Abstract
The intergenerational influence of adverse childhood experiences on individual outcomes demonstrates a need for research that considers both personal and environmental contributors to risk. As such, the current study explored how maternal cumulative risks influence the relationship between maternal Adverse Childhood Experiences (ACEs) and their children’s behavioral dysregulation among families involved with the child welfare system (N = 314). The importance of child age is also considered. The sample was stratified by age groups of children (1.5–5 years and 6–18 years), and the relationship between maternal ACEs, cumulative risk, and child behavior was assessed using OLS regressions. For younger children, maternal ACEs were only associated with externalizing behaviors when not controlling for cumulative risk, but cumulative risks were independently and significantly associated with both internalizing and externalizing behaviors. For older children, ACEs were independently associated with both types of behavior, but controlling for cumulative risk attenuated the strength of this relationship. Cumulative risks were also independently associated with older children’s internalizing and externalizing behaviors. Findings suggest the need to consider both individual and environmental risks for parents and children involved in the child welfare system, and the developmental timing and stability of that risk, in order to adequately support parent-child relationships as well as caregiving environments.
Keywords
Introduction
It is well established that the intergenerational transmission of trauma is detrimental to development. Previous literature suggests that caregivers’ trauma exposure affects parenting approaches and downstream child development (Clarkson Freeman, 2014; Fonagy, Steele, Moran, Steele, & Higgit, 1993; Fraiberg, Adelson, & Shapiro, 1975; Main & Goldwyn, 1984). Previous caregiver trauma may result in mental health challenges for parents and difficulties within the parent–child relationship (Kessler et al., 2010; Steele et al., 2016; Thomson & Jaque, 2017). Decades of research has established the link between adverse childhood experiences (ACEs) and negative mental and physical health outcomes (Anda et al., 2006), but the intergenerational link between caregivers’ ACEs and children’s clinical symptoms has only recently emerged (Dube et al., 2003; Madigan, Wade, Plamondon, Maguire, & Jenkins, 2017; Stepleton et al., 2018). Generally, there has been a lack of attention to the developmental nature of trauma exposure, the intersection with broader ecological risk factors, and the continuity in risk over time.
Through an almost exclusive focus on caregivers’ ACEs, the literature largely assumes that caregivers’ early trauma exposures remain in the past and the effects on adult and child outcomes are direct and linear. For highly vulnerable caregivers, however, who are more likely to have high ACEs scores, additional trauma and broader ecological risks may be ongoing and cumulative. Such risks would include elevated anxiety or depression, and socioeconomic risks, which occur later and affect both the individual and their environments. A parallel literature in developmental science on cumulative risk has highlighted the continuity and stability of social–ecological risk across the life span, meaning that even as circumstances change across the life course the experience of risk has a tendency toward stability (Sameroff, Seifer, Zax, & Barocas, 1987). For families in the child welfare system (CWS), exposure to ACEs and concurrent cumulative risks (i.e., risks that occur outside of childhood traumas and are ecological in nature) are both likely to be high, which may exacerbate children’s behavior problems due to the chronic nature of risk and trauma exposure. The ACEs perspective is therefore likely to be incomplete when identifying the mechanisms between caregivers’ childhood trauma exposure and child clinical or behavioral symptoms across development.
The cumulative risk perspective, unlike ACEs, accounts for adverse risk factors that proximally and concurrently affect an entire family or household (MacKenzie, Kotch, Lee, Augsberger, & Hutto, 2011; Sameroff et al., 1987). Little is known, however, about the relationship between caregiver ACEs, later cumulative risks, and intergenerational outcomes. Therefore, this study examines how cumulative risks may be associated with intergenerational outcomes previously attributed primarily to ACEs. In doing so, this study provides a transactional developmental framework for considering how ACEs and later cumulative risks may intersect to impact development across the life span and intergenerationally. This study also explores the developmental role of age with regard to maternal ACEs, maternal cumulative risk factors, and children’s behavioral dysregulation among a child welfare involved sample.
ACEs and intergenerational transmission of trauma
ACEs refer to traumatic events that occur in early childhood (e.g., parental substance misuse, physical abuse, sexual abuse). ACEs have been linked to a broad range of physical and psychological health outcomes (Anda et al., 2002; Brown et al., 2009; Chapman et al., 2004; Dube et al., 2001; Dube, Anda, Felitti, Edwards, & Croft, 2002; Dube et al., 2006; Felitti et al., 1998). The likelihood of negative outcomes across a number of domains are higher with each ACE (Anda et al., 2006; Dube et al., 2003). The association of ACEs on developmental outcomes is likely due in part to increased exposure to toxic stress (Shonkoff et al., 2012). Toxic stress refers to the “strong, frequent or prolonged activation of the body’s stress response systems in the absence of the buffering protection of a supportive adult relationship” (Shonkoff et al., 2012). Experiencing toxic stress in childhood alters brain functioning and development, potentially causing long-term impairments across physical, cognitive, emotional, and relational domains (Center for the Developing Child at Harvard University, 2010).
The connection between the intergenerational transmission of trauma (Cook et al., 2005; Lieberman, 2004) and child maltreatment (MacKenzie, Kotch, & Lee, 2011; Oliveira, Maroco, & Pais, 2012) is also well developed in the literature. Less is known, however, about the mechanistic process(es) underlying the intergenerational transmission of maternal ACEs to children’s development. Emerging studies examining the intergenerational link between caregiver ACEs and child outcomes have found increased maternal ACEs to be associated with increased physical and psychological risks for mothers and their children (Le-Scherban, Wang, Boyle-Steele, & Pachter, 2018; Madigan et al., 2017; Stepleton et al., 2018). The attachment literature has elucidated the ways in which parents’ understandings of relationships and parenting behaviors are connected to their own experiences with caregivers in childhood, and that childhood adversity is linked to later parenting difficulties (Bowlby, 1988; Fraiberg et al., 1975; Steele et al., 2016). Less understood, however, are how these parenting behaviors function as mechanisms for the intergenerational transmission of the consequences of ACEs (Schickedanz, Halfon, Sastry, & Chung, 2018).
One important mechanism for mitigating the impact of trauma on a child is the presence of a consistent and nurturing caregiver (Shonkoff et al., 2012; Shonkoff, Richter, van der Gaag, & Bhutta, 2012). When caregivers’ nurturing capacities are limited, the ability to provide relational stability to mitigate developmental consequences of past trauma is impeded (Lieberman, 2004). Relational dysfunctions themselves are also associated with increased exposure to trauma (Shonkoff et al., 2012). Conversely, sensitive and supportive caregiving buffers the harmful impact of environmental risks.
Cumulative risks
The ACEs literature has not substantively accounted for the significant continuity in risk exposure after childhood, nor has it contended well with teasing apart the contributions of the adverse events to later development from the broader developmental risks in the family’s context, which contribute to the onset of adverse events such as maltreatment, and also contribute to later health and well-being outcomes (MacKenzie, Kotch, Lee, Augsberger, & Hutto, 2011). Although developments in research concerning ACEs and intergenerational impacts have contributed greatly to understandings of development (Madigan et al., 2017; Stepleton et al., 2018), further research is needed to understand the intersecting contributions of both ACEs and ecological risks that occur later in life on outcomes for individuals. Prospective longitudinal research suggests that the combination of child maltreatment and other later risk factors are associated with increased behavioral difficulties in children, which further warrants the continued exploration of cumulative risks in conjunction with ACEs, but also suggest that across development the more proximal ongoing ecological risk may be a larger driver of child behavioral dysregulation than more temporally distal early child maltreatment (MacKenzie, Kotch, Lee, Augsberger, & Hutto, 2011). Substantial literature demonstrates that cumulative risk within a child’s ecological context plays an important role in development and behaviors (Evans, Li, & Whipple, 2013; Sameroff, Bartko, Baldwin, Baldwin, & Seifer, 1998). Similar to findings that cumulative ACEs are more predictive of negative outcomes than isolated adversities, experiencing multiple environmental risks outside of childhood is more harmful to development than exposure to any one risk factor (Evans et al., 2013; MacKenzie, Kotch, Lee, Augsberger, & Hutto, 2011; Sameroff et al., 1998).
The cumulative risk perspective in developmental science emerged in response to the limitations of studies focusing on single risk factors which were inadequate in predicting developmental outcomes and drew on work emerging on the cumulative nature of risk for heart disease (Evans et al., 2013; Rutter, 1979; Sameroff et al., 1998; Sameroff, Seifer, Baldwin, & Baldwin, 1993). The cumulative risk perspective draws on two environmental theories of child development: the transactional model (Sameroff, 1975; Sameroff & Mackenzie, 2003), which conceptualizes a child’s development in terms of the reciprocal effects of a child’s interaction with environmental contexts. This perspective also draws on the ecological model of child development (Bronfenbrenner, 1979), which places the child’s development in the context of the interaction of complex processes within and across immediate and distal environmental systems.
Cumulative risk indices have included environmental risk factors from various ecological systems levels (Bronfenbrenner, 1979), for example, maternal mental health, parent education level, housing conditions, and neighborhood-level indicators (MacKenzie, Kotch, & Lee, 2011; MacKenzie, Kotch, Lee, Augsberger, & Hutto, 2011; Sameroff et al., 1998). Cumulative risk is generally measured using indices of relevant risk factors, and cumulative risk scores are calculated by summing binary risk factors on these indices (MacKenzie, Kotch, & Lee, 2011; MacKenzie, Kotch, Lee, Augsberger, & Hutto, 2011; Sameroff et al., 1998), in a similar manner as the ACE measures were developed.
Cumulative risk and child behaviors
The predictive relationship between cumulative risk and increased behavioral dysregulation among children and adolescents is well documented, providing ample support for understanding adverse experiences through a transactional and continuous lens (Evans et al., 2013; Gerard & Buehler, 2004; MacKenzie & McDonough, 2009; Sameroff et al., 1993). Studies have shown that children with higher levels of cumulative risk were more likely to have dysregulated behavior (MacKenzie, Nicklas, Brooks-Gunn, & Waldfogel, 2014). Cumulative risks have even been shown to predict externalizing and internalizing behaviors more so than reported early child maltreatment (MacKenzie, Kotch, Lee, Augsberger, & Hutto, 2011). Recent research showed an intergenerational association between maternal ACEs and children’s behaviors among a CWS involved sample, indicating that maternal ACEs are linked to greater behavior dysregulation in their children (Stepleton et al., 2018). Because cumulative risks are temporally distinct from ACEs and may independently affect children’s behaviors, an important next step is to identify how cumulative risk factors influence the relationship between maternal ACEs and child behaviors within a CWS context.
Influence of age
Taking a developmental perspective is critical when considering the impact of cumulative risk and caregiver ACES on children’s behavior. Normative developmental trajectories for children’s behaviors have been established in prior research with internalizing behaviors increasing and externalizing behaviors decreasing over time (Bongers, Koot, van der Ende, & Verhulst, 2003). This may be in part due to the fact that the proximity of other influences on children’s lives changes as they develop. Peers, school, and community contexts play more important roles in development in adolescence compared to more proximal influences (e.g., quality of caregiving) among younger children (Zielinski & Bradshaw, 2006). Developmental research on the early experience of maltreatment has also demonstrated that in the temporally proximal period, close to the experience of maltreatment trauma, it contributes to child behavior over and above the cumulative socioeconomic and family risk factors present, but across development as the child ages the cumulative risk picture has a greater contribution to child behavioral regulation than the earlier maltreatment itself (MacKenzie, Kotch, Lee, Augsberger, & Hutto, 2011). Studies such as these highlight the importance of substantively considering the influence of children’s ages when considering the relationship between intergenerational transmissions of trauma, later cumulative risks, and behavioral dysregulation.
Research questions
Bridging literature on intergenerational transmission of trauma, ACEs, and cumulative risk, this study seeks to understand the relationship between early maternal adversity, later cumulative risk, and children’s behavioral regulation. The study is guided by two research questions. First, does cumulative risk influence the relationship between maternal ACEs and behavioral dysregulation for children in the CWS? We hypothesize that cumulative risk will play a role in predicting child behaviors separate from the impact of the number of maternal ACEs for children in CWS. This study expands on the findings of (Stepleton et al., 2018) who found that caregiver exposure to ACEs is associated with higher levels of internalizing and externalizing behaviors among child welfare involved children. The second research question explores the importance of development stage in examining how maternal cumulative risk impacts behavioral dysregulation among children. We hypothesize that age across the school transition will be relevant in determining whether maternal cumulative risk plays a later role on child behaviors due to differences in developmental trajectories of behavior, the caregiving context, as well as the importance of proximal and distal factors differ in development for older and younger children.
Method
Data
As part of a wider comprehensive needs assessment of the CWS in a populous state located in the northeastern U.S., caregivers involved with the state CWS were surveyed about services as well as individual and child characteristics. Caregivers provided information about one target child, even if parenting multiple children. Surveys were conducted by trained interviewers via phone, using Computer Assisted Telephone Interviewing software. Participants received a US$25 gift card. This study was approved by the Rutgers University Institutional Review Board.
Caregivers with an open, active case with the CWS for a child still in their home at least 2 months prior to the survey period were eligible. A list of 10,207 in-home caregivers were identified from state administrative data, maintained in accordance with federal standards for statewide automated child welfare information systems. For families with more than one child, the included target child was selected based on the child whose birthday fell on the earliest in the month, regardless of month or year, to ensure limitation of age or seasonal effects in child selection. If the caregiver had twins, the child whose name was first alphabetically was selected. The sampling frame was comprised of caregivers listed as the primary caregiver.
Caregivers were not informed about the survey prior to participation. Once contacted, caregivers were informed that participation in the survey was voluntary and confidential, and informed about incentives. Consent was secured prior to participation and affirmed throughout the interview. Caregivers were contacted in no particular order, and quotas were utilized to ensure that responses were dispersed comparably across counties.
Of the 10,207 eligible caregivers, 2,221 caregivers were successfully contacted by the survey team, and 476 agreed to survey participation, a response rate of 21%. Given the limited literature regarding response rates among an in-home CWS population, the rate for this study aligned with expectations, particularly given the challenges faced by many of the families and the challenges the state faced in maintaining up to date contact information. It is understood that caregivers unsatisfied with the services they are receiving from the state CWS, or whom are dissatisfied clients could be more willing to participate in a survey; as well as that some eligible respondents, despite confidentiality assurances, may have concerns about survey participation negatively impacting their open cases (Alpert, 2005; Baker, 2007). Response rates for telephone have also diminished over time, evidenced by declining response rates for well-established political polling (Keeter, Hatley, Kennedy, & Lau, 2017), which is likely amplified for a child welfare involved population facing higher rates of poverty and contextual risk.
Study sample
The study sample was selected based upon availability of data for key variables. Figure 1 demonstrates the construction of the study sample.
Flowchart of sample selection process.
For this study, the 314 mothers in the final sample were divided by whether the participant was the mother of a child ages 1.5–5 or 6–18, as responses to the Child Behavior Checklist (CBCL) scales were based upon the age of the target child for the two versions of the instrument.
Measures
Child behaviors
The primary outcome variables in this study were internalizing and externalizing behaviors. To measure these, caregivers were asked to respond to one of two sets of scales based upon the age of their child. There were two CBCL scales for the target age groups (Achenbach & Rescorla, 2001). The CBCL asks caregivers to rate challenging behaviors from 0 (not true) to 2 (very true or often true). Caregivers of children ages 1.5–5 were asked to complete the withdrawing and aggressive subscales. The withdrawing subscale corresponds to internalizing behaviors and the aggressive subscale corresponds to externalizing. For children ages 6–18 years, two subscales (withdrawn/depressed and aggressive) were included on the survey. The withdrawn/depressed subscale corresponds to internalizing behaviors and the aggressive subscale corresponds to externalizing. Sum scores for the respective scales were used in analysis.
ACEs
The ACEs module from the Behavioral Risk Factor Surveillance System (Centers for Disease Control and Prevention, n.d.) served as the primary independent variable in the study. The ACEs module includes 11 items listing challenging experiences related to maltreatment and household dysfunction, and asks mothers to select each experience that occurred for them before the age of 18. A sum score was generated for each participant based on their responses to the 11-item questionnaire.
Controls
Additional variables of interest were included in analyses as controls. Sociodemographic variables included maternal age, race/ethnicity, as well as child age and sex. Race/ethnicity was recoded from seven distinct categories and condensed to four categories of White (non-Hispanic), Black or African American (non-Hispanic), Hispanic, and others (included Asian, multiracial, and American Indian). Ages for both caregivers and children were recorded continuously, and child sex was coded as male/female with female as the indicator. In prior studies, both parent and child ages were associated with differences in child behaviors over time (Tearne et al., 2015). Child behaviors have also been found to differ by child sex (Gerard & Buehler, 2004; Hunt, Slack, & Berger, 2017; Rescorla et al., 2007).
Cumulative risk
To measure cumulative risk, a 10-item risk index was constructed, modeled after the existing literature regarding risks and outcomes in children (reviewed by MacKenzie, Kotch, & Lee, 2011). Risk indicators in this study included: (1) Mother not married, recoded from a measure of relationship status in which “never married,” “separated,” “divorced,” and “widowed” were considered indicators of risk. (2) Any public assistance utilization, constructed from a series of items related to utilization of TANF, SNAP, WIC, SSI, public housing or Section 8, and ACA subsidies. Any utilization of one of these programs accounted for risk. (3) Mother’s health less than very good/excellent, a measure of health in which “fair,” “poor,” “very poor,” and “don’t know” were considered indicators or risk. (4) Any indication of maternal anxiety/depression, determined from responses to the Generalized Anxiety Disorder scale and the Depression scale from the Patient Health Questionnaire 9 (Spitzer, Kroenke, Williams, & Lowe, 2006; Kroenke, Spitzer, & Williams, 2001). Scores meeting the criterion for a provisional diagnosis of anxiety or depression were considered an indicator of risk. (5) Any maternal drug use, constructed from items concerning services received for alcohol or substance use including during pregnancy in which any service receipt was considered a risk factor. (6) Any reported economic hardship, derived from an item asking about receipt of services concerning finances. (7) Four or more children living in the home, constructed from a continuous variable asking the number of children in the home. (8) Maternal employment risk, which, based on developmental literature and the current data, was coded differently for the two age groups. For younger age group (1.5–5 years), full-time maternal employment was associated with greater behavioral dysregulation, so full-time work was included as the indicator of risk. For the older age group (6–18 years), maternal unemployment or underemployment was associated with greater behavioral dysregulation, so those two statuses comprised the risk indicator. (9) Any indication of violent/controlling behavior, determined by responses to the question “Do you think it would have been helpful or not helpful if you had received services for domestic violence” in which responses of “helpful” were considered an indication of risk. (10) Household income of less than US$15,000, constructed from reported income.
Analytic strategy
To examine the contributions of cumulative risk and maternal ACEs to child behavior, two Ordinary Least Squares (OLS) regression models were run predicting internalizing and externalizing CBCL scores for children ages 1.5–5 and 6–18, based upon respective scale scores. Behavioral dysregulation was examined among younger and older children as distinct groups to account for developmental sensitivities associated with age. First, the associations between maternal ACEs and CBCL scores were tested, controlling for covariates. In the second model, cumulative risk score was included as an additional predictor of CBCL scores. Changes in the weight and significance of the coefficients from the first to second models were examined to determine the influence of cumulative risk.
Results
Descriptive statistics
Descriptive statistics for the study are reported in Table 1.
Descriptive statistics.
Note. CBCL = Child Behavior Checklist. ACE = adverse childhood experience. SD = standard deviation.
Mothers of children ages 1.5–5
There were 101 mothers included in analysis whose children were 1.5–5 years old. For these mothers, the average internalizing behavior score of their children was 1.37 (SD = 2.08) ranging from 0 to 11. The average externalizing behavior score of children ages 1.5–5 was 9.38 (SD = 8.50) ranging from 0 to 31. The average number of ACEs for mothers in this group was 2.92 (SD = 2.72) ranging from 0 to 11. Mothers in this group had an average current cumulative risk score of 4.22 (SD = 1.22) ranging from 1 to 7 risk indicators. The average age for mothers of younger children was 28.61 years (SD = 5.34), with the youngest mother in the sample being 19 years and the oldest 43 years. Approximately 35% of this sample group was White (non-Hispanic), 33% Black or African American (non-Hispanic), 26% Hispanic, and 7% some other race (non-Hispanic). The average age of children in this sample group was 3.21 years (SD = 1.15) ranging from 1.5 to 5 years. Approximately 58% of the children in this sample group was identified as female.
Mothers of children ages 6–18
There were 213 mothers included in the analysis whose children were 6–18 years old. For these mothers, the average internalizing behavior score of their children was 3.09 (SD = 3.41) with a sample range of 0–13. The average externalizing behavior score of their children was 7.40 (SD = 8.40) with a sample range of 0–35. The average number of ACEs for mothers in this group was 3.09 (SD = 2.99) and a range of 0–11. The distribution of ACEs in this sample group was also skewed positively. Mothers in this group had an average current cumulative risk score of 3.83 (SD = 1.56) ranging from 1 to 8 risk indicators. The average age for mothers of older children was 35.47 years (SD = 6.86), with the youngest mother in the sample being 21 years and the oldest 56 years. Approximately 30% of this sample group was White (non-Hispanic), 31% Black or African American (non-Hispanic), 30% Hispanic, and 10% some other race (non-Hispanic). The average age of children in this sample group was 10.81 years (SD = 3.49) ranging from 6 to 18 years old. Approximately 49% of the children in this sample group was identified as female.
Regression results
The two OLS regression models for internalizing and externalizing behaviors of children ages 1.5–5 and 6–18, respectively, were tested, and results from these regression models are reported in Table 2.
OLS regression results of internalizing/externalizing behavior scores on maternal ACEs.
Note: t-statistics are in parenthesis. CBCL = Child Behavior Checklist. ACE = adverse childhood experience.
***p < .001. **p < .01. *p < .05.
Models testing association between ACEs and CBCL scores
The initial models demonstrated that when controlling for demographic characteristics, maternal ACEs are associated with elevated internalizing behavior scores in their children only at ages 6–18, with every increase in total number of ACEs being associated with a 0.37 higher internalizing behavior score (p < .001).
Maternal ACEs were associated with elevated externalizing behavior scores for both children ages 1.5–5 and 6–18. For younger children, each increase in ACE score for their mothers was associated with a 0.75 higher externalizing behavior score (p < .05). For older children, each increase in the number of ACEs was associated with almost a full point higher externalizing behavior score (0.9; p < .001).
The only demographic co-variate independently associated with an increase in behavior score was child age, which was only significantly associated with increased internalizing behavior scores in older children. With each additional year of age, internalizing behavior scores were higher by 0.15 (p < .05).
Models controlling for cumulative risk
After adding cumulative risk scores to the models, maternal ACE scores were no longer significantly associated with either the internalizing or externalizing behavior scores of children ages 1.5–5. However, when controlling for all other variables, every increase in cumulative risk score was associated with a 0.38 higher internalizing behavior score (p < .05), and 1.43 higher externalizing score (p < .05) for younger children
For older children, maternal ACEs continued to be significantly associated with both internalizing and externalizing behavior scores. After adding cumulative risk scores to the models, the coefficient for internalizing scores reduced to 0.33 (p < .001). For externalizing scores, the coefficient was reduced to 0.83 (p < .001). In the models concerning older children, cumulative risk scores were independently associated with higher internalizing and externalizing behavior scores, with every increase in maternal cumulative risk being associated with a 0.34 higher internalizing behavior score (p < .05), and 0.72 higher externalizing score (p < .05). In the internalizing model, increases in child age were again independently associated with increased internalizing behavior scores (b = .15; p < .05).
Discussion
This study contributes to our thinking about risk and trauma by drawing attention to the importance of integrating the, to date, parallel and distinct literatures on cumulative risk and ACEs. It also highlights the importance of taking age into account when examining the impact of the intergenerational transmission of trauma. The ACEs scale is constructed to examine various individual risk factors that occur in childhood; however, it does not take into account the developmental continuity, or stability, of risk in an individual’s environment beyond childhood. The cumulative risk model is founded on the tenets of two environmental theories of child development, the transactional model (Sameroff & Chandler, 1975; Sameroff & MacKenzie, 2003) and the ecological model of child development (Bronfenbrenner, 1979). Together, these theories account for the ways in which an ecology of risk affects both parents and children, which shapes their interactions with one another. Accounting for cumulative risks in addition to ACEs when considering intergenerational transmissions of trauma is an important step, as it offers a more robust understanding of the factors involved in the transmission of risks from caregivers to their children.
This study examined the impact of cumulative risks and maternal ACEs on childhood experiences and children’s behavioral dysregulation among a sample of child welfare involved families. As hypothesized, cumulative risk was significantly and independently associated with child behaviors. The inclusion of cumulative risk in the models also diminished the association between maternal ACEs and younger children’s externalizing behaviors. As hypothesized, children’s ages also contributed to behavioral symptomology. In this sample, internalizing behavior problems increased with age, perhaps reflecting developmental advancements in personal and caregiver awareness (Carroll & Steward, 1984). Older children often demonstrate stronger cognitive and verbal capacities than younger children, which may also influence parent’s recognition of withdrawing behaviors such as lack of enjoyment, isolation, and sadness. Maternal cumulative risks were also found to be stronger predictors for younger than older children. This finding is supported by Sroufe, Egeland, Carlson, and Collins (2005) who contend that experiences in infancy and early childhood may lose their power to predict adjustment in later development. However, maternal ACEs remained significantly associated with both internalizing and externalizing behaviors in older children, even after controlling for cumulative risks, suggesting that early maternal trauma and ongoing socioecological risk may offer unique and intersecting contributions to child health and well-being. Comparable levels of maternal ACEs and maternal cumulative risks for mothers of younger and older children suggest that relationships between maternal ACEs, cumulative risks, and child outcomes are salient across children’s development.
Strengths and limitations
This study has several strengths to be considered. The data collected come directly from over 300 mothers whose children were engaged in the CWS and at risk for cumulative disadvantage. Mothers reported on experiences of child maltreatment, household dysfunction, and several areas of risk in their own lives. Participants also represented children across the developmental continuum from 1.5 years to 18 years. The inclusion of two generations allowed for the exploration of intergenerational risk transmission. Families in this study also represented a sample of those involved in in-home placements, providing an exploration of existing risks that may benefit from child protective services in order to improve child behaviors.
This study also has limitations that should be acknowledged. In the direct reporting of personal difficulties, mothers may have been subject to reporter bias if they were fearful or hopeful that identifying personal risks would negatively or positively affect perceptions of their parental fit. That is, risks could be construed as limitations to current functioning and future parenting ability, but also as explanations for the challenges a mother faces encouraging an empathic response. Some mothers may have censored their replies based on their comfort acknowledging particular circumstances. For example, some mothers may have felt pressured to minimize their reporting of child psychopathology following reports of higher ACE levels. Thus, data may offer a conservative estimate of trauma and risk exposure. Mothers who participated in the study came from a large population of families in this state’s CWS. They do not represent the experiences of those whose children were placed in out-of-home services. Mothers with in-home child placements could have risk profiles that are distinct from mothers with children in alternative placements, but are a largely understudied population of child welfare involved families despite the growing focus on in-home services in many jurisdictions. Although all children included in this study experienced maltreatment, the nature of abuse or neglect cannot be specifically characterized and is not examined in the association of mothers’ risks and children’s behaviors. The risk factors considered in this study are based on high risk and low risk determinations made in the existing literature, but could have alternative thresholds in other studies; the reader is encouraged to use caution in making comparisons of risk across samples.
Implications
Practice and policy
Identification of a mother’s personal risks, including exposure to ACEs, has implications for how the CWS interacts with families. Children’s risks may be associated with previously existing maternal risks and the accumulation of current family risks. Cumulative risks, which are reflective of personal experiences of hardship, lack of support, limited resources, and structural inequality, may influence children’s withdrawing and aggressive behaviors. This association suggests that caseworkers have an opportunity to explore a family’s situational challenges in order to develop a comprehensive case plan that will improve a child’s environment. To adequately understand the child’s situation, an examination of risks associated with both child and parents is needed. Supporting child well-being necessitates that maternal risks be alleviated as much as possible to interrupt their consequences, which may be amplified for mothers who themselves experienced early adversity and trauma, including their potential influence on parenting interactions. In recognizing and responding to ecological constraints, the CWS must expand the focus of current intervention standards. This shift requires transitioning from a singular emphasis on improving child and caregiver behaviors to a broader emphasis on improving the contextual circumstances that influence and maintain disruptive behaviors, including those related to the environment and family.
Further, cumulative risk indicates the necessity of taking a broader view of assessment and intervention. Aspects of well-being such as maternal health and household functioning are related to both caregivers’ and children’s circumstances, while ACEs are restricted to a caregiver’s childhood. The persistence of risk though all developmental stages from a mother’s own childhood to her parenthood provides many potential ports of entry for the systems of care to support individuals in negotiating personal and situational improvements. As children age, their development of personal insight may also increase their awareness of family and environmental dynamics influencing depressive and withdrawal behaviors. Likewise, children are progressively exposed to new interpersonal contacts external to their immediate families, from peers to educational systems, which may mitigate or exacerbate behavioral responses to family difficulties. This persistence in internalizing and externalizing behavioral challenges among children of different developmental stages suggests that mothers engaging with the CWS have a high likelihood of transitioning from their own adverse childhood exposures to stable adult cumulative risk exposures.
Child well-being and safety cannot be achieved in isolation from family well-being and safety. For CWSs to achieve the strongest safety and permanency plans possible, policies must support assessment of parental risk factors through identification of stressors and service delivery to ease these pressures. Furthermore, in order to fully alleviate cumulative risks, strong economic and social policies are needed to address poverty, structural inequalities, and violence. Policies that address the consequences of these issues are important for attending to individual and family functioning; however, expanding access and delivery of services at an individual level is incomplete. Cumulative risks cover a wide range of ecological factors that are outside the purview of the CWS to address. Strong public policies supporting prevention efforts, including offering home visiting services to families at risk of child welfare involvement, are critical to responding to the wide variety of family challenges. Moreover, public policies that improve opportunities for individuals to participate in quality education, earn fair wages, and live in adequate housing provide the strongest prevention efforts to promote family success. Only a comprehensive community response that both manages symptoms such as housing instability and eliminates the sources of these burdens such as economic inequality is likely to adequately address the intersection of cumulative risk, child welfare involvement, and child behavioral outcomes.
Research
The current research identified an association between maternal risks and children’s behaviors. Longitudinal research is needed to understand this relationship over time. Further, research that considers the nature of child maltreatment on the relationship between risks and behaviors is needed. Future work concerned with maternal ACEs and children’s behaviors should include specifics about the number, type, severity, and frequency of maltreatment experienced. Additionally, research examining mechanisms for risk transmission should include the quality of parent–child relationships, family dynamics, and family social support and resources.
Conclusion
This study contributes to the growing knowledge base about the intergenerational effects of ACEs, while also acknowledging the important role of continuities and discontinuities in cumulative risk in shaping these outcomes across development. Research on ACEs, cumulative risks, and child outcomes is especially critical for child welfare research and demonstrates the need to address both individual and ecological factors affecting parenting-child relationships and caregiving environments over time. By situating this research within a child welfare sample, this study is able to generate meaningful implications for research, practice, and policy. We are likely to learn more in the coming years about potential mechanisms for direct effects of ACEs on downstream outcomes, including biological stress processes, epigenetic expression patterns, and stability in early behavioral and neurological organization, but our results highlight the importance of not losing sight of a developmental perspective and the unfortunately stable nature of risk. When it comes to studying risk, simply associating early adversity with later outcomes may lead us to false assumptions about direct causal pathways if we fail to appreciate in our approach the tendency for harsh social environments to be stable and for risk exposure to exhibit developmental continuity and cumulative effects.
Footnotes
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: This research was supported in part by a William T. Grant Foundation Scholar award to Michael J. MacKenzie.
