have been reported in the perinatal period to identify parents with a history of childhood maltreatment and/or assess symptoms of complex trauma?

Methods
PRISMA guidelines for systematic reviews informed the reporting of this scoping review (see S2 Appendix for checklist) and the protocol on which it is based (S3 Appendix). We have also referred to extension statements for scoping reviews [26] and equity-focussed reviews [27].

Criteria for inclusion
Participants: Prospective (pre-pregnancy), pregnant and new parents (mothers and/or fathers) or families caring for children up to two years after birth. Where mean ages only were reported (and it was unclear if children were two years of age or younger), studies reporting a mean age of less than five years only (i.e. preschool age) were included. Studies which reported a proportion of participants as parents of children aged two years or less were also included, to err towards inclusivity. Despite the primary aim of the research being to inform co-design of strategies to support Aboriginal parents, we have not restricted the inclusion criteria to Aboriginal or Indigenous parents for several reasons. Firstly, we know there will be very limited published Indigenous-specific evidence available. Second, we will incorporate evidence from other population groups in a comprehensive co-design process which draws on the knowledge of Aboriginal parents and key-stakeholders, and enables Aboriginal Australian parents to consider the relevance of this evidence for them. Interventions/exposures: Any parental report of childhood maltreatment. There is no broadly accepted consensus on a definition for complex PTSD, and this lack of consensus will be reflected in previously published included studies in this review. Therefore, for the purposes of this review, we are using the WHO definition of a key antecedent, child maltreatment: "abuse and neglect that occurs to children under 18 years of age. It includes all types of physical and/or emotional ill-treatment, sexual abuse, neglect, negligence and commercial or other exploitation, which results in actual or potential harm to the child's health, survival, development or dignity in the context of a relationship of responsibility, trust or power. Exposure to intimate partner violence is also sometimes included as a form of child maltreatment." [1] We also include 'proxy measures' such as child protection substantiations or removal from their family of origin.
Study type/comparisons: Any study design, including randomised controlled studies (RCTs), cluster RCTs, cohort studies (including measurement/assessment studies), economic evaluations or qualitative studies (see S1 Appendix for an overview of the study types considered relevant to each question). We excluded reviews, guidelines, discussion and opinion papers, government reports and non-peer reviewed reports of primary studies. Review articles were screened for additional primary studies.
Outcomes: Theories; risk or protective factors which mediate or moderate parental or child outcomes; experiences, perspectives and strategies parents use for healing or preventing intergenerational transmission of trauma; acceptability, effectiveness and cost of current perinatal interventions; and perinatal tools used for assessing symptoms of complex trauma or exposure to childhood maltreatment. Studies which only reported associations between parental childhood maltreatment history and outcomes, without any investigation of mediating/moderating factors, were not included in this review as these associations have been well-established in other reviews [28].

Search methods
We searched the following databases: Psychinfo, Medline, Cinahl and Embase up to 30/11/ 2016. The search terms included both thesaurus (MeSH) and keyword synonyms for 'child abuse' AND 'intergenerational' AND 'prevention' AND 'parent', using a search strategy developed and piloted in Psychinfo (see S4 Appendix) and subsequently modified for use in the remaining databases. We checked reference lists from relevant reviews identified from the search for additional potentially relevant primary studies.
References were downloaded into bibliographic reference management software (Endnote) and de-duplicated. One reviewer (CC) independently screened titles and abstracts to identify potentially relevant studies using an over-inclusive approach. The full texts of all potentially relevant studies were independently assessed by two reviewers (CC/GG) for inclusion, based on the pre-specified criteria for inclusion. Discrepancies were resolved by discussion.

Data extraction
A data extraction tool was developed and piloted in Microsoft Excel by two reviewers (CC/SB). Two reviewers (CC & GG/SG/YV/SC/SB) independently extracted data on the items that follow.
Population: study setting; selection criteria; recruitment process and sample size; parenting stage (pre-pregnancy, pregnancy to six weeks postpartum, six weeks to one year postpartum, one to two years postpartum); type of childhood trauma reported; parent characteristics that relate to progress-plus 'equity' criteria relevant to this review population ('at risk' status, age, place of residence, race/ethnicity, gender, language, religion, socio-economic status, social capital (e.g. marital status), and other (e.g. mental illness)) [29].
Intervention/study detail coding was based on the TiDIER framework [30]: aims; brief description; mode of delivery; who conducted study/intervention; duration and frequency of contacts; theoretical basis; analysis framework; individual tailoring; modifications/fidelity; collaboration/engagement.
Outcomes: how assessed (e.g. mail, face to face); unit of analysis; results summary; conclusion summary; detailed results under each of the main outcome categories (theories; mediating/moderating factors; parent experiences/perceptions; interventions; assessment tools and 'other').

Assessment of risk of bias within studies
For each included study, the risk of bias was assessed independently by two reviewers using domains from one of the following tools as appropriate for the study design (S5 Appendix). Overall confidence in study findings was assessed using an adaption of the GRADE approach (S6 Appendix). This provides a transparent and systematic method for incorporating information about study limitations (risk of bias), the validity of outcome measures (indirectness/relevance), and the adequacy of the sample (imprecision) in a summary of findings for each study. The overall assessment was conducted by one reviewer (CC), with the first ten assessments checked and discussed with a second reviewer (SB) before completing the remaining assessments. All studies started with an assessment of 'high' confidence, and were downgraded one level for serious concerns or two levels for very serious concerns about: study/ methodological limitations; indirectness/relevance; and imprecision/adequacy to attain a final assessment of high, moderate, low or very low confidence. Overall confidence across study findings (e.g. consistency and publication bias) was not assessed in this scoping review.

Data synthesis
Outcome data were synthesized narratively under each of the main outcome categories (theories, risk and protective factors, parents' views, interventions, assessment tools). Outcomes are summarised only briefly in this scoping review to map and illustrate the breadth of evidence, rather than offer a comprehensive assessment of strengths and limitations of the evidence. Evidence map templates were generated using primary and secondary outcomes specified in the scoping review protocol. Each study providing relevant information was illustrated numerically in the map/table, using a number which corresponds to an allocated study ID number in the Characteristics of Included Studies (COIS) Table (Table 1). The COIS table (Table 1) provides summary study-level information on the country, setting, participant number, parenting stage, parental childhood maltreatment history, study type, study aim, main results and overall confidence in study findings. In subsequent systematic reviews, we will use synthesis methods appropriate for specific study designs and explore the respective evidence and strengths and limitations of the evidence in more depth (S1 Appendix).

Description of included studies
Settings: Over 20 000 parents/parent-child dyads participated in the 57 included studies. The majority of studies were conducted in the United States (n = 44), with other studies in Australia (n = 3), Canada (n = 3), Germany (n = 1), Netherlands (n = 2) and the United Kingdom (n = 4). Most studies have been published in the most recent decade (2010+) (n = 27), with fewer studies published in earlier decades; 2000-2009 (n = 17), 1990-1999 (n = 11) and before 1990 (n = 2). Maternal age moderated the relation between a maternal history of neglect and infant neglect, and social support moderated the relation between childhood neglect and maternal empathy. Neglected mothers had considerably higher levels of parenting empathy when they had frequent access to social support than when they had less frequent support, whereas the protective effect of social support was not nearly as strong for non-maltreated mothers.

Moderate
(Continued) Cross-sectional study (mixed quantitative/ qualitative) Investigate intergenerational pathways: to identify types of social support utilized by resilient individuals from abusive backgrounds; and to explore the relationship between perceived social support and the ability to interrupt the cycle of violence. No one type of social support was identified as being more prevalent than another. There was a positive relationship between perception of one's own abuse after the age of 13 and type of support labelled as belonging social support. Similarly, there was a positive correlation between experienced abuse after the age of 13 and belonging and self-esteem support. There was a negative correlation between perception and experience of one's own abuse prior to age 13 and the presence of a caring adult in childhood and a negative correlation between the perception and experience of one's own abuse after age 13 and the presence of a caring adult and friend in childhood. Two thirds of the participants were affirmative in their belief that social support was helpful in breaking the cycle of violence. Narrative themes of social support, internal aspects of self and spirituality emerged as factors which played a role in breaking the cycle of violence. Awareness of alternatives and non-judgmental support were identified as aspects of social support which were deemed helpful.  (Chung 2008). (1) Mothers exposed to childhood physical abuse and verbal hostility were more likely to report IS use than those not exposed. In the adjusted analyses, maternal exposure to physical abuse, other ACEs, and valuing CP were independently associated with IS use. Attitudes that value CP did not mediate these associations (Chung 2009). (2) For each ACE, affected women were more likely to have depressive symptoms than their counterparts. There was a dose-response effect in that a higher number of ACEs was associated with a higher likelihood of having depressive symptoms. PICs, on the other hand, were associated with a lower likelihood of having risk. Efforts to prevent ACEs and to promote PICs might help reduce the risk of depressive symptoms and their associated problems in adulthood (Chung 2008 (Dixon 2009). Within 13 months after birth, 6.7% AP families were referred for maltreating their own child in comparison to.4% NAP families. AP families had significantly higher numbers of risk factors. Mediational analysis found that intergenerational continuity of child maltreatment was explained to a larger extent (62% of the total effect) by the presence of poor parenting styles together with the three significant risk factors (parenting under 21 years, history of mental illness or depression, residing with a violent adult). The three risk factors alone were less explanatory (53% of the total effect) (Dixon , 2005b. Protective factors of financial solvency and social support distinguished Cycle Breakers from Maintainers and Initiators. Therefore, it is the presence of protective factors that distinguish Cycle Breakers from families who were referred to Child Protection professionals in the first year after birth (Dixon 2009 Qualitative (Interviews) To describe, explain, and predict women's and men's responses to sexual violence throughout their lives and construct a healing model. The CSA Healing Model is a model that represents the multifaceted and dynamic process of healing from CSA over the lifespan. The model does not focus on discrete variables associated with positive or negative coping but rather captures some complex healing processes that culminate in the experience of laying claim to one's life. The model includes four stages of healing, five domains of functioning, and six enabling factors that facilitate movement from one stage to the next. The model indicates that clinicians should focus on how clients might move from grappling with the meaning of the CSA, to figuring out its meaning, to tackling its effects, and ultimately, to laying claim to their lives, and they should be ready to discuss healing in any one of several domains. Because parenting was of great concern to most participants, it should routinely be addressed as a therapeutic issue. The model suggests that parenting should not be considered a dichotomous factor; that is, one either abuses or one nurtures one's children. Rather, the processes of wishing to stop the cycle of abuse and attempting to stop the cycle of abuse were critical steps that need to be acknowledged and fostered. Prospective cohort Investigate intergenerational pathways (1): to determine whether two groups of mothers, those who broke the cycle of abuse compared to those who did not, differ on dissociative process and symptomatology (Egeland 1996); (2) contrast the incidence of supportive relationships experienced by mothers who continued the cycle by abusing their own child versus those who broke the cycle and provided the child with adequate care (Egeland 1988); and (3) examine whether a maternal history of maltreatment in childhood has a detrimental impact on young children's mental health and to test theoretically and empirically informed pathways by which maternal history may influence child mental health (Bosquet 2016). (1) Mothers who were abused and are abusing their children were rated higher on idealization, inconsistency, and escapism in their description of their childhood and they scored higher on the Dissociative Experience Scale compared to mothers who broke the cycle. Mothers who were abused and abused their children recalled the care they received as children in a fragmented and disconnected fashion whereas those who broke the cycle integrated their abusive experience into a more coherent view of self. Even after partialling out the effects of IQ, large differences were found indicating that dissociative process plays a part in the transmission of maltreatment across generations (Egeland 1996).
(2) Abused mothers who were able to break the abusive cycle were significantly more likely to have received emotional support from a nonabusive adult during childhood, participated in therapy during any period of their lives, and to have had a nonabusive and more stable, emotionally supportive, and satisfying relationship with a mate. Abused mothers who re-enacted their maltreatment with their own children experienced significantly more life stress and were more anxious, dependent, immature, and depressed (Egeland 1988   Longitudinal/ cross-sectional study Investigate intergenerational pathways: examine evidence of the continuity in abusive discipline across two generations (G1 and G2) and the role of Safe, Stable, and Nurturing Relationships (SSNRs) as protective factors. Significant predictive association between physical discipline by G1 and G2 parents, after accounting for childhood socioeconomic status and gender. While being physically disciplined as a child was inversely related to reports of having had a caring relationship with one's mother, only care and support from one's father predicted a lower risk of harsh physical discipline by G2s. None of the SSNR variables moderated the effect of G1 discipline on G2 discipline. A case-level examination of the abusive histories of G2 harsh discipliners found they had in some cases been exposed to physical and emotional abuse by multiple caregivers and by adult partners. the process of reorganization on the transmission of attachment across generations; associations between unresolved trauma and mother-child attachment, while testing whether reorganization is associated with more secure child attachment. Mothers with unresolved trauma had insecure attachment themselves and were more likely to have infants within secure attachment. However, the one exception was that all of the mothers with unresolved trauma who were reorganizing toward secure attachment had infants with secure attachment. Cross-sectional study Investigate intergenerational pathways: whether women who report abuse in childhood adapt in less resilient ways to challenges to their sense of parenting competence due to infant difficult behaviour than women who report no childhood abuse. Pregnant women who reported childhood abuse decreased more in Parenting Self Efficacy (PSE) in response to the difficult-to-soothe infant (i.e., failure condition) than pregnant women who reported no abuse, whereas no differences were found in women's adjustment of PSE to the easyto-soothe infant (i.e., success condition) or with respect to PSE at baseline. Effects did not vary according to resolution of trauma. Physical and sexual abuse and neglect.
Case study Psychotherapy description: to illustrate the diagnostic and therapeutic issues in the course of an intervention designed to prevent intergenerational abuse. Drawing upon an ecological, transactional model of development, the case study utilized a multimethod, longitudinal approach to assess the mother's history and current psychosocial functioning, the infant's developmental competence and attachment status, patterns of motherinfant interaction and components of the family's social ecology. The treatment involved two weekly therapy sessions; one, an individual session for the mother and the other, a session in which mother and infant were seen together. The findings at the one-year evaluation showed improved maternal psychosocial functioning, the infant's shift from an insecure to a secure attachment classification and improved patterns of mother-infant interactions. Low (Continued) that might stand out as sources of strength in the parents' sense of themselves and ability to care for their children. Authors propose that angels in the nursery-care-receiving experiences characterized by intense shared affect between parent and child in which the child feels nearly perfectly understood, accepted, and loved-provide the child with a core sense of security and self-worth that can be drawn upon when the child becomes a parent to interrupt the cycle of maltreatment. Uncovering angels as growth-promoting forces in the lives of traumatized parents is as vital to the work of psychotherapy as is the interpretation and exorcizing of ghosts. Longitudinal/ prospective cohort Investigate intergenerational pathways: (associations between childhood trauma and rates of disorganized infant attachment strategies as well as less involved and more hostile maternal caregiving. and mediating role of the continued presence of trauma-related symptoms in adulthood, such as dissociative or post-traumatic symptoms, unresolved trauma on the caregiving and attachment systems.) A history of physical abuse was associated with increased hostile-intrusive behaviour toward the infant, increased infant negative affect, and a decreased tendency to act on trauma-related symptoms. A history of sexual abuse was associated with decreased involvement with the infant, more restricted maternal affect, and more active reporting of trauma-related symptoms. Infants of mothers who had experienced childhood violence or abuse were not more likely to display insecure attachment strategies than infants of mothers who had not experienced trauma. However, the form of insecure behaviour was significantly different. Insecure infants of violenceexposed mothers displayed predominantly disorganized strategies, whereas insecure infants of mothers with benign childhoods or neglect only displayed predominantly avoidant strategies. case study Parent strategies to break cycle: to describe the reactions, attitudes and interactional characteristics of a young mother with a history of abuse, illustrating projective mechanisms in the transmission of a 'potential for violence'. Maternal perception of her infant was distorted to the extent that the mother was re-experiencing encounters with her own intrusive and traumatizing mother in the face of her screaming child.
She also perceived the infant's motor impulses as physical attacks on herself and expressed intense anxieties about her daughter's future aggressive potential. She tried to ward off these anxieties by employing a rigid scheme of rules and obsessively controlling the father's and grandmother's interaction with the child. For the success of therapy in this case, 'containment' of Mrs L's needs turned out to be important. However, this also underlined the need for including the partner in the therapy in order to make the abused woman's striving for acceptance, her narcissistic vulnerability and her mistrust of relationships understandable to the husband. Qualitative (Interviews) Parent strategies to break cycle: investigate the process of resisting the legacy of intergenerational maltreatment as well as explore the ways in which helpers help or fail to help these families as they struggle with the challenges of parenting. Results suggest that some of the difficulties in helping at risk or maltreating parents may come from an insufficient respect for how the themes developed through a childhood of maltreatment interfere with the establishment of a positive relationship. Key qualities in building a helping alliance were described by the participants, and the good match between client needs and a self/ mutual help model were noted. Important aspects of resilient outcomes involved childhood sibling bonds and the successful negotiation of earlier critical events.

Moderate
(Continued)  Participants described ambivalence about seeking help but also a sincere desire for healing, coupled with hope for the future. This tension was apparent in the discussions highlighting the importance of access to experienced, non-judgmental, and knowledgeable health and social care staff and volunteers, the wish for both formal, integrated physical and mental health services, and for informal opportunities to meet other trauma-exposed mothers in a non-stigmatizing, childfriendly setting. Finally, positive relationship-building, respect, and safety were identified as key elements of services critical to counteract trauma-related shame and mistrust in others. (2) All women independent of risk status increased in bonding to their infant over the first 6 months postpartum; however, women with postpartum psychopathology (depression and PTSD) showed consistently greater bonding impairment scores at all times points. The 6 months assessment bonding impairment and observed parenting behaviours were significantly associated (Muzik 2013b) (3) General PTC were more likely to include negative and positive changes; parenting PTC were more likely to be exclusively positive. Indicators of more severe CM (parent perpetrator, more CM experiences) were related to parenting but not general PTC. Concurrent demographic risk moderated associations between number of CM experiences and positive parenting PTC such that among mothers with more CM experiences, demographic risk was associated with stronger positive parenting PTC (Fava 2016). (4) Controlling for domestic violence (DV), distorted prenatal representations were associated with higher rates of self-reported childhood physical neglect. DV moderated the relationship between representations and CM, such that women who were exposed to DV during pregnancy and had distorted prenatal representations were least likely to report childhood physical and sexual abuse (Malone 2010). (5) Posttraumatic stress in pregnancy, alone, or comorbid with depression, is associated with postpartum depression. Postpartum depression alone, or comorbid with posttraumatic stress, was associated with impaired bonding. In both models, higher quality of life ratings in pregnancy were associated with better outcomes, while reported dissociation in labor was a risk for worse outcomes. The effect of a history of childhood maltreatment on both postpartum mental health and bonding outcomes was mediated by pre-existing mental health status (Seng2013). (6) Resilience, childhood trauma severity, and their interaction predicted postpartum PTSD and MDD. In those with highest resilience, no mothers met criteria for postpartum MDD, irrespective of childhood trauma, while for those with lowest quartile of resilience, 25% with lowest CTQ severity and 68% of those with highest CTQ severity were depressed. The CD-RISC, but not the CTQ, was predictive of postpartum sense of competence. The CD-RISC and the CTQ were predictive of postpartum family functioning, though no moderating influence of resilience on childhood trauma was found. (Sexton 2015). (7) Participants who reported childhood neglect or physical abuse (regardless of sexual abuse) were significantly more likely to endorse difficulty falling asleep and staying asleep relative to participants who were not exposed to childhood trauma. Furthermore, PTSD was associated with sleep problems. (Swanston 2014). (8) Infants of PTSD+ mothers scored higher on the CSHQ and had more separation anxiety around bedtime than PTSD-and CON, and the severity of their symptoms was correlated with the degree of sleep disturbance. Maternal postpartum depression symptoms mediated impaired mother-infant bonding, while infant sleep disturbance contributed independently to impaired bonding. Poorer mother-infant bonding at 4 months predicted more behavioural problems at 18 months (Hairston 2011 Qualitative (Interviews) To examine the phenomenon of anger and its role in the recovery process of 6 midlife women. A typology was constructed, depicting 5 types of anger ((1) self-castigating anger; (2) displaced anger; (3) the anger of indignation; (4) self-protective anger; and (5) righteous anger). Anger ranged from non-productive, self-castigating behaviour to empowering, righteous anger that enabled women to protect themselves from further abuse and to advocate for abused children.
Anger intensity waxed and waned across the years of the healing trajectory, peaking for some women in their 20s and 30s. in the change process. The highest rates of 'negative' coping behaviours reported were overeating (65%), smoking (50%), compulsive work (50%), and oversleeping (40%). Participants also reported engaging in positive coping behaviours such as counselling/therapy (65%), volunteering (65%), spiritual or religious activities (55%), meditation (50%), and exercise (50%). In the "other" category of positive coping behaviours, several participants listed reading and learning about abuse.  (Zuravin 1996). (2)  neglected were about 20% less likely to perpetuate the cycle than mothers who reported no or infrequent severe beatings (Zuravin 1996). (2) PCT predicated higher depression, lower social support & self-esteem. Social support and a less external locus of control were protective in function (Banyard 1999). (3) Higher rates of trauma exposure were related to decreased parenting satisfaction, reports of child neglect, use of physical punishment, and a history of protective service reports. These links were partially mediated by the relationship between trauma exposure and increased maternal depression. Social support and taking care of one's own needs was protective against for parental satisfaction, while maternal depression was a mediator (Banyard 2003 Parents were recruited from a wide range of settings, including; prenatal/maternity clinics (n = 12), paediatric/special care nurseries (n = 2), community health and primary care clinics (n = 5), general communities (n = 7), Indigenous communities (n = 1), mental health programs (n = 5), child welfare agencies (n = 5), parenting programs (n = 4), financial assistance programs (n = 3), youth foster programs (n = 2), schools (n = 2), relationship support centre (n = 1), parents self-help group (n = 1), correctional centre (n = 1), health visitor program (n = 1) and an addiction treatment centre (n = 1). The setting was unclear in four studies.
Participants: Forty-five of the 57 included studies comprised only mothers and ten studies included both parents. In the remaining two studies it was unclear whether the participants were mothers and/or fathers. No studies included only fathers, extended family, partners of same-sex couples or other carers. Forty-eight studies reported sociodemographic characteristics of the participants. The mean age of parents was 28 years, with an age range of 12-62 years. Forty-three studies reported parent ethnicity; 23 studies reported a majority (>50%) of Caucasian parents and 20 studies a majority of African-American and/or Latino parents. Indigenous parents were identified in six studies (<7%), and included Native American, First Nation Canadian and Aboriginal and Torres Strait Islander Australian parents.
Parent education level was reported in 30/57 studies. In eight studies the majority (>50%) of parents were college graduates, in 12 studies the majority were high school graduates, and in 10 studies the majority reported not completing high school. In 14 of these 30 studies there was a relatively high proportion of parents (>20%) who had not completed high school. A measure of socio-economic status (SES) of parents was reported in 35 studies, and in 22 studies the majority of parents reported experiencing socio-economic disadvantage. The marital/relationship status of parents was reported in 34 studies. The majority (>50%) were classified as single/sole parents in 14 studies; with a further three studies reporting more than 20% parents as single/sole parents.
Thirty-eight studies identified parents during the perinatal period (pregnancy to two years postpartum). In 19 studies the age of the children was unclear. No studies explicitly included parents before pregnancy. Twenty-three studies involved parents during pregnancy and up to six weeks postpartum, 30 studies involved parents from 6 weeks postpartum to 12 months after birth, and 23 studies involved parents from 13 to 24 months postpartum. Thirty-five studies included parents both with and without a history of childhood maltreatment, 20 exclusively included parents with histories of childhood maltreatment, and two were unclear (reports of resolved and unresolved trauma and childhood maltreatment as one of several risk factors required for inclusion). Childhood maltreatment experiences reported by parents included; emotional abuse (17 studies), sexual abuse (39 studies), physical abuse (44 studies), neglect (emotional (9 studies), physical (9 studies) or otherwise unspecified (14 studies)), exposure to domestic/intimate partner violence (IPV) (14 studies) or witnessing other violence (2 studies) and 'Adverse Childhood Experiences (ACEs)' (5 studies). Experiences and definitions of childhood maltreatment within included studies also incorporated; 'harsh parenting' (2 studies), parental suicide and 'bizarre punishments' (1 study), household dysfunction (1 study), having a guardian in trouble with the law or in jail or homeless (1 study), living in poverty or apart from parents before 16 years of age (1 study), unresolved trauma (1 study), and adult sexual and/or physical assault (2 studies) or 'adult single event trauma' (1 study). Five studies relied on retrospective reports of child protection substantiations, and one study exclusively included parents who had themselves been removed from their families of origin during childhood.
Types of studies: The majority of studies used a quantitative descriptive design (36/57 studies). Of the remaining studies, 16 were qualitative, three were RCTs of interventions (and one associated article), and two evaluated measurement/assessment tools (one associated article).
Outcomes reported by included studies: Most outcomes were measured at the level of individuals (or parent-infant dyads) (52 studies), one study collected outcomes at a family level and in two studies this was unclear. Only five studies reported outcomes separately according to socio-demographic factors (based on PROGRESS-plus equity criteria), including maternal age (4 studies), ethnicity (4 studies), educational level (3 studies), relationship status (4 studies), socioeconomic status (2 studies). However many studies conducted analyses in which the associations with outcomes that were estimated were adjusted for socio-demographic factors. See Table 1.
A summary of evidence under each of the main outcomes (theories, risk and protective factors, parents' views, interventions and assessment tools) are outlined briefly below to provide an 'evidence map'. Evidence will the synthesized in detail using methods appropriate to each study type in subsequent comprehensive systematic reviews (S1 Appendix).

Theoretical frameworks
Sixteen different theoretical frameworks were identified across 14 studies in this review. A brief description of the main theories are listed in Table 2 and outlined narratively below.
Attachment theory [138] was the most common theory identified among studies in this review [45,129,132]. Attachment theory was used to try to understand effects of adversity and protective factors on continuation or discontinuation of intergenerational cycles of maltreatment [41,72,73,132]. Iyengar et al. [86] examined the role of attachment reorganisation, where parents were actively changing their understanding of past and present experiences and moving toward 'earned attachment security' and 'resolved trauma' as assessed on the Adult Attachment Interview (AAI) and they found all mothers who were 'reorganising' had infants with secure attachment. Lyons-Ruth and Block [94] found infants with insecure attachment born to mothers with a history of childhood maltreatment displayed predominantly disorganised attachment, while infants with insecure attachment born to mothers without a history of childhood maltreatment displayed predominantly avoidant strategies. Madigan et al. [96] examined the stability of disorganised attachment in a sample of adolescent mothers transitioning to parenthood and concluded adolescents seemed to find it harder to re-organise attachment and 'resolve trauma' than older mothers.
Belsky's sociological model [139] was used in several studies as a theoretical framework to investigate the dynamic nature of factors affecting discontinuity and continuity of childhood maltreatment [43,45,132].
A ghosts in the nursery [140] metaphor has been used to describe the way parents re-enact their childhood histories from unremembered early relational experiences [37].
Malone et al. [111] suggested that relational theories provide support for how the "ghosts" of one's past develop and may become a persistent presence in future relationships. They suggest that a safe supportive relationship or "holding environment" will be needed before parents can begin to reflect on how abuse in their own childhood influences their lives now, and argues interventions that help parents recognize and resolve the negative experiences from their past will enable the formation of a positive relationship with their child [111]. These schemas, or internal working models, initially develop through interpersonal interactions with primary caregivers, including experiences of loss and abuse [138], and are then kept outside of consciousness by defensive strategies that protect an individual.
An Angels in the nursery [93] metaphor was proposed to reframe 'ghosts in the nursery' to also consider positive factors. Leiberman et al. described the ways in which early benevolent experiences with caregivers can work as protective forces even in the face of overwhelming trauma, and how examining these consciously can be used as a powerful healing tool, by placing the traumatic cues within the larger perspective of nurturing and growth-promoting experiences [93].
Relational development theories were used to consider multiple aspects of a developmental system (e.g., individual history, social relationships, and environmental context), as well as the parents current social context. These may be particularly important given that parents with a history of childhood maltreatment may draw on fewer support networks and experience higher levels of parenting stress [43].
Object relations theory was used to help understand aspects of psychological functioning that support adaptive relational development [45]. Baumgardner suggests that this is, in essence, what Bowlby [138] referred to as an 'internal working model', a cognitive framework comprising mental representations for understanding the world, self, and others [45].
Resilience theories and frameworks [141] were used as a way of understanding protective factors associated with discontinuities in the intergenerational transmission of trauma [43,57,101]. Authors argued that understanding risk factors is only half of the picture, and Belsky's sociological model (Belsky, 1978) 8a, 9, 55c Ghosts in the nursery (Fraiberg et al., 1975)  understanding factors associated with resilience and exceptions is critical for designing effective interventions and policies [43,57]. Bysom suggested these can be viewed as existing on a continuum, with variability exhibited throughout a lifetime [57]. Social learning theory [142] was used to explain how some parents replicate the behaviour of their own parents [45,57,72,73]. However, Bysom [57] argued it was limited in its ability to explain the exceptions and discrepancies seen by most parents with a history if child maltreatment and suggested social cognitive theory [143] may be more helpful to distinguish and explain some of these exceptions, but is inadequate for explaining resilience.
The family systems model [85] was used in one study [85,108] to help explain differences between maltreating and non-maltreating families in a special care nursery setting, with 'differentiation of self' identified as an important factor in discontinuing the cycle of violence.
A Child Sexual Assault (CSA) Theoretical Healing Model was constructed from qualitative research [71]. In developing the model, the authors also drew on literature about post-traumatic growth, arguing that survivors not only cope but engage in dynamic processes that include growth and recovery. Their model incorporates constructs from positive psychology, including perceived self-efficacy, personal control, altruism and empathy, and spirituality. Fava et al. [14] explored themes of post-traumatic change and found that parents' reports of post-traumatic changes during the postpartum period (four to 18 months) were more likely to be exclusively positive, compared to other periods of the life-course when post-traumatic changes were likely to include both negative and positive changes.
Anger theories were explored in a qualitative study to help understand the role of anger in the healing trajectory from childhood maltreatment [125]. The authors constructed a typology depicting five types of anger ranging from non-productive self-castigating and displaced behaviour to empowering indignant, self-protective and righteous anger that enabled women to protect themselves from further abuse and to advocate for abused children [125].

Life-course and intergenerational pathways (risk and protective factors)
Thirty-eight studies described risk and protective factors that mediate and/or moderate pathways from a parental history of childhood maltreatment to parental and infant outcomes in the perinatal period. These are briefly outlined below and depicted in Fig 2 which includes the study ID numbers for the COIS Table 1.

Perinatal experiences described by parents with a history of childhood maltreatment and strategies used during the perinatal period to heal from childhood maltreatment and/or prevent intergenerational transmission of trauma
Eleven studies described perinatal experiences and/or strategies of parents with a history of childhood maltreatment. One study described maternal healing after CSA [71] and six studies reporting parents' perceptions of factors that influence healing and/or ability to 'break the cycle' of intergenerational trauma [39,57,67,71,98,101,128]. Two studies reported women's experiences of perinatal services [84,102,103] and one reported experiences of breastfeeding [65], specifically among women who had experienced CSA. A brief outline of the main themes are summarised and mapped to the COIS Table 1 study number in Table 3 below.
Parents' experiences. Parents described experiences of post-traumatic growth [71] and the unique opportunities for healing through parenting for adolescents transitioning out of Table 3. Parents experiences in the perinatal period: Evidence map of qualitative studies. (Table 1) Parent descriptions about 'breaking the cycle' of childhood trauma Parents descriptions of perinatal care after Childhood Sexual Assault (CSA)

Experiences parents described regarding effects of childhood trauma in the perinatal period
Post-traumatic growth 24 Parenting a healing opportunity/living family legacy 24 Healing process is complex, including parenting transition 4,24 Birth and parenting journey difficult for most survivors of CSA 4 Disclosure not always beneficial/stigma 22,24 Fear of child protection agencies 4 Feeling of lack of parenting knowledge/skills 44 Many aspects of care re-traumatising 32,45a Disbelief/betrayal of confidence/invalidation of experiences 32 Experiencing punitive responses to self-harm from care-providers 32 Unexpected triggers from intimate procedures, pain, encounters with strangers, lack of control, breastfeeding 45a, 20 Breastfeeding important to help improve maternal-infant attachment after CSA (but can trigger trauma responses) 20 Ability to breastfeed validates body as 'good or bad ' 20 System-level barriers to care (lack of services, poor access) 32 foster care [39]. Parents' also described challenges related to disclosure of child sexual assault [71] and child maltreatment, such as; stigma, awareness of the time pressures on staff in maternity services [67], [101]. Parents' described difficulties negotiating boundaries which was perceived as contributing to being unable to pay attention to selfcare, and taking responsibility for the behaviour of others, particularly their children [101]. Parents described the impact of having a negative sense of self-worth on parenting [101] which included feelings of guilt and shame that were shaped by previous experiences [39]. Experiencing a lack of control within the context of parenting was identified as one of the most re-traumatising aspects of perinatal care and breastfeeding [65,103]. Gaining a sense of control was seen as critical to overcoming these types of stressors [65,103]. CSA survivors described how breastfeeding could validate their views of their bodies as 'good' or 'bad' depending on the 'success' of their breastfeeding [65].

fear of child protection agencies [39], lack of parenting knowledge and skills, and mental and physical health problems
Lack of trust in relationships was a common issue that impacted parents' ability to make friends which frequently led to feelings of isolation [67]. These difficulties in trusting others were also seen to impact on developing therapeutic relationships with perinatal care providers [39], often exacerbated by fears of child protection agencies [84]. Parents also described challenges related to dealing with feelings of loss and abandonment [101] which also impacted on experiences of maternity care [84].
Parents described how some experiences could be re-traumatising [84,103]. These included encountering responses of disbelief and invalidation of their past experiences, punitive responses to self-harming and betrayal of confidence [84]. Women described a range of unexpected triggering of trauma responses in birth and parenting, including intimate procedures, encounters with strangers, pelvic pain [103], and distress at being unable to soothe their crying baby [84]. A range of system level barriers were also identified, including; lack of services (particularly psychotherapy), access problems (lack of transport and childcare), infrequent sessions with limited opening hours, short term contracts for staff, use of waiting lists and long waiting times in clinics [84]. There were mixed reports about the value of psychological therapies [39,128] and concerns about the lack of alternative approaches to therapy (such as art therapy). Stigma and the chaotic lives of some parents were identified as individual level barriers [39]. Monaghon-Blout [101] suggested that critical parenting experiences often increased motivation for change, and if coupled with access to resources 'at the right time' could help to support positive healing.
Strategies parents describe using to heal or break the cycle of trauma. Understanding the experience of trauma, 'storying' and 'meaning making' were common strategies used by parents to heal and prevent intergenerational transmission of trauma [67,71,128], conceptualised by Draucker et al. [71] as a specific healing phase. Other phases involved enacting conscious strategies to tackle the effects of trauma [71], such as leaving violent partners to ensure child safety [84] and promoting positive communication [67]. Draucker et al. [71] described a final healing phase where parents make a commitment to not pass on the effects of trauma to their children. Parents who had experienced maltreatment themselves consistently reported a strong desire and determination to parent differently with their own children [39,67,71,84,98,101,103]. However, some parents continued to disclose behaviours that were harmful, despite expressing these intentions [57,98].
Reducing isolation and increasing social support was also identified as critical to healing [67]. Parents reported that an important aspect of social support was increasing awareness of alternative methods of parenting, and that non-judgemental social support was the most helpful [57]. Parents also highlighted the ways in which issues such as lack of trust and a negative sense of self-worth could contribute to difficulties in establishing and maintaining friendships [101]. An important source of support for parents who had challenging relationships with intimate partners or their family of origin was having a 'family of friends'. Also important were supportive interpersonal relationships [67,101] and mentors [39].
Spirituality, self-care and 'centering on self' were identified as key aspects of healing [57, 71,128]. Self-care activities included taking 'time out' [39], meditation, exercise and reading [128]. Parents also described helping others and volunteering as having a healing effect [71,128].
Thomas et al. [125] explored the role of anger in healing, and suggested that 'righteous anger' could assist healing as it often motivated people to help others. 'Self-protective' and 'indignant' anger were seen as positive, while self-castigating and displaced anger were considered harmful [125]. There was consensus that 'hanging on' to any type of anger was harmful [125]. Parents also reported that speaking up about trauma could add to self-empowerment [67].

Perinatal interventions to support parents with a history of childhood maltreatment
Eleven studies provided evaluations or descriptions of strategies/interventions to support parents with a history of maltreatment in their own childhood. The main outcomes from these studies are briefly described below and summarised in Table 4. Details about the setting, population and overall confidence in the study findings are summarised in COIS Table 1. A comprehensive systematic review of intervention studies is planned and will include detailed synthesis of study outcomes and assessment of heterogeneity, barriers and facilitators (S1 Appendix). No economic evaluations were found.
One study analysed cross-sectional data from a parenting program intervention and found high rates of trauma exposure among the participating parents. Service providers found the associated screening process both feasible and beneficial Three formative studies asked parents whether they would like support and/or what type of support would be helpful. Adams [36] mail-surveyed 89 first-time mothers in the US, and found that of 15 percent who disclosed a history of childhood maltreatment, two thirds expressed interest in receiving support. Muzik et al. [108] undertook qualitative interviews with 52 trauma-exposed mothers and found that although women were often ambivalent and anxious about receiving formal help, they consistently expressed a desire for healing and hope. Women in the study identified a need for experienced, non-judgmental and knowledgeable support, which includes formal integrated mental and physical health services, and opportunities to meet other parents informally in a non-stigmatising child-friendly environment. Positive relationship building, respect and safety were described as key elements of support to counteract trauma-related effects of shame and mistrust [108]. In-depth interviews with seven members of a 'parents anonymous' self-help group [101] were analysed to develop a recommendations for 'building a helping alliance' within parent self-help groups. These included helping parents to understand how childhood maltreatment can interfere with establishing positive relationships, and enabling access to 'the right support' during critical periods [101].
The effects of five perinatal interventions were evaluated in RCTs. However, none of these parenting interventions were specifically designed for parents with a history of childhood maltreatment and the results reported here represent findings from subgroup analyses only. Armstrong et al.
[40] evaluated an Australian nurse home-visiting intervention compared to standard community child health services (from 6 weeks postpartum) with 181 'vulnerable parents' that included 33% mothers and 17% fathers with a history of childhood maltreatment. No differences were reported between groups for breast feeding initiation or duration, knowledge of SIDS, or use of health services. Those receiving nurse home visits were reported as  [133] examined the moderating effects of child maltreatment history on depression, social functioning, and parenting in mothers participating in a home-visiting trial for mothers with depression in the United States. They found a number of main effects in which experiences of different types of trauma were associated with poorer parent functioning regardless of whether they were receiving the home-visiting intervention, suggesting a need for targeted strategies to support parents who have experienced child maltreatment. A study evaluating the effects of 'the 'Incredible Years Parenting program' compared to the regular 'Head Start' curriculum (Puget Sound, US) with 421 parents, reported lower baseline parenting skills among parents with a history of childhood maltreatment, and presented cross-sectional data suggesting that this history did not impact on participation rates in the program [46]. Parents' receiving the intervention demonstrated significant improvements in parenting behaviours compared to those receiving the standard 'Head Start' curriculum [46]. An RCT evaluating the effects of cognitive behavioural therapy (CBT) aiming to support smoking cessation in pregnancy reported improvements in postnatal depression scores, but no differences in smoking cessation for women receiving CBT compared to standard health education [52]. However, subgroup analyses suggests women with a history of childhood maltreatment were more likely to experience improvements in depression and less likely to stop smoking than women without [52]. Finally, an RCT of brief interpersonal therapy (IPT-B) for parents with a history of childhood maltreatment reported improvements in postnatal depression, anxiety, social dysfunction, interpersonal problems and insecure attachment among parents receiving the intervention, compared to parents allocated to the control group who received 'usual care' [20].
Three articles described case studies of individual participants receiving psychotherapy. Leifer and Smith [91] presented quantitative data suggesting improvements in maternal psychosocial functioning, parent-infant interactions and secure attachment following therapy. Two qualitative case studies presented rich descriptions demonstrating the complexity of adult psychotherapy following severe trauma in infancy with thoughtful therapist reflections on the shared journey, including challenges for the therapist [41, 100].

Tools used during the perinatal period to identify parents with a history of childhood maltreatment and/or effects
We found 22 tools used to identify parents with a history of childhood maltreatment. These are briefly outlined below and summarised in Table 5. A more detailed assessment of Child Abuse and Trauma Scale (n = 1) 38 items about the frequency and extent of negative childhood experiences, including the general atmosphere of their home and the way they felt they were treated in childhood by caregivers.

30
Early Trauma Inventory Self-Report Form (n = 1) General trauma, and physical, emotional and sexual abuse. 50a Trauma History psychometric values, the populations that tools have been validated within (including fathers), and barriers and facilitators for using tools will be addressed in a subsequent comprehensive review (S1 Appendix). The tools are categorised as those which; (a) exclusively assess an experience of traumatic event history (i.e. exposure), (b) assess experience of traumatic event history as part of a broader psychosocial needs assessment, (c) assess the impact of experience of traumatic event history as symptoms, (d) assess the impact of experience of traumatic event history on parenting, or (e) 'other'. a) Tools which exclusively assess experience of traumatic event history (ie exposure). The Childhood Trauma Questionnaire (CTQ) was reported in nine studies (20 articles) and measures emotional, physical and sexual abuse, and emotional and physical neglect domains [14, 20, 37, 51-53, 96, 105, 108-110, 113, 117, 118, 133, 135, 144, 145]. Blalock et al.
[51] cites use of the tool in psychiatric, community, substance using, adolescent and adult populations. Test-retest reliability from birth to five years postpartum was also reported [55].
The Assessing Environments III-History of Parents' Childhood Parenting Experiences tool [146] was used in two studies [46,77]. Baydar et al.
The Adverse Childhood Experiences (ACE) study questionnaire [147] was used in three studies (5 articles) in this review [61,87,[104][105][106]. It retrospectively assesses exposure to ten forms of abuse, neglect, and household dysfunction before the age of 18. It was adapted and used in conjunction with the AAI in one study [104][105][106], with correlations reported between ACEs and the unresolved trauma category of the AAI [87,104].
The Childhood Experience of Care and Abuse (CECA) [148] was used in one study [88] for comparing measures on the Parental Bonding Instrument (PBI) and includes assessments of CSA, physical abuse, psychological abuse, antipathy and neglect. The Brief Physical and Sexual Abuse Questionnaire (BPSAQ) [149] was used in one study [122,150] to quantify the severity of maternal violent trauma history. The Child Abuse and Trauma Scale (CATS) [151] was used in one study [82] and consists of 38 items about the frequency and extent of negative childhood experiences, including the general atmosphere of their home and the way they felt they were treated in childhood by caregivers. The Early Trauma Inventory Self Report Short Form (ETI-SF) [151] was used in one study [123] and includes assessment of general trauma, and physical, emotional and sexual abuse.
The Trauma History Table Interview (THTI) was developed for the Maternal Anxiety in Childbearing Years (MACY) project [108][109][110] and assesses information about the type, frequency, duration, and perpetrator identity for any childhood maltreatment experienced prior to age 16 years [14]. The Life Events Checklist [152] as used in one study [122,150] and is a 17-item checklist covering a range of potentially traumatic events from natural disasters to accidents, sudden losses to combat and interpersonal violent events. The Life Stressors Checklist [153] was used in one study [110] to assess trauma history. It is designed for use with women, and includes five items about childhood maltreatment (physical abuse, molestation, completed rape, emotional abuse, and physical neglect occurring prior to age 16). The Childhood History Questionnaire (CHQ) [154] was used by Bysom [57] to assess presence and frequency of physical abuse, as well as witnessing abuse. The Antecedent Experiences Questionnaire [155] was used by one study [94] with an adapted version of the AAI.
The Trauma section of the American Indian Service Utilization and Psychiatric Epidemiology Risk and Protective Factors Project (AI-SUPERPFP) interview [156] was used in a study with an American Native Indian community [92] and asks about 16 traumas; how often they occurred and at what age(s).
(b) Tools which assess experiences of traumatic event history as part of a broader psychosocial needs assessment. The Index of Need questionnaire developed and used in one study [68][69][70], asks parents to identify if "you or your partner were physically and/or sexually abused as a child". The Brisbane Evaluation of Needs Questionnaire was used in another study [40] and includes childhood maltreatment screening questions.
(c) Tools which assess the symptoms of trauma. The Post-traumatic Stress Disorder section of the Structured Clinical Interview for DSM-IV (SCID) was used in conjunction with other tools in three studies (four articles) [9,56,122,150] and assesses exposure to and effects of events involving actual or threatened death, serious injury, a threat to one's physical integrity, or witnessing somebody else be killed or seriously harmed. Juul et al. [56] cites studies using the tool with mental health patients [157], college students and primary care patients [158].
The Posttraumatic Stress Diagnostic Scale (PDS) [159] was used in one study reporting screening burden and utility of the tool [59]. The PDS measures both trauma exposure and trauma-related symptoms, with many items corresponding to the PTSD section of the SCID.
The Adult Attachment Interview (AAI) [160] was identified in six studies [75,86,87,94,96,97,104]. The AAI assesses childhood relationships with attachment figures, usually parents. Attention is also paid to experiences with loss, abuse or other trauma, and a lack of resolution of trauma. Madigan et al. [96] assessed the stability of AAI attachment representations at three perinatal time points (prenatally, at infant age 6 to 9 months and at infant age 12 to 15 months) and found relatively low internal consistency for reporting the same experiences of loss, but higher consistency for reporting experiences of trauma.
(d) Tools that assess the impact of childhood maltreatment history on parenting. The Trauma Meaning Making Interview (TMMI) [161] was used in one study [14] to assess posttraumatic change and explore maltreatment in childhood experiences and feelings, with questions added about the perceived impact of their child maltreatment experiences. The Parent-Child Conflict Tactics Scale (PC-CTS) [162] was used in three studies [49,50,58] and includes questions about the types of discipline parent's received as a child, and if they ever experienced any of 11 parental disciplinary or neglectful actions. The Conceptual Change Questionnaire (CCQ) (1999) was used in one study [128] to assess how parents conceptualise abuse.
e. Other tools. The Child Abuse Potential (CAP) inventory [163] was used in one study [58] and is a 160-item questionnaire designed to assist in the assessment of physical child abuse reports.

Discussion
We located 57 studies conducted with participants during the perinatal period involving more than 20,000 parents. Most studies were conducted in the US and involved mothers only. No studies exclusively involved fathers, other family members or same-sex/gender diverse partners, and few included Indigenous parents. More than 75% of studies were categorised as 'descriptive observational' studies and we found no perinatal interventions specifically designed for parents with a history of childhood maltreatment and/or complex trauma.
Theories identified in our review are similar to those outlined in a clinical review for practitioners by [164]. However, our review included additional constructs that incorporated socioecological models to help explain some of the interactions and differences in parental outcomes. Studies in this review found post-traumatic changes were perceived by parents as predominantly positive during the perinatal period [14], supporting previous studies that suggest the parenting transition [165] offers a window of opportunity to transform the 'vicious cycle of trauma' into a 'virtuous cycle of healing' [166]. Most parents with a history of childhood maltreatment are able to provide nurturing environments for their children [44] and examining these 'cycles of discontinuity' is a promising place of exploration to illuminate innovative strategies for supporting parents [4]. Strategies parents reported using to support healing included; understanding and 'meaning making', conscious strategies to keep children safe and 'parent differently', increasing social support, spirituality and helping others. There were mixed reports about the value of psychotherapy and some parents expressed a desire for better access to more alternative therapies. A focus on these positive aspects and careful use of strengths-based language may also help to counteract some of the concerns about stigma, lack of trust and shame reported by parents in this review.
Studies in this review have reinforced National Trauma Guideline recommendations [6] by emphasizing the importance of considering childhood maltreatment and the sequelae of complex trauma within a socioecological context. This seems particularly important for Indigenous peoples with a histories of colonisation and ongoing discrimination. Our review also highlighted a number of system-level (e.g. lack of services, lack of transport, limited opening hours, use of waiting lists and long waiting times in clinics) and individual-level (e.g. fear of child protection agencies) contextual barriers.
Our findings support those of previous reviews that have identified the significance of the perinatal period in transmission of intergenerational trauma regarding the effects of stress [167], and a wide range of protective factors which are outside the mother-child relationship [168]. Studies in our review also have similarities with other reviews of intergenerational transmission of self-regulatory capacities and development of executive functioning skills in the first year of life [169,170]. While much of the focus of childhood maltreatment is on the impact of experiences within the family, some studies suggest there are particular stages of development where children are more vulnerable to trauma from people outside the immediate family [171]. There is also a need to understand the impact of cumulative trauma as children start to explore and seek acceptance outside the family, in the broader community and society, and as they become parents; and the intersections between parental experiences of childhood maltreatment or complex trauma and societal racism that compound trauma [172].
No perinatal intervention studies were found specifically designed for parents with a history of childhood maltreatment. However, our findings are consistent with reviews not specific to the perinatal period or parents with a history of childhood maltreatment, that suggest parenting interventions can improve parent-infant interactions and attachment security [173][174][175]. Intervention studies of nurse home-visiting, parenting programs, IPT-B and CBT all suggested positive effects on parent wellbeing. However, more comprehensive meta-analysis of outcomes is needed in a full systematic review to articulate effect size estimates and explore sources of heterogeneity.
Parents with a history of childhood maltreatment often face specific social and emotional challenges during the perinatal period, and these understandings of social and emotional wellbeing are strongly shaped by culture. Thus it is essential that tools for measuring and assessing wellbeing are developed and validated among populations they are intended to be used by. There is a dearth of appropriate tools to measure concepts of social and emotional wellbeing among Indigenous people [176]. Often, Indigenous understandings of health and wellbeing strongly emphasize the collective and relational, incorporating dimensions of connectedness to family, community, culture, country and spirituality; alongside physical and mental wellbeing [177]. We found only one measurement tool and observational study involving Indigenous people [177]. We also found no studies involving only fathers [178] or extended family members. There is also a need to consider validity, reliability across the perinatal period, and utility [179]. There are multiple domains to measure, but practical considerations include feasibility [180], cost, time to administer, scoring ease, and sensitivity of questions [181]. We should also consider acceptability [181], parents' perceptions [182], preferences [182], barriers and facilitators [183,184].
To our knowledge, this is the first review to map evidence specific to the perinatal period involving parents with a history of childhood maltreatment. We used systematic methods including double screening all full-text articles, double data extraction and assessed risk of bias to inform the development of comprehensive protocols for in-depth reviews in the second phase of this work (S1 Appendix). The results of this scoping review have enabled development of a targeted search strategy for future reviews. It has mapped all index terms of included studies, searched similar studies on PubMed and developed word trees from included abstracts and titles. This scoping review has provided a summary 'evidence map' across all study types which is not characteristic of targeted reviews of single study designs.
Importantly, while full systematic reviews will take several years, this scoping review has provided important timely information that we have incorporated into a co-design process for developing perinatal strategies to identify and support Aboriginal Australian parents with a history of childhood maltreatment and/or complex trauma. For example, we have summarised protective and risk factors from the epidemiological studies, and themes from parent experiences and strategies on cards. Following interactive discussion groups with Elders and parents as part of an Intervention Mapping [185] framework, the card summaries were shared with an explanation that these issues emerged from studies with parents in other population groups, and asked parents to identify issues which may be relevant for them. In addition, we extracted and synthesized constructs from the 22 measurement tools which informed an interactive workshop activity with key stakeholders. In the workshop setting, we discussed the degree of importance of each of the main constructs, as well as issues associated with asking parents who have experienced childhood maltreatment about them during the perinatal period. Another major strength of this review is the inclusion of Aboriginal people in all stages of the review process to ensure culturally-relevant insight, in alignment with the United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP) and Ethical conduct in research with Aboriginal and Torres Strait Islander Peoples and communities [186], and importantly balancing Indigenous worldviews with Western science.
There are limitations to this scoping review. First, our search may have missed intervention studies evaluating support strategies for parents that are embedded in general parenting programs. Second, only one reviewer conducted the preliminary title and abstract screen, which increases the likelihood of error. Third, we could not include detailed syntheses of findings specific to each study design, for example, thematic synthesis of qualitative studies and pooled analyses of epidemiological and intervention studies with sensitivity analysis to identify sources of heterogeneity or sub-group analyses to examine differential effects for subpopulations. We aim to address these limitations when we undertake in-depth systematic reviews (S1 Appendix).
There is little evidence that has demonstrated application of existing perinatal interventions for parents with a history of childhood maltreatment. Perinatal care services need to be 'trauma-informed' (aware) to minimise the risks of 'triggering' trauma responses for parents with a history of childhood maltreatment. The perinatal period offers a unique life-course 'window of opportunity' to recognise and assess parents who have experienced childhood maltreatment (and/or complex trauma) and enable access to trauma-specific support during frequent scheduled contacts with health services. For many predominantly healthy young people, starting a family may be the first contact with health services since their own childhood. It is also an important period for relational development, when many parents are motivated to make positive changes for the benefit of their child, and developing a nurturing relationship with the infant can generate positive responses that reinforce healing and 'earned security'.
Health experts have identified childhood maltreatment as a major public health challenge in mental health [187], with similar effects to those of tobacco on physical health [188]. In addition to significant associations with risk behaviours, it is likely that the effects of childhood maltreatment in the proposed criteria for complex PTSD (i.e. affective dysregulation, negative self-concept and interpersonal disturbances) [5], will have a differential and important impact public health program effectiveness. This includes how programs are received, engaged with and the capacity or agency that individuals have to respond to public health advice. There is an urgent need for all public health professionals to consider the impact of child maltreatment on health and health inequities more broadly [189].

Conclusions
Growing observational and qualitative evidence and theories are available to inform development of trauma-informed perinatal services, recognition, assessment and trauma-specific support for parents who have experienced childhood maltreatment and/or complex trauma. There is continuing interest in developing theoretical constructs to explain phenomena, growing evidence of protective and risk factors and rich descriptions of parents experiences and strategies they use. However, evidence involving fathers, extended families, same-sex or gender diverse partners and Indigenous people is lacking. There has been little activity in applying existing evidence to support parents in the perinatal period. Incorporating socioecological and cultural contexts seems essential to informing effective perinatal strategies for recognising and supporting parents during this critical window of opportunity. Given the paucity of applied evidence, it is critical that strategies are developed in collaboration with families and communities, particularly for Indigenous peoples and fathers. Program design and evaluation strategies should incorporate short reflective cycles and flexible evaluation approaches to enable early detection of unforeseen circumstances and tailoring to respond to needs and reduce the risks of harm for vulnerable parents.
Supporting information S1 Appendix. Perinatal awareness, recognition, assessment and support for parents who have experienced maltreatment in their own childhoods: Overview of reviews.