Abstract
Adverse childhood experiences (ACEs) can negatively impact physical, emotional, cognitive, and social development, consequently affecting the next generation. The aim of the present study was to systematically review evidence from empirical studies on the association between maternal history of adversity in childhood (maltreatment and household dysfunction) and subsequent mother–child interactions at an early age. A search was performed in the PubMed, Web of Science, PsycINFO, EMBASE, Scopus, LILACS, and SciELO databases to identify studies, including measures of maternal childhood adversities and mother–child interaction, published between 2016 and 2022. Twenty-nine studies met the inclusion criteria. The results showed that in 90% of the studies, maternal childhood adversities negatively impacted subsequent mother–child interactions in early childhood, reducing maternal displays of affection, emotional availability, sensitivity, mother–child communication, and bonding. Biological factors (e.g., genetic and hormonal) and maternal emotional recognition moderated these associations. In addition, biological factors (i.e., neurobiological and hormonal) and psychosocial factors (e.g., depression, executive functioning, and violence) acted as mediators. Preventive interventions should be implemented to break out of the intergenerational cycle of violence that impacts mother–child interactions.
Keywords
Introduction
Adverse childhood experiences (ACEs) are potentially traumatic events that include abuse (physical, sexual, and psychological), neglect (physical and emotional), and household dysfunction (exposure to domestic violence, substance abuse, mental illness, criminal behavior, and divorce) (Center for Disease Control and Prevention [CDC], 2019). ACEs negatively impact a person’s life, leading to potential physical, emotional, cognitive, and social impairments, increasing health-risk behaviors, and ultimately leading to serious chronic physical and emotional health problems and developmental disturbances (Kalmakis & Chandler, 2015). However, ACEs’ influence may go beyond the person who experienced them firsthand, impacting the next generation (Su et al., 2020; Sun et al., 2017). Parents with a history of childhood maltreatment are more likely to maltreat their children than parents without such histories (Assink et al., 2018; Madigan et al., 2019). The studies about the intergenerational transmission of childhood adverse experiences have shown that maternal exposure to ACEs is associated with increased risk of children’s developmental delays in several domains (Lotto et al., 2021; Sun et al., 2017), such as motor, social, communication, and problem-solving skills (Folger et al., 2018), maladaptive socio-emotional symptoms (McDonnell & Valentino, 2016), and an increased risk of psychopathology (Su et al., 2020).
One of the ways, ACEs impact later child development and behavior is through dysfunctional mother–child interactions (Cicchetti & Toth, 2016; Fuchs et al., 2015). Mothers exposed to maltreatment in childhood are at a greater risk of struggling with their subsequent parenting behaviors (Greene et al., 2020). These parenting difficulties could arise because parents, who were abused as children, could not have had an adequate parenting model, making it harder for them to later adopt this role themselves (Fuchs et al., 2015).
Contrary to negative parenting, positive mother–child interactions involving responsive and sensitive parenting can contribute to secure attachment, which in turn positively impacts child development (Bilgin & Wolke, 2015; Roggman et al., 2013). Highly responsive mothers are able to respond to a child’s cues in an adequate and timely manner, regardless of how subtle the cues may be (Shire et al., 2016). Stable and responsive care protects early childhood development in an optimal period considered a “window of opportunity” for human development, shaping the brain architecture, strengthening, and preparing it to function properly in a variety of circumstances throughout the child’s life (Center on the Developing Child at Harvard University, 2016; WHO, 2018).
Childhood maltreatment is a considerable risk factor for negative parenting; however, there are still some inconsistencies. In the meta-analysis by Savage et al. (2019), childhood maltreatment significantly impacted relationship-based parenting outcomes, with greater effects when parenting measures were relationship-based or focused on negative behaviors and in samples with a higher proportion of boys. However, in a review by Vaillancourt et al. (2017), the findings were inconsistent, probably because of the diversity of measurements and characteristics of the study samples. In addition, in this review, only studies on sexual and physical abuse were included in the types of maternal childhood maltreatment, and they did not examine the effects of other types of maltreatment (Vaillancourt et al., 2017).
There is still a gap in the literature related to the examination of the impact of a maternal history of ACEs, including both childhood maltreatment and exposure to household dysfunction, on subsequent mother–child interactions in early childhood (zero-to-6-years old) development. The present study aimed to systematically review empirical studies that examined the associations between adverse maternal childhood experiences and subsequent mother–child interactions at early ages.
Method
The present review was registered in the PROSPERO database (CRD42021258204).
Search Strategy
This systematic review was conducted based on the principles of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA; Liberati et al., 2009). The review was conducted by searching the PubMed, Web of Science, PsycINFO, EMBASE, Scopus, LILACS, and SciELO databases, using the following search strategy: (maternal childhood trauma) OR (adverse childhood experience*) OR (adverse life experience*) OR (childhood maltreatment) OR (maltreatment history) OR (maternal childhood maltreatment) AND (mother-child interaction*) OR (mother-infant interaction*) OR (mother–toddler interaction*) OR (maternal interaction*). The first search was conducted on May 11, 2021, and the second, which aimed to update the review, was conducted on January 31, 2022.
Selection Criteria
Inclusion criteria: (a) studies with associations between adverse maternal childhood experiences and subsequent mother–child interactions; (b) empirical studies; (c) studies with exclusively mothers; (d) studies from January 2016 to 2022; and (e) studies in English, Portuguese, or Spanish. Exclusion criteria: (a) studies without those associations; (b) studies that did not assess maternal and dyadic interactive behaviors; (c) studies involving other caregivers; (d) studies with children older than 6 years; (e) qualitative, psychometric or theoretical studies, reviews, meta-analysis, protocols; (f) intervention studies; (g) other languages; and (h) animal studies.
Selection of the Studies
As seen in Figure 1, the initial search generated 1,206 articles. First, 566 duplicate articles indexed in more than one database were excluded. Second, after applying the criteria, 506 articles were excluded. Then, 45 remaining articles were fully read, which resulted in sixteen articles being excluded for following reasons: (a) children over 6 years old (n = 5), (b) did not analyze maternal childhood adversities separately from current adversities (n = 3), (c) other caregivers (n = 2), (d) did not include maternal childhood adversities in the results (n = 2), (e) no mother–child interactions (n = 2), (f) intervention studies (n = 1), and (g) did not examine associations between ACEs and mother–child interactions (n = 1). Finally, 29 articles were reviewed.

Flowchart of the selection of articles.
Data extraction
All articles were reviewed by following the PRISMA checklist (Liberati et al., 2009; Moher et al., 2009), and key study characteristics were identified. Information extracted from the studies included the following topics: authors and year of publication, country, objectives, design, sample, instruments and measures, maternal childhood adversities variable, interactive behavior variable, data analysis, and the main findings of the studies. The first author read all articles and performed data extraction coding according to these items, which were later reviewed by the second and third authors to ensure the accuracy of the analysis.
Methodological quality assessment
The methodological quality of the studies was assessed using Strengthening the Reporting of Observational Studies in Epidemiology (STROBE; von Elm et al., 2008). The final score was the sum of all items (maximum score = 22); higher scores indicate greater methodological care. An index level was created with the items present in each study, showing the overall percentages of items achieved by the articles.
Results
Overview of the Studies
The 29 studies reviewed represented a total sample of 3,855 mother–child dyads, with sample sizes ranging from 28 to 511 participants. The studies were conducted in ten different countries (United States, Canada, Australia, Denmark, Spain, Italy, Germany, China, Turkey, and Palestine) across four continents: North America (62%), Europe (21%), Oceania (7%), and Asia (7%), and in Turkey, which is a transcontinental country (3%). Regarding the study design, 15 articles (52%) were prospective longitudinal studies (two cohort studies), and 14 studies (48%) used a cross-sectional design.
In the 29 studies, we detected that in 15 studies (52%) the characteristics of the study sample comprised ethnic-racial diversity (e. g., White/Black, Caucasian/African, American/European American, Asian/Latin American/Indigenous Australian, Hispanic/Not Hispanic). It is important to note that in nine studies (31%) the sample was recruited exclusively in one specific ethnical group (e.g., Caucasian, Palestine, Asian, or Latin samples), and in 17% of the studies the ethnic-racial aspects of the sample were not outlined clearly.
Mother–Child Interactions
Appendix A presents a table of mother–child interactions. The age of the children when mother–child interaction was assessed varied between studies. In cross-sectional studies, the sample of children was between 0 and 5 years old, and in 64% of them, the children were under 1 year old. Twelve out of the 15 longitudinal studies included (80%) assessed mother–child interaction only once; in 83% of them, the children were under 1 year old (the age of the children varied between 0 and 3 years). In the remaining three longitudinal studies, dyadic interactions were assessed at different time points as follows: 6 months and 1 year old (Liu et al., 2021); 1 and 2 years old (Dayton et al., 2016); 18 months and 3 and 5 years old (Harris et al., 2021).
Regarding the instruments used for the mother–child interaction assessment, most studies used only one instrument (86%), and only four studies used more than one instrument combined. The majority of articles exclusively used observational procedures to assess mother–child interactions (86%), two studies used self-report measures (Kankaanpää et al., 2020; Lara-Cinisomo et al., 2018), and two studies used both types of measures combined to assess dyadic interactions (Julian et al., 2019; Williams et al., 2021). The 27 studies that used observational procedures assessed mother–child interactions in different contexts: 48% used a free-play situation, 26% applied both free- and structured-play situations, and three studies used only a structured-play situation (Ensink et al., 2017; Julian et al., 2019; Karakaş et al., 2020), two studies used a structured-play situation and observed the mother–child interaction during routine activities at home (Ludmer et al., 2018; Toepfer et al., 2019), one study assessed the mother–child interaction in routine activities in the family’s home (Dayton et al., 2016), and one study did not specify how the observation was conducted (Olsavsky et al., 2021).
Regarding the observational measures, the most used was the Emotional Availability Scale in seven (26%) studies, followed by the MACY Infant–Parent Coding System, used in three (11%) studies, and the Atypical Maternal Behavior Instrument for Assessment and Classification in three studies. The remaining 14 studies (52%) used a great diversity of 15 observational measures, most of which were applied in one study each. The most commonly used self-report measure was the Postpartum Bonding Questionnaire, which has been used in three different studies.
Maternal history of childhood adversities
Of the 29 articles, 69% (n = 20) exclusively assessed maltreatment experiences, whereas 31% (n = 9) evaluated both maltreatment and household dysfunctions. The types of maternal childhood maltreatment evaluated were physical, sexual, and emotional abuse and physical and emotional neglect. The physical abuse was assessed in 100% of the studies. The most frequently assessed household dysfunctions were caregivers’ mental illness, domestic violence, parental divorce, and parental death.
Appendix B presents the table on adverse maternal childhood experiences. They were assessed using self-report measures in all studies. Most articles exclusively measured childhood maltreatment experiences (69%), and nine studies (31%) evaluated both childhood maltreatment and household dysfunction. The majority of studies used a single instrument to assess the maternal childhood adversities (93%), and only two studies used more than one instrument for assessment. Both questionnaires and interviews were used to assess the maternal history of childhood adversities, and the most commonly used instrument was the Childhood Trauma Questionnaire (CTQ short- or full-form) in 16 different studies (55%). The Adverse Childhood Experiences Questionnaire (ACE-Q) was used in four studies, and the Structured Clinical Interview for DSM IV (SCID) was used in three studies. The remaining six studies (21%) assessed maternal childhood adversities using eight diverse instruments.
Main Findings of the Studies
Appendix C summarizes the main findings of the 29 studies; 90% showed statistically significant associations between adverse maternal childhood experiences and subsequent mother–child interactions.
Direct effects of maternal ACEs on mother–child interactions
Statistically significant direct effects of adverse maternal childhood experiences on mother–child interactions were found in most of the studies. Mothers with histories of ACEs (maltreatment and household dysfunction) reported more impaired bonding, rejection and anger, and anxiety about care at 4 and 8 weeks postpartum than women without such histories (Lara-Cinisomo et al., 2018). Mothers with a history of childhood maltreatment showed less maternal sensitivity and more intrusiveness and hostility in interactions with their 1-year-old children, in comparison to mothers who had not been exposed to previous maltreatment (Fuchs et al., 2017). Mothers who had greater exposure to both childhood maltreatment and household dysfunction showed less maternal positive engagement and social monitoring and more intrusiveness with their 3-month-old children (Crugnola et al., 2019) and lower maternal emotional availability toward their 18 months of age children (Harris et al., 2021).
Concerning the severity of maternal history of childhood maltreatment, the moderate-to-severe childhood trauma was associated to low maternal emotional availability in interactions with children at 6 months of age (MacMillan et al., 2020). After controlling for the time since the mother ingested caffeine, mothers’early-life trauma (maltreatment and exposure to household dysfunction) significantly predicted an increased neutral effect during the mother–infant interaction at 6 months of age (Juul et al., 2016). High exposure to multiple types of childhood maltreatment most strongly predicted subsequent disrupted parenting behavior (role/boundary confusion and maternal affective communication error) in mother–child interactions at age of 1 year (Guyon-Harris et al., 2020). Mothers who experienced more childhood maltreatment showed lower maternal gentle control during mother–child interaction at 14 months of age, regardless of educational status, and maternal gentle control at 6 months of age (Liu et al., 2021).
In mother–child interactions at 4 months old, mothers who suffered multiple types of childhood maltreatment displayed more role confusion and fearful/disoriented behaviors than mothers who experienced emotional abuse and mothers with no experience of childhood maltreatment (Khoury et al., 2021). Unexpectedly, in this study, mothers who experienced childhood emotional abuse showed less role confusion in mother–child interactions than mothers with a low probability of having experienced childhood maltreatment.
Adolescent or young mothers who had childhood experiences of maltreatment showed more violent and intrusive behaviors and less sensitive behaviors in interactions with their 3-month-old children in comparison to mothers of the control group (Moioli et al., 2022). In addition, in this study, more cumulative maternal childhood experiences of maltreatment predicted more maternal violent behaviors and less sensitive behaviors in interaction. In comparison to the non-maltreated group, mothers with early-life moderate to severe maltreatment and higher dissociative symptoms at five months postpartum presented significantly more self-reported parental bonding impairment and stress, higher physical abuse risk, less observed maternal sensitivity, and a lower emotional availability during mother-child interactions with 12- month-age children.(Williams et al., 2021).
Regarding dyadic interactive behaviors, higher levels of maternal childhood physical and emotional abuse have been linked to lower levels of close and positive mother–child interactions with 1-year-olds (Kankaanpää et al., 2020). However, mothers who experienced higher levels of childhood maltreatment showed lower levels of dyadic reciprocity in interactions with children aged less than 3 years old; however, the direct effect of the maltreatment experience on reciprocity was no longer statistically significant after including attentional bias in the model (Thompson-Booth et al., 2019).
Maternal ACEs and maternal moderators impacting subsequent mother–child interactions
In eight studies, there were associations between maternal childhood adversities and mother–child interactions moderated by other predominantly biological maternal variables. Oxytocin was the most assessed biological moderator in five studies. Maternal experience of childhood maltreatment in association with the presence of the oxytocin receptor gene (OXTR) rs1042778 TT genotype negatively affects mother–child interactions, increasing the occurrence of frightened/frightening behavior in 7-month-old children (Julian et al., 2019). The mothers’ oxytocin secretion assessment showed different results depending on maternal childhood maltreatment and household difficulty level, in which higher oxytocin production among mothers with high ACEs was associated with lower positive parenting at 34–48 months of age (Julian et al., 2018). In mothers with high expression of the OXTR T-allele, there was an association between maternal childhood maltreatment (regardless of severity or type of maltreatment) and high maternal insensitivity during an interaction at 6 months of age (Toepfer et al., 2019). The experience of maternal childhood sexual abuse, specifically when associated with lower levels of oxytocin, impacted the mother–child interaction by increasing maternal reports of impaired bonding at 4 and 8 weeks postpartum (Lara-Cinisomo et al., 2018).
Additionally, oxytocin (specifically, alleles of the oxytocin gene receptor) and cortisol (secretion assessed during the Strange Situation procedure) were examined together, showing that maternal childhood maltreatment more strongly predicted mother–infant attachment disorganization (score and classification) at 16–17 months of age for mothers with more plasticity alleles of OXTR than mothers with fewer plasticity alleles and for mothers with higher cortisol levels than mothers with lower cortisol levels (Ludmer et al., 2018). Higher levels of maternal childhood maltreatment and household dysfunction was associated with the presence of the dopamine receptor gene DRD4’s long allele and led to lower levels of maternal sensitivity during mother–child interactions with 2-year-old children (Su et al., 2018). The increased volume in the right inferior longitudinal fasciculus tract predicted more positive emotional availability only in maltreated mothers, and this neurobiological alteration appeared to represent a protective factor for neglectful mothers being able to establish a more sensitive and responsive interaction with their children aged less than 5 years (Rodrigo et al., 2016). Focusing only on the psychological moderator analyzed, results showed that among mothers exposed to severe childhood maltreatment, higher emotional recognition was related to fewer sensitive behaviors toward 2-to-5-year-old children (Bérubé et al., 2020).
Mediation effects on the associations between maternal ACEs and subsequent mother–child interactions
In 10 studies, the mediation effects were examined. Regarding biological mediators, maternal experiences of childhood abuse (emotional, physical, and sexual) impacted maternal cortisol levels, elevating them at 4 months postpartum, which, in turn, resulted in disrupted maternal behavior in interactions with the child at the same age (Nyström-Hansen et al., 2019). In addition, maternal history of childhood adversities (maltreatment and household dysfunction) was associated with lower mean maternal cortisol levels, which, in turn, significantly predicted maternal neutral affection during mother–child interactions at 6 months of age, after controlling for infant sex and premature childbirth (Juul et al., 2016). Mothers with higher levels of childhood adversity (physical and verbal abuse and exposure to domestic violence) had both heightened amygdala activation (in response to sounds of infant distress) and increased connectivity with prefrontal areas, which in turn was associated with fewer intrusive maternal behaviors in mother–child interactions with 1–10-month-old infants (Olsavsky et al., 2021).
Concerning the psychosocial mediators, maternal childhood maltreatment leads to higher environmental risk composite score (e.g., overcrowding of four or more children in the same household, receipt of government benefits as major source of income, maternal low social support, high maternal stress), which, in turn, results in lower emotional availability in mothers of 12–42-month-old children, controlled for borderline personality features (Hatzis et al., 2019). Additionally, mothers with higher levels of childhood maltreatment had a lower attentional bias toward infant faces, which, in turn, led to lower levels of dyadic reciprocity in mother–child interactions with children younger than 3 years of age (Thompson-Booth et al., 2019). Mothers who experienced early-life maltreatment showed dissociative symptoms, which in turn led to higher bonding impairment in mother–child interactions at 12 months of age (Williams et al., 2021).
Considering the mediation of the mental health, the mothers who experienced childhood maltreatment had greater depression levels, which, in turn, led to less responsive parenting behaviors toward their 3-year-old children. The experience of childhood maltreatment was also associated with higher maternal-reported child behavior problems, which, in turn, seemed to lead to higher levels of harsh parenting behaviors in mother–child interactions at 3 years of age (Michl-Petzing et al., 2019). Mothers who experienced more severe childhood maltreatment developed greater symptoms of depression, which then led to lower maternal positive affection in interaction with 6-month-old children; and greater severity of maternal childhood maltreatment was also linked to the development of greater posttraumatic stress disorder (PTSD) symptoms, which in turn led to higher maternal positive effect and lower maternal negative effect in interaction with children aged 6 months (Morelen et al., 2016).
Focusing the chained mediational effects, there was a mediation pathway starting from the maternal history of childhood maltreatment, assessed when the children were 6 months old, to negative control in mother–child interactions at 14 months of age, through maternal executive function, and then marital satisfaction (Liu et al., 2021). Another long mediation chain was found, showing that higher levels of maternal history of childhood maltreatment had a negative effect on mothers' relationships, leading to higher levels of exposure to intimate partner psychological aggression and negative perceptions of adult romantic relationships, which, in turn, caused more negative perceptions of infants’ emotions at 1year of age, which, finally, led to lower levels of sensitive parenting in mother–child interactions when children were 1 year old and later at 2 years old (Dayton et al., 2016).
Moderator and mediation effects on the associations between maternal ACEs and subsequent mother–child interactions
Both mediation and moderation effects were detected by examining the associations between adverse maternal childhood experiences and subsequent mother–child interactions. Higher levels of maternal childhood adversities (including maltreatment and household dysfunction), in association with the D4 dopamine receptor gene's long allele, resulted in lower levels of maternal emotional regulation, which in turn predicted lower maternal sensitivity in mother–child interactions at 2 years of age (Su et al., 2018).
Methodological quality of the studies
All studies obtained an index score of at least 50%, indicating that at least half of the applicable items were adequately presented in the studies. Also, two studies obtained an index score above 75%, demonstrating excellent methodological quality. The weaknesses detected in the studies were as follows: no description of the reasons for nonparticipation (n = 24), reports of the number of individuals at each stage of the study (n = 23), and the absence of a flowchart diagram (n = 24) (Appendix D).
Discussion
In the present review, most studies were conducted in developed and high-income countries. Despite the lack of socioeconomic differences between the countries, the study samples were quite diverse, including mothers of different ages, ethnicities, recruitment settings, and from general or targeted populations. Regarding the inclusive aspects of the studies, we detected that 51% of the studies had diverse ethnic-racial characteristics and 31% of studies examined exclusively a specific ethnic sample (e.g., Caucasian or Latin). It is important to highlight that the description of the samples in the studies did not present a homogenous classification. There is a great heterogenous mix of categories of participants like mixed race, ethnicity, skin-color, and country of origin (e.g., Italian, Caucasian, Palestinian, White/Black, biracial, non-Hispanic White). As suggested by Tajima (2021), we have to pursue equity, focusing on the diversity and inclusion, in sampling, measure, data analysis, and interpreting results. The composite of diverse population comprises gender, racial, ethnicity, and social characteristics, and in the data analysis the desegragated data analysis of samples is recommended (Tajima, 2021). (Bent- Goodley, 2021; Tajima 2021). In the present review, only one study considered the ethnic-racial as covariate in the data analysis with non-significant variable (Juul et al., 2016). Another study considered that the ethnic-racial variable is relevant to include in the analyzing; however, the small size sample does not allow to desegregated groups in the data analysis. Consequently, the external validity of findings depends on the diversity of the samples with several genders, ethnic-racial, and cultural characteristics, to better understand to whom could be they generalized the findings, or not (Ludmer et al., 2018).
The current review showed that 90% of the studies revealed the negative effects of maternal childhood adversities on subsequent mother–child interactions at 0–6 years of age. The results of the present review confirm previous empirical findings on the influence of maternal childhood adversities on subsequent dysfunctional parenting and mother–child interactions (Cicchetti & Toth, 2016; Fuchs et al., 2015; Greene et al., 2020). In addition, our findings were aligned with the meta-analysis by Savage et al. (2019) and the systematic review by Vaillancourt et al. (2017). However, our review study advanced in relation to these two previous studies because we summarized the findings of maternal childhood adversities of both maltreatment and household challenge factors. Additionally, the present review focused specifically on mother–child interactions and not parenting outcomes, as examined by Savage et al. (2019). Unlike Vaillancourt et al. (2017), the current review comprised findings of mother–child interactions evaluated using both systematic observational methods and self-report measures. It is important to note that none of the articles in our review overlapped with the previous review of Vaillancourt et al. (2017), and only two studies (Dayton et al., 2016; Sexton et al., 2017) were included in the meta-analysis by Savage et al. (2019).
Definition of Maternal ACEs and Their Measures
Evidence concerning the differential impacts of different types of childhood adversities on mother–child interactions is scarce in the literature. Considering that different adverse experiences in childhood were significantly associated with each other (Dong et al., 2004; Racine et al., 2018), studies using the combination of maltreatment and household challenges that focus on specific types of adversities could better help understand their impacts on subsequent mother–child interactions. As previously seen in the study by Bailey et al. (2012), sexual and physical abuse in mothers’ childhood did not significantly impact emotional availability in mother–child interactions; however, maternal childhood neglect, emotional abuse, and exposure to domestic violence were associated with higher hostility in mother–child interactions. If only a few types of childhood adversities are assessed while several others are not considered, it remains unclear whether those specific adversities measured are directly or indirectly associated with parenting outcomes. However, exposure to multiple types of adversities is very common in samples with a history of maltreatment (Finkelhor et al., 2007), and it is sometimes difficult to differentiate between these different types of adversities.
Regarding the assessments of maternal ACEs, both interviews and self-report questionnaires were used in the present review, relying on the mothers’ recollections of the adverse experiences they faced when they were children. Retrospective reports are a valid and accurate method to assess childhood adversities, particularly if the adversities in question are well defined, and the measurement tools used to assess them are well structured and of high quality (Hardt & Rutter, 2004; Krinsley et al., 2003). However, there are potential limitations to assess childhood experiences through recollection of events in adulthood. Retrospective reports of maternal childhood adversities may underestimate their actual occurrence (Hardt & Rutter, 2004). Studies assessing the consistency of retrospective reporting of adverse events in childhood have found that inconsistencies in reporting and poor stability of the reports across time are common (Colman et al., 2015; Fergusson et al., 2000). There is potential inherent difficulty in accurately remembering and reporting childhood events many years after they occur, as memory can be imperfect (Brennen et al., 2010; Fergusson et al., 2000). In addition, failure to accurately and consistently recall and report childhood adversities could be a protective mechanism induced by cognitive processes that support the inhibition of distressing or uncomfortable memories or the production of inaccurate memories (Fergusson et al., 2000). A person could minimize a traumatic childhood experience or even deny it (Church et al., 2017). Therefore, the limitation of the recollection of adverse events should be considered when examining studies of adult childhood maltreatment experiences.
The CTQ was used in 60% of the studies. The CTQ has a minimization/denial scale to measure potential recall and reporting bias (MacDonald et al., 2014), but only one study in the present review reported the score of the minimization/denial scale and showed a score within an acceptable range (Hatzis et al., 2019). The under-use of the CTQ minimization/denial scale has been previously mentioned in the literature (Church et al., 2017; MacDonald et al., 2014).
Measures of Mother–Child Interactions in Early Childhood
Concerning the mother–child interaction assessments, 86% of the studies (n = 25) exclusively used observational measures. Self-report questionnaires and interviews can be useful to better understand the interactions between mother and child, but direct observations of parental behaviors are considered the “gold standard” to evaluate interactions between caregivers and their children (Hawes & Dadds, 2006). Observational procedures reduce potential parental biases, allow a better understanding of specific interactive behaviors, and help capture bidirectional processes in the parent–child relationship (Aspland & Gardner, 2003; Funamoto & Rinaldi, 2014). In the studies reviewed, the observational procedures included diverse contexts, such as free- and structured-play situations (exclusively or combined), and routine activities in the family’s home.
It is important to highlight that the context of observation of mother–child interactions impacts maternal interactive behaviors with their children (Dittrich et al., 2017). In comparison to free-play situations, caregivers interacting with children in a structured challenging context (e.g., problem-solving tasks or stress-induced situations) showed less sensitivity, positive behaviors (Blacher et al., 2013; Kwon et al., 2013), and more intrusiveness (Blacher et al., 2013; Dittrich et al., 2017; Kwon et al., 2013). Thus, considering that interactive contexts affect parenting and interactive behaviors (Biringen et al., 2005; Dittrich et al., 2017), observational contexts should be carefully chosen and described clearly according to the purpose of the study.
Regarding the self-report method to assess mother–child interactions, its use is more convenient, less expensive, and less time-consuming than observational procedures (Aspland & Gardner, 2003; Herbers et al., 2017). In self-reports, mothers offer a personal perspective about their relationship with their children, and this perception, in turn, can impact their interactive behaviors (Weinfield et al., 2002). In the present review, two studies used self-report measures exclusively to assess mother–child interactions (Kankaanpää et al., 2020; Lara-Cinisomo et al., 2018) and two used both self-report and observational measures (Julian et al., 2019; Williams et al., 2021). In this review, Julian et al. (2019) found few significant correlations between observed and self-reported measures, confirming the findings of previous study (Potharst et al., 2021). In addition, maternal childhood maltreatment mediated by mothers’ dissociative symptoms negatively impacted mother–child interactions when the interactions were measured by self-report but not when they were observed (Williams et al., 2021).
Several reasons have been raised to understand the different results depending on the measure used to assess mother–child interactions. In self-report assessments, maternal desirability bias or mothers’ difficulties acknowledging and remembering their own behaviors when interacting with their children may interfere with maternal reports (Arney, 2004). In addition, self-report measures can assess several behaviors in diverse situations, whereas observational procedures are context-specific (Gardner, 2000). Regarding observational measures, the presence of an observer could interfere with the mothers’ interactive behaviors. In addition, differences between self-reported and observed mother–child interactions may occur in samples of mothers who have experienced childhood maltreatment, as they tend to have lower self-awareness (Bailey et al., 2007). These mothers could be less aware of possible difficulties in interactions with their children and, consequently, self-report fewer problems in mother–child interactions but show negative interactive behaviors when observed (Julian et al., 2019). The use of mixed methods of observational and self-report measures could be a complementary form to better understand mother–child interactions in their different dimensions and associations with mediators and moderators’ variables.
Associations Between Maternal ACEs and Mother–Child Interactions
The main findings of the present review showed that most studies presented direct negative effects of maternal childhood adversities on mother–child interactions. These results are consistent with those of previous studies (Bailey et al., 2012; Fuchs et al., 2015). It is important to note that the severity and type of maternal childhood maltreatment interfere with the impacts detected in maternal interactive behaviors. Moderate-to-severe maternal childhood maltreatment provoked lower emotional availability in mother–child interactions than minimal and minor maltreatment (MacMillan et al., 2020). In addition, maternal childhood physical and emotional abuse results in fewer positive mother–child interactions (Kankaanpää et al., 2020). However, multiple types of maternal childhood maltreatment affect maternal withdrawal, role confusion, and fearful/disoriented behaviors in interactions with their children, in comparison to exclusively emotional abuse (Khoury et al., 2021). In turn, non-optimal mother–child interactions can lead to several negative consequences for children’s development and behavior (Folger et al., 2018; Sun et al., 2017).
Biological factors moderated the associations between maternal history of childhood adversities and mother–child interactions, contributing to a better understanding of the mechanisms of action of these factors. Maternal ACEs, when associated with higher cortisol secretion, impacted maternal disorganized attachment in mother–child interactions (Ludmer et al., 2018), and when associated with dopamine receptor gene long alleles, reduced maternal sensitivity (Su et al., 2018). Focusing on maternal genetics and physiological stress responses can help clarify the mechanisms involved in the association between maternal childhood adversities and mother–child interactions. Maternal genotypes, for example, are related to other maternal aspects that are strongly related to their behaviors toward their children, influencing the quality of mother–child interactions (Mileva-Seitz et al., 2016).
Concerning cortisol, the direction of the associations between maternal childhood adversities, cortisol, and mother–child interactions was inconsistent. In the present review, both low and high levels of cortisol secretion were found to mediate the relationship between mothers’ childhood adversities and mother–child interactions, as seen in Juul et al. (2016) and Nyström-Hansen et al. (2019), respectively. These conflicting findings were also found in previous studies, which showed that a maternal history of childhood adversities was associated with heightened cortisol secretion (Bublitz & Stroud, 2013; Schechter et al., 2004), and these stress reactions, in turn, led to less sensitive maternal interactive behaviors with their children (Gonzalez et al., 2012). Conflicting findings have shown that mothers with a history of childhood adversities have lower cortisol secretion levels (Brand et al., 2010). The effects of cortisol secretion in mothers with childhood maltreatment and its impact on mother–child interactions remain inconclusive and require further investigation.
Regarding oxytocin as a moderator variable in the association between adverse maternal childhood experiences and mother–child interactions, OXTR acted as a moderator (specifically its plasticity alleles, the TT genotype, and the expression of the T-allele of the receptor gene), exacerbating the negative effects of maternal childhood maltreatment on mother–child interactions (Julian et al., 2019; Ludmer et al., 2018; Toepfer et al., 2019). When maternal oxytocin secretion is associated with maternal childhood sexual abuse, lower levels of oxytocin lead to higher scores of impaired bonding in subsequent mother–child interactions (Lara-Cinisomo et al., 2018). In contrast, higher levels of maternal oxytocin secretion associated with a low maternal history of childhood adversities increased mothers’ positive interactive behaviors whereas the opposite occurred when the mother had a high level of childhood adversities (Julian et al., 2018). Previous studies have shown the importance of oxytocin for its effects on social cognition and behavior, highlighting its function as a neurotransmitter in brain regions associated with diverse aspects of social functioning (Herpertz & Bertsch, 2016). Notably, there are associations between oxytocin secretion and parenting behaviors, showing that oxytocin is associated with attachment and parent–child synchrony (Bakermans-Kranenburg & van IJzendoorn, 2014; Feldman & Bakermans-Kranenburg, 2017). The OXTR is associated with maternal sensitivity (Bakermans-Kranenburg & van IJzendoorn, 2008).
Concerning psychosocial risk factors that moderated the associations between maternal childhood adversities and subsequent mother–child interactions, severe maternal childhood maltreatment associated with higher maternal emotional recognition impacted less maternal-sensitive interactive behaviors (Bérubé et al., 2020). Emotion recognition could vary in mothers exposed to childhood adversities in comparison to mothers who were not exposed; it does not activate the same regions in their brain, and consequently, the way they process the emotions they recognize is different, which may lead to less positive behaviors toward their children (Neukel et al., 2019).
Exposure to childhood adversity is associated with mental health problems in adulthood (Danese et al., 2009; Schilling et al., 2007). Maternal history of childhood maltreatment impacts higher levels of depression, which, in turn, negatively affects mother–child interactions (Michl-Petzing et al., 2019; Morelen et al., 2016). Mothers with depressive symptoms may have difficulty responding to interactions with their children, showing lower sensitivity to children’s cues, less involvement in positive interactions, and use fewer positive strategies to manage their children’s behaviors (Chung et al., 2018; Howard et al., 2014). Additionally, maternal childhood adversities were associated with more reports of child behavior problems (Michl-Petzing et al., 2019), attentional bias to infant faces (Thompson-Booth et al., 2019), and maternal emotional regulation (Su et al., 2018).
In the current review, few studies found moderators that reduced the associations between maternal adversity in childhood experiences and mother–child interactions. The volume of the right inferior longitudinal fasciculus tract in neglectful mothers in association with maternal childhood maltreatment predicted more positive maternal emotional availability in mother–child interactions (Rodrigo et al., 2016). Additionally, high levels of maternal childhood adversity, when mediated by heightened amygdala activation and increased connectivity with prefrontal areas, impact less intrusive maternal interactive behaviors (Olsavsky et al., 2021). Unexpectedly, in the present review, symptoms of PTSD acted as mediators of the association between maternal childhood adversities and subsequent mother–child interactions, leading to positive interactive behaviors (Morelen et al., 2016). This finding differs from previous studies that did not find significant associations between PTSD symptoms and mother–child interactions (Ammerman et al., 2012; Lang et al., 2010; Martinez-Torteya et al., 2014). However, the positive mediating effect of PTSD symptoms could be explained by the fact that the mothers did not perceive the mother–child interactions as triggers, reminding them of the past trauma; as the interaction was seen as a non-threatening activity, the mothers were able to show positive interactive behaviors (Muzik et al., 2017). In addition to the analysis of biological and psychosocial risk factors, we need to focus on protective factors that could moderate the negative impact of maternal childhood experiences on subsequent mother–child interactions.
Regarding the methodological quality of the studies reviewed, all studies obtained an index score of at least 50% in the STROBE statement, while six of them obtained scores higher than 75%. It is worth emphasizing that while most reviewed studies were below the 75% mark, only one obtained index scores below 60%, demonstrating the overall good methodological quality of the articles included in the present review.
In conclusion, this review showed the negative effects of adverse maternal childhood experiences on subsequent mother–child interactions, reducing positive maternal displays of affection, emotional availability, sensitivity, mother–child communication, and bonding. Maternal variables play a very important role in mediating or moderating this association. Most moderating variables (e.g., emotion recognition, oxytocin, cortisol), when associated with maternal childhood maltreatment, heightened negative interactive behaviors, such as frightened/frightening behavior and impaired bonding, or lowered positive interactive behaviors, such as dyadic reciprocity and sensitivity. There are numerous mediating variables, including mental health symptoms (e.g., depression), psychological and cognitive aspects (e.g., emotional regulation, executive function), social factors (e.g., environmental risk, marital satisfaction), and biological aspects (e.g., cortisol and neurobiological alterations). Similarly, to the moderated results, the indirect effects of maternal childhood maltreatment on mother–child interaction were mostly negative, decreasing interactive behaviors such as maternal sensitivity, emotional availability, and responsiveness, and increasing negative interactive behaviors such as maternal negative control and bonding impairment.
This review has several practical implications. First, to reduce childhood maltreatment, preventive intervention programs aimed at parents and caregivers should be considered. Reducing childhood adversities would mean reducing its negative effects on maternal health and behavior, which would then mean reducing at least one risk factor for the next generation. Second, considering that, regrettably, not all childhood adversities can be avoided or reduced, parenting programs aimed at improving the quality of mother–child interactions should be an option, as the interventions could help mothers learn more positive interactive behaviors.
The present review detected some limitations in the studies reviewed. Most of them were performed in developed and high-income countries; therefore, the generalization of these findings to low and middle income countries should be done with caution. In addition, the diversity of samples was not clearly stated in some studies, which could also impact in the interpretation and generalization of their findings. There were few longitudinal studies with different time points; only three longitudinal studies assessed mother–child interactions at different ages. As mother–child interactions change as children grow older, additional data from longitudinal studies would be an interesting way to assess how the association between maternal childhood adversity and mother–child interactions can change with time.
Future studies should address these issues as well. First, to assess the maternal history of childhood adversities, studies should implement, alongside their self-report instruments, a tool to evaluate potential biases that may skew results. Second, potential protective factors should be examined to better understand the buffer association between maternal childhood adversity and subsequent mother–child interactions. Third, studies using observational measures should clearly define the contexts of the observations of mother–child interaction to assess possible differences in the interactive behaviors shown by the mothers depending on the interactive situation. Fourth, future studies on the associations between maternal childhood adversities and subsequent mother–child interactions should be examined in LMICs. Fifth, special investment in pursuing diversity issues should be carefully addressed in future studies. Sixth, the maternal experience of community adversity is still scarce in the literature, given it could overlap with household and individual adversities.
Supplemental Material
sj-docx-1-tva-10.1177_15248380221130355 – Supplemental material for Maternal History of Adversity and Subsequent Mother–Child Interactions at Early Ages: A Systematic Review
Supplemental material, sj-docx-1-tva-10.1177_15248380221130355 for Maternal History of Adversity and Subsequent Mother–Child Interactions at Early Ages: A Systematic Review by Fernanda Leite Ongilio, Cláudia Maria Gaspardo and Maria Beatriz Martins Linhares in Trauma, Violence, & Abuse
Supplemental Material
sj-docx-2-tva-10.1177_15248380221130355 – Supplemental material for Maternal History of Adversity and Subsequent Mother–Child Interactions at Early Ages: A Systematic Review
Supplemental material, sj-docx-2-tva-10.1177_15248380221130355 for Maternal History of Adversity and Subsequent Mother–Child Interactions at Early Ages: A Systematic Review by Fernanda Leite Ongilio, Cláudia Maria Gaspardo and Maria Beatriz Martins Linhares in Trauma, Violence, & Abuse
Supplemental Material
sj-docx-3-tva-10.1177_15248380221130355 – Supplemental material for Maternal History of Adversity and Subsequent Mother–Child Interactions at Early Ages: A Systematic Review
Supplemental material, sj-docx-3-tva-10.1177_15248380221130355 for Maternal History of Adversity and Subsequent Mother–Child Interactions at Early Ages: A Systematic Review by Fernanda Leite Ongilio, Cláudia Maria Gaspardo and Maria Beatriz Martins Linhares in Trauma, Violence, & Abuse
Supplemental Material
sj-docx-4-tva-10.1177_15248380221130355 – Supplemental material for Maternal History of Adversity and Subsequent Mother–Child Interactions at Early Ages: A Systematic Review
Supplemental material, sj-docx-4-tva-10.1177_15248380221130355 for Maternal History of Adversity and Subsequent Mother–Child Interactions at Early Ages: A Systematic Review by Fernanda Leite Ongilio, Cláudia Maria Gaspardo and Maria Beatriz Martins Linhares in Trauma, Violence, & Abuse
Footnotes
Appendix
Methodological quality assessment of the studies based on STROBE Statement.
| Methodological quality (STROBE Index level) | References | STROBE Index (%) |
|---|---|---|
| ≥ 75% | Nyström-Hansen et al. (2019) | 87 |
| Kankaanpää et al. (2020) | 80 | |
| Harris et al. (2021) | 77 | |
| Liu et al. (2021) | 76 | |
| Guyon-Harris et al. (2020); Karakas et al. (2021) | 75 | |
| 74%-50% | MacMillan et al. (2020); Fuchs et al. (2017); Lara-Cinisomo et al. (2018) | 74 |
| Crugnola et al. (2019) | 73 | |
| Su et al. (2018); Khoury et al. (2021) | 71 | |
| Hatzis et al. (2019) | 70 | |
| Juul et al. (2016); Julian et al. (2018); Morelen et al. (2016) | 69 | |
| Michl-Petzing et al. (2019) | 68 | |
| Ludmer et al. (2018); Bérubé et al. (2020); Olsavsky et al. (2021); Sexton et al. (2017) | 67 |
|
| Toepfer et al. (2019) | 66 | |
| Julian et al. (2019); Dayton et al. (2016) | 64 | |
| Thompson-Booth et al. (2019); Williams et al. (2021) | 62 | |
| Moioli et al. (2022) | 61 | |
| Ensink et al. (2017) | 60 | |
| Rodrigo et al. (2016) | 53 |
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors thank the Coordination for the Improvement of Higher-Level Personnel—CAPES (grant number 88887.510383/2020-0) for supporting FL Ongilio and the Brazilian National Council for Scientific and Technological Development—CNPq (senior scientific investigator, grant number 310726/2021-2) for supporting MBM Linhares.
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
