Disorganized and controlling parent–child attachments reflecting severe disturbances in the parent–child relationship are critical antecedents of maladjustment and psychopathology across childhood and late adolescence (e.g., Lyons-Ruth & Jacobvitz, 2016; Macfie et al., 2015; Madigan et al., 2016). Whereas disorganization reflects a lack of a coherent interacting strategy, controlling attachment occurs after the age of three years when the child takes over parental roles in a nurturing or hostile way. Some previous studies suggest that these controlling behaviors mostly develop in children who previously exhibited disorganization in infancy (e.g., Moss et al., 2005). Further, controlling behaviors may reflect children’s attempts to reduce stress levels unaddressed by caregivers (Solomon et al., 1995). Parent–child attachments may change due to new attachment experiences and life circumstances (Pinquart et al., 2013). However, little is known about disorganized and controlling attachments in young adultsFootnote 1 who undergo separation-individuation (Erikson, 1968) and experience unique life transitions (e.g., adjusting to college/work; Arnett, 2000). Therefore, it is important to evaluate disorganized and controlling parent–child attachments’ associations with mental health and adjustment in this age group.

One reason the relations between disorganized and controlling attachments and young adults’ mental health are less understood is the lack of well-validated and cost-effective means of assessing these patterns, particularly controlling attachments, during this period. The Childhood Experiences of Disorganization and Role Reversal Scale (CDRR; Meier & Bureau, 2018), a self-report of young adults’ perceptions of childhood disorganized and controlling attachment patterns with two parents, when available, addresses this need. However, measure development is an iterative process, and the CDRR has only been evaluated in one other study and further evidence for the suitability of its internal structure is needed to support it as a good measure of young adults’ experiences of disorganization and role reversal. Thus, this paper aims to further validate the CDRR by evaluating whether its factor structure replicates in a new sample and testing its construct validity by investigating conceptually meaningful associations (i.e., childhood adversity, internalizing problems, and subjective well-being).

Disorganized and Controlling Attachments

Disorganized/disoriented attachment originated from Main and Solomon's (1990) classification of infants who displayed odd, disoriented, or contradictory behaviors in the presence of their parent and reflects the lack of a consistent strategy for organizing attachment responses with a caregiver when under stress. Main and Hesse (1990) suggested that disorganization emerges from infants experiencing the paradoxical situation of the parent being a source of unresolved fear and the available figure to provide a sense of safety simultaneously (fear without solution). Others suggest that disorganized attachment may emerge from children’s fear related to the caregiver’s inability to protect them, abdication of parental responsibilities, and hostile or helpless parental behavior (e.g., Lyons-Ruth et al., 1999). Around three years old, some disorganized children remain disorganized, while others adopt controlling strategies (Deneault et al., 2023; Moss et al., 2005), increasing the predictability of caregiver interactions (Solomon et al., 1995). Children with controlling stances “actively attempt to control or direct the parent's attention and behavior, and assume a role, which is usually considered more appropriate for a parent with reference to a child” (Main & Cassidy, 1988, pp. 418–419). These stances include controlling-punitive and controlling-caregiving patterns (Main & Cassidy, 1988). As a major goal of attachment is to keep a child organized and safe, disorganized children who face an unpredictable and unreliable caregiver may seek to organize their daily interactions to ensure their safety and reduce stress that cannot be regulated through dependency on their caregiver (Moss et al., 2011). In support of this proposition, mothers of controlling children report feeling helpless to control themselves or their children (George & Solomon, 1996). Notably, although children organize their behaviors into a controlling pattern, in doll play narratives, to an extent, their representations remain disorganized, dominated by distress/vulnerability (Bureau & Moss, 2010; Solomon et al., 1995).

Controlling-caregiving children set aside personal needs to attend to those of the caregiver through supportive, entertaining, and organizing behaviors (Moss et al., 2011). This concept is somehow similar to the term “parentification” (Boszormenyi-Nagy & Sparks, 1973), from family systems theory, which describes parent–child interactions where the child is expected to assume a parental role and engage in instrumental and/or expressive caregiving (Boszormenyi-Nagy & Sparks, 1973). Bowlby (1980) also theorized that role inversion might result in a child showing compulsive caregiving (i.e., providing excessive care) when a caregiver pressures their child to care for them, provides inadequate caregiving, or experiences depression or disability. Empirical evidence showed that caregiving strategies may indeed arise when experiencing a withdrawn, uninvolved parent reliant on the child for support (e.g., Bureau et al., 2009a). Controlling-punitive offspring control a parent through degrading, authoritarian, and hostile behaviors and are likely to experience a hostile or intrusive parent (Moss et al., 2011). Bureau et al.’s (2009b) longitudinal study found that punitive behaviors in middle childhood are related to maternal hostility, disrupted communication in infancy (e.g., withdrawal, intrusiveness, parental fearful disoriented behavior), and maternal reports of physically abusing their child. This suggests that, although both controlling patterns may develop in the context of helpless parents (e.g., George & Solomon, 1996; Obsuth et al., 2014), punitive behaviors may be more likely to emerge in a chaotic home environment where children have to protect themselves from their parents, whereas caregiving behaviors may be more relevant to caring for less capable/unavailable parents. In brief, disorganized, controlling-punitive, and controlling-caregiving attachments are conceptualized as comprising the disorganized (D) attachment spectrum, as all three patterns share similar origins (i.e., disorganization in infancy), correlates (i.e., stressful/unpredictable parental relationships), and disorganized attachment representations.

Previous studies have identified disorganized and controlling patterns in preschool (ages 3 to 5 years; e.g., Moss et al., 2004; Deneault et al., 2023), early middle childhood (ages 7 to 9 years; Bureau et al., 2009a), later middle childhood (ages 10 to 12 years; Brumariu et al., 2018), and late adolescence (ages 18 to 23 years; Obsuth et al., 2014). However, little is known about controlling patterns in young adulthood, primarily due to the lack of a reliable and convenient measure to simultaneously evaluate all three disorganized/controlling patterns. Further, most studies assessing all patterns rely on samples of mothers, with a few exceptions (e.g., Deneault et al., 2020). However, the quality of individuals’ relationships with each parent may take different forms, as mothers and fathers may serve different attachment-related functions (i.e., a safe haven and secure base, respectively; Kerns et al., 2015). Societal changes in Western countries (e.g., women’s increased participation in the workforce) have led to increased recognition of the fathers’ crucial role in their children’s lives (Fagan et al., 2014). Thus, advancing measures of disorganized and controlling stances toward fathers is essential.

Disorganized and Controlling Attachment in Young Adulthood

In young adulthood, individuals continue re-evaluating their relationships with parents and balancing their needs for independence with family connection (see Institute of Medicine & National Research Council, 2015). Autonomy and exploring possible life directions become critical as young adults attempt to define their identity and decide on a course for their future (Arnett, 2000; Institute of Medicine & National Research Council, 2015). Non-familial social networks expand (Institute of Medicine & National Research Council, 2015), and young adults attend college (Arnett, 2000), face increased responsibilities, and separate from familiar support systems. Navigating these changes may have implications for long-term life trajectories, as many of these tasks represent preparations for later adult roles (e.g., mature long-term romantic relationships; Arnett, 2000). Disorganized and controlling attachments may be especially disruptive at this age. For example, prioritization of parental over personal needs may hinder caregiving young adults' ability to balance their needs for independence and family obligations adaptively. Because attachment security should facilitate autonomy (e.g., Macfie et al., 2015), disorganized and controlling stances may hinder this capacity and the relationship/identity formation tied to young adults’ exploration of their social world (see Koepke & Denissen, 2012).

Current Measurement Approaches

Attachment measures used in young adulthood vary in which disorganized/controlling patterns they capture. Some measures capture the quality of the relationship with each parent, while others capture integrated representations of the attachment relationships. The Adult Attachment Interview’s (AAI; George et al., 1996) Unresolved category contains elements of disorganization (e.g., fragmented internal working models and unaddressed experiences of fear). However, the AAI was not designed to capture and code for controlling stances. The Goal Corrected-Partnership in Adolescence Coding System (GPACS; Obsuth et al., 2014) is based on coding adolescent–parent conflict discussions and yields factors capturing aspects of the three disorganized stances (punitive, caregiving, and disoriented interaction). However, parents may not be readily available for such assessments in young adulthood. Parentification assessments (e.g., Parentification Inventory; Hooper & Doehler, 2012) capture instrumental aspects of controlling caregiving—what kind of parent-like responsibilities the child took on. Despite their strengths, these measures either do not specifically evaluate controlling patterns (e.g., AAI), are resource-demanding (e.g., GPACS), or capture overlapping but not all aspects of disorganized/controlling attachments (e.g., parentification measures).

The CDRR, a 95-item self-report questionnaire grounded in attachment theory, fills the gap for an efficient measure for all three patterns by including scales capturing disorganization, controlling caregiving, and controlling punitive/hostile stances based on children’s experiences growing up. Meier and Bureau (2018) independently developed scales for mothers and fathers (i.e., conducted separate factor analyses with the same item pools) and scale names reflect the factor structures they observed. Thus, the measure captures shared and unique aspects of young adults' experiences with both mothers and fathers. Using three samples of Canadian college students ages 18–25 (n = 750, n = 656, n = 96), Meier and Bureau (2018) identified patterns of disorganization with mothers (Disorganized/Punitive, Affective Caregiving, and Mutual Hostility) and fathers (Disorganized, Affective Caregiving, and Punitive), of the CDRR, in addition to the Appropriate Boundaries subscale. The Disorganized/Punitive scale reflects young adults' experiences of a combination of attachment disorganization with their mother (e.g., unresolved fear and dealing with a helpless, frightening, or withdrawn/unavailable caregiver) and punitive strategies. The Affective Caregiving scale captures aspects of controlling caregiving such as orienting, supporting, and attending to parental emotional needs. Mutual Hostility captures aspects of punitiveness related to aggressive and emotionally dysregulated behaviors on the part of both the mother and the child. Appropriate Boundaries captures the caregiver maintaining an appropriate parental role (i.e., an authoritative stance and not abdicating responsibilities). For the father version of the scale, the Disorganized and Punitive scales represent disorganized and punitive strategies, respectively, and the Affective Caregiving and Appropriate Boundaries scales are analogous to their mother version counterparts. Notably, the CDRR showed slightly different patterns for mothers and fathers regarding the attachments related to fear, hostility, and resentment. This finding may not be surprising in a society where most fathers are commonly viewed as authority figures and mothers, more generally, endorse a caregiving role. Meier and Bureau provided evidence for the measure’s four-factor structure, test–retest reliability and criterion, convergent, and discriminant validity through associations with parentification measures and childhood risk factors (e.g., domestic violence).

Overall, the CDRR adds to emerging measures of young adult disorganized/controlling stances due to its target age, inclusion of controlling and disorganized constructs, and rating of attachment stances toward mothers and fathers. Moreover, it fills the need for an easy-to-administer self-report measure that complements representational and observational-based approaches such as the AAI and GPACS. However, the CDRR’s psychometric properties were only assessed in one study and further validation is necessary. To add to the literature, we aimed to evaluate whether the CDRR’s factor structure replicates and further test its construct validity through associations with childhood adversity, internalizing problems, and subjective well-being.

Validating Criteria

Childhood Adversity

Experiences of childhood adversity, including abuse and maltreatment, are likely to contribute to disorganized and controlling stances as children may attribute these events to their parents’ helplessness and inability to protect them (Liotti, 2004) or by fostering representations of parental figures as hostile and frightening (Main & Hesse, 1990). Meta-analytic evidence suggests that childhood adversity (including physical, sexual, and emotional abuse and neglect) may increase the likelihood of disorganized attachment, and in many cases, disorganized children have experienced multiple types of adversities/maltreatment (e.g., Cyr et al., 2010). Lyons-Ruth and Jacobvitz (2016) pointed out that multiple social-risk factors (e.g., maltreatment, marital conflict, loss) function as cumulative risks that, together, may increase the incidence of attachment disorganization, particularly in combination with changes in parenting behavior. Adverse experiences of sexual and/or physical abuse and marital conflict have also been associated with controlling behaviors (Macfie et al., 2015). Khoury, Rajamani, et al. (2020) further showed that the severity of childhood maltreatment by age 18 is associated with disorganization in middle childhood and hostile-punitive, caregiving, and odd, out-of-context behaviors as assessed in late adolescence using the GPACS, supporting the assumption that all of the disorganized and controlling patterns are associated with a form of maltreatment, though odd-disoriented behavior (i.e., disorganization) showed the strongest association. Meier et al. (2014) also observed that youths who reported experiences of physical or verbal abuse, SES risk, or parental domestic violence tended to feel burdened by caregiving responsibilities for mothers and fathers.

Some specific associations have been found in previous studies, such as marital conflict and disorganization (Moss et al., 2011), maternal loss of their own attachment figure and child caregiving (Moss et al., 2004), and maternal physical abuse of the child and child punitiveness (Bureau et al., 2009b). These findings should be interpreted with caution due to the few studies assessing the disorganized and controlling subtypes. Nonetheless, as discussed above, there is growing evidence that all three disorganized/controlling attachments are associated with some form of childhood adversity. Further, unresolved fear/distress is conceptualized as a central emotion in the development of disorganized and controlling attachments (Lyons-Ruth & Jacobvitz, 2016; Main & Hesse, 1990). Thus, it is likely that all kinds of adversities relate to these patterns, given their potential to induce fear/distress in relationships with caregivers. Therefore, we evaluated whether CDRR dimensions are associated with childhood adversity.

Internalizing Problems

Another essential claim of attachment theory is that a secure attachment promotes healthy development (Bowlby, 1982). Consequently, insecure attachments, including disorganized and controlling patterns, should be associated with greater maladjustment, including anxiety and depression. Children with disorganized attachments see themselves as helpless when facing frightening situations and attachment figures as failing to protect them, which, combined with a lack of a consistent strategy to deal with distress, would then leave their fearful arousal unresolved—perpetuating or increasing the risk for anxiety (e.g., Brumariu et al., 2013). Further, they may perceive difficulties as overwhelming and the self as unable to cope with challenges, beliefs typically associated with depressive symptoms (Brumariu & Kerns, 2010). Meta-analytic evidence suggests that disorganized attachment beyond infancy (i.e., across childhood and adolescence) relates to higher levels of internalizing problems (Madigan et al., 2016). In late adolescence, GPACS disorientation was associated with depressive symptoms (Obsuth et al., 2014) and with anxiety disorders and their associated conditions (Brumariu et al., 2013).

The emerging literature on controlling behaviors in late adolescence, primarily assessed with the GPACS, is consistent with the notion that these behaviors pose risks for psychopathology. More role confusion (and more disoriented behavior) in interaction with the parent was associated with borderline personality disorder (BPD) traits, suicidality/self-injury (Lyons-Ruth et al., 2015), dissociation (Byun et al., 2016), and less attention to comforting attachment-related images (Schneider, et al., 2022). Further, individuals with a diagnosis of BPD were more likely to show disorganized/controlling interactions with their mothers than those with other diagnoses (anxiety, depression, or substance use diagnoses) or no diagnosis (Khoury, Zona, et al., 2020). Disorganized, punitive, and caregiving interactions in late adolescence were related to more dissociative symptoms, and punitive and caregiving interactions were also associated with more overall psychiatric comorbidities on Axis-I (Obsuth et al., 2014).

Regarding internalizing problems specifically, punitive attachment relates to greater depression or internalizing problems (e.g., Bureau et al., 2009a; Brumariu et al., 2018; but see Brumariu et al., 2013 for exceptions; Macfie et al., 2015 for a review). While Moss et al. (2004) found greater internalizing problems in controlling caregiving than in secure preschool children, studies with older children/adolescents did not find such a relation (e.g., Bureau et al., 2009a; Brumariu et al., 2013; Brumariu et al., 2018, but see Obsuth et al., 2014 and Meier et al., 2014 for exceptions). Children relying on caregiving strategies may not outwardly exhibit maladjustment so as not to undermine their presentation as a capable support for their parent (e.g., Bureau et al., 2009a). Brumariu et al. (2021) suggested that associations of controlling patterns with psychopathology, particularly caregiving, might emerge later than in middle childhood. In young adulthood, detrimental effects of caregiving strategies may appear when attending to caregivers’ needs that might compete with developmental tasks such as separating/individuating from parents and constructing a personal life course (see Arnett, 2000). Thus, we expect all disorganized/controlling patterns to relate to greater depression and anxiety.

Subjective Well-Being

A measure of parent–child attachment should relate to key aspects of adjustment (Solomon & George, 2008). In this paper, we focused on adjustment as reflected by broad indicators of subjective well-being (i.e., subjective happiness, life satisfaction, self-ratings of meaning in life), which Diener (2006) defines as “all of the various types of evaluations, both positive and negative, that people make of their lives” (p.153). It is likely that pathways linking disorganized/controlling stances to well-being are similar to those for internalizing problems (but in the opposite direction), as well-being is negatively associated with depressive symptoms (e.g., Newcomb-Anjo et al., 2017). Children with greater disorganization show poor emotion regulation (Brumariu, 2015) and more difficulties in peer relationships (Groh et al., 2014). These difficulties, combined with experiencing a caregiver who is fragile/helpless or hostile and frightening (e.g., Lyons-Ruth & Jacobvitz, 2016), are likely to influence disorganized/controlling children’s later propensities to experience happiness. They may negatively color individuals’ perceptions of events, affect evaluations of dispositional happiness independent of experiences (subjective happiness), disrupt their quality of life, and, by extension, their life satisfaction. Due to these difficulties, young adults with greater disorganized and controlling attachments might struggle to crystallize a sense of meaning in life (i.e., life has a purpose, is understandable, and worth living; Martela & Steger, 2016). For example, since affective caregiving may impede young adults’ exploration of personal goals due to prioritization of caregivers’ needs, it may relate to a lessened sense of meaning in life. No studies investigated disorganized/controlling attachments’ relation to subjective well-being. Thus, we tested the relations of the CDRR scales with subjective happiness, life satisfaction, and meaning in life.

The Current Study

This study’s goals were to evaluate whether Meier and Bureau’s (2018) factor structures of the CDRR replicate and to add to the literature by further evaluating the CDRR’s validity by testing its associations with internalizing problems (depression and anxiety) and subjective well-being (happiness, life satisfaction, and meaning in life). We followed the original scale development by Meier and Bureau (2018) and evaluated separate scales for disorganized stances toward mothers and fathers so the two studies would be comparable. We hypothesized that (a) the factor structure for the mother and father versions of the CDRR would replicate in our sample; (b) all CDRR scales capturing disorganized or controlling attachments (all scales except appropriate boundaries) would relate positively to childhood adversity; (c) all CDRR subscales capturing disorganized or controlling attachments (all scales except appropriate boundaries) would relate positively to depression and anxiety; (d) all CDRR scales capturing disorganized or controlling patterns will relate negatively to happiness, life satisfaction, and meaning in life. Although not our main focus, because the CDRR appropriate boundaries scales represent caregivers maintaining appropriate parental roles (unlike the other disorganized/controlling patterns), we expected that it would relate negatively to depression and anxiety and positively to happiness, life satisfaction, and meaning in life.

Methods

Procedure and Participants

Ethics approval was obtained from the institutional review board prior to participant recruitment (IRB #102,220). Participants were undergraduate students at a Northeastern university in the US and were eligible for the study if they were 18 to 25 years old. After completing informed consent, participants answered the survey through the online Qualtrics platform and received course credit. Of the 313 who expressed interest, 286 (91.37%) participants completed the measures. Most participants were in their first year of college (n = 146; 51.05%), 58 (20.28%) were in their second year, 44 (15.38%) were in their third year, 32 (11.19%) were in their fourth year, 3 (1.05%) were in their fifth year, and 3 (1.05%) were graduate students. Participants were between 18 and 25 years old (M(SD) = 19.3(1.6); ~ 63.73% psychology majors). Participants provided information regarding their gender identity; 213 (74.48%) females, and 68 (23.78) males. Of the participants, 163 (56.99%) identified as White, 49 (17.13%) as Asian, 45 (15.73%) as Black or African American, 20 (6.99%) as Other (Not Hispanic/Latino/Spanish origin), 3 (1.05%) as Native American or Alaska Native, and 2 (0.70%) as Native Hawaiian or Other Pacific Islander. In addition, 63 (22.03%) participants were of Hispanic, Latino, or Spanish origin. Most participants grew up with two primary caregivers (n = 226; 79.02%). Most participants who grew up in a single-parent household (n = 40; 13.99%) identified their primary caregiver as female (n = 33; 82.50%). Lastly, 206 participants reported living with family (72.03%), and 80 (27.97%) reported living alone or with non-family members.

Measures

Disorganized/Controlling Stances with Parents

The CDRR (Meier & Bureau, 2018) is a 95-item questionnaire that assesses young adults’ perceptions of their childhood (birth to age 12) disorganized/controlling attachments toward their primary caregiver(s). Items are rated on a 5-point Likert scale from 1 (not at all/never) to 5 (completely agree/always; see supplemental material Table S1 for items and instructions). As described above, Meier and Bureau (2018) obtained two slightly different factor solutions for mothers (disorganized/punitive, affective caregiving, mutual hostility, appropriate boundaries) and fathers (disorganized, affective caregiving, punitive, appropriate boundaries). Scores are mean-based, with higher scores representing a higher level of a specific pattern. Meier and Bureau (2018) reported initial evidence for the scale’s criterion, convergent/discriminant validity, and test–retest reliability.

Childhood Adversity

The 10-item Adverse Childhood Experiences Scale (ACES; Felitti et al., 1998) assesses experiences of childhood trauma, stress, and household dysfunction. Participants answered each question with 1 (yes) or 0 (no). A sample item is “Did a parent or other adult in the household often swear at you, insult you, put you down, or humiliate you? Or act in a way that made you afraid that you might be physically hurt?” Scores are sum-based and higher scores reflect a higher incidence of childhood adversities (in our sample α = 0.68). Studies have reported evidence of strong psychometric properties of the ACES (e.g., Ford et al., 2014).

Internalizing Problems

Depression The Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977) is a 20-item questionnaire assessing depressive symptoms experienced in the past week. Items are rated on a 4-point scale from 0 (rarely or none of the time) to 3 (most or all of the time). A sample item is “I was bothered by things that usually don't bother me.” The total score is sum-based, and higher scores indicate a higher frequency of symptoms (in our sample α = 0.92). The CES-D has adequate psychometric properties (e.g., Obsuth et al., 2014; Radloff, 1977).

Anxiety Participants completed the 20-item State-Trait Anxiety Inventory for Adults, Form Y-2, Trait Version Only (Spielberger et al., 1983). They rated how they “generally” feel on a 4-point scale from 1 (almost never) to 4 (almost always). A sample item is “I feel nervous and restless.” Items are summed and higher scores indicate higher anxiety levels (in our sample α = 0.94). The STAI-Y2 has demonstrated strong psychometric properties (Spielberger et al., 1983).

Subjective Well-Being

Happiness The Subjective Happiness Scale (SHS; Lyubomirsky & Lepper, 1999) is a 4-item questionnaire assessing global subjective happiness. Items are rated on a 7-point scale. A sample item is “In general, I consider myself not a very happy person/a very happy person.” Items are averaged and higher scores reflect greater happiness (in our sample α = 0.87). The SHS has shown good validity and reliability (Lyubomirsky & Lepper, 1999).

Life Satisfaction The Satisfaction with Life Scale (SWL; Diener et al., 1985) is a 5-item questionnaire that assesses quality of life, an aspect of well-being. Items are rated on a 7-point scale from 1 (strongly disagree) to 7 (strongly agree). A sample item is “In most ways, my life is close to my ideal.” Items are summed and higher scores reflect greater life satisfaction (in our sample α = 0.87). The SWL has good convergent and construct validity (e.g., Diener et al., 1985).

Meaning in Life The presence of meaning subscale of the Meaning in Life Questionnaire (MLQ; Steger et al., 2006) includes five items assessing the presence of a sense of meaning in life. Items (e.g., “I understand my life’s meaning”) are rated on a 7-point scale from 1 (absolutely untrue) to 7 (absolutely true). Scores are sum-based, with higher scores reflecting a greater meaning in life (in our sample α = 0.89). The MLQ has consistently demonstrated good psychometric properties (Schulenberg et al., 2011).

Control Variables

Demographics Participants reported their age, gender, race/ethnicity, academic year, the number of caregivers in their household during childhood, and current living situation.

Power Analysis

For our confirmatory factor analyses (CFA’s) of the four-factor models, we conducted Monte Carlo simulations in MPlus with loadings of 0.80 and varying numbers of observations. With small-to-medium factor covariances (0.25) and missing data set to 30% to be conservative, 200 participants should be acceptable to reach 0.80 power (range = 0.78–1.00) for parameter estimates. For the goal of evaluating the CDRR’s association with childhood adversity, internalizing problems, and subjective well-being, a-priori power analysis in G*Power 3.1 (α = 0.05, power = 0.80, two-tailed) showed that we would need 29 participants to detect a large effect (r = 0.50), 84 for a small-to-medium effect (r = 0.30), and 782 for a small effect (r = 0.10). Thus, we expected our sample (N = 286) to be adequately powered for our analyses.

Overview of Analyses

First, to address our goal of evaluating whether the original four-factor structures of the mother and father versions of the CDRR replicate (and verify the scales we would be relying on in our subsequent analyses), we performed two CFAs, using a maximum likelihood estimation with robust standard errors (MLR) in MPlus v.8.8. We followed Meier and Bureau’s (2018) model specifications, not specifying cross-loadings or residual covariances, and their use of parceling for CFAs with the same item aggregates (supplemental material, Table S1). Latent variables were allowed to covary. We evaluated model fit using comparative fit index (CFI), standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA) indices, as they do not depend on sample size and identify misspecified models well (Jackson et al., 2009). We used the following cutoffs for good fit: CFI ≥ 0.95, SRMR < 0.05, and RMSEA < 0.05 (Hu & Bentler, 1999). Second, we conducted preliminary analyses to evaluate whether demographic variables (gender, age, etc.) relate to the CDRR and whether childhood adversity relates to internalizing problems and subjective well-being variables. Finally, we completed zero-order correlations to address our second goal of evaluating disorganized and controlling attachments’ relations to childhood adversity, internalizing problems, and subjective well-being. For the purpose of replicability, we conducted correlations separate from CFA models using the mean-based scores specified in the CDRR’s scoring instructions (correlations using latent CDRR variables had similar results and are available upon request). Missingness across our measures was 6–36 participants (% missing: Mother CDRR Scale = 13.62%, Father CDRR scale = 15.81%, ACES = 2.10%, CES-D = 2.80%, STAI = 10.10%, SHS = 5.24%, SWL = 2.80%, MLQ = 11.09%). We used full-information maximum likelihood missing data handling, as its assumptions are less narrow than alternatives (Enders, 2012).

Results

Confirmatory Factor Analysis and Internal Consistency of CDRR Subscales

Figures 1 and 2 present the results of CFAs for the four-factor models of the mother and father version of the CDRR identified by Meier and Bureau (2018). All indicators loaded significantly onto their respective factors (all loadings > 0.50, all p's < 0.001; cutoff = 0.40 Stevens, 1992). The four-factor model of disorganized/controlling attachments with mothers fits the data well (CFI = 0.96, SRMR = 0.05, RMSEA = 0.08, 90% CI = 0.07, 0.10) and the model with fathers had marginal/acceptable fit (CFI = 0.93, SRMR = 0.07, RMSEA = 0.11, 90% CI = 0.09, 0.13). For the mother version of the CDRR (see Fig. 1), the disorganized/punitive factor was positively associated with the affective caregiving and mutual hostility factors. Disorganized/punitive and appropriate boundaries had a small but marginally significant association (p = 0.05). Affective caregiving was positively associated with appropriate boundaries and mutual hostility. Appropriate boundaries and mutual hostility were unrelated. For the father version of the scale (see Fig. 2), all CDRR factors except appropriate boundaries, which did not relate to the other factors, were positively associated. Coefficient alphas for disorganized/punitive, affective caregiving, mutual hostility, and appropriate boundaries subscales of the mother CDRR and the disorganized, affective caregiving, punitive, and appropriate boundaries subscales for the father CDRR were 0.96, 0.80, 0.92, 0.79, and 0.96, 0.92, 0.81, 0.68, respectively. Follow-up CFAs combining highly correlated factors showed worse model fits for models that combined factors than the initial four-factor solutions (all Satorra-Bentler ∆X2, p < 0.001; see supplemental material, Table S2 & 3 for model details/comparisons).

Fig. 1
figure 1

Factors Loadings and Factor Correlations of the CDRR Mother Version Using Confirmatory Factor Analysis (CFA) (N = 244). Note: DPP = Disorganized/Punitive Parcel, ACP = Affective Caregiving Parcel, ABP = Appropriate Boundaries Parcel, MHP = Mutual Hostility Parcel *p < .05, **p < .01, ***p < .001

Full size image
Fig. 2
figure 2

Factors Loadings and Factor Correlations of the CDRR Father Version Using Confirmatory Factor Analysis (CFA) (N = 216). Note: DOP = Disorganized Parcel, ACP = Affective Caregiving Parcel, ABP = Appropriate Boundaries Parcel, PUP = Punitive Parcel. *p < .05, **p < .01, ***p < .001

Full size image

Descriptive Statistics and Preliminary Analyses of Demographic and CDRR Variables

Table 1 presents the descriptive statistics for all main variables. Preliminary analyses revealed that age was not related to any of the disorganized stances with mothers or fathers, and there were no significant gender differences in disorganized stances toward mothers and fathers (all p's > 0.05), with two exceptions. Male participants (M(SD) = 3.2(0.7) and M(SD) = 2.4(0.8)) endorsed greater appropriate boundaries with their mothers and punitive stances toward fathers than females did (M(SD) = 2.8(0.8) and M(SD) = 2.1(0.8); t(237) = 2.80, p = 0.01, Cohen’s d = 0.36 and t(211) = 2.51, p = 0.01, Cohen’s d = 0.35, respectively). Participants living alone or with non-family members (M(SD) = 2.5(0.9)) reported higher levels of disorganization with fathers than participants living with family members (M(SD) = 2.2(0.8); t(214) = 2.08, p = 0.04, Cohen’s d = 0.28). Therefore, the analyses involving appropriate boundaries with mothers/punitive stances toward fathers and disorganized stances toward fathers presented below were repeated, controlling for participant gender and living situation, respectively, using partial correlations. For ease of readability, partial correlations were reported only when there were significant changes in results. As seen in Table 2, childhood adversity was positively and significantly correlated with depressive symptoms and anxiety. Childhood adversity was negatively and significantly correlated with happiness, life satisfaction, and meaning in life. Internalizing problems (depression and anxiety) and well-being variables (happiness, life satisfaction, and meaning in life) were significantly positively correlated among themselves and negatively correlated with each other.

Table 1 Descriptive Statistics for CDRR Variables and Main Study Variables
Full size table
Table 2 Bivariate Correlations Among All Variables
Full size table

Associations Between CDRR Variables and Childhood Adversity

All disorganized/controlling stances toward mothers and fathers were significantly (and positively) associated with childhood adversity, except for punitive stances toward fathers (Table 2). Appropriate boundaries with fathers, but not with mothers, were negatively and significantly correlated with childhood adversity.

Associations Between CDRR and Internalizing Problems

Young adults’ disorganized/punitive and mutually hostile stances toward mothers were significantly and positively associated with depression and anxiety. The association between affective caregiving and depression, but not anxiety, was significant and positive. All three disorganized stances toward fathers were significantly and positively related to depression and anxiety. Appropriate boundaries with mothers or fathers did not relate to depression or anxiety.

Associations Between CDRR Variables and Subjective Well-Being

Young adults’ disorganized/punitive stances toward mothers and their subjective happiness and life satisfaction were negatively and significantly correlated. Mutual hostility with mothers was negatively and significantly associated with happiness, life satisfaction, and meaning in life. Affective caregiving and appropriate boundaries were unrelated to subjective well-being (happiness, life satisfaction, and meaning in life). After controlling for participant gender, appropriate boundaries' association with subjective happiness and life satisfaction became significant, partial r(197) = 0.16, p = 0.02, and partial r(197) = 0.17, p = 0.02, respectively). Disorganization with fathers was significantly and negatively associated with subjective happiness, life satisfaction, and meaning in life. Affective caregiving toward fathers was also significantly and negatively related to happiness and life satisfaction. Punitive stances were negatively and significantly associated with life satisfaction only. Appropriate boundaries with fathers was significantly and positively related to happiness, life satisfaction, and meaning in life.

Discussion

Given the extensive evidence showing that disorganization at earlier ages is the most detrimental parent–child insecure attachment (e.g., Lyons-Ruth & Jacobvitz, 2016; Madigan et al., 2016), it is imperative to rely on measures that capture its stances in young adulthood. We aimed to add to the literature and address the need for a well-validated, efficient, and cost-effective questionnaire measure of three disorganized and controlling attachment patterns for young adults by further validating the CDRR scale. Our first goal was to evaluate whether the factor structure of the CDRR identified by Meier and Bureau (2018) replicates. The results of our CFA for the mother and father versions of the CDRR provide additional support for the suitability of the previously identified factor structure. Further, the associations among the CDRR's factors and their strength in our sample closely mirrored those of Meier and Bureau. The associations among the disorganized and controlling patterns ranged between 0.38 and 0.74, suggesting that the CDRR factors are, as expected, distinct but related. Importantly, the associations among the disorganized and controlling patterns in our young adult sample are expected and in line with the previous studies in younger middle childhood (Bureau et al., 2009a; disorganization and punitive), later middle childhood (Brumariu et al., 2018; disorganization and punitive), and late adolescence (Obsuth et al., 2014; disorganization, controlling caregiving, and punitive). They also point out that some young adults might show elevated scores on multiple factors and possibly in their associated struggles. Brumariu et al. (2021) noted that such individuals may be at particular risk for difficulties, and future work should test this hypothesis.

In our sample, the associations between appropriate boundaries and other factors were relatively small, ranging from 0.01 to 0.25, and most were non-significant and relatively lower than those observed by Meier and Bureau (2018). Meier and Bureau found small-to-moderate associations between appropriate boundaries with mothers and fathers and all other CDRR variables except for punitive stances toward fathers. Studies should further test whether associations with the appropriate boundaries factor vary by sample characteristics.

Our second goal was to further extend the construct validity of the CDRR by testing its associations with young adults’ history of childhood adversity and aspects of internalizing problems and subjective well-being. In support of our second hypothesis, most CDRR scales capturing disorganized and controlling attachment patterns with mothers and fathers were significantly related to a more extensive history of childhood adversity. Our results align with the literature suggesting adverse childhood experiences may set the stage for these attachment patterns to some extent by disrupting the parent–child relationship quality and fostering frightening, hostile, or helpless representations of caregivers (see Khoury, Rajamani, et al., 2020; Liotti, 2004; Lyons-Ruth et al., 1999). Our findings also extend the evidence for the validity of the CDRR scales. The non-significant relation between young adults’ trauma history and punitive stances toward fathers deserves attention. Whereas our findings show that young adults recall displaying punitive stances toward fathers, it is possible that experiencing familial adversity may not necessarily translate into punitive stances with fathers but rather with mothers (as we found) due to more reliance on mothers for the safe haven role when experiencing distress (Kerns et al., 2015). Young adults who experienced childhood adversity might have greater expectations for protection from mothers, and this idea should be further tested.

Further, while different types of childhood adversity (i.e., emotional or physical threat and deprivation) may pose unique developmental consequences (McLaughlin et al., 2019), previous authors suggest a multi-risk model where cumulative adverse events may contribute to attachment disorganization, in association with parental responses (e.g., helplessness, withdrawal, and hostility; Lyons-Ruth & Jacobvitz, 2016). While we assessed cumulative risks using the ACES, future research should evaluate how a broader spectrum of adversities with specific transgressors, not captured by this scale, may interact with parental behavior to predict disorganized and controlling patterns. For example, in the context of marital discord, it is possible that children may adopt caregiving strategies toward a parent reliant on them for support not provided by their partner. Some children might adopt punitive stances to defend against the more intrusive or controlling parent in the relationship (Bureau et al., 2009b). Alternatively, other children may become disorganized due to the unpredictability in the family dynamic (see Moss et al., 2011). Addressing such questions may elucidate if there are associations between each disorganized/controlling pattern and various types of adverse events.

We also evaluated whether the CDRR disorganized and controlling scales relate to young adults’ depressive and anxiety symptoms. Consistent with our expectations and some existing studies (e.g., Brumariu et al., 2018; Bureau et al., 2009a; Madigan et al., 2016 for a meta-analysis), CDRR scales capturing disorganized and punitive attachment patterns were all related to greater depressive and anxiety symptoms. In line with some past studies (e.g., Meier et al., 2014; Obsuth et al., 2014), affective caregiving toward mothers was associated with greater depression, and affective caregiving toward fathers was associated with greater depression and anxiety. These findings support the notion that controlling caregiving may be more relevant to psychopathology at periods later than middle childhood (e.g., Brumariu et al., 2018; Bureau et al., 2009a) when, for example, young adults have to reconcile previously subordinated needs/goals with the competing needs of carving out a life course/identity. Overall, results support the validity of the CDRR’s subscales for disorganized and controlling behaviors.

Regarding subjective well-being, disorganized and affective caregiving stances toward fathers and disorganized/punitive and mutually hostile stances toward mothers were related to less happiness and life satisfaction. Punitive stances toward fathers were also associated with less life satisfaction. Notably, this study is the first to evidence that childhood experiences of disorganized and controlling attachment may impede young adults’ subjective well-being. Theoretically, difficulties with emotion regulation (e.g., Brumariu et al., 2021), peer relationships (e.g., Groh et al., 2014), and poor self-esteem (see, e.g., Lecompte et al., 2014) are good candidates to explain these relations, as such impairments are also associated with well-being (e.g., Moreira et al., 2021). Future research should test these hypotheses to provide a more nuanced picture of how disorganized and controlling attachments interfere with the capacity to experience well-being, aside from signaling risk for psychopathology.

Greater affective caregiving, however, did not relate to less meaning in life, possibly because young adults may still garner rewards for caregiving behaviors (e.g., continuing to be well-liked for being a good child who cares for a struggling parent; see Moss & Lecompte, 2015), that help them compensate and not yet tackle life meaning. However, as expected, mutual hostility toward mothers and disorganized stances toward fathers did. Elements of these patterns, such as chronic conflict or unresolved fear, may be detrimental to the sense that life is worth living or comprehensible, defining components of meaning in life (Martela & Steger, 2016). Thus, our findings show that young adults engaged in hostile relationships with their mothers and disorganized relationships with their fathers are at risk for struggling with meaning-making and extend support for the CDRR scales’ utility.

We found partial support for our hypothesis that appropriate boundaries would relate to less internalizing problems and greater subjective well-being. Appropriate boundaries with mothers and fathers did not relate to depression and anxiety. Thus, while a necessary aspect of adaptive attachment relationships, maintaining appropriate boundaries may not mitigate the risks of internalizing problems on its own. However, appropriate boundaries with fathers were related to greater happiness, life satisfaction, and meaning in life. Previous work suggests that fathers may uniquely contribute to children’s secure exploration or typically function more as a secure base (e.g., Kerns et al., 2015). This may promote autonomy by fostering a sense that support is available in novel situations and persistence in new disappointing contexts. Since autonomy is essential in young adulthood (Koepke & Denissen, 2012), fathers having maintained appropriate parental roles would likely relate to well-being at this age. After controlling for participant gender, appropriate boundaries with mothers was related to happiness and life satisfaction. Thus, mothers not abdicating parental responsibilities may similarly support well-being.

Overall, disorganized/punitive and mutually hostile stances toward mothers and disorganized and affective caregiving stances toward fathers appeared most salient to young adults’ mental health. Distinct and similar patterns of associations suggest that there may be overlapping and unique aspects of young adults’ experience of attachment relationships with each parent (Meier & Bureau, 2018). Future research should pay attention to disorganized patterns for mothers and fathers separately, as disorganized/controlling stances with both parents do not appear identical in their adjustment implications.

Several limitations qualify our findings. Model fit was marginal for the father version of the CDRR. While we lacked a theoretical justification for specifying residual covariances, such covariances could account for this limited model fit. Further, this study's cross-sectional design precludes causal/directional inferences. We relied on participants' self-reports, which could be prone to bias and shared-method variance. Thus, future studies should employ longitudinal designs and evaluate how disorganized and controlling attachments assessed with the CDRR relate to young adults’ adjustment assessed with other methods (e.g., interviews). Additionally, our sample consisted of a convenience sample of college students, mostly females and majoring in psychology. Thus, it would be valuable to test if these results replicate in a community sample more representative of the general population and in high-risk/clinical samples where results may be more powerful (Cyr et al., 2010). Since parent–child relationships may change over time, future studies might explore specific periods in childhood using the CDRR. Validating the CDRR against observational measures, such as the GPACS, is an essential next step. Future research should also assess the CDRR's relation with constructs that may differentiate the disorganized stances such as externalizing problems and familial loss (Lyons-Ruth & Jacobvitz, 2016), and outcomes identified as relevant to these patterns in young adulthood, including suicidality (Lyons-Ruth et al., 2015), dissociation (Byun et al., 2016), partner abuse (Obsuth et al., 2014), features/diagnoses of BPD (Khoury, Zona, et al., 2020; Lyons-Ruth et al., 2015), and attention to comforting attachment-related images (Schneider et al., 2022). Overall, the CDRR opens the door to novel prospective studies on disorganized/controlling attachments’ developmental trajectories from infancy to young adulthood, addressing questions regarding factors that influence their development, sequelae, and opportunities for intervention.

In summary, our findings are consistent with the notion that the CDRR is a promising measure of young adults' perceptions of their childhood disorganized and controlling attachments and provide evidence of its structural stability. The CDRR’s relations to childhood adversity, internalizing problems, and subjective well-being support its construct validity. Associations between internalizing problems and controlling caregiving, unique to this age group, underscore young adulthood’s relevance as an area of focus for future attachment research and the value of the CDRR. Lastly, our study advances the literature by evidencing that disorganized/controlling attachments with each parent signal risk for maladjustment and may interfere with well-being.