human resource management 6100

Maternal Childhood Abuse and Children’s Emotional-Behavioral Difficulties: Intergenerational Transmission via Birth Outcomes and

Psychosocial Health

Rebecca Giallo and Deirdre Gartland Murdoch Children’s Research Institute, Parkville, Victoria,

Australia, and The University of Melbourne

Monique Seymour and Laura Conway Murdoch Children’s Research Institute, Parkville, Victoria,

Australia

Fiona Mensah Murdoch Children’s Research Institute, Parkville, Victoria,

Australia, and The University of Melbourne

Lorraine Skinner and Ali Fogarty Murdoch Children’s Research Institute, Parkville, Victoria,

Australia

Stephanie Brown Murdoch Children’s Research Institute, Parkville, Victoria, Australia, and The University of Melbourne

Understanding the mechanisms and psychosocial pathways potentially underlying the association be- tween maternal childhood abuse exposure and poor child health and wellbeing is important to inform opportunities for support and intervention early in the period of becoming a parent. The aim of the study was to investigate whether adverse birth outcomes and psychosocial health issues (maternal depressive symptoms, exposure to intimate partner violence in the first postnatal year) are potential mechanisms underlying the association between maternal childhood abuse and children’s emotional– behavioral functioning at 10 years. Data were drawn from 1,507 first-time mothers and their 10-year-old children participating in the Maternal Health Study, a prospective study of women’s health during pregnancy and after birth. One in four women reported that they had experienced physical or sexual abuse in childhood. Children whose mothers had experienced either of these types of childhood abuse had significantly higher emotional– behavioral difficulties than children whose mothers had not. Psychosocial health pathways via maternal depressive symptoms and exposure of mothers to intimate partner violence in the first 12 months postpartum, but not adverse birth outcomes, mediated the association between maternal childhood abuse and children’s emotional-behavioral difficulties. These findings underscore the need to support mothers who have experienced childhood abuse and psychosocial health issues, both as a means of promoting maternal health and mitigating the potential intergenerational risks for children’s emotional and behavioral health.

Keywords: maternal childhood abuse, children, mental health, depression, intimate partner violence

This article was published Online First December 19, 2019. Rebecca Giallo and Deirdre Gartland, Healthy Mothers Healthy Families

Research Group, Murdoch Children’s Research Institute, Parkville, Victo- ria, Australia, and Department of Paediatrics, The University of Mel- bourne; Monique Seymour and Laura Conway, Healthy Mothers Healthy Families Research Group, Murdoch Children’s Research Institute; Fiona Mensah, Healthy Mothers Healthy Families Research Group, Murdoch Children’s Research Institute, and Department of Paediatrics, The Univer- sity of Melbourne; Lorraine Skinner and Ali Fogarty, Healthy Mothers Healthy Families Research Group, Murdoch Children’s Research Institute; Stephanie Brown, Healthy Mothers Healthy Families Research Group, Murdoch Children’s Research Institute, and Department of Paediatrics and Department of General Practice and Primary Health Care Academic Cen- tre, The University of Melbourne.

This study was approved by the following human research ethics com- mittees: La Trobe University (2002/38); Royal Women’s Hospital, Mel- bourne (2002/23); Southern Health, Melbourne (2002-099B); Angliss Hos-

pital, Melbourne (2002), Royal Children’s Hospital, Melbourne (27056A). This research was supported by project grants from the Australian National Health and Medical Research Council and from the Victorian Govern- ment’s Operational Infrastructure Support Program. Stephanie Brown was supported by a VicHealth Public Health Research Fellowship, a NHMRC Career Development Award and Senior Research Fellowship, and an ARC Future Fellowship; and Rebecca Giallo and Fiona Mensah were supported by a NHMRC Career Development Fellowship. We are extremely grateful to the women taking part in the study, to members of the Maternal Health Study Collaborative Group, and to members of the Maternal Health Study research team who have contributed to data collection and coding. The ideas expressed and data appearing in this article have not been previously disseminated.

Correspondence concerning this article should be addressed to Rebecca Giallo, Healthy Mothers Healthy Families Research Group, Murdoch Children’s Research Institute, 50 Flemington Road, Parkville, VIC 3052, Australia. E-mail: [email protected]

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

Journal of Family Psychology © 2019 American Psychological Association 2020, Vol. 34, No. 1, 112–121 ISSN: 0893-3200 http://dx.doi.org/10.1037/fam0000623

112

Childhood physical and/or sexual abuse is associated with in- creased risks of poor physical and psychosocial health across the life span (Paolucci, Genuis, & Violato, 2001; Wegman & Stetler, 2009) and across generations (McDonnell & Valentino, 2016). A better understanding of how the adverse effects of maternal child- hood abuse are transmitted to the next generation is critical to informing prevention and early intervention approaches to break intergenerational cycles of poor health. Drawing upon Buss and colleagues’ (2017) transdisciplinary model of the intergenerational transmission of maternal childhood maltreatment exposure, the present study investigated adverse birth outcomes (preterm birth, low birth weight) and psychosocial pathways (maternal mental health, mother exposure to intimate partner violence in first post- natal year) as potential mechanisms underlying the association between maternal childhood abuse and children’s emotional– behavioral functioning at 10 years.

Prevalence and Potential Health Consequences of Childhood Abuse Among Women and Their Children

Between 15– 40% of women in high income countries such as Canada (MacMillan et al., 1997), the United States (Centers for Disease Control and Prevention (CDCP), 2010), and Australia (Gartland et al., 2016) have experienced childhood physical, emo- tional, and/or sexual abuse. Child abuse often co-occurs with other experiences such as neglect, parental mental health issues, sub- stance use problems, family conflict, and violence (CDCP, 2010), increasing risks for poor health in adulthood including headaches, gastrointestinal symptoms and cardiovascular disease (Wegman & Stetler, 2009), mental health problems such as depression, post- traumatic stress symptoms, substance use and self-harm (Paolucci et al., 2001; Wegman & Stetler, 2009), and the experience and/or use of interpersonal violence (Widom, Czaja, & Dutton, 2008).

There is a growing body of evidence also pointing to potential intergenerational effects of maternal childhood abuse on children’s health and development (Bosquet Enlow, Englund, & Egeland, 2018; Collishaw, Dunn, O’Connor, & Golding, 2007; Madigan, Wade, Plamondon, & Jenkins, 2015; Racine, Plamondon, Madi- gan, McDonald, & Tough, 2018; Roberts, O’Connor, Dunn, Gold- ing, & The ALSPAC team, 2004). For example, in a U.K. population-based study (n � 5,619), maternal childhood abuse was associated with children’s emotional-behavioral difficulties at ages 4 and 7 years (Collishaw et al., 2007). In a more recent study of 490 mother-child dyads, maternal childhood physical abuse (but not sexual abuse) was associated with children’s internalizing difficulties at age 3 (Madigan et al., 2015). Given this evidence, there is increasing focus on understanding how the effects of maternal childhood abuse are transmitted to the next generation.

Understanding the Potential Intergenerational Pathways of Maternal Childhood Abuse on Children

A model of the intergenerational transmission of maternal child- hood maltreatment exposure has been proposed by Buss and colleagues (2017). Described as a transdisciplinary model, it draws upon theory and research into (a) maternal neurobiological re- sponses to stress across the life span to preconception, (b) the fetal or developmental origins of health and disease, and (c) the psy- chosocial sequelae of the postnatal environment. The first mech-

anism recognizes that child maltreatment can result in dysregula- tion of the hypothalamic-pituitary-adrenal (HPA) axis, which can threaten maternal health outcomes, and may also lead to epigenetic alterations that are transmitted during the conception process. These changes may predispose offspring to a range of health problems in childhood and across the life span. This mechanism was not investigated in the current paper.

The second mechanism focuses on the prenatal endocrine and immune/inflammatory environment, highlighting that maternal bi- ological and psychosocial stress in utero as well as poor maternal health (i.e., diabetes, hypertension, chronic renal disease, mental health problems; Racine, Madigan, Plamondon, McDonald, & Tough, 2018) may compromise fetal neurodevelopment. They can also increase risks for adverse birth outcomes, which can have long-term consequences for children’s neurodevelopment (Saigal & Doyle, 2008). Although a systematic review of six studies revealed mixed findings (Wosu, Gelaye, & Williams, 2015), half of the included studies found that mothers who had experienced childhood sexual abuse had three to five times the odds of preterm birth compared to women with no abuse history. In a more recent study of 398 women and children in the US, maternal childhood maltreatment was associated with lower birth weight, which in turn was associated with higher infant social– emotional problems at 6 months old (McDonnell & Valentino, 2016). Further research is still needed to better understand the mediating role of adverse birth outcomes.

The final mechanism focuses on the influence of the postnatal environment and the psychosocial sequelae often experienced by women with a history of childhood abuse on children’s outcomes. For example, in a large population-based study of 8,292 families in the U.K., childhood sexual abuse was associated with increased maternal mental health problems and parenting difficulties, which in turn were associated with children’s emotional– behavioral dif- ficulties (Roberts et al., 2004). In another study, childhood phys- ical abuse was associated with postnatal depressive symptoms, which were in turn associated with parenting difficulties and children’s internalizing difficulties at age 3 (Madigan et al., 2015). Maternal mental health in the first postnatal year was also of particular interest in the current study, along with women’s expe- rience of intimate partner abuse, which has also been associated with poor outcomes for children (Bair-Merritt, Blackstone, & Feudtner, 2006; Holt, Buckley, & Whelan, 2008).

Approximately one in four women experience violence in an intimate relationship in the early years after having a baby (Gart- land et al., 2016). This can include physical violence (i.e., beating, pushing, choking), emotional or psychological abuse (i.e., humil- iation, intimidation, and threats), and sexual violence (Hegarty, Bush, & Sheehan, 2005). Although research into childhood abuse as a risk factor for IPV has been mixed (Stith, Smith, Penn, Ward, & Tritt, 2004), some studies have shown that child maltreatment (including physical and sexual abuse) is associated with increased risk of IPV in adulthood (Gartland et al., 2016; Widom et al., 2008). It has been posited that child abuse can disrupt children’s attachments to their caregivers, leading to insecure and disorga- nized attachment styles and emotional regulation difficulties (Cyr, Euser, Bakermans-Kranenburg, & Van Ijzendoorn, 2010). Such difficulties may predispose women (and men) to difficulties in forming and maintaining healthy adult relationships, making them vulnerable to conflict and violence (Madigan, Bakermans-

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

113MATERNAL CHILDHOOD ABUSE

Kranenburg, Van Ijzendoorn, Pederson, & Benoit, 2006). Al- though children’s exposure to IPV is associated with adverse health outcomes (Bair-Merritt et al., 2006; Holt et al., 2008), few studies have explored IPV as a potential mechanism underlying the associations between maternal childhood abuse and children’s outcomes.

Recognizing the complex interplay between physical and psy- chosocial health, two recent studies have explored these mecha- nisms simultaneously in mediating models (Madigan, Wade, Pla- mondon, Maguire, & Jenkins, 2017; Racine et al., 2018). In a US study (n � 510), maternal health (e.g., gestational diabetes, hy- pertension) and infant birth outcomes (e.g., low birth weight and preterm birth) mediated the association between maternal adverse childhood experiences and infant physical health, but not emo- tional health, at 18 months old. In contrast, psychosocial factors (e.g., maternal depression, interparental conflict, and single par- enthood) mediated the relationship with infant emotional problems but not physical health problems (Madigan et al., 2017). In con- trast, a Canadian study (n � 1,994) reported that both maternal physical health (e.g., diabetes, heart disease, hypertension) and psychosocial factors (mental health and hostile parenting) medi- ated the associations between maternal adverse childhood experi- ences and infant gross and fine motor skills, problem-solving, and personal–social domains at age one (Racine et al., 2018). These studies underscore the value of investigating multiple mechanisms simultaneously, reflecting the complexity of women’s and chil- dren’s lives. They also highlight that there may be different path- ways of effect depending upon the children’s health outcomes of interest.

Aims of the Current Study

To address critical gaps in our understanding about how mater- nal childhood abuse exposure can affect children in the next generation, the current study investigated two potential mecha- nisms embedded within Buss and colleagues’ (2017) model. Draw- ing upon data from a large Australian prospective pregnancy cohort study of women and their first-born children, we investi- gated whether adverse birth outcomes (low birth weight and pre- term birth) and exposure to maternal depressive symptoms and IPV in the first postnatal year mediated the relationship between maternal childhood abuse and children’s emotional-behavioral functioning at 10 years. We focused on psychosocial pathways during the first year of children’s lives as this is a time of signif- icant infant neurodevelopment. It is also a time when women have a high level of contact with health services for themselves and their children, presenting opportunities for support and intervention early in the period of becoming a parent.

The second aim of the study was to assess whether the strength of the relationships between maternal childhood abuse, birth out- comes, maternal depressive symptoms and IPV, and children’s outcomes differed for boys and girls. Given some evidence that boys can be more sensitive to adversity in the family environment (Mensah & Kiernan, 2010), and that boys exposed to IPV are more likely to experience behavioral difficulties than girls (Evans, Da- vies, & DiLillo, 2008), we hypothesized that the strength of the model relationships would be stronger for boys than girls. A more nuanced approach to understanding how these mechanisms operate

for boys and girls may inform the monitoring and tailoring of interventions for families at risk.

Method

Study Design

Data were drawn from the Maternal Health Study, a prospective longitudinal study of women’s health during pregnancy and after childbirth, with follow-ups of their children’s health at 4 and 10 years of age. Research ethics committee approvals for the Maternal Health Study were obtained from La Trobe University (2002/38), Royal Women’s Hospital (2002/23), Southern Health (2002-099B), Angliss Hospital, and the Royal Children’s Hospital (27056A, 27056B, 33127A/B, 34058A-D, 36189A/B). Study design and sam- pling are detailed in the study protocol (Brown, Lumley, McDon- ald, Krastev, & the Maternal Health Study collaborative group, 2006). Briefly, women registered to give birth at six public hos- pitals in metropolitan Melbourne, Australia between April 2003 and December 2005 were invited to participate. Eligibility criteria were: (a) 18 years or older, (b) nulliparity, (c) gestation of up to 24 weeks at time of enrolment, and (d) proficiency in English to complete written questionnaires. Approximately 6,000 invitation packages were distributed to women attending the six study hos- pitals for maternity care, shortly after their first (booking) visit between 10 and 20 weeks’ gestation. Invitation packs were dis- tributed to women during clinic visits and at antenatal education classes and were mailed to women at their home address. It is not possible to determine precisely how many women received more than one invitation or how many ineligible women received an invitation. We conservatively estimate that around a third of eli- gible women enrolled in the study. Women completed self- administered questionnaires at 10 –24 weeks’ gestation; at 3, 6, 12 and 18 months postpartum; and at 4 and 10 years postpartum.

Participants

There were 1,507 women who met the eligibility criteria and returned a baseline questionnaire. Mothers’ demographic charac- teristics at enrolment were compared with routinely collected Victorian data for nulliparous women giving birth as public pa- tients in Victoria during the recruitment period (see Table 1). The majority of women were aged between 30 and 34 years, born in Australia, university educated, in paid employment, and married or living with a partner.

The number of surveys returned at the 3, 6, 12, 18 months and 4 and 10 year follow-ups were 1,431, 1,400, 1,357, 1,327, 1,102, and 950, respectively. Selective attrition was observed. Using data collected in early pregnancy or in the first year postpartum, women who did not participate in the 10-year follow-up were significantly more likely to be: of non-English speaking background, younger at birth of first baby, in the lowest income brackets, reporting sec- ondary school education or less, and having reported depression and/or IPV in the first year postpartum.

Measures

Maternal childhood abuse. Maternal childhood abuse was retrospectively reported in the 4-year follow-up questionnaire us-

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

114 GIALLO ET AL.

ing the Child Maltreatment History Self Report (MacMillan, Tanaka, Duku, Vaillancourt, & Boyle, 2013). Women were asked six questions about how often an adult may have physically abused them (e.g., “Pushed, grabbed or shoved you”, “Kicked, bit or punched you”, “Physically attacked you in some way”) when growing up on a 3-point scale ranging from 0 � Never to 3 � Often. Women were also asked four questions about whether an adult or someone older than them had performed sexual acts (e.g., “Touched sex parts of your body”, “Tried to have sex with you or sexually attacked you”) when growing up. Scores for the physical and sexual abuse items were summed to create subscale scores. Physical and sexual abuse were defined as a positive response to any of the items. The scale has excellent reliability and validity (MacMillan et al., 2013). Cronbach’s alpha for the physical and sexual abuse subscales was .85 and .83, respectively.

Children’s emotional-behavioral functioning. Children’s emotional– behavioral functioning at 10 years of age was assessed using the parent-report of the Strengths and Difficulties Question- naire (SDQ; Goodman, 2001). The SDQ is comprised of 25 items assessing a range of emotional and behavioral symptoms such as sadness, worries, attention difficulties, fighting, and problems with peers on a 3-point scale (0 � Not True to 2 � Certainly True). There are 4 subscales representing emotional symptoms, conduct problems, hyperactivity/inattention, and peer problems. There is also a Total Difficulties score where higher scores indicate more emotional and behavioral difficulties. Cut-off scores for the “nor- mal” (0 –13), “at risk/borderline” (14 –16), and “clinical” (17�) ranges based on Australian norms are available. The categorical variable based on the Total Scale cut-off scores was used to describe the proportion of the children exposed and not exposed to IPV experiencing clinically significant emotional– behavioral dif-

ficulties. The continuous scores for the subscales were used in the modeling. The SDQ has had extensive psychometric evaluation with Australian samples, with moderate to strong internal consis- tency and test–retest reliability reported (Hawes & Dadds, 2004). Cronbach’s alpha for the Total Difficulties scale for the current sample was .79.

Maternal depressive symptoms. Maternal depressive symp- toms were assessed using the 10-item self-report Edinburgh Post- natal Depression Scale (EPDS; Murray & Cox, 1990) at 12 months postpartum. Women reported on the extent to which they had experienced a range of symptoms such as depressed mood, feeling anxious, crying, and thoughts of self-harm in the previous week on a 4-point scale. A cut-off of �13 is recommended when screening for probable major depression in community samples (Murray & Cox, 1990). Cronbach’s alpha for the current sample was .88.

Intimate partner violence. Intimate partner violence (12 month period prevalence) was assessed using the short 18 item version of the Composite Abuse Scale (CAS; Hegarty, Bush, & Sheehan, 2005). The 18 items measure how often emotional (e.g., “My partner told me I was crazy”) and physical violence (e.g., “My partner threw me”) from an intimate partner occurred in the last 12 months, rated on a six-point scale ranging from 0 � Never to 5 � Daily. Women were identified as experiencing physical IPV if they scored 1 or more on any of the physical violence items, and emotional IPV if they scored three or more on the emotional violence items. Any IPV (physical and/or emotional) was coded as 1 � Yes, and no abuse as 0 � No. The CAS has excellent reliability and is well-validated with clinical and general popula- tions (Hegarty et al., 2005). Cronbach’s alpha for the current sample was .73.

Table 1 Characteristics of Study Participants at Enrollment and Nulliparous Women Giving Birth in Victoria as Public Patients During the Study Recruitment Period

Participants Victorian Patientsa

(N � 1,507) (N � 40,905) Characteristics n (%) n (%)

Maternal age (in years) �24 years 190 (12.6) 12,216 (29.9) �25 years 1246 (82.4) 28,679 (70.1) Not reported 76 (5.0)

Maternal country of birth Australia 1,074 (71.3) 29,791 (73.3) Overseas—English speaking background 138 (9.7) 2330 (5.7) Overseas—Non-English speaking background 212 (14.1) 8,517 (21.0) Not reported 83 (5.5)

Relationship status Married 876 (58.1) 22,790 (55.9) Living with partner 490 (32.5) n/ab

Divorced/separated/not partnered 65 (4.3) n/ab

Not reported 76 (5.0) Highest educational attainment n/ab

Post-high school qualification 1,036 (68.7) High school completion (Year 12) or less 387 (25.7) Not reported 84 (5.6)

In paid employment 1,172 (77.8) n/ab

Low income (recipient of health care card) 366 (24.3) n/ab

a Victorian Perinatal Data Collection Unit (was provided upon request). b n/a indicates that this information was not routinely collected by Victorian hospitals and therefore not available.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

115MATERNAL CHILDHOOD ABUSE

Birth outcomes. Birth outcomes including number of weeks’ gestation and infant birth weight were reported by mothers. Pre- term birth was defined as being born less than 37 weeks gestation, and low birth weight was defined as being born less than 2500 g. A composite variable was derived where either preterm birth and/or low birth weight were identified as an adverse birth out- come.

Demographic information. Demographic information re- ported in this paper included maternal age, country of birth, education, employment status, and household income, along with child age and gender.

Data Analysis

Exploratory data analysis, descriptive statistics, and correlations using Pearson’s r(phi for pairs of dichotomous variables; point- biserial for pairs of continuous and dichotomous variables) were conducted in SPSS V22. Logistic regression analysis was also conducted to assess the relationships between maternal childhood abuse and children’s emotional– behavioral difficulties using the clinical cut points on the SDQ Total Difficulties scale.

A series of models: (a) direct effects model of maternal child- hood abuse on children’s emotional– behavioral functioning, (b) a partially mediating model where both the direct effect of maternal child abuse on children’s emotional– behavioral functioning (latent construct with the SDQ subscales for emotional, conduct, hyper- activity/inattention, and peer problems as indicators) and the indi- rect effects via the intervening variables (adverse birth outcomes, maternal depressive symptoms and IPV) were estimated, and (c) the hypothesized model where the effect of maternal child abuse on children’s emotional– behavioral functioning is fully mediated by the intervening variables (adverse birth outcomes, maternal depressive symptoms and IPV). Each model was adjusted for maternal age and education, and the association between the in- tervening variables was accounted for by correlating the error terms. Please note that the covariates and correlations have not been drawn in the figures for ease of interpretation.

The models were estimated in Mplus V7.4 (Muthen & Muthen, 1998-2011) using both maximum likelihood estimation with ro- bust standard errors (MLR) and maximum likelihood estimation (ML) with bootstrapping to obtain bootstrap confidence intervals. The results yielded similar estimates and only the ML results are presented. Model fit of the hypothesized model was assessed using a range of model fit statistics (Satorra-Bentler chi-square test, Tucker Lewis Index (TLI), Comparative Fit Index (CFI), and Root Mean Square Error Approximation (RMSEA)). While a nonsig- nificant chi-square indicates good model fit, this is generally unexpected for large samples. Therefore, given our sample size, greater attention was given to other fit indices. TLI and CFI should exceed .90 for an acceptable fit, and values closer to or below .05 for the RMSEA are acceptable. Missing data in all models were managed using Full Information Maximum Likelihood (FIML).

To assess for evidence of mediation, we assessed for changes in the magnitude and significance of the direct path between maternal childhood abuse and children’s emotional– behavioral functioning after the introduction of the intervening variables (adverse birth outcomes, IPV, maternal depression). If the direct path decreases and becomes nonsignificant, this provides evidence of full medi- ation, but if the direct path decreases and remains significant, this

provides evidence of partial mediation. We also assessed the specific indirect pathways between maternal childhood abuse and children’s emotional– behavioral functioning via each of the inter- vening variables. The indirect effects were estimated using the product of coefficients approach with bootstrapping to obtain bootstrap confidence intervals in MPlus (MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). Evidence of mediation (i.e., an intervening variable transmits the effect of an independent variable to a dependent variable) was provided if the confidence interval for the estimate for the indirect effect did not cross zero and was significant.

Finally, multigroup analyses were conducted to assess for mod- eration by child gender. This involved testing a model with all parameters freely estimated (unconstrained model) and comparing it to a model where the path estimates for relationships between the model variables were constrained to be equal (constrained model) among boys and girls using the chi-square difference test. As above, MLR and ML with bootstrapping to obtain bootstrap con- fidence intervals were used to estimate the models. They yielded similar results, and the ML results are presented.

Results

Descriptive Statistics and Missing Data

Missing data was approximately 20% across all variables, and highest for the SDQ Total Difficulties at 10 years (�30%), which is to be expected given attrition associated with an 11-year cohort study. These missing data were managed using FIML when esti- mating the models. Descriptive statistics for the key model vari- ables are presented in Table 2. Statistical (see Table 2) and graphical measures of normality indicated that the maternal de-

Table 2 Descriptive Statistics for the Study Variables

Study variable Statistic

Children’s emotional-behavioral difficulties (10yrs) Range 0–28 M (SD) 6.85 (5.39) Skewness 1.13 At risk/Borderline and Clinical range, n (%)a 108 (7.2)

Maternal depressive symptoms (1yr pp) Range 0–27 M (SD) 4.81 (4.74) Skewness 1.23 Clinical range, n (%)b 109 (7.2)

Maternal childhood sexual abuse, n (%)c 241 (16.0) Maternal childhood physical abuse, n (%)c 294 (19.5) Any maternal childhood abuse, n (%) 428 (28.4) Preterm birth (�37 weeks gestation), n (%) 96 (6.4) Low birthweight (�2500g), n (%) 80 (5.3) Adverse birth outcomes (pre-term birth and/or low

birthweight), n (%) 119 (7.9) Intimate partner violence (first 12mths pp), n (%)d 234 (15.5)

Note. pp � postpartum; sample size ranged from N � 877–1325 due to missing data. a Strengths and Difficulties Questionnaire Total Difficulties � 14�. b Edinburgh Postnatal Depression Scale Clinical Range � 13�. c Child Maltreatment History Self-report. d Composite Abuse Scale.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

116 GIALLO ET AL.

pression scores on the EPDS and children’s emotional– behavioral scores on the SDQ Total Difficulties scale were positively skewed. Therefore, maximum likelihood estimation with robust standard errors was used to estimate the models.

Table 3 presents the Pearson’s r correlations among all key analysis variables in the hypothesized model and potential medi- ator variables. While most associations between the model vari- ables were significant, adverse birth outcomes were not associated with the independent (maternal childhood abuse) or dependent (children’s emotional– behavioral difficulties) variables. Despite this, we proceeded with testing our theoretically/conceptually driven research aims and hypotheses.

Maternal Childhood Abuse and Children’s Emotional- Behavioral Difficulties

Table 4 reveals that compared to children of mothers who had not experienced childhood abuse, children of mothers who had experi- enced abuse had 1.6 times higher odds of emotional– behavioral difficulties in the At-risk/Borderline and Clinical ranges.

Testing the Intergenerational Pathways Models

Prior to testing the hypothesized model, the latent construct of children’s emotional– behavioral difficulties represented by the observed SDQ subscales was tested and found to be a good fit to the data, �2(2, N � 1507) � 16.29 p � .001; CFI � .98; TLI � .94; RMSEA � .09 (90% CI [.05, .13]); and SRMR � .02. This was used in subsequent modeling. The direct effects model of the relationship between maternal childhood abuse and the latent construct of children’s emotional– behavioral functioning was es- timated. The model was a reasonable fit to the data, �2(11, N � 1507) � 32.17, p � .001; CFI � .97; TLI � .95; RMSEA � .04 (90% CI [.02, .05]); and SRMR � .03. The standardized estimate for the direct path was significant (.12, p � .003). This model accounted for 3% of the variance in children’s emotional– behavioral functioning, (R2 � .03, p � .074). Next, the partially mediated model including the mediators (IPV, depressive symp- toms, adverse birth outcomes) was tested and found to be a

reasonable fit to the data, �2(20, N � 1507) � 48.34, p � .001; CFI � .97; TLI � .94; RMSEA � .03 (90% CI [.02, .04]); and SRMR � .02. The standardized estimate between children’s emotional– behavioral difficulties and maternal childhood abuse decreased but remained statistically significant (.09, p � .031), providing evidence of partial mediation by one or more of the intervening variables. This model accounted for 6% of the variance in children’s emotional– behavioral functioning, (R2 � .06, p � .006). Finally, the hypothesized fully mediated model (without the direct effect of maternal childhood abuse on children’s emotional– behavioral difficulties) was tested and found to be a reasonable fit to the data, �2(21, N � 1507) � 53.33 p � .001; CFI � .97; TLI � .93; RMSEA � .03 (90% CI [.02, .04]); and SRMR � .03. This model also accounted for 5% of the variance in children’s emotional– behavioral functioning, (R2 � .05, p � .013). A chi- square difference test revealed that the fully mediated model was a better fit to the data than the partially mediated model, �diff2 (1, N � 1507) � 6.99, p � .008, and therefore was accepted as the final model.

The standardized parameter estimates for the mediated model are presented in Figure 1. Maternal childhood abuse was signifi- cantly associated with IPV and higher maternal depressive symp-

Table 3 Correlations Among All Model Variables

Study variables 1 2 3 4 5 6 7 8 5 6 7

1. SDQ Total Scale scorea — 2. SDQ emotional problems .69��� — 3. SDQ conduct problems .73��� .34��� — 4. SDQ hyperactivity/inattention .81��� .30��� .53��� — 5. SDQ peer problems .69��� .35��� .41��� .39��� — 6. Maternal childhood abuse (0 � no; 1 � yes)c .11�� .06 .08� .10�� .08�� — 7. Adverse birth outcomes (1 � no; 2 � yes) �.01 �.02 �.06 .03 .01 .04 — 8. Intimate partner violence (first 1yr pp) (1 � no; 2 � yes)d .12��� .04 .12��� .11�� .09� .14��� .05 — 9. Maternal depressive symptoms (1yr pp)b .15��� .13��� .11�� .11�� .07� .14��� .04 .29��� —

10. Child gender (1 � girl; 2 � boy) .12��� �.08� .11�� .22��� .09�� �.03 .02 .02 �.001 — 11. Maternal age at first birth (1 � 24yrs; 2 � 25yrs) �.03 .05 �.08� �.05 �.004 �.03 .01 �.11��� �.06� .01 — 12. Maternal education (1 � tertiary, 2 � high school or less) .06 �.02 .06 .06 .07 .06 .05 .09�� .02 .04 �.23���

Note. SDQ � Strengths and Difficulties Questionnaire; pp � postpartum; sample size ranged from N � 877–1325 due to missing data. a Strengths and Difficulties Questionnaire Total Difficulties. b Edinburgh Postnatal Depression Scale. c Child Maltreatment History Self- report. d Composite Abuse Scale. � p � .05. �� p � .01. ��� p � .001.

Table 4 Logistic Regression Analyses Estimating the Relationship Between Maternal Childhood Abuse and Child Emotional- Behavioral Difficulties at 10 Years (N � 877)

SDQ Total Difficulties Clinical Rangea

Maternal childhood

abuseb Normal Borderline/Clinical n (%) n (%) OR [95% CI], p

No 497 (90.7) 51 (9.3) Ref. Yes 282 (85.7) 47 (14.3) 1.62 [1.07, 2.48], .024

Note. SDQ � Strengths and Difficulties Questionnaire. a Score of 14 � reflects classification into the Borderline/Clinical Range on the SDQ Total Difficulties Scale. b Child Maltreatment History Self- report.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

117MATERNAL CHILDHOOD ABUSE

toms at 12 months postpartum, but not with adverse birth out- comes. Maternal childhood abuse, IPV, and higher maternal depressive symptoms were associated with higher children’s emotional– behavioral difficulties at 10 years, while adverse birth outcomes was not.

The indirect effects of maternal childhood abuse on children’s emotional– behavioral difficulties were also assessed. The total indirect effect via the intervening variables was also significant, standardized estimate � .04 (95% CI [.02, .06]), p � .001. The specific indirect effects are provided in Table 5. The indirect pathways via maternal depressive symptoms and IPV were signif- icant, providing evidence of mediation. There was no evidence of mediation via adverse birth outcomes given the nonsignificant indirect effects.

Testing for Moderating Effects by Child Gender

Multigroup analysis was conducted to assess moderation by child gender. The unconstrained model where all parameters were freely estimated among boys and girls (�2(49, N � 1507) � 139.97) was compared to the constrained model where the path estimates between the model variables were constrained to be equal among boys and girls (�2(55, N � 1507) � 145.30). The chi-square difference test was not significant (�diff2 (6, N � 1507) � 5.33, p � .502), providing little evidence of moderation by child gender.

Discussion

Findings from this population-based study of first-time Austra- lian mothers provide support for the intergenerational association between maternal childhood abuse and their children’s emotional– behavioral difficulties. Children of mothers with a history of childhood abuse had higher odds of experiencing emotional– behavioral difficulties than children of mothers who had not ex- perienced childhood abuse. One in four mothers had a history of childhood abuse into parenthood, heightening their risk of poor mental health and exposure to IPV.

Our findings also provided support for several mechanisms proposed in Buss and colleagues’ model (2017) of the intergen- erational transmission of maternal childhood maltreatment expo- sure to children. In addition to the direct risk of maternal childhood abuse to children’s emotional– behavioral functioning, we found evidence for mediation via the psychosocial health pathways of exposure to IPV and maternal depressive symptoms in the first year postpartum but not for adverse birth outcomes (low birth weight, preterm birth). Although the model accounted for a small proportion of the variance in the outcome, this highlights how the sequalae of psychosocial issues for women who have experienced childhood abuse can contribute to their children’s emotional– behavioral difficulties. There are several ways in which the psy- chosocial pathways may work to influence children’s mental health. For example, exposure to IPV and maternal mental health

Figure 1. Standardized parameter estimates for the mediated model of the relationship between maternal childhood abuse and children’s emotional-behavioral difficulties at age 10 years. Covariates (maternal age and education) and correlated error terms for the mediator variables are not shown; dotted lines indicate nonsignif- icant pathways. �� p � .01. ��� p � .001.

Table 5 Standardized Estimates for Specific Indirect Pathways for the Overall Sample and by Child Gender (N � 1,507)

Overall sample

Specific direct and indirect pathways Standardized estimate

[95% CI] p

Maternal childhood abuse ¡ Adverse birth outcomes ¡ Children’s emotional-behavioral difficulties �.001 [�.006, .002] .513 Maternal childhood abuse ¡ Intimate partner abuse ¡ Children’s emotional-behavioral difficulties .02 [.004, .03] .025 Maternal childhood abuse ¡ Maternal depressive symptoms ¡ Children’s emotional-behavioral difficulties .02 [.01, .04] .015

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

118 GIALLO ET AL.

problems in the first years of life can contribute to marked stress and physiological arousal among infants (Levendosky et al., 2016). When the biological systems responsible for regulating stress (i.e., HPA axis) are overactivated and become dysregulated, the devel- opment of the neural systems responsible for emotional regulation may be threatened (De Bellis & Zisk, 2014). Therefore, these children may find it harder to manage strong emotions and their behavior, particularly in stressful situations.

Although parenting behavior was not investigated in the current study, these findings can also be considered within the broader context of parent– child relationships. In a meta-analysis of 55 studies, childhood maltreatment was found to be associated with insecure and disorganized attachment styles (Cyr et al., 2010). These styles can lead to challenges in adulthood, including mental health difficulties, problems managing stress, and parenting diffi- culties (Madigan et al., 2006). Mothers who have significant trauma histories of abuse, IPV, and poor mental health may find it harder to be sensitive and responsive to their children’s emotional experiences (Howell, Graham-Bermann, Czyz, & Lilly, 2010). Mothers who find it difficult to regulate their own emotions may find it hard to role-model effective coping strategies to their children. Furthermore, compared to women who have not experi- enced childhood abuse, women with a history of childhood sexual abuse are more likely to use harsh parenting practices (DiLillo & Damashek, 2003) and have anxious, overprotective attachment styles (Kwako, Noll, Putnam, & Trickett, 2010), which are both risk factors for children’s emotional– behavioral difficulties.

Our study contributes to the research effort focused on biomed- ical impacts on children’s outcomes, where there is still inconclu- sive evidence. We found little evidence for adverse birth outcomes underlying the associations between maternal childhood abuse and children’s emotional-behavioral difficulties, which is consistent with Madigan et al. (2017), who found that infant birth outcomes (e.g., low birth weight, preterm birth, loss of fetal movements during pregnancy) did not mediate the association between mater- nal childhood abuse and children’s emotional health at 18 months old. These findings are reassuring and positive for children who have had a difficult start to early life.

The final aim of the study was to assess for model differences by child gender. No evidence of moderation was found, suggesting that the pathways from maternal childhood abuse to children’s emotional– behavioral difficulties via IPV and maternal depression are similar for boys and girls. This is somewhat in contrast with research indicating that IPV may differentially affect boys and girls, with stronger negative impacts on externalizing behavior for boys (Evans et al., 2008). In addition to understanding risks for boys and girls, another area for future research is to identify and deepen our understanding of whether there are differences in protective factors or mechanisms by child gender.

Study Strengths, Limitations and Future Research Directions

The availability of intergenerational data drawn from a preg- nancy cohort of more than 1,500 women and their children was a key strength of the study. Our conceptually driven approach and sophisticated longitudinal analyses enabled the simultaneous mod- eling of three potential mechanisms by which maternal childhood abuse can have a long-term impact on children. A well validated

multidimensional measure of IPV was used, and a nuanced inves- tigation into the effects of maternal childhood abuse and potential mechanisms on boys and girls was conducted.

There are also several limitations to note. The majority of women were Australian-born, over 25 years of age, had a tertiary education, were in paid employment, and were partnered. More- over, selective attrition was observed among younger women, those with lower educational attainment, and those with higher levels of depressive symptoms during pregnancy, which may limit generalizability of the findings to women who were underrepre- sented in this study. It is noted, however, that while low initial response fractions and selective attrition are likely to result in biased prevalence estimates, measures of association are unlikely to be affected (Mealing et al., 2010).

All measures were self-reported by mothers, and it is possible that reporter bias could have influenced the associations be- tween the study variables. Mothers reported on their children’s emotional– behavioral functioning, and ratings may have been influenced by mental health difficulties or heightened concern for their children who may have been exposed to IPV. The retrospec- tive assessment of maternal childhood abuse was limited to in- stances of physical and sexual abuse by an adult, which did not capture details about the relationship of the perpetrator to the mother or the frequency and duration of the abuse. The items pertaining to sexual abuse also reflected a broad range of sexual abuse experiences, from potentially protracted experiences of abuse within the child’s immediate caregiving environment to infrequent or isolated experiences outside the caregiving context. It is possible that associations with maternal depression, IPV, and children’s emotional– behavioral functioning may be stronger for repeated experiences of childhood sexual abuse by a family mem- ber or family friend than for isolated experiences outside the family environment. We also did not ask about other forms of childhood maltreatment such as emotional abuse or neglect. Fur- thermore, we only asked about women’s experience of IPV toward them and did not ask about other forms of violence that children might be exposed to, such as the women’s use of violent or abusive behaviors toward their partner or child maltreatment by a parent.

Finally, it is important to acknowledge alternative models ex- ploring different pathways as areas for future research. For in- stance, we did not examine other aspects of maternal mental health such as anxiety or posttraumatic stress symptoms that are common among women who have experienced abuse and trauma in child- hood (Paolucci et al., 2001). Similarly, other aspects of child development and functioning might be affected by maternal his- tory of childhood abuse, including physical health (i.e., asthma, allergy), language, and cognitive functioning. For example, Madi- gan et al. (2017) found that adverse birth outcomes (e.g., low birth weight, preterm birth, loss of fetal movements during pregnancy) did mediate the association between maternal childhood abuse and children’s physical health. Therefore, a key area for future research is to investigate poor birth outcomes as mechanisms underlying the association between maternal childhood abuse and children’s physical health. It is also worth noting that the prevalence of adverse birth outcomes in our study was �10% and may have lacked power to detect small associations between the variables. Another area for research is to investigate pathways via an accu- mulation of physical health risks for both mothers (e.g., high and low BMI, gestational diabetes, hypertension) and children (e.g.,

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

119MATERNAL CHILDHOOD ABUSE

very preterm birth, small for gestational age, low birth weight). The impacts of maternal childhood abuse on children’s emotional– behavioral functioning may be greater for those infants and moth- ers with a greater accumulation or specific pattern of health risk factors during pregnancy and at time of birth. Moreover, there are likely to be other pathways involved in the processes of intergen- erational transmission that were not investigated in this study. For example, genetics and epigenetic mechanisms and pathways via maternal physical health and poor health behaviors (i.e., smoking, substance use, poor nutrition). It is also likely that there are protective mechanisms that may disrupt the intergenerational transmission processes such as access to professional and social support, positive parenting behaviors, and children’s engagement in school.

Implications and Conclusions

Improving maternal and child health is a key priority of the United Nation’s Sustainable Development Goals. Childhood abuse is a serious threat to the health and wellbeing of women and their children across the life course. Importantly, our findings show that a history of childhood abuse is associated with maternal depressive symptoms and IPV in the first year after having a baby, and these are associated with emotional– behavioral difficulties for children at 10 years. This underscores the importance of early intervention to disrupt the potential for intergenerational impacts and cycles of abuse and poor health within families.

Pregnancy and the early years of parenting when women have a high level of contact with maternal health and early childhood services provide multiple and sustained opportunities for early identification and intervention to disrupt cycles of intergenera- tional trauma. For some women, this might be the first time that they have ever been asked about experiences of childhood abuse, encouraged to share their experiences, and explore options for support. For women who have accessed professional support in the past, the transition to parenthood might be another pivotal time in their lives when accessing support could be beneficial. Access to intervention and support can provide opportunities to promote the health of women in the early years of their children’s lives, as well as identify and support women experiencing relationships difficul- ties, partner conflict, and IPV.

Given that the psychosocial health pathways of maternal child- hood abuse have the potential to affect children, efforts to promote children’s health and wellbeing are critical. Maternal and early childhood services are in an ideal position to facilitate timely referral pathways for children showing early signs of emotional and behavioral regulation difficulties. Building the capacity of mothers and other caregivers to recognize and respond to these early signs of emotional– behavioral dysregulation and difficulties is also important. Several studies have shown that children who are exposed to IPV are more likely to display positive emotional– behavioral functioning when their mother or another caregiver is able to provide emotion coaching to help children manage distress, regulate emotions, and use positive coping skills (Manning, Da- vies, & Cicchetti, 2014). The effectiveness of mother– child dyadic interventions such as Child-Parent Psychotherapy (CPP) and Fo- cused Cognitive Behavioral Therapy to support healing and recov- ery and promote maternal and child health has been demonstrated

(Cohen, Mannarino, & Iyengar, 2011; Lieberman, Ghosh Ippen, & Van Horn, 2006).

Finally, our results have implications for all professionals work- ing with children in the middle years of childhood. Intergenera- tional trauma is common, and many children presenting with emotional– behavioral difficulties in schools, hospitals, and pri- mary care settings may have been exposed to a range of psycho- social stressors in their caregiving environment, including parental depression, IPV, and other adversities. The earliest possible iden- tification and support in these settings has the potential to change the trajectories of children and families who have been affected by intergenerational cycles of abuse and poor health.

References

Bair-Merritt, M. H., Blackstone, M., & Feudtner, C. (2006). Physical health outcomes of childhood exposure to intimate partner violence: A systematic review. Pediatrics, 117, e278 – e290. http://dx.doi.org/10 .1542/peds.2005-1473

Bosquet Enlow, M., Englund, M., & Egeland, B. (2018). Maternal child- hood maltreatment history and child mental health: Mechanisms in intergenerational effects. Journal of Clinical Child and Adolescent Psy- chology, 47, S47–S62. http://dx.doi.org/10.1080/15374416.2016 .1144189

Brown, S. J., Lumley, J. M., McDonald, E. A., & Krastev, A. H., & the Maternal Health Study collaborative group. (2006). Maternal health study: A prospective cohort study of nulliparous women recruited in early pregnancy. BMC Pregnancy and Childbirth, 6, 12. http://dx.doi .org/10.1186/1471-2393-6-12

Buss, C., Entringer, S., Moog, N. K., Toepfer, P., Fair, D. A., Simhan, H. N., . . . Wadhwa, P. D. (2017). Intergenerational transmission of maternal childhood maltreatment exposure: Implications for fetal brain development. Journal of the American Academy of Child & Adolescent Psychiatry, 56, 373–382. http://dx.doi.org/10.1016/j.jaac.2017.03.001

Centers for Disease Control and Prevention. (2010). Behavioral risk factor surveillance system ACE module data. Atlanta, GA: Author.

Cohen, J. A., Mannarino, A. P., & Iyengar, S. (2011). Community treat- ment of posttraumatic stress disorder for children exposed to intimate partner violence: A randomized controlled trial. Archives of Pediatrics & Adolescent Medicine, 165, 16 –21. http://dx.doi.org/10.1001/archpediat- rics.2010.247

Collishaw, S., Dunn, J., O’Connor, T. G., & Golding, J., & the Avon Longitudinal Study of Parents and Children Study Team. (2007). Ma- ternal childhood abuse and offspring adjustment over time. Development and Psychopathology, 19, 367–383. http://dx.doi.org/10.1017/ S0954579407070186

Cyr, C., Euser, E. M., Bakermans-Kranenburg, M. J., & Van Ijzendoorn, M. H. (2010). Attachment security and disorganization in maltreating and high-risk families: A series of meta-analyses. Development and Psychopathology, 22, 87–108. http://dx.doi.org/10.1017/S0954579 409990289

De Bellis, M. D., & Zisk, A. (2014). The biological effects of childhood trauma. Child and Adolescent Psychiatric Clinics of North America, 23, 185–222. http://dx.doi.org/10.1016/j.chc.2014.01.002

DiLillo, D., & Damashek, A. (2003). Parenting characteristics of women reporting a history of childhood sexual abuse. Child Maltreatment, 8, 319 –333. http://dx.doi.org/10.1177/1077559503257104

Evans, S. E., Davies, C., & DiLillo, D. (2008). Exposure to domestic violence: A meta-analysis of child and adolescent outcomes. Aggression and Violent Behavior, 13, 131–140. http://dx.doi.org/10.1016/j.avb.2008 .02.005

Gartland, D., Woolhouse, H., Giallo, R., McDonald, E., Hegarty, K., Mensah, F., . . . Brown, S. J. (2016). Vulnerability to intimate partner violence and poor mental health in the first 4-year postpartum among

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

120 GIALLO ET AL.

mothers reporting childhood abuse: An Australian pregnancy cohort study. Archives of Women’s Mental Health, 19, 1091–1100. http://dx .doi.org/10.1007/s00737-016-0659-8

Goodman, R. (2001). Psychometric properties of the Strengths and Diffi- culties Questionnaire. Journal of the American Academy of Child & Adolescent Psychiatry, 40, 1337–1345. http://dx.doi.org/10.1097/ 00004583-200111000-00015

Hawes, D. J., & Dadds, M. R. (2004). Australian data and psychometric properties of the Strengths and Difficulties Questionnaire. Australian and New Zealand Journal of Psychiatry, 38, 644 – 651. http://dx.doi.org/ 10.1080/j.1440-1614.2004.01427.x

Hegarty, K., Bush, R., & Sheehan, M. (2005). The composite abuse scale: Further development and assessment of reliability and validity of a multidimensional partner abuse measure in clinical settings. Violence and Victims, 20, 529 –547. http://dx.doi.org/10.1891/vivi.2005.20.5.529

Holt, S., Buckley, H., & Whelan, S. (2008). The impact of exposure to domestic violence on children and young people: A review of the literature. Child Abuse & Neglect, 32, 797– 810. http://dx.doi.org/10 .1016/j.chiabu.2008.02.004

Howell, K. H., Graham-Bermann, S. A., Czyz, E., & Lilly, M. (2010). Assessing resilience in preschool children exposed to intimate partner violence. Violence and Victims, 25, 150 –164. http://dx.doi.org/10.1891/ 0886-6708.25.2.150

Kwako, L. E., Noll, J. G., Putnam, F. W., & Trickett, P. K. (2010). Childhood sexual abuse and attachment: An intergenerational perspec- tive. Clinical Child Psychology and Psychiatry, 15, 407– 422. http://dx .doi.org/10.1177/1359104510367590

Levendosky, A. A., Bogat, G. A., Lonstein, J. S., Martinez-Torteya, C., Muzik, M., Granger, D. A., & von Eye, A. (2016). Infant adrenocortical reactivity and behavioral functioning: Relation to early exposure to maternal intimate partner violence. Stress: The International Journal on the Biology of Stress, 19, 37– 44. http://dx.doi.org/10.3109/10253890 .2015.1108303

Lieberman, A. F., Ghosh Ippen, C., & Van Horn, P. (2006). Child-parent psychotherapy: 6-month follow-up of a randomized controlled trial. Journal of the American Academy of Child & Adolescent Psychiatry, 45, 913–918. http://dx.doi.org/10.1097/01.chi.0000222784.03735.92

MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects. Psychological Methods, 7, 83–104. http:// dx.doi.org/10.1037/1082-989X.7.1.83

MacMillan, H. L., Fleming, J. E., Trocmé, N., Boyle, M. H., Wong, M., Racine, Y. A., . . . Offord, D. R. (1997). Prevalence of child physical and sexual abuse in the community. Results from the Ontario Health Sup- plement. Journal of the American Medical Association, 278, 131–135. http://dx.doi.org/10.1001/jama.1997.03550020063039

MacMillan, H. L., Tanaka, M., Duku, E., Vaillancourt, T., & Boyle, M. H. (2013). Child physical and sexual abuse in a community sample of young adults: Results from the Ontario Child Health Study. Child Abuse & Neglect, 37, 14 –21. http://dx.doi.org/10.1016/j.chiabu.2012.06.005

Madigan, S., Bakermans-Kranenburg, M. J., Van Ijzendoorn, M. H., Mo- ran, G., Pederson, D. R., & Benoit, D. (2006). Unresolved states of mind, anomalous parental behavior, and disorganized attachment: A review and meta-analysis of a transmission gap. Attachment & Human Devel- opment, 8, 89 –111. http://dx.doi.org/10.1080/14616730600774458

Madigan, S., Wade, M., Plamondon, A., & Jenkins, J. (2015). Maternal abuse history, postpartum depression, and parenting: Links with pre- schoolers’ internalizing problems. Infant Mental Health Journal, 36, 146 –155. http://dx.doi.org/10.1002/imhj.21496

Madigan, S., Wade, M., Plamondon, A., Maguire, J. L., & Jenkins, J. M. (2017). Maternal adverse childhood experience and infant health: Bio- medical and psychosocial risks as intermediary mechanisms. The Jour-

nal of Pediatrics, 187, 282–289.e1. http://dx.doi.org/10.1016/j.jpeds .2017.04.052

Manning, L. G., Davies, P. T., & Cicchetti, D. (2014). Interparental violence and childhood adjustment: How and why maternal sensitivity is a protective factor. Child Development, 85, 2263–2278. http://dx.doi .org/10.1111/cdev.12279

McDonnell, C. G., & Valentino, K. (2016). Intergenerational effects of childhood trauma: Evaluating pathways among maternal ACEs, perina- tal depressive symptoms, and infant outcomes. Child Maltreatment, 21, 317–326. http://dx.doi.org/10.1177/1077559516659556

Mealing, N. M., Banks, E., Jorm, L. R., Steel, D. G., Clements, M. S., & Rogers, K. D. (2010). Investigation of relative risk estimates from studies of the same population with contrasting response rates and designs. BMC Medical Research Methodology, 10, 26. http://dx.doi.org/ 10.1186/1471-2288-10-26

Mensah, F., & Kiernan, K. (2010). Gender differences in educational attainment: Influences of family environment. British Educational Re- search Journal, 36, 239 –260. http://dx.doi.org/10.1080/0141192090 2802198

Murray, D., & Cox, J. (1990). Screening for depression during pregnancy with the Edinburgh Depression Scale (EDDS). Journal of Reproductive and Infant Psychology, 8, 99 –107. http://dx.doi.org/10.1080/ 02646839008403615

Muthen, L., & Muthen, B. (1998 –2011). Mplus user’s guide. Los Angeles: Author.

Paolucci, E. O., Genuis, M. L., & Violato, C. (2001). A meta-analysis of the published research on the effects of child sexual abuse. The Journal of Psychology: Interdisciplinary and Applied, 135, 17–36. http://dx.doi .org/10.1080/00223980109603677

Racine, N. M., Madigan, S. L., Plamondon, A. R., McDonald, S. W., & Tough, S. C. (2018). Differential associations of adverse childhood experience on maternal health. American Journal of Preventive Medi- cine, 54, 368 –375. http://dx.doi.org/10.1016/j.amepre.2017.10.028

Racine, N., Plamondon, A., Madigan, S., McDonald, S., & Tough, S. (2018). Maternal adverse childhood experiences and infant develop- ment. Pediatrics, 141, e20172495. http://dx.doi.org/10.1542/peds.2017- 2495

Roberts, R., O’Connor, T., Dunn, J., & Golding, J., & the ALSPAC Study Team. (2004). The effects of child sexual abuse in later family life; mental health, parenting and adjustment of offspring. Child Abuse & Neglect, 28, 525–545. http://dx.doi.org/10.1016/j.chiabu.2003.07.006

Saigal, S., & Doyle, L. W. (2008). An overview of mortality and sequelae of preterm birth from infancy to adulthood. Lancet, 371, 261–269. http://dx.doi.org/10.1016/S0140-6736(08)60136-1

Stith, S., Smith, D., Penn, C., Ward, D., & Tritt, D. (2004). Intimate partner physical abuse perpetration and victimisation risk factors: A meta- analytic review. Aggression and Violent Behavior, 10, 65–98. http://dx .doi.org/10.1016/j.avb.2003.09.001

Wegman, H. L., & Stetler, C. (2009). A meta-analytic review of the effects of childhood abuse on medical outcomes in adulthood. Psychosomatic Medicine, 71, 805– 812. http://dx.doi.org/10.1097/PSY.0b013e3181 bb2b46

Widom, C. S., Czaja, S. J., & Dutton, M. A. (2008). Childhood victimiza- tion and lifetime revictimization. Child Abuse & Neglect, 32, 785–796. http://dx.doi.org/10.1016/j.chiabu.2007.12.006

Wosu, A. C., Gelaye, B., & Williams, M. A. (2015). Maternal history of childhood sexual abuse and preterm birth: An epidemiologic review. BMC Pregnancy and Childbirth, 15, 174. http://dx.doi.org/10.1186/ s12884-015-0606-0

Received May 1, 2019 Revision received October 25, 2019

Accepted November 8, 2019 �

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

121MATERNAL CHILDHOOD ABUSE