Co-development of internalizing and externalizing behavior problems during early childhood among child welfare-involved children
Introduction
Although prior research suggests that internalizing and externalizing behavior problems commonly co-occur, both concurrently and consecutively (Gilliom & Shaw, 2004), this has not been studied with child welfare–involved children who are at elevated risk for developing internalizing and externalizing behavior problems (McCrae, 2009). Internalizing and externalizing behavior problems in early childhood are of serious concern as early behavior problems tend to persist throughout childhood and adolescence (Ashford, Smit, Van Lier, Cuijpers, & Koot, 2008), serving as a marker for later psychopathology (Pihlakoski et al., 2006). Furthermore, children exhibiting continuous co-occurring internalizing and externalizing behavior problems are at heightened risk for experiencing subsequent negative outcomes, including peer rejection, association with deviant peers, engagement in risky behaviors, and substance use during adolescence (Colder et al., 2013, Fanti and Henrich, 2010). Examining cross-domain associations and early risk factors of internalizing and externalizing behavior trajectories during early childhood, defined here as the period from birth to 5 years (Votruba-Drzal, 2006), among children in the child welfare system is important to develop effective early intervention strategies that can promote positive development for these vulnerable children. Using parallel-process growth curve modeling, the current study aimed to extend the current knowledge base by investigating the developmental course of internalizing and externalizing behavior problems in early childhood and identifying risk factors among children involved in the child welfare system.
Children involved in the child welfare system are at a heightened risk for experiencing internalizing and externalizing behavior problems (Burns et al., 2004, McCrae, 2009). Internalizing behavior problems are characterized by inward-directed feelings and include symptoms of depression, anxiety, withdrawal, and somatic complaints, whereas externalizing behavior problems refer to acting-out behaviors directed toward others, such as aggressive and rule-breaking behaviors (Achenbach & Edelbrock, 1978). Despite a high volume of research indicating elevated levels of internalizing and externalizing behavior problems among children involved in the child welfare system (Aarons et al., 2010, McCrae, 2009), studies examining the developmental trajectories of internalizing and externalizing behavior problems in maltreated children are scarce, and the few existing studies have yielded inconsistent findings (e.g., Godinet et al., 2014, Kim et al., 2009, Thompson and Tabone, 2010). Furthermore, no research has examined trajectories during toddlerhood despite empirical evidence suggesting that most persistent and serious behavior problems develop as early as the toddlerhood/preschool period (Moffitt & Caspi, 2001). For internalizing behavior problems, two studies reported a similar trajectory pattern (a curvilinear, smooth inverse U-shaped trajectory), in which internalizing behavior problems peaked around ages 8 and 9 and slightly decreased afterward, following children from age 6 to age 14 (Kim et al., 2009) and age 4 to age 10 (Thompson & Tabone, 2010). In contrast to these studies, one study found a flat trajectory of internalizing behavior problems (i.e., no significant increase or decrease over time) from age 4 to age 12 (Godinet et al., 2014). Findings related to externalizing behavior problems have been more mixed, with Kim et al. (2009) reporting a smooth inverse U-shaped trajectory, Thompson and Tabone (2010) reporting a flat trajectory, and Godinet et al. (2014) reporting a declining trajectory. While these studies offer valuable information regarding internalizing trajectories during childhood, the samples used in the studies were geographically limited (Kim et al., 2009: a Northeastern city; Thompson & Tabone, 2010: 5 study sites in the U.S.). The current study addresses such limitations and extends the existing literature by using a nationally representative sample of children involved in the child welfare system.
Identifying early predictors of internalizing and externalizing trajectories is critical to preventing and reducing the development of serious internalizing and externalizing behavior problems over time. Previous studies have identified several early risk factors, including child maltreatment, intimate partner violence, caregivers' mental health problems, caregivers' substance use problems, out-of-home placement, and caregiver-child attachment (Dance et al., 2002, Godinet et al., 2014, Mian et al., 1996).
Early childhood maltreatment, including physical abuse (Egeland et al., 2002, Lansford et al., 2007), sexual abuse (Mian et al., 1996), emotional abuse (Dance et al., 2002), and neglect (Kotch et al., 2008) occurring in the first 5 years of life has been found to have a strong association with children's internalizing and externalizing behavior problems. However, longitudinal studies examining the influence of early childhood maltreatment on the growth factors (intercept, slope) of trajectories of internalizing and externalizing behavior problems have produced inconsistent findings. Thompson and Tabone (2010) found that early maltreatment (birth–age 4) predicted only the slope of the internalizing trajectory, but not the intercept, meaning there were no baseline differences of internalizing problem rates among maltreated and non-maltreated children, though maltreated children exhibited significantly greater increases in internalizing problems over time. On the contrary, Godinet et al. (2014) found that early maltreatment (birth–age 4) was significantly related to the intercept but not the slope of behavior trajectories in boys. Maltreated boys exhibited higher levels of internalizing and externalizing behavior problems at the initial assessment (i.e., age 4), but the rate of change in their behavior problems did not differ from those without early maltreatment experience.
Exposure to intimate partner violence (IPV) has also been related to elevated internalizing and externalizing behavior problems among young children (Evans, Davies, & DiLillo, 2008). Children younger than age 6 are particularly vulnerable for exposure to IPV, because they are more likely to be present in the home when violence occurs (Fantuzzo & Fusco, 2007). Using a sample of child welfare–involved children, Holmes, Yoon, and Berg (2017) found that exposure to IPV during toddlerhood was significantly related to increased aggression in toddlerhood and the preschool/kindergarten period.
Caregivers' mental health problems, including depression and anxiety disorders, are another well-known risk factor for children's behavior problems in early childhood (Avan et al., 2010, Edwards and Hans, 2015). A review of studies regarding the association between caregiver’ mental health and child's behavior problems suggested that parenting disruption was the central mechanism through which parental mental disorders transmitted to child's behavior problems. Caregivers' substance use also tends to increase the risk of children's behavior problems, partly due to compromised parenting capacity (Child Welfare Information Gateway, 2014). Parental substance use has been linked to children's internalizing (Fusco & Cahalane, 2013) and externalizing (Osborne & Berger, 2009) behavior problems during early childhood. Additionally, caregiver substance use was found to be predictive of children's worsening externalizing behavioral trajectory in a sample of 827 maltreated children (Tabone et al., 2011).
Early behavior problems among children in the child welfare system may be largely influenced by the arrangement of placement (Halfon, Mendonca, & Berkowitz, 1995). Studies have suggested higher levels of internalizing and externalizing behavior problems for children younger than 6 in out-of-home care (Simms, Dubowitz, & Szilagyi, 2000), with approximately 56% of children aged 2 and 40% of children aged 3–5 showing clinical levels of behavior problems in a nationally representative child welfare sample (Leslie et al., 2005).
Finally, caregiver-child attachment may exert strong influence on early childhood internalizing and externalizing behavior problems. Meta-analysis studies indicated a relatively large effect size for the association between insecure attachment and externalizing behavior problems, especially among high-risk boys (Fearon, Bakermans-Kranenburg, van IJzendoorn, Lapsley, & Roisman, 2010), and a significant association between insecure attachment and internalizing behavior problems, with small to medium effect size (Madigan, Atkinson, Laurin, & Benoit, 2013).
Prior studies enhanced the understanding of risk factors for early childhood internalizing and externalizing behavior problems, but these studies have usually used cross-sectional design, and rarely considered co-development of internalizing and externalizing behavior problems. Longitudinal cross-domain trajectory research focusing on young children in the child welfare system is needed to gain a better understanding of the course of behavioral development among these high-risk children and to provide them with most effective interventions to ensure their healthy and positive behavioral functioning.
A substantial body of research has indicated that internalizing and externalizing behavior problems commonly co-develop (Gilliom and Shaw, 2004, Keiley et al., 2000, Lee and Bukowski, 2012, Oland and Shaw, 2005), yet directional explanations of these relationships have been mixed. A robust body of literature supports a mutually reinforcing relationship, suggesting that the increase or decrease in one domain leads to the same trend of symptoms in the other domain (Gilliom and Shaw, 2004, Keiley et al., 2000). In these studies, as symptoms in one domain increase or decrease, so do symptoms in the other domain (Gilliom and Shaw, 2004, Keiley et al., 2000, Lee and Bukowski, 2012, Lee and Stone, 2012). A number of studies, in contrast, suggest a unidirectional relationship, with externalizing behavior problems leading to the development of internalizing problems (Boylan et al., 2007, Patterson and Capaldi, 1990). For example, Capaldi (1991) reported that early conduct problems predicted later depressed mood symptoms, but early depressed mood symptoms did not predict later conduct problems in an at-risk sample of 203 preadolescent boys. Conversely, others have indicated that internalizing behavior problems precede externalizing behavior problems (Bornstein et al., 2010, Ritakallio et al., 2008). A study of 117 children from an East Coast metropolitan area found that internalizing behavior problems at age 4 predicted greater externalizing behavior problems at age 14 (Bornstein et al., 2010). To date, the co-development and cross-domain relations of internalizing and externalizing behavior problems have not been studied with a child welfare–involved population, during early childhood in particular, which is a research gap worth future investigation.
In summary, little research has examined the co-development and cross-domain relations of internalizing and externalizing behavior trajectories among child welfare-involved children, despite the high prevalence of internalizing and externalizing behavior problems in this population (Burns et al., 2004, McCrae, 2009). Previous findings from general population studies on cross-domain relationships of internalizing and externalizing behavior problems have been mixed, with some studies reporting mutually reinforcing bi-directional relationship (Gilliom & Shaw, 2004) and others reporting one domain preceding the other domain (externalizing as an antecedent: Boylan et al., 2007; internalizing as an antecedent: Bornstein et al., 2010). While child maltreatment, child exposure to IPV, out-of-home placement, caregiver mental health, substance use, and caregiver-child insecure attachment have been identified as early risk factors for internalizing and externalizing behavior problems, the longitudinal impact of such early risk factors on the co-development of internalizing and externalizing behavior problems over time remains unclear. To fill the gaps in the literature, we sought to examine the cross-domain associations and early risk factors of internalizing and externalizing behavior trajectories in young children involved in the child welfare system.
We applied the developmental psychopathology perspective, which concerns the origins, course, and outcomes of an individual's functioning (Sroufe & Rutter, 1984). This perspective seeks to understand how children's early life experiences and subsequent environmental challenges affect their later developmental outcomes by emphasizing the concept of ‘developmental pathways,’ underscoring the importance of a process-level investigation of developmental outcomes (e.g., longitudinal trajectories) (Cicchetti, 1993). Drawing from developmental psychopathology, this study seeks to answer the following research questions: (a) What are the initial levels and growth rates of internalizing and externalizing behavior problems in early childhood among children who have been involved in the child welfare system? (b) How are the trajectories of internalizing and externalizing behavior problems associated across domains? (c) How are early risk factors related to the developmental trajectories of internalizing and externalizing behavior problems? The conceptual model is depicted in Fig. 1.
Section snippets
Sample
This study is a secondary data analysis using data from the National Survey of Child and Adolescent Well-Being (NSCAW-II), which is a nationally representative longitudinal study of children and families involved in the child welfare system. The NSCAW-II baseline sample includes 5872 children, ages birth to 17.5 years at the time of sampling, who were subjects of child abuse or neglect investigations conducted by child protective services (CPS) closed between February 2008 and April 2009.
Sample characteristics
A little less than half of the children were girls. Children were 2 years old at W1, 3.5 years old at W2, and 5 years old at W3. About 42% of the sample were White/non-Hispanic, 28% were Black/non-Hispanic, 27% were Hispanic, and 5% were other race. Approximately one quarter of the caregivers had less than a high school education. Descriptive statistics of study variables are presented in Table 1.
Unconditional parallel-process model
Fig. 2 presents the unconditional parallel-process model of internalizing and externalizing behavior
Discussion
Guided by the developmental psychopathology perspective, the present study examined the developmental trajectories of behavior problems (both internalizing and externalizing) and assessed early risk factors for those trajectories among child welfare-involved children from ages 2 to 5. It contributes to the field of child psychopathology by expanding our knowledge on co-development and cross-domain relations of internalizing and externalizing behavior trajectories in early childhood among
Conclusion
We found a positive correlation between internalizing and externalizing behavior problems in toddlerhood as well as positively correlated rates of change in symptoms throughout the early childhood period, with a group of early risk factors predicting these problem behavior trajectories. Taken together, the results of this study highlight the importance of comprehensive screening, assessment and treatment for young children, as they are likely to have co-occurring internalizing and externalizing
Acknowledgments
This document includes data from the National Survey on Child and Adolescent Well-Being, which was developed under contract with the Administration on Children, Youth, and Families, U.S. Department of Health and Human Services (ACYF/DHHS). The data were provided by the National Data Archive on Child Abuse and Neglect.
Conflict of interest
The authors declare that they have no conflict of interest.
References (85)
- et al.
Behavior problems and placement change in a national child welfare sample: A prospective study
Journal of the American Academy of Child & Adolescent Psychiatry
(2010) Reliability and validity of temperament scales of the NLSY child assessments
Journal of Applied Developmental Psychology
(1995)- et al.
Mental health need and access to mental health services by youths involved with child welfare: A national survey
Journal of the American Academy of Child & Adolescent Psychiatry
(2004) Developmental psychopathology: Reactions, reflections, projections
Developmental Review
(1993)- et al.
Exposure to domestic violence: A meta-analysis of child and adolescent outcomes
Aggression and Violent Behavior
(2008) Prenatal depression effects on early development: A review
Infant Behavior and Development
(2011)- et al.
Primary care clinicians' use of standardized tools to assess child psychosocial problems
Ambulatory Pediatrics
(2003) - et al.
Early childhood maltreatment and trajectories of behavioral problems: Exploring gender and racial differences
Child Abuse & Neglect
(2014) Depressive disorders in maltreated children
Journal of the American Academy of Child & Adolescent Psychiatry
(1991)- et al.
Co-development of internalizing and externalizing problem behaviors: Causal direction and common vulnerability
Journal of Adolescence
(2012)