Recent contributions for understanding Inhibited Reactive Attachment Disorder

ABSTRACT

This special issue brings together studies representing the current state of the art research on inhibited reactive attachment disorder (RAD) from developmental psychology and developmental attachment research. This editorial introduces and discusses conceptual topics of the development and  the assessment of inhibited RAD. Next, five empirical studies of this special issue cover timely topics from developmental trajectories and longitudinal stability of RAD to associations with caregiving experiences in institutions, IQ, and psychopathology, moreover aspects of internal working models of children with inhibited RAD are examined. The commentary meta-analytically integrates the empirical evidence of all five studies and offers theoretically derived suggestions for further research. In this editorial, we discuss unsettled issues and future directions for attachment oriented research in this field.

Inhibited reactive attachment disorder: perspectives from developmental psychology and attachment research

The detrimental effects of prolonged deprivation, low-quality institutional care or neglect and maltreatment within families on children’s social, emotional, cognitive, and biological development have been demonstrated clearly in many cross-sectional and longitudinal studies within the last 40 years (e.g. Tizard & Rees, 1974; Zeanah et al., 2009; Loman, Wiik, Frenn, Pollack, & Gunnar, 2009; Kumsta et al., 2015; Cicchetti, 2016; Millward, Kennedy, Towlson, & Minnis, 2006; Soares et al., 2014; Bovenschen et al., 2016). One of the key developmental outcomes of such extreme negative caregiving experiences can be attachment disorders. According to ICD-10, attachment disorders encompass an inhibited reactive and a disinhibited subtype. The changes from DSM-IV to DSM-5 more clearly separate reactive attachment disorder (RAD) comparable to the inhibited subtype in the ICD-10 classification from disinhibited social engagement disorder (Zeanah & Gleason, 2015). However, some attachment researchers also regretfully question this loss of an attachment framework for the disinhibited form of attachment disorders in DSM-5, as a valuable theoretical framework for intervention (Lyons-Ruth, Zeanah, & Gleason, 2015). Compared to research on the disinhibited or DESD subtype, the number of studies that exclusively focus on inhibited RAD is limited (Gleason et al., 2011) and the impressive English and Romanian Adoptee (ERA) study also only reports data on disinhibited/DESD subtype (O’Connor & Rutter, 2003; Kumsta et al., 2015).

This special issue exclusively brings together studies on inhibited RAD coming from an attachment perspective that try to give answers to some research lacunas in the field. In a former special issue on attachment disorders, O’Connor and Zeanah (2003) stated that at that time the empirical evidence on RAD was smaller than the amount of reviews on that topic. Since then, two major longitudinal studies on the consequences of early deprivation have been followed up until early adolescence and emerging adulthood (Kumsta et al., 2015; Zeanah, Humphreys, Fox, & Nelson, 2017), and many empirical studies on attachment patterns or RAD assessment in deprived children from different countries and cultures have been published or reviewed (e.g. Vorria et al., 2003; Bakermans-Kranenburg, Dobrova-Krol, & Van IJzendoorn, 2012 ; Zeanah, Gunnar, McCall, Kreppner, & Fox, 2011; Oliveira, Fearon, Belsky, Fachada, & Soares, 2015). Thus, empirical evidence has increased, but there still are debates regarding the assessment of inhibited RAD and the possible comorbidity of inhibited RAD and disinhibited/DESD subtype. Gleason and colleagues (2011) have shown that the DAI scales for inhibited and disinhibited RAD are significantly moderately positively correlated. However, when using a clinical cut-off both disorders do not overlap. Zeanah and Gleason (2015) convincingly reviewed the empirical evidence on this distinction using the DAI. Researchers using other assessment approaches report more overlap between both subtypes of RAD (Mayes, Calhoun, Waschbusch, & Baweja, 2017) in small samples. Thus, similarly to the state of the art as reported in the first special issue on attachment disorders (O’Connor & Zeanah, 2003), there still are different approaches for the assessment of RAD that mainly are based on caregiver reports or clinical impression.

Supported by the Köhler-Stiftung, teams of Gottfried Spangler, Isabel Soares, Helen Minnis, and Peter Zimmermann discussed about observation assessment tools for signs of inhibited RAD. Based on these meetings they developed observational tools to assess signs of RAD guided by attachment theory and the rich empirical database on RAD. Empirical results of this discussion are reported in this special issue beside other attachment-based studies on inhibited RAD. We prefer the term “inhibited RAD” here to avoid mixing the term with other studies using RAD for both the inhibited and disinhibited type.

The special issue addresses three research lacunas in the field of inhibited RAD: (1) longitudinal developmental trajectories, (2) age-specific observation of signs of inhibited RAD, and (3) aspects of internal working models (IWMs) of children with inhibited RAD.

Longitudinal development of inhibited RAD

One understudied area in the field of attachment disorder is the topic of longitudinal (dis)continuity of inhibited RAD over time. There is little empirical evidence regarding the development of attachments in children with severely negative and deprived caregiving experiences and signs of inhibited RAD over time and whether earlier signs of RAD in infancy and toddlerhood are predictive of later signs of RAD. The manuscript by Guyon-Harris, Humphreys, Degnan, Fox, Nelson, & Zeanah (2018) reports results of the well-known Bucharest Early Intervention Project (BEIP) on longitudinal development of RAD from infancy to age 12 applying a person-centered approach. This study highlights different trajectories of inhibited RAD development, differentiating children with prolonged elevated inhibited RAD signs over time from pathways of children who are quickly responding to high-quality foster care. As a substantial number of children in the foster care group and the care-as-usual group change in their caregiving environment over time (e.g. from foster care to institutions or institutions to the biological parents) this approach reveals the significant influence of early care and time spent in institutions on long-term development of both inhibited RAD and the suggested increase of symptoms in early adolescence.

Observation of signs of inhibited RAD

A second major lacuna in research on inhibited RAD is the assessment within an attachment framework. Most developmental attachment studies use observation of children’s behavior either in attachment-relevant, emotion-eliciting contexts like the separation–reunion paradigms or during long observations at home (Solomon & George, 2016). The assessment of inhibited RAD either is based on systematic caregiver reports with specific interviews (Disturbances of Attachment Interview (DAI); Smyke & Zeanah, 1999) or direct reports in questionnaires by caregivers (e.g. RPQ, Minnis, Rabe-Hesketh, & Wolkind, 2001). However, direct systematic observation of children’s behavior, a standard for attachment research, is rarely applied for the assessment of inhibited RAD. In some highly informative studies, observation during separation–reunion paradigms is used to assess children’s attachment patterns or signs of disinhibited RAD, but not to assess inhibited signs of RAD (Kreppner, Rutter, Marvin, O’Connor, & Sonuga‐Barke, 2011; McLaughlin, Zeanah, Fox, & Nelson, 2012). In this special issue, two observational approaches for the assessment of inhibited RAD for different age groups are presented. The first coding system for infancy and early childhood, the RInAB developed by Corval, Belsky, Baptista, Mesquita, and Soares (2018), assesses three dimensions of inhibited RAD derived from attachment theory. The authors clearly support the assumed internal structure of their assessment approach in a psychometric analysis and examine core attachment related aspects of validity by exploring associations with caregiver sensitivity in institutions in Portugal. The cross-cultural applicability of this assessment approach is reported in the studies by Spangler and colleagues (2018) in Germany and Georgia (in Eurasia) and by Bruce and colleagues (2018) in the UK. The studies from Germany and Georgia show that the RInAb can validly be used in both countries, especially in the institutional context. Moreover, the existence of a primary caregiver in institutions also is associated to a lower rate of observed signs of inhibited RAD. In addition, the study by Bruce and colleagues (2018) found moderate stability of observed inhibited RAD symptoms based on a preliminary version of the RInAB in a sample of at-risk foster children in the UK.

A second observational approach for the assessment of inhibited RAD in older children, the CADB-C, is reported by Zimmermann and Iwanski (2018a). In their study, a RAD-risk group shows an increased number of reliably assessed observed signs of inhibited RAD during emotionally challenging social interactions and increased caregiver reported DAI and RPQ scores of inhibited RAD compared to non-clinical controls.

All these studies suggest that inhibited RAD can be reliably and validly assessed by systematic observation with different assessment approaches adapted to children’s age. Such observational approaches can be additionally informative to clinicians as sometimes requested (Woolgar & Baldock, 2015), especially when including an attachment framework.

IWMs in children with inhibited RAD

A third topic that has not yet been explored in detail are the characteristics of IWMs of self and others in children with inhibited RAD. Based on caregiving experiences, children build up IWMs that influence information processing and self-regulation in attachment-relevant contexts (Dykas & Cassidy, 2011; Zimmermann & Iwanski, 2015). Securely attached children will expect support and availability from specific, self-selected caregivers, whereas children classified as insecurely-avoidant attached consistently expect rejection of emotional needs or emotional unavailability from specific selected caregivers. However, the empirical basis for the functioning of IWMs in children with RAD is poor or even non-existent. Children with such extreme negative caregiving experiences will have difficulties in developing consistent expectations towards one selected caregiver. Bowlby (1973) speculated that children growing up in nurseries would have even more generally negative and distrustful IWMs of others. Indeed, children with inhibited RAD do not seek proximity of a caregiver when distressed. Moreover, they have difficulties in differentiating their caregivers as potential attachment figures from other adults. Thus, their IWMs may be characterized by generalized negative expectations and distrust towards everyone independent of even supportive prior experiences and a highly negative view of their self. Although IWMs of attachment figures are also expected to be generalized to other emotionally close relationships in case of organized or disorganized attachment patterns (Bowlby, 1973; Bretherton, 1996) there is a clear lacuna in empirical attachment research whether similar processes can be found in children with inhibited RAD. The study by Bosmans and colleagues (2018) clearly shows that children in middle-childhood with high scores on the inhibited RAD scale of the DAI report less felt trust towards their teachers and that they show more observed distrust while interacting with them. Similarly, Zimmermann and Iwanski (2018a) found that children of a RAD-risk group showed more observable signs of negative IWMs (e.g. distrust, false positive affect) towards caregivers and strangers when emotionally aroused compared to non-clinical controls. Both studies also explored associations between inhibited RAD and children’s self-concept as one component of the IWM of self in middle-childhood with one study finding significant associations (Zimmermann & Iwanski, 2018a) and one did not (Bosmans et al., 2018), suggesting some diversity, maybe depending on sample characteristics.

Resume

Current research on inhibited RAD somehow seems to be divided in the two domains of clinical assessment traditions in child and adolescent psychiatry and developmental attachment research with little overlap. This may contribute to the diversity of results found in the literature. A profound understanding of similarities and differences in assessment of inhibited RAD and of the comorbidity and differentiation with other disorders is a timely issue. In a comprehensive commentary, Atkinson (2018) provides meta-analytic evidence of moderate but significant associations between the observational assessment approaches and the widely accepted standard assessment with the DAI (Smyke & Zeanah, 1999). Moreover, in his second meta-analysis he reports evidence of significant associations between inhibited RAD and caregiver-rated internalizing problems, but also some comorbidity with externalizing behaviors. Nevertheless, inhibited RAD is clearly different from other disorders characterized mainly by intense signs of anxiety or depression in children. Atkinson (2018) recommends an attachment perspective on signs of inhibited RAD in socially and emotionally deprived children with a closer look at the “not to classify” or “unclassifiable” or “insecure-other” categories that have been reported in studies on inhibited RAD as well as attachment studies (Guyon-Harris et al., 2018; Kreppner et al., 2011).

Future directions

The number of systematic empirical studies on RAD clearly has increased over the last years. However, there still may be some more steps to be done, as pointed out below.

One future direction that needs more expansion and replication may be including a developmental perspective as shown in the study by Guyon-Harris and colleagues (2018). Besides using the same assessment tools to test homotypic continuity over time, adjustment of assessment of RAD to the development of cognition, self-control, and emotion regulation that may change the phenotype of RAD in middle-childhood and adolescence should be examined systematically. Similarly to developmental changes in the expression of attachment needs and the changes in the scoring of attachment security from infancy to adolescence, there may be developmental changes in the way inhibited RAD looks like in older children and adolescents compared to infants and toddlers. Proximity seeking as a sign of attachment security transforms into seeking psychological security after emotional arousal in adolescence (Zimmermann & Iwanski, 2018b). Thus, controlling behavior in middle-childhood may be a developmental follow-up of earlier fearful hypervigilance as older children can be more proactive compared to infants (Atkinson, 2018). Longitudinal follow-ups of deprived children like in the BEIP are necessary. However, studies on children with a long “career” in different institutions or foster families are equally important to observe age-specific signs of RAD in those children who have enduring difficulties in developing new attachments. This can be examined cross-sectional as well.

Another topic for further research is the differential diagnosis given the comorbidity of RAD with other disorders or symptoms. For example, the high comorbidity with Attention Deficit Hyperactivity Disorder (ADHD) in the ERA study (Kennedy et al., 2016)and the BEIP might lead some researchers or clinicians to favor diagnosing the more frequent disorders (Woolgar & Baldock, 2015). However, this is no rationale for excluding RAD as diagnosis. Indeed, not all children growing up in deprived environments develop inhibited RAD or other forms of psychopathology and insecure attachment patterns beside RAD also are is prominent (Lionetti, Pastore, & Barone, 2015). An attachment figure within an institution or high-quality foster care can reduce inhibited RAD and psychopathology in childhood (Corval et al., 2018; Zeanah et al., 2017). Moreover, genetic polymorphisms (e.g. 5-HTTLPR) can moderate the pathogenic effect of early deprivation on later psychopathology or attachment problems (Bakermans-Kranenburg et al., 2012; Baptista, Belsky, Mesquita, & Soares, 2017; Humphreys, Zeanah, Nelson, Fox, & Drury, 2015), similarly as in case of insensitive parenting in non-deprived families (Spangler, Johann, Ronai, & Zimmermann, 2009).

In addition, future research needs to identify observable signs of RAD and signs of other disorders in socially deprived and non-deprived children, including social information processing. This may help to better distinguish insecure attachment from the clinical diagnosis of inhibited RAD and also might help to avoid preferring more frequent diagnosis like ADHD instead of RAD in cases where ADHD symptoms are observed but RAD may be more prominent.

Although strong empirical evidence showing that adverse contextual conditions – like institutional care – promote attachment disorders, not all institutionalized children develop disturbances. This calls attention to the need to consider non-institutional factors as contributors to the development of inhibited RAD. Following this, the focus on the links between genes, brain, and social contexts is crucial for understanding its emergence despite deprivation. In addition, the role of epigenetics as a mechanism by which early adversity influences DNA methylation is a promising avenue of empirical inquiry that could illuminate the evolution of inhibited RAD. Indeed, the reversibility of environmentally induced epigenetic marks already evinced (Weaver et al., 2004; Weaver, Meaney, & Szyf, 2006) seems compatible with the remission of RAD symptoms once the child is placed in high quality social environments (Gleason et al., 2011; O’Connor, Rutter, & Romanian Adoptees Study Team, 2000).

The contributions of this special issue represent some aspects of current research on inhibited RAD all informed from an attachment perspective and with a developmental background. We hope that this special issue attracts attention to new and interesting directions and topics in the domain of inhibited RAD and stimulates further research in this field.

Additional information

Funding

We are very grateful to the Köhler-Stiftung, Germany [Grant number S 112/10161/12] who supported the organization of meetings on inhibited RAD.