Adverse childhood experiences (ACEs) have a powerful impact on adolescents’ development. The greater the number of stressful life events experienced, the more likely adolescents are to experience internalizing symptoms, such as anxiety. Due to the negative effects of early adversity, it is essential that research examine which protective factors might mitigate this relationship, such as emotion regulation (ER) and attachment. In the current study, the researchers examined the relationships between ACEs, ER, attachment, and anxiety among adolescents. We hypothesized that ER would mediate the relationship between ACEs and anxiety and attachment would moderate the relationship between ACEs and ER within the mediation model. A sample of adolescents (n = 105) completed measures of adolescent anxiety, ACEs, ER, and attachment. Using the PROCESS macro, a moderated mediation model was analyzed. Mother and father attachment moderated the mediation of difficulties in ER on ACEs and anxiety, suggesting that mean levels of mother and higher levels of father attachment function as a mechanism that strengthens the relationship between early negative life events and current anxiety. These findings suggest that the mother/father attachment relationship for adolescents might be of particular importance for youth who have experienced adverse life experiences and have trouble regulating their emotions.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Adverse stressors early in life have a powerful impact on adolescents’ development. The greater the number of stressful life events experienced, the more likely adolescents are to experience internalizing symptoms, such as depression and post-traumatic stress (Suliman et al., 2009). Additionally, certain types of early life stressors (e.g., abuse and neglect) are associated with specific internalizing symptoms, such as anxiety (Suliman et al., 2009). Early adverse life experiences have been shown to lead to a variety of poor outcomes that can last long after childhood (Anda et al., 2010; Felitti et al., 1998; Salokangas et al., 2024). Due to the negative effects of early adversity, it is essential that research examine potential protective factors, such as emotion regulation and attachment, that might mitigate the impact of early adversity on adolescent anxiety.
Adverse Childhood Experiences
Adverse childhood experiences (ACEs) are defined as exposure to physical, sexual, and emotional abuse, physical and emotional neglect, and dysfunction in the household (i.e., inter-personal violence, divorce or separation of parents, mental illness, criminal acts, or substance abuse; see Supplemental Material; Felitti et al., 1998) that are experienced prior to eighteen years of age (Poole et al., 2017). The seminal research studies on ACEs (Dube et al., 2001; Felitti et al., 1998) yielded a theoretical model that conceptualized childhood adversity based on cognitive, emotional, and social problems, subsequently producing high-risk behaviors. With the addition of chronic effects of physiological stress, these high-risk behaviors increase the risk of disability, disease, and premature death. The healthcare burden of ACEs (Cameron et al., 2018) is estimated to cost an average of over $200,000 in additional fees over the lifetime of someone who experienced child maltreatment (Fang et al., 2012). Results from the 2016 National Survey of Children’s Health revealed that approximately half of the children (46.3%) and adolescents (55.7%) in the United States have experienced at least one ACE (Bethell et al., 2017). While some studies have focused on ACEs and related outcomes in adolescents, most of these studies focus on a few specific behaviors in relation to ACEs (e.g., violence and substance use; Meeker et al., 2021). Additional research on ACEs and related outcomes within adolescent populations is necessary for a better understanding of the present impact ACEs have on mental health (Meeker et al., 2021). As the number of ACEs experienced increases, adolescents experience more internalizing and externalizing problems (Oei et al., 2023). Youth with two or more ACEs have a greater likelihood of needing health care support, repeating a school year, and other negative outcomes (Bethell et al., 2017). These negative outcomes display the lower thresholds with which child development is affected by trauma (Bethell et al., 2017). Meeker et al. (2021) indicated that the relationship between trauma and risky behaviors emerges during adolescence, and thus early interventions should target this period of development. Studies have reliably shown that adolescents with a history of ACEs have an increased risk of a wide variety of poor health outcomes, including violence, suicidality, and risky behaviors (Duke et al., 2010; McLaughlin et al., 2011; Meeker et al., 2021). Due to ACEs’ high prevalence, substantial health difficulties, and significant healthcare costs, research should address the increasing need for intervention (Cameron et al., 2018), especially in youth due to negative impacts on their social-emotional, neurological, and behavioral health (Meeker et al., 2021).
ACEs and Anxiety Among Adolescents
Of the studies that have examined the relationship between ACEs and anxiety within adolescent populations, many studies indicate that ACEs have a large impact on a number of challenges that adolescents face within school systems (Lee et al., 2020). A few of these challenges that adolescents face include emotion dysregulation (Bradley et al., 2011), decreased academic performance (Bethell et al., 2014), and internalizing disorders (Kieling et al., 2011). Of the internalizing disorders related to experiences of ACEs, anxiety and depression are the most prevalent mental health concerns for adolescents (Kieling et al., 2011; Lee et al., 2020). Anxiety disorders have a significant influence on physical health, social and interpersonal functioning, and cognition (Teubert & Pinquart, 2011). Additionally, adolescents with anxiety disorders are likely to have greater adverse psychosocial consequences later in adulthood. The more ACEs an adolescent experiences, the more they are likely to have severe anxiety (Lee et al., 2020).
Emotion Regulation and Anxiety for Adolescents Experiencing ACEs
In this study, the researchers conceptualized emotion regulation (ER) as a multidimensional construct including the assessment of and the behavioral and expressive responses to specific stimuli (Gratz & Roemer, 2008; Naragon-Gainey et al., 2017; Rudenstine et al., 2019). In other words, ER is defined as the ability to efficiently evaluate, monitor, and regulate emotional responses concerning valence, duration, and intensity throughout differing contexts (Gratz & Roemer, 2004). ER theories indicate that as adolescents develop executive functioning, the efficacy of their ER approaches increases across abilities and emotional contexts (Fombouchet et al., 2023). ER difficulties may be one risk factor for anxiety (Mennin et al., 2005; Poole et al., 2017). Stimuli that typically evoke little to no negative emotional responses in the average population can elicit intense or frequent negative emotions in anxious adolescents (McLaughlin et al., 2007; Carthy et al., 2010). Anxious adolescents also exhibit an inability to successfully regulate negative emotions when they appear. This relationship between ER and anxiety may be especially important during adolescence when individuals experience a period of crucial biopsychosocial development (Casey, 2015) as well as maturation and refinement of ER skills (Gresham & Gullone, 2012; Desatnik et al., 2021). Given that anxiety disorders typically appear early in life (median age of onset is 11 years; Kessler et al., 2005), examining the relationship between ER and anxiety symptoms in adolescents may indicate a critical pathway for anxiety development (Schneider et al., 2018).
A potential pathway between poor health outcomes and ACEs may appear as maladaptive stress coping and emotional dysfunction (Poole et al., 2017; Salinas-Miranda et al., 2015). When stimulated by ACEs, deficits in emotional, cognitive, and social development may inhibit the ability to regulate anxiety, depression, and anger through adaptive dispositional mindfulness, cognitive reappraisal, expression, and processing (Boyes et al., 2016; Whitaker et al., 2014). Indeed, the experience of more ACEs is associated with more maladaptive ER tendencies (e.g., poor mindfulness, cognitive reappraisal, high rumination, and suppression; Cameron et al., 2018). Given that ER skills can be taught and improved, these skills indicate an encouraging intervention goal, specifically when targeted at decreasing the harmful effects of ACEs (Cameron et al., 2018; Jazaieri et al., 2014).
Attachment in Adolescence
Not every individual with a history of ACEs will experience ER deficits or anxiety symptomology (Poole et al., 2017). Adolescents who report ACEs vary significantly in regard to their mental health outcomes (Meeker et al., 2021). Protective factors exist to mitigate some of the harmful effects and some individuals are more resilient than others. One possible protective factor that may exhibit this effect is attachment, or a normative and fundamental process that contains affective and behavioral regulation starting in early development (Bowlby, 1969). In order for appropriate psychological adjustment, an individual should have an emotional bond with at least one other person across their lifetime (MacDonald, 1992). Additionally, this bond needs to be warm, available, trustworthy, and responsive. For individuals confronted with ACEs, research has shown that a bond like this can be an essential element in buffering against negative mental health outcomes (Walker & Venta, 2023). Adolescence may be an especially important time to examine attachment because it is a crucial period of development in terms of developing new peer attachments, self-regulation skills, and competency (Watters et al., 2024). For children and adolescents, relationships with parents and peers containing higher levels of warmth, trust, and acceptance are associated positively with self-esteem (Greenberg et al., 1983; Walker & Green, 1986). Adolescents with secure attachments (e.g., those higher in warmth, trust, and acceptance) have greater social support and are more likely to overcome adversity and develop lower levels of internalizing disorders (Watters et al., 2024). Secure attachments can provide a foundation for abilities in regulation in adolescents as well as buffer them from dysregulation (Rogers et al., 2022).
When positive, secure attachments are available, this support can attenuate the harmful impact of ACEs (e.g., child maltreatment) and other types of adverse events on brain circuit functioning related to ER and threat processing (Wymbs et al., 2020). The sensitivity of parents can shield children from an allostatic load, or exposure to chronic stress that results in physiological deterioration (McEwen et al., 2015). An individual with a secure attachment defined by high levels of warmth, trust and acceptance, may be provided with a psychological resource that supports the internal regulation of anxiety and facilitates resilience following traumatic experiences (Charuvastra & Cloitre, 2009; Pierrehumbert et al., 2012). Without a secure attachment, one can be particularly vulnerable to life stressors (Dagan et al., 2018). Attachment has been observed to moderate the relationship between a history of childhood adversity and physiological indications (Dagan et al., 2018). Attachment may have a role in intensifying or reducing the consequences of early stress and trauma on physical well-being (Cicchetti & Blender, 2006; Gunnar & Quevedo, 2007). Moreover, a continual perception of a secure bond that is reliably safe, comfortable, and supportive would lead to a secure-autonomous attachment (Waters & Cummings, 2000). This attachment style may permit being more open to social support as well as flexible in terms of problem-solving (Dagan et al., 2018).
In comparison to the comprehensive literature on attachment in infancy as well as adulthood (Kokkinos et al., 2019), fewer studies have examined the attachment of adolescents to both parents and peers. Adolescents with a secure attachment who also have increased ER skills (effectively regulating emotional responses to events) and emotional competence are more likely to be accepted by their peers due to higher levels of social competence and popularity (Collins & Steinberg, 2006; Kokkinos, 2013). In addition, securely attached adolescents can understand that they have the ability to control negative emotions and express them in a positive manner (Kokkinos et al., 2019). Parents of children with secure attachments are likely to socialize their children to ER skills through talking, assistance with coping, and provision of strategies for regulation (Brumariu & Kerns, 2010; Waters et al., 2010). Therefore, these adolescents have a higher likelihood of using cognitive reappraisal. However, adolescents with insecure attachments may find it challenging to establish friendships with peers due to experiences of peer rejection and a lack of prosocial behaviors (Groh et al., 2014; Sroufe et al., 2005). Studies have indicated that ER is undermined by insecure attachments, which are associated with restricted caregiver emotional security (Cassano et al., 2007; Waters et al., 2010).
During adolescence, individuals develop significant attachments to a variety of people (e.g., mothers, fathers, and peers). As these attachments increase in meaning and integration for the adolescent, attachments categorized as less secure are directly related to increased internalizing symptoms, such as anxiety (McGinley & Evans, 2020). While adolescents seek greater autonomy and more time with friends, the quality of their attachment to peers increases (Allen, 2008; McGinley & Evans, 2020). These peer attachments are typically more reciprocal than those with parents, in that both individuals in the relationship serve the attachment needs of the other. Therefore, similar to parental attachment, peer attachments have been shown to provide a protective factor against internalizing symptomatology (McGinley & Evans, 2020). Even though attachments to peers become more significant during adolescence, parent attachment remains important for their needs (Allen, 2008). Within parent attachment, adolescents differ in their relationships with their mothers and fathers. Attachment to mothers appears to remain higher in relationship quality across adolescence, whereas father attachment appears to be rated higher in terms of autonomy support (McGinley & Evans, 2020). However, both mother and father attachment are directly related to lower levels of internalizing symptomatology. Peer, mother, and father attachments are simultaneously related to internalizing symptomatology throughout a variety of adolescent samples (Armsden & Greenberg, 1987; McGinley & Evans, 2020). However, some studies indicate that parent but not peer attachment is important for this relationship (Oldfield et al., 2016); therefore, this association appears to warrant further investigation.
Current Study
In the current study, we examined the relationships between ACEs, ER, anxiety, and attachment in a sample of adolescents (n = 105). Research indicates that adolescents that experience ACEs also face a variety of challenges including internalizing symptoms such as anxiety (Kieling et al., 2011). The greater the number of ACEs an adolescent experiences, the more likely they are to experience significant anxiety (Lee et al., 2020). Thus, hypothesis 1 was that increases in ACEs would be associated with increased levels of anxiety. Additionally, individuals with reported ACEs indicate higher levels of emotion dysregulation (Gratz et al., 2007). An increased number of ACEs is associated with more maladaptive ER tendencies (Cameron et al., 2018). Based on these findings, hypothesis 2 was that increases in ACEs would be associated with increased levels of difficulties in ER skills. Also, poor ER skills are associated with difficulties in psychological well-being across the life cycle (Kuster et al., 2012). Deficits in ER are commonly reported by people with anxiety disorders (Poole et al., 2017). Examining the relationship between ER and anxiety symptoms in adolescents may indicate a critical pathway for anxiety development (Schneider et al., 2018). Therefore, hypothesis 3 was that increases in difficulties in ER skills would be associated with increased levels of anxiety.
As the initial research model on ACEs described (Felitti et al., 1998), a potential pathway between poor health outcomes and ACEs may appear as emotional dysfunction (Poole et al., 2017). When stimulated by ACEs, deficits in emotional, cognitive, and social development may inhibit the ability to regulate anxiety (Boyes et al., 2016). ER skills have been shown to mediate the interaction between ACEs and poor mental health (Cameron et al., 2018). Taking into account this research and the first three hypotheses, hypothesis 4 was that increased difficulties in ER would mediate the relationship between ACEs and anxiety. When positive attachments are available, support can improve the harmful impact of ACEs on ER (Wymbs et al., 2020). Attachment may have a role in intensifying or reducing the consequences of early stress and trauma on an individual’s well-being (Cicchetti & Blender, 2006). In addition, securely attached adolescents can understand that they have the ability to control negative emotions and express them in a positive manner (Kokkinos et al., 2019). Thus, given the previous hypotheses and these research indications, hypothesis 5 was that attachment would moderate the relationship between ACEs and difficulties in ER as well as ACEs and anxiety. Three models were evaluated for each type of attachment (i.e., peer, mother, and father). These variables were examined in a moderated mediation model (see Fig. 1A) in order to evaluate why and under what conditions ACEs and anxiety are related. Moderated mediation has been shown to significantly relate ACEs and child mental health with maternal compassion moderating the relationship via ER difficulties (Zhu et al., 2023). A moderated mediation model allowed for examination of whether the indirect effect of ER is moderated by attachment (Hayes & Preacher, 2013).
Fig. 1
Moderated mediation models. (A): Hypothesized Moderated Mediation Model; (B): Peer Attachment Moderated Mediation Model; (C): Mother Moderated Mediation Model; (D): Father Moderated Mediation Model. alow = a at low attachment; ahigh = a at high attachment; c’low = c’ at low attachment; c’high = c’ at high attachment
×
Method
Participants
A total of 105 participants completed the study (see Table 1). The participants who completed this study were adolescents attending an introductory psychology course, through their high school, in a large midwestern university in The United States of America. Inclusion criteria included being between 13 and 17 years of age (Mage = 16.4, SD = 0.87), enrolled in the psychology course, and able to read in English. Exclusionary criteria included being younger than 13 or older than 17 years of age (i.e., 7 removed from original 112). Participants received 1 credit towards a class research requirement. The data were collected through a survey accessed on a participant management software (i.e., SONA) across three semesters. The posted survey included several questionnaires that assess demographics, adverse and stressful life events, anxiety, attachment, and ER strategies. Bootstrapping was utilized to account for the modest sample size (i.e., 5000 bootstrap samples with sample sizes of 105) as recommended to account for the dataset’s failure to meet the assumption of normality (Preacher et al., 2007). The majority of the participants were female (i.e., 86.7%; 12.4% male and 1% nonbinary) and White (i.e., 77.1%). The institutional review board at the author’s university approved the study. Participants were recruited through Sona Systems (i.e., an online participant management software) at the university or through an email sent from their high school. They were permitted to select from a range of studies or complete an alternative project for class credit. All students were required to obtain blanket parental consent at the start of the semester to participate in research. Those students who indicated interest in this study received a link to participate. Before they started, they were given the opportunity to provide their assent. Then, they completed a series of questionnaires, including demographic questions and four measures on anxiety, ACEs, ER, and attachment. After their participation, they were compensated with one class credit toward the four needed to fulfill their course research requirement. Given the sensitive content, participants were provided mental health resources upon completion of the study.
Table 1
Demographics
Attribute
n (%)
Mean (SD)
Attribute
n (%)
Mean (SD)
Age
16.4 (0.87)
Number of Siblings
2.09 (1.39)
13
2 (1.9)
0
13 (12.4)
14
2 (1.9)
1
42 (40.0)
15
9 (8.6)
2
25 (23.8)
16
31 (29.5)
3
10 (9.5)
17
61 (58.1)
4+
15 (14.3)
Gender
Family Income
Male
13 (12.4)
$0–$50,000
Female
91 (86.7)
$50,000–$100,000
Other
1 (1.0)
Over $100,000
Ethnicity
Grade in School
11.48 (0.96)
White
81 (77.1)
10th grade
13 (12.5)
Black
9 (8.6)
11th grade
34 (32.4)
East Asian
1 (1.0)
12th grade
49 (46.7)
Multiracial
9 (8.6)
13th grade
8 (7.6)
Other
5 (4.8)
14th grade
1 (1.0)
Usual Academics
Current Academics
A’s
87 (82.9)
A’s
87 (82.9)
B’s
16 (15.2)
B’s
16 (15.2)
C’s
2 (1.9)
C’s
2 (1.9)
Parents’ Relationship
Living with Parents
Married
79 (75.2)
Both Mom and Dad
78 (74.3)
Divorced/Separated
16 (15.2)
Mostly Mom
19 (18.1)
Living Together
2 (1.9)
Mostly Dad
2 (1.9)
Living Separately
8 (7.6)
Other
3 (2.9)
Father Ethnicity
Mother Ethnicity
White
86 (81.9)
White
86 (81.9)
Black
8 (7.6)
Black
9 (8.6)
East Asian
1 (1.0)
East Asian
0 (0.0)
Multiracial
5 (4.8)
Multiracial
4 (3.8)
Latinx
1 (1.0)
Latinx
2 (1.9)
Other
4 (3.8)
Other
4 (3.8)
Father’s Age
48.56 (7.21)
Mother’s Age
45.27 (5.58)
30 – 39
9 (8.6)
30–39
12 (11.6)
40 – 49
50 (47.8)
40–49
72 (68.6)
50 – 59
38 (36.4)
50–59
20 (19.3)
60 – 69
7 (6.8)
60–69
1 (1.0)
70 – 79
1 (1.0)
70–79
0 (0.0)
Father’s Education
Mother’s Education
no high school diploma
6 (5.8)
no high school diploma
6 (5.8)
Trade school
32 (30.5)
Trade school
10 (9.5)
Some college
15 (14.3)
Some college
10 (9.5)
Bachelor’s degree
26 (24.8)
Bachelor’s degree
47 (44.8)
Master’s degree
13 (12.4)
Master’s degree
21 (20.0)
Doctorate degree
4 (3.8)
Doctorate degree
6 (5.7)
Measures
Demographics
A wide range of demographic variables were examined in order to assess how different aspects of the adolescents’ identities may impact the variables of interest. The demographic variables assessed include age, sex, ethnicity, mother ethnicity, father ethnicity, living with parents, number of siblings, parents’ relationship, grade in school, usual and current academic performance, father age, father job, father education, mother age, mother job, mother education, and estimated family income.
Adolescent Anxiety
The Revised Children’s Anxiety and Depression Scale (RCADS; Chorpita et al., 2000) was employed to assess anxiety in adolescents. The RCADS consists of 47 items organized into 6 subscales that evaluate anxiety and depression (i.e., Generalized Anxiety Disorder (GAD), Separation Anxiety Disorder, Major Depressive Disorder (MDD), Social Phobia, Panic Disorder, and Obsessive Compulsive Disorder). Each item is rated on a 4-point Likert scale ranging from 0 (never) to 3 (always) to calculate the frequency of each item arising in the participant’s life. An example of items included in the questionnaire is “I worry about things.” The RCADS has established high convergent and discriminant validity, good internal consistency, and an adequate factor structure in samples of children and adolescents in community and clinical settings (Chorpita et al., 2000; Chorpita et al., 2005).
Adverse Childhood Experiences
The Adverse Childhood Experiences (ACEs) Questionnaire (Dube et al., 2003; Felitti et al., 1998) was utilized to assess the ACEs experienced by the participants. The ACEs questionnaire was retrospectively adapted to measure forms of neglect, abuse, and household dysfunction (Dube et al., 2003; Murphy et al., 2014). This questionnaire consists of 10 items that are rated as 0 (no) or 1 (yes). Example items include “Were your parents ever separated or divorced?” and “Did a household member go to prison?”. The ACEs scale has acceptable internal consistency and satisfactory convergent validity (Karatekin & Hill, 2019). Additionally, research has shown that the different types of ACEs are interrelated within child populations (Scott et al., 2013).
Emotion Regulation
The Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004) was used to assess ER skills. Additionally, it includes six component scales, including difficulty engaging in goal-directed behavior, nonacceptance of emotion responses, lack of emotional awareness, impulse control difficulties, lack of emotional clarity, and limited access to ER strategies. This scale includes 36 items. Each item is scored on a 5-point scale ranging from 1 (almost never) to 5 (almost always) with higher scores indicating more difficulties with ER. Example items include “I am clear about my feelings” and “I pay attention to how I feel.” Research has shown that the DERS has strong psychometric properties within nonclinical populations (Gratz & Roemer, 2004) and specifically acceptable to high internal consistency and construct validity in adolescents (Neumann et al., 2010).
Attachment
Attachment was assessed using the Inventory of Parent and Peer Attachment-Revised (IPPA-R; Gullone & Robinson, 2005). The questionnaires utilized include 24 peer items (e.g., “I like to get my friends’ opinions on things I’m worried about”) and 23 parent items (e.g., “My parents respect my feelings”). The parent questions were modified to ask about the participant’s mother and father separately. Thus, three attachment relationships were investigated: mother, father, and peer. The IPPA-R provides three attachment scores defined by trust, communication, and alienation; however, this study focuses on the global score, combining the trust, communication, and reverse-coded alienation subscales, for each participant. Items are rated on a 3-point scale: 1 (never true), 2 (sometimes true), and 3 (always true) denoting the degree to which every item indicates a characteristic of their relationship with a parent or peer. Research has shown that the IPPA-R has strong psychometric properties, including moderate convergent validity and adequate to good internal consistency (Gullone & Robinson, 2005).
Data Management
The data were analyzed with IBM SPSS Statistics 27 (including PROCESS 3.5). Participants were excluded prior to bootstrapping if they completed less than 50% of the items overall or were older than 17 years of age (n = 7). Evaluation of the data revealed that the only missing data appeared when appropriate (i.e., no data for mother (n = 3) or father (n = 6) attachment when that participant reported not having a respective parental figure to fill that role). Therefore, no steps were taken to replace missing data. All the primary variables of interest were approximately normally distributed (skew and kurtosis both < |2.00 |). While the ACEs questionnaire appears positively skewed, based on recommended statistical cut-offs, the measure is not significantly skewed (Curran et al., 1996). Descriptive analyses were conducted to present the means, standard deviations, and correlations of the variables of interest (See Table 2). All variables of interest were significantly correlated in the expected directions (i.e., anxiety, problems with ER, and ACEs were positively correlated with each other and negatively correlated with all three types of attachment, and the types of attachment were positively correlated with each other). Of note, the only nonsignificant relationship appears between ACEs and peer attachment, r = −0.16, p = 0.11. Some concerns of multicollinearity with ACEs were apparent based on the standardized difference in fit and beta values (i.e., VIF > 5). However, since it was structural multicollinearity based on an interaction term, this concern was addressed by centering the variables. The relationships between the variables of interest and the demographic variables were examined using Pearson correlation analyses and ANOVAs. Based on the significant analyses the following covariates were included: mother’s age, family income, sex, living with parents, parents’ relationship, mother’s ethnicity, father’s ethnicity, and current academic achievement. The frequencies of the ACEs items endorsed were examined (see Supplemental Material).
Table 2
Correlations
1
2
3
4
5
6
1. Anxiety
–
2. Emotion Regulation
0.74**
–
3. ACEs
0.22*
0.25*
–
4. Peer Attachment
−0.27**
−0.31**
−0.16
–
5. Mother Attachment
−0.59**
−0.53**
−0.55**
0.28**
–
6. Father Attachment
−0.39**
−0.43**
−0.59**
0.47**
0.55**
–
Mean
74.64
94.68
1.55
59.41
56.15
52.86
Standard Deviation
22.24
28.38
1.81
8.91
10.74
12.05
Cronbach’s Alpha
0.96
0.96
0.69
0.94
0.95
0.96
*Correlation is significant at the 0.05 level (2-tailed)
**Correlation is significant at the 0.01 level (2-tailed)
Internal Consistency and Reliability
The internal consistency and reliability of the observed estimates were evaluated through the use of commonly accepted recommendations (i.e., < 0.5 – unacceptable; > 0.5 – poor; > 0.6 – questionable; > 0.7 – acceptable; > 0.8 – good; > 0.9 – excellent; George & Mallery, 2005) with further support from a meta-analysis on Cronbach’s alpha coefficients (Peterson, 1994). All variables of interest fell in the excellent range (see Table 2) except ACEs which fell in the questionable range (i.e., α = 0.69). Thus, the ACEs questionnaire may be of some concern due to potentially poor inter-relatedness. However, the ACEs questionnaire is constructed differently than the other measures in that it is a checklist of items rather than a true measure of a construct, namely adverse experiences. Therefore, Cronbach’s alpha may not be the best measure of reliability (Zinbarg et al., 2005). Rather another measure of reliability may be more accurate (e.g., ρ or ω). Using McDonald’s omega, the ACEs questionnaire falls in the acceptable range, ω = 0.70.
Analyses
Zero-order correlations were examined between the variables of interest (Hypotheses 1–3; see Table 2). The correlations were defined using the recommended guidelines for effect sizes that are small (r = 0.1–0.3), medium (r = 0.3–0.5), and large (r > 0.5) (Cohen, 1989). A simple mediation model was tested with ACEs as the predictor variable, anxiety as the outcome variable, and ER as the mediator (Hypothesis 4; PROCESS model 4). The covariates sex, living with parents, parents’ relationship, and mother’s education were included in the model. For the proposed model (see Fig. 1A), the PROCESS macro was used to analyze three moderated mediation models with the mediation effect of ER and the moderating role of attachment on the mediation path and between ACEs and anxiety (Hypothesis 5; PROCESS model 8; Hayes, 2018). Covariates (i.e., sex, living with parents, parents’ relationship, mother’s ethnicity, father’s ethnicity, and current academic achievement) were incorporated within the analysis based on which variables are significantly related to the variables of interest.
Results
Correlations & Mediation
Hypothesis 1 was supported by the positive relationship between ACEs and anxiety, r(103) = 0.22, p = 0.03, suggesting that a greater experience of adverse early life events was associated with a greater experience of current symptoms of anxiety. Hypothesis 2 was supported by the positive relationship between ACEs and difficulties in ER, r(103) = 0.25, p = 0.01, suggesting that a greater experience of adverse early life events was also associated with greater current difficulties with ER. Hypothesis 3 was supported through the large positive relationship between difficulties in ER and anxiety, r(103) = 0.74, p < 0.001, suggesting that greater challenges with ER was associated with more endorsement of current feelings of anxiety.
Hypothesis 4 was supported by a simple mediation model. The total effect of ACEs on anxiety was not significant, t(90) = 1.00, p = 0.32, β = 1.32. However, the significant indirect effect of ACEs on anxiety through ER, β = 1.73, 95% CI [0.05, 3.55], showed a complete mediation based on a nonsignificant direct effect of ACEs on anxiety, t(90) = −0.42, p = 0.67, β = −0.42. These results suggested that difficulties in ER explained the relationship between early adverse events and current anxiety, with ER acting as a mechanism through which recollection of early adverse events were related to reports of anxious symptomatology.
Moderated Mediation Analyses
Peer Attachment Model
The moderated mediation model through which peer attachment moderates the mediational relationship indicated a significant association between ACEs and difficulties in ER but was not moderated by peer attachment (See Table 3). While a significant association between ACEs and anxiety was not observed, the association was moderated by peer attachment (See Table 3). Tests of simple slopes indicated that the positive prediction from ACEs to difficulties in ER was not significantly stronger for adolescents with high levels of peer attachment than for adolescents with mean or low levels of peer attachment (Fig. 1B). Additionally, the positive prediction from ACEs to anxiety was relatively stronger for adolescents with high levels of peer attachment than for adolescents with mean or low levels (Fig. 1B). The analyses indicate that the indirect effect of ACEs on anxiety via difficulties in ER was not moderated by peer attachment (See Table 3). In particular, for adolescents with low levels of peer attachment, the indirect relationship between ACEs and anxiety was not significant (β = 1.06, SE = 1.08, 95% CI = [−1.06, 3.27]). However, for adolescents with high levels of peer attachment, a significant indirect relationship between ACEs and anxiety was found (β = 2.93, SE = 1.45, 95% CI = [0.47, 6.26]). These results suggested that high levels of peer attachment function as a mechanism that strengthens the positive relationship between recollection of early adverse events and reports of anxious symptomatology.
Table 3
Moderated mediation models
Peer moderated mediation model
Emotion regulation (M)
Anxiety (Y)
Predictor
B
SE
95% CI
B
SE
95% CI
ACEs (X)
3.99*
1.85
0.30; 7.67
1.41
1.51
−1.59; 4.41
Emotion Regulation (M)
0.50**
0.08
0.33; 0.67
Attachment (W)
−1.26*
0.44
−2.13; −0.39
−0.22
0.27
−0.75; 0.31
X * W | M * W
0.22
0.21
−0.21; 0.64
0.11
0.10
−0.09; 0.32
Model R2
0.31, F(11, 75) = 2.76, p = 0.005
0.56, F(12, 74) = 21.78, p < 0.001
Interaction Δ2
0.01, F(1, 75) = 1.03, p = 0.31
0.02, F(1, 74) = 4.71, p = 0.03
Mother Moderated Mediation Model
ACEs (X)
0.09
2.84
−5.57; 5.74
−0.86
1.50
−3.83; 2.12
Emotion Regulation (M)
0.42**
0.08
0.25; 0.58
Attachment (W)
−1.57**
0.34
−2.25; −0.90
−0.72**
0.25
−1.22; −0.22
X * W | M * W
0.27
0.15
−0.02; 0.56
0.11*
0.07
0.002; 0.27
Model R2
0.32, F(9, 75) = 5.31, p < 0.001
0.61, F(10, 74) = 15.16, p < 0.001
Interaction Δ2
0.02, F(1, 75) = 3.35, p = 0.07
0.01, F(1, 74) = 1.91, p = 0.17
Father Moderated Mediation Model
ACEs (X)
3.94
2.54
−1.12; 9.00
0.76
1.97
−3.16; 4.68
Emotion Regulation (M)
0.50**
0.08
0.34; 0.66
Attachment (W)
−1.11**
0.33
−1.75; −0.46
−0.28
0.23
−0.74; 0.18
13X * W | M * W
0.28
0.17
−0.05; 0.61
0.14*
0.07
0.02; 0.30
Model R2
0.28, F(8, 77) = 4.96, p < 0.001
0.56, F(9, 76) = 15.67, p < 0.001
Interaction Δ2
0.03, F(1, 77) = 2.89, p = 0.09
0.03, F(1, 77) = 2.89, p = 0.09
Peer model (N = 87); Mother model (N = 85); Father model (N = 86)
*p < 0.05; **p <0.01
Mother Attachment Model
The moderated mediation model through which mother attachment moderates the mediation process indicated that the association between ACEs and difficulties in ER was not significant, and the association was not moderated by mother attachment (see Table 3). Moreover, a significant association between ACEs and anxiety was not shown, and the association was also not moderated by mother attachment (see Table 3). Within this model the covariant adolescent sex was significant, (t(85) = 2.26, p = 0.03, β = 14.27). Tests of simple slopes indicated that the positive prediction from ACEs to difficulties in ER was not significant for adolescents with high, mean, or low levels of mother attachment (Fig. 1C). Additionally, the positive prediction from ACEs to anxiety was not significant for adolescents with high, mean, or low levels of mother attachment (Fig. 1C). The analyses indicate that the indirect effect of ACEs on anxiety via difficulties in ER was moderated by mother attachment (see Table 3). However, for adolescents with low levels of mother attachment, the indirect relationship between ACEs and anxiety was not significant (β = −1.12, SE = 1.03, 95% CI = [−3.36, 0.73]). For adolescents with high levels of mother attachment, an indirect relationship between ACEs and anxiety was also not found (β = 1.19, SE = 1.65, 95% CI = [−1.68, 4.93]). These results suggest that mean levels of mother attachment function as a mechanism that strengthens the positive relationship between recollection of early adverse events and reports of anxious symptomatology.
Father Attachment Model
The moderated mediation model through which father attachment moderates the mediation process indicated that the association between ACEs and difficulties in ER was not significant and the association was not moderated by father attachment (See Table 3). In addition, a significant association between ACEs and anxiety was not shown, and the association was also not moderated by father attachment (See Table 3). Tests of simple slopes indicated that the positive prediction from ACEs to difficulties in ER was not significant for adolescents with high, mean, or low levels of father attachment (Fig. 1D). Additionally, the positive prediction from ACEs to anxiety was not significant for adolescents with high, mean, or low levels of father attachment (Fig. 1D). The analyses indicate that the indirect effect of ACEs on anxiety via difficulties in ER was moderated by father attachment (See Table 3). In particular, for adolescents with low levels of father attachment, the indirect relationship between ACEs and anxiety was not significant (β = 0.29, SE = 1.10, 95% CI = [−2.21, 2.16]). However, for adolescents with high levels of father attachment, a stronger indirect relationship between ACEs and anxiety was found (β = 3.63, SE = 1.81, 95% CI = [0.16, 7.45]). These results suggest that high levels of father attachment function as a mechanism that strengthens the positive relationship between recollection of early adverse events and reports of anxious symptomatology.
Discussion
Overall, this study suggested support for the notion that mother and father attachment relationships might be of particular importance for youth who have experienced adverse life experiences and who have difficulty regulating their emotions. Specifically, mother attachment attenuated the negative effects of ACEs on anxiety via difficulties in ER, suggesting that adolescents’ attachment to their mothers may serve as a protective factor against the harmful effects of experiencing ACEs and having trouble regulating their emotions on their likelihood of developing symptoms of anxiety. Similarly, father attachment attenuated the negative effect of ACEs on anxiety via difficulties in ER suggesting that attachment to father figures may serve as a protective factor for adolescents with anxiety symptoms, especially if they have experienced ACEs or have trouble regulating their emotions. Below, we review findings related to each of our hypotheses in turn and the potential implications of these observations.
Firstly, our results supported hypothesis 1 with higher levels of ACEs demonstrating a positive relationship with higher levels of anxiety. Additionally, hypothesis 2 was supported by a positive relationship between levels of ACEs and difficulties in ER. Hypothesis 3 was also supported by the large positive relationship observed between difficulties in ER and levels of anxiety. The results of this study in an adolescent sample, specifically supporting a positive relationship between difficulties in ER and anxiety symptoms, fill a gap in the literature. Taken together, these findings suggest that adolescents who have trouble regulating their emotions are likely to have a higher risk of anxiety. The data also indicate that difficulties in ER mediated the relationship between ACEs and anxiety; therefore, hypothesis 4 was supported. These findings are consistent with past research (Berking & Wupperman, 2012; Cameron et al., 2018; Kieling et al., 2011; Kuster et al., 2012; Lee et al., 2020; Mennin et al., 2005; Poole et al., 2017) and add to this literature by examining the mediation within an adolescent sample and focusing on difficulties in ER skills. These findings expound upon the notion that ER skills explain the relationship between ACEs and poor mental health by demonstrating this relationship in an adolescent sample.
Secondly, hypothesis 5 predicted that three different types of attachment would moderate the mediation effect of ACEs and difficulties in ER as well as ACEs and anxiety. This study indicated that peer attachment moderated the relationship between ACEs and anxiety, and this positive prediction was relatively stronger for adolescents with high levels of peer attachment than for average or low levels, but the overall moderated mediation was not significant. While the overall peer attachment moderated mediation model was not significant, the model provided some interesting results. The results indicate that peer attachment buffers the relationship between ACEs and anxiety but not ACEs and difficulties in ER. Specifically, higher levels of peer attachment in adolescents affect the strength of the positive relationship between ACEs and anxiety results of this study provide novel findings on the relationship between peer attachment, ACEs, and anxiety by identifying high levels of peer attachment as a protective factor against anxiety, especially for adolescents with the additional risk factor of experiencing ACEs. While literature suggests peer attachment is more influential during adolescence (McGinley & Evans, 2020), there is evidence that adolescents are more likely to suppress their emotions around their peers in comparison to their parents especially when they have a closer attachment with their parents (Wylie et al., 2023). Additionally, research suggests parent attachment has a greater importance than peer attachment on protecting against internalizing symptoms (Oldfield et al., 2016). These findings support the results of this study that suggest that peer attachment does not moderate the relationship between ACEs and ER difficulties.
The other paths in this study’s models were also supported. While the relationships between ACEs, ER, and anxiety are not significant individually within the mother attachment model, when all of the variables are examined together, mother attachment moderates the mediation of difficulties in ER on ACEs and anxiety. These results suggest that an indirect effect of difficulties in ER on ACEs and anxiety is conditional on the value of attachment adolescents have with their mothers. This finding expounds upon previous studies’ findings that indicate that attachment can lessen the relationship between ACEs and ER (Wymbs et al., 2020). Thus, connecting all four variables of mother attachment, ACEs, ER, and anxiety improves upon previous literature by providing a more intricate potential relationship between them. One additional result of this model is that the sex of the adolescent is a significant covariant. This finding suggests that boys and girls may differ in how attachments to their mothers buffer the mediation between difficulties in ER on ACEs and anxiety. While results are ambiguous (e.g., Del Giudice, 2019; Sroufe et al., 2005), the majority of attachment research indicates no significant gender differences exist (Bakermans-Kranenburg & Van Ijzendoorn, 2009). However, the results of this study indicate that when mother attachment is included in the model, the adolescents’ gender may influence the results, such that the relationship between ACEs, ER, and anxiety may depend on mother attachment differently for girls than for boys.
Similarly, to the mother attachment model, the relationships between ACEs, ER, and anxiety are not significant individually within the father attachment model. However, when all the variables are examined together, father attachment moderates the mediation of difficulties in ER on ACEs and anxiety. This finding expands on previous studies’ results that indicate that attachment can lessen the relationship between ACEs and ER (Wymbs et al., 2020). Thus, connecting all four variables of father attachment, ACEs, ER, and anxiety improves upon previous literature by providing a more intricate potential relationship between them. Recently research is focusing more on the father-child relationship due to the lack of insight into this relationship (Wang, 2020). Therefore, the results of this model add to the literature in terms of learning more about father attachment and its relationship to other variables (i.e., ACEs, difficulties in ER, and anxiety). The results suggest that an indirect effect of difficulties in ER on ACEs and anxiety is conditional on a higher value of attachment adolescents have with their fathers. Additionally, father attachment moderated the relationship between ACEs and difficulties in ER but only for adolescents with low levels of attachment. However, this result contrasts with previous research that suggests that fathers may have a higher level of influence on adolescents’ internalizing problems, such as anxiety (Lux & Walper, 2019). Father attachment may serve as a risk factor when adolescents have a low level of attachment to their fathers, such that there is an association between ACES and difficulties in ER when father attachment is low.
Limitations
The current study was limited in several ways. When examining the mother and father attachments, while the participants were given the option to select the type of mother/father they had (i.e., biological, step, other), the questionnaire was set up to assume that each participant had one mother figure and one father figure. In the future, it may be more inclusive to change this, so the participants identify the number and type of parent figures they have and base the questionnaire responses on those answers. Another limitation of this study design is that the data were collected only through adolescent self-report. While online responding allowed for anonymous data collection, it may have impeded the reliability of the data through response bias. Additionally, there may be an issue with power evidenced by the sample size and effect sizes. A larger sample size may allow for the detection of these effects when present. Due to the nature of the specific sample of participants assessed in this study, there is likely a smaller number of ACEs endorsed in this sample compared with the general population. The adolescents sampled were educated and high-achieving individuals as indicated by their enrollment in college credit courses while still in high school. Research has shown that children with higher levels of ACEs are much more likely to not achieve their expected academic ability level (Evans et al, 2020). Therefore, it is possible that the students who can achieve higher levels of academic performance (e.g., earning college credit in high school) have not experienced the same number of ACEs as their peers. Additionally, since the sample is in adolescence, they have not yet experienced the same duration of life experience as an adult sample and might therefore report fewer ACEs than a comparable sample of older individuals. Another concern presented by the ACEs questionnaire is the low Cronbach’s alpha coefficient. However, as mentioned above, it is possible that the questionnaire is constructed differently than the other measures as a checklist rather than a true measure of the construct (Flora, 2020). Thus, Cronbach’s alpha is expected to be lower especially as it was originally designed for use in an epidemiological research study (Felitti et al., 1998). Even in the absence of a non-significant skew, given the visual appearance of skew, the researchers applied transformations to the ACEs measure however the transformations did not fix the appearance of skew.
Future Directions
Regardless of these limitations, the results of this study indicate potential intervention implications. Interventions targeting the mitigation of ACEs effects could focus on increasing mother and father attachment. A strong parental attachment may be a protective factor that helps adolescents with ACEs thrive and lessen anxiety development. Programs evaluating ACEs should include investigation into ER and parent attachment as influential factors affecting adolescents and their mental health development. Research in this area could benefit from continued investigation into ACEs within adolescent populations as well as how gender differences interact with parental attachment.
The mother attachment model indicated that the adolescent’s gender identity may affect the moderated mediation relationship. Therefore, this covariant may have some effect on the model and account for some of the variance shown. In the future, studies may focus on the mother attachment moderated mediation model separated between different gender identities. It is possible that girls and boys have different attachments to their mothers, or that their attachments moderate the regulation skills differently. Specifically, it should be investigated if sex differences are more related to the outcome of anxiety or the attachment with their mother and if mother attachment is more significant than father or peer attachment. However, there are limitations to using demographic variables as covariates (Miller & Chapman, 2001). Miller and Chapman concluded that when demographic variables are included in the model, they may remove variance that is difficult to separate (e.g., gender), and this removal could change the nature of the impact the variables have. In the future, it may be helpful to examine adolescent sex as a moderator in the models rather than a covariate for these reasons (Miller & Chapman, 2001). Much of the current literature on ACEs and related outcomes is dependent on retrospective reports of adults (Hardt & Rutter, 2004) which can lead to issues with recall (Meeker et al., 2021). In order to resolve this issue, research can focus on reports from adolescents who currently are experiencing an ACE or may remember a recent ACE more clearly (Meeker et al., 2021). Future studies may want to include a parent’s perspective or corroboration from other reporters on the significant life events that the adolescents have experienced.
Conclusions
In the present study, we observed that mother and father attachment moderate the mediation of difficulties in ER on ACEs and anxiety within an adolescent sample. The findings of these advanced statistical analyses illustrate how the variables all relate together in a more intricate manner. The findings of this study further the literature by highlighting how attachment to different important figures in an adolescent’s life may decrease the harmful effects of ER difficulties and adverse experiences in childhood and serve as protective factors against internalizing problems, specifically anxiety. In particular, these findings suggested that it may be important for adolescents who have experienced adverse life experiences and have trouble regulating their emotions to have a strong attachment with their mother or father but not a peer. Having a strong attachment to a parent may provide an adolescent with protective factors that help them thrive and lessen their chance of developing subsequent psychopathology, specifically anxiety. Additionally, if a child experiences an adverse relationship with one parent, having another protective parent figure may mitigate the adverse effects.
All procedures performed in the study that involved human participants was in accordance with the ethical standards of the Institutional Review Board at The University of Toledo and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
Informed consent
Written informed content was obtained fromeach child’s parent or legal guardian before participation in this study commenced. In addition, assent forparticipation was obtained from each child. All the authors contributed to the creation and design of the study.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Met BSL Psychologie Totaal blijf je als professional steeds op de hoogte van de nieuwste ontwikkelingen binnen jouw vak. Met het online abonnement heb je toegang tot een groot aantal boeken, protocollen, vaktijdschriften en e-learnings op het gebied van psychologie en psychiatrie. Zo kun je op je gemak en wanneer het jou het beste uitkomt verdiepen in jouw vakgebied.
