Introduction
Trauma represents a significant public health issue given its current prevalence in the United States (Kilpatrick et al.
2013). Typically, traumatic events are assessed through structured tools, including the Diagnostic and Statistical Manual of Mental Disorders (DSM) interview and the Traumatic Life Events Questionnaire (Peirce et al.
2009). Examples of traumatic events included on the DSM-5 interview for post-traumatic stress disorder (PTSD) include disasters, accidents, war exposure, experiencing or witnessing assault, and the death of, or threat of injury to, a close individual due to disasters, accidents or violence (Kilpatrick et al.
2013). While trauma has been assessed and defined in a number of ways, recent research using the DSM-5 found that 89.7% of individuals in the United States have experienced a traumatic event at some point in their lives (Kilpatrick et al.
2013).
Adverse childhood experiences (ACEs) are one conceptualization of trauma that have gained increasing attention in recent years. While the conceptualizations of trauma, such as the one found in the DSM-5, are fairly broad and include a number of experiences, such as war and disasters (Kilpatrick et al.
2013), typical scales used to assess ACEs do not encompass all of these experiences. Instead, typical assessments of ACEs include three general categories—childhood abuse, neglect, and household dysfunction (Centers for Disease Control and Prevention
2016). Additionally, ACEs are specific to a certain timeframe, as they only include experiences from when an individual was 18 years of age or younger (Centers for Disease Control and Prevention
2016). This is a critical component of the conceptualization of ACEs, as this early exposure to trauma, and the subsequent stress related to this repeated exposure, has been found to be associated with the disruption of a child’s developing brain (National Scientific Council on the Developing Child
2005/2014). For example, over time, this repeated exposure to stressful experiences (toxic stress), can alter stress responses and an individual’s ability to regulate themselves (Anacker et al.
2014; Loman and Gunnar
2010).
ACEs can result in a number of adverse outcomes for individuals. For example, ACEs have been associated with behavioral difficulties (Moore et al.
2014), adolescent pregnancy (Hillis et al.
2004), poorer educational outcomes and unemployment (Liu et al.
2013; Moore et al.
2014), substance use and abuse (Anda et al.
2006; Dube et al.
2003,
2006), poorer physical health outcomes (Anda et al.
2006; Chartier et al.
2010; Felitti et al.
1998), and poorer mental health outcomes or symptoms (Afifi et al.
2008; Anda et al.
2006; Chapman et al.
2004; Danese et al.
2009; Dube et al.
2001; Edwards et al.
2003; Whitfield et al.
2005), with a graded (dose–response) relationship being found between ACEs and a number of these outcomes (Afifi et al.
2008; Anda et al.
2006; Chapman et al.
2004; Chartier et al.
2010; Danese et al.
2009; Dube et al.
2001,
2003,
2006; Edwards et al.
2003; Felitti et al.
1998; Hillis et al.
2004; Moore et al.
2014; Whitfield et al.
2005). Further, research with low-income women has shown that specific ACEs, such as abuse and neglect, have been associated with reduced social support (Vranceanu et al.
2007).
Given the pernicious outcomes for both adults and children that result from trauma, several research studies have sought to examine how traumatic experiences may affect parenting practices. One study with low-income, parenting women found that “physical abuse was associated with increased hostile-intrusive behavior toward the infant,” while “sexual abuse was associated with decreased involvement with the infant” (Lyons-Ruth and Block
1996). Further, a number of studies have shown that outcomes associated with trauma, including mental illness, are associated with parenting behaviors, such as insecure parent–child attachment (bond) and decreased maternal sensitivity (responding to a child’s signals) (Downey and Coyne
1990; Lovejoy et al.
2000).
While specific types of trauma, such as physical abuse, sexual abuse, or neglect, have been studied in relation to future parenting practices (Bert et al.
2009; Hughes and Cossar
2016; Lyons-Ruth and Block
1996), there has been limited research specifically on the totality of ACEs in relation to parenting stress and practices (Chung et al.
2009; Steele et al.
2016). Additionally, though studies have focused on specific forms of childhood trauma and later parenting practices, the maximum age at which these acts of trauma can occur often vary by study. Thus, research focusing specifically on trauma a mother experienced before the age of 18 may be important. As such, this research aimed to (1) understand the relationship between a mother’s own experience of ACEs before the age of 18 and her current parenting stress and practices, as measured by several parenting scales; and (2) to determine if a dose–response relationship existed between a mother’s experience with ACEs and parenting stress and practices. Determining whether a dose–response relationship exists is especially important, as understanding the incremental nature of the effect of ACEs and other forms of trauma on parenting can allow us to better understand the mechanism whereby cumulative stress, trauma and adversity impact parenting and could potentially help to target interventions based on the moderating effect of the number of ACEs a mother has experienced.
Discussion
Prior to this study, research had shown a relationship between specific traumatic events and parenting practices (Lyons-Ruth and Block
1996). However, limited research has been conducted specifically on the totality of a parent’s own early experience of ACEs in relation to a parent’s current parenting stress and practices (Chung et al.
2009; Steele et al.
2016). As such, this study aimed to understand this association, including whether a dose–response relationship exists, using measures of parenting stress (PSI-SF) and Positive Parenting Practices (PPP), in relation to a parent’s own ACEs.
Previous research has shown that ACEs have been associated with a number of negative long-term outcomes. This study was able to add to the existing literature by showing that more ACEs experienced early in life by a mother are positively associated with a mother’s current parenting stress, and that this association follows a dose–response relationship. Specifically, significant, dose–response relationships were observed at the p < 0.05 level for the PSI Total Stress score, where each additional ACE was associated with a 3.19 increase in percentile, and for the Difficult Child subscale, where each additional ACE was associated with a 3.69 increase in percentile. Additionally, at the p < 0.10 level of significance, each additional ACE was associated with a 2.45 increase in percentile on the Parental Distress subscale. However, no statistically significant relationship was observed on the Parent–Child Dysfunctional Interaction subscale of the PSI-SF. Further, no statistically significant relationship was observed for the PPP scale in relation to ACEs.
One potential mechanism for the increased parenting stress found in this study might be the dysregulation of the stress-response system caused by traumatic experiences as a child. Research has shown that early traumatic experiences can have an adverse effect on a number of important biological functions, and subsequent long-term outcomes, as they can cause “cumulative damage over time” and can embed “adversities during sensitive developmental periods” (Shonkoff et al.
2009). Further, recent research has shown that changes occur to the adult brain during the transition to parenthood (Kim et al.
2010,
2014), and these changes could be influenced by early childhood experiences, such as trauma.
For example, over time, biological mechanisms related to stress, including processes related to the hypothalamic-pituitary-adrenocortical (HPA) axis, are dysregulated, which prevent the body from “returning to homeostatic balance” (Shonkoff et al.
2012). Thus, individuals who have had these repeated exposures are likely to experience difficulty when faced with subsequent stressful experiences, as their stress systems have already been unduly burdened, and will not be able to as effectively regulate bodily processes related to stress. This may explain why parenting women in this study who experienced ACEs, and in particular, multiple ACEs, were shown to have increased levels of parenting stress, as measured by the PSI-SF.
Due to the increased levels of stress that parenting women who have experienced ACEs may have, and due to the subsequent dysregulation of the stress system, it is possible that the parenting styles of these women could be affected. Baumrind posits that there are four distinct parenting styles—authoritarian, authoritative, neglectful, and permissive (Baumrind
1991). One recent study has shown that “mothers who experience high levels of trauma symptoms are more likely to parent using authoritarian or permissive behaviors,” as measured by the Parenting Practices Questionnaire (which does not include neglectful parenting styles) (Leslie and Cook
2015). Further, a study looking at poor parenting practices, which includes factors such as neglect and aggression toward the child, found that maltreatment as a child was associated with poor parenting practices for mothers, and that childhood sexual abuse specifically was associated with aggressive parenting behaviors (Newcomb and Locke
2001). These findings are of critical importance, given that certain parenting behaviors captured within these parenting styles, such as neglect, are considered to be ACEs (Centers for Disease Control and Prevention
2016). Thus, it is possible for an inter-generational transmission of trauma to occur (Newcomb and Locke
2001).
As discussed, statistically significant associations on the Total Stress score, Parental Distress, and Difficult Child subscale with ACEs were found. The Parental Distress subscale is designed to assess factors, including stress, relationships, or mental health issues, that could affect an individual’s parenting practices (Abidin
2012). Research has shown that ACEs may make it difficult for individuals to cope with added stressors (American Academy of Pediatrics
2014), and as such, it is likely that increased ACE scores were associated with elevated percentiles on this scale, as parenting women may have had difficulty responding to additional parental stressors, such as parental conflict. Additionally, previous research has shown that early traumatic experiences are associated with reduced social support in adulthood (Vranceanu et al.
2007). Thus, parenting women may have seen elevated percentiles on the Parental Distress subscale due to more limited social networks, as more robust social networks, which could have helped buffer against stress, were not available. However, this study cannot confirm this association, as mediating factors were not measured.
The Difficult Child subscale is designed to capture difficulties that may occur for parents related to their child’s behavior (Abidin
2012). Recent research with parenting women has shown that mothers with mental health and substance use challenges were more likely to have children who experienced behavioral difficulties (Whitaker et al.
2006). Given that studies have shown that ACEs are associated with a number of adverse mental health outcomes (Afifi et al.
2008; Anda et al.
2006; Chapman et al.
2004; Danese et al.
2009; Dube et al.
2001; Edwards et al.
2003; Whitfield et al.
2005), it is possible that the increase in percentiles on the Difficult Child subscale, associated with the experience of each additional ACE, may be due to mental health outcomes related to the experience of trauma.
A significant association between ACEs and the parent–child dysfunctional interaction subscale was not found. This subscale is designed to measure the whether the child is seen as a negative part of a parent’s life (Abidin
2012). This finding could be due to the homogeneity observed in the sample on this measure, as only 8 individuals had scores indicating high or significant stress, with the majority of the sample having scores indicative of normal or below normal stress in this domain. Additionally, this result may be due to small sample size.
Finally, no significant association was found between the PPP scale and ACEs. This result is likely due to the significant homogeneity in results on this scale, as 72 of the 81 participants had an average score of 5, which is the highest possible score on the scale, and indicates that the mother almost always reports engaging in all behaviors (Dahlberg et al.
2005). It is possible that due to social desirability bias, women could have over-reported the rewarding behaviors described in the survey, perhaps believing that not endorsing some items, such as giving a “hug, pat on the back, or kiss” (Dahlberg et al.
2005), could be viewed unfavorably.
Given the relationship found between ACEs and parenting stress, it is important to develop both psychosocial and policy interventions to address these issues. Psychosocial research examining potential parenting interventions for adults will be especially critical, because research has shown that the negative effects of trauma can be lessened with early intervention (NGA Center for Best Practices, National Conference of State Legislatures, & Center on the Developing Child at Harvard University
2007), and as such, many interventions have been developed for younger populations (Wethington et al.
2008), while limited research has been completed related to adult populations.
Additionally, policies could be created, which seek to develop environments that are more sensitive and aware of the potential consequences of trauma. Policies creating more sensitive and aware school environments have been advocated for by the trauma-informed school movement (Cole et al.
2013), while policies creating more sensitive and aware community environments and service systems have been advocated for by the trauma-informed systems movement (Ko et al.
2008). Further, policies could be created to help aid adults who have experienced ACEs, such as policies that promote affordable access to appropriate mental health and parenting services.
Additionally, given that a dose response-relationship has been found in relation to ACEs and parenting stress within this study and in past studies related to ACEs and adverse outcomes (Afifi et al.
2008; Anda et al.
2006; Chapman et al.
2004; Chartier et al.
2010; Danese et al.
2009; Dube et al.
2001,
2003,
2006; Edwards et al.
2003; Felitti et al.
1998; Hillis et al.
2004; Moore et al.
2014; Whitfield et al.
2005), this research could be used to help inform genetic arguments. For example, Reiss et al. have posited a relationship between genes and the social environment, suggesting, among other mechanisms, that “genes can influence an individual’s response to environmental stress, genes may enhance an individual’s sensitivity to both favorable and adverse environments…” (Reiss et al.
2013). Thus, future studies should look not only at adverse outcomes of ACEs, but also at potential genetic influences.
Strengths and Limitations
This study was able to address a critical gap in the literature on the effect that the totality of ACEs have on parenting stress and practices. This study benefited from the use of the PSI-SF, which is a validated measure that has been used in a number of studies with parents (Abidin
2012; Ispa et al.
2004; Lecavalier et al.
2006; Smith et al.
2001).
While the study was able to address this gap, there are a number of limitations that must be considered. Many of these limitations stem from the fact that this was a secondary analysis of data for a study not initially designed to assess the effect of ACEs on parenting. First, the sample size for this study was small. Given additional statistical power, it is possible that additional associations could have been identified. Second, this study focused on a homogenous population of low-income, parenting women living in public housing who scored highly on the CES-D. Thus, results may not be generalizable to other populations, including parenting men, those not living in public housing, those of other socio-economic statuses, and those not currently experiencing depressive symptoms.
Third, all measures of parenting relied on the self-report of mothers and thus single-source bias could exist. Additionally, these measures could be confounded by a mother’s own psychological status. Further, as with all parenting measures, social desirability bias may be present, which could lead to a lack of variability in responses, as was found with the PPP scale. It is possible that additional measures, such as the independent observation of parenting behaviors, could have helped triangulate the findings. While many studies investigating the associations between specific ACEs and parenting have relied on participant self-report, others have included participant observation to assess parenting behaviors (Bailey et al.
2012; Baumgardner
2007; Koren-Karie et al.
2008; Pasalich et al.
2016), and this could be an important avenue for future research. Additionally, it may be important to measure potential factors that may be affecting the parenting of mothers. Specifically, the behavior of children could affect the parenting stress a woman experiences. Thus, future studies may benefit from assessing the behavior of children through multiple sources. Further, potential mediators between ACEs and parenting, such as mental health, substance use, and financial issues, should be examined in future research.
Fourth, there are potential limitations related to the measurement of ACEs that must be considered. For example, questions related to ACEs, such as sexual abuse, are sensitive topics. As such, it is possible that women under-reported their experiences. Additionally, as described in the “
Methods” section, ACEs were measured through eight yes/no questions. However, the full ACEs battery consists of additional questions, which may have more accurately assessed ACEs (Centers for Disease Control and Prevention
2016). Additionally, it is important to consider the distribution of ACEs within this sample. Specifically, the majority of the sample had ACE scores of three and under, with only 4 individuals having ACE scores of 6 and above. As such, there may be some uncertainty regarding the estimates in the higher range of the distribution. Finally, though there are benefits to measuring the totality of ACEs, some research has suggested that it is beneficial to instead examine the characteristics of specific adverse experiences (Dunn et al.
2018).