disorganized attachment (Vondra et al., 2001). Similarly, a meta‐analytic study conducted by Fearon and colleagues (2010) found a significant relation between attachment insecurity and externalizing symptoms across studies. A complimentary meta‐analytic study conducted by Groh et al. (2012) found a significant link between attachment insecurity, particularly avoidance, and internalizing symptoms. This conclusion agrees with many other studies linking avoidant attachments to increased anxiety symptoms in school‐aged children (Warren, Emde, & Sroufe, 2000; Shamir‐Essakow, Ungerer, & Rapee, 2005). Importantly, these findings suggest that while attachment relates to various forms of psychopathology, the relation to externalizing problems is strongest.
Adolescence
Among adolescents, research has linked insecure attachment to higher symptoms of both internalizing and externalizing disorders in clinical samples (Lacasa, Mitjavila, Ochoa, & Balluerka, 2015; Venta, Shmueli‐Goetz, & Sharp, 2014), as well as elevated trauma symptoms (Jardin, Venta, Newlin, Ibarra, & Sharp, 2015), suicidal ideation (Venta, Mellick, Shatte, & Sharp, 2014), and borderline personality disorder (covered later in Chapter 13) traits. While these studies point to non‐specific relations between attachment insecurity and psychopathology, the picture becomes a bit clearer when looking just at studies that explore subtypes of insecure attachment. For example, preoccupied attachments have been linked to depression, social problems, aggression, and thought problems (e.g., Lacasa et al., 2015) as well as suicidal ideation (Adam, Sheldon‐Keller, & West, 1996) and eating disorders (Miljkovitch, Pierrehumbert, Karmaniola, Bader, & Halfon, 2005). In contrast, avoidant attachments have generally been linked to externalizing behaviors like oppositional defiant behaviors, conduct problems, social problems, and substance use (Lacasa et al., 2015; Venta et al., 2014). The broad array of psychopathology associated with insecure attachment in adolescents is mirrored among adults, in which studies have linked nearly every form of psychopathology to attachment insecurity (see Mikulincer & Shaver, 2007).
Developmental Continuities and Discontinuities
Overall, research has been mixed regarding the continuity of attachment. For instance, some research suggests that attachment classifications at 15 months are not associated with attachment classifications at age 14 (Groh et al., 2014). However, other research shows greater stability in attachment security from 24 months to 18 years (Groh et al., 2014). One category that does seem stable is disorganization, in which significant links to adult disorganization are seen as early as 15 months.
Research has sought to understand discontinuities in attachment. Individuals who were reclassified from secure in early childhood to insecure at 18 years of age were likely to have experienced lower and declining maternal sensitivity; were less likely to be living with their fathers; and had mothers who reported increased negative life events, compared to individuals who remained securely attached across development (Booth‐LaForce et al., 2014). Individuals who changed from early insecure attachments to secure attachments experienced higher maternal sensitivity during the timeframe in question. Continuous insecure attachment was associated with less maternal sensitivity and greater paternal depressive symptoms (Booth‐LaForce et al., 2014). Thus, this research indicates that insecure attachment is not a fixed property of an individual but clearly malleable by the environment and experience.
A Developmental Psychopathology Model of the Etiology of Attachment Security
Figure 4.4 depicts a developmental psychopathology model of attachment security, laying out how risk and protective factors interact in the development of attachment security.
Genetic Factors (Distal Risk Factors)
Family and twin studies
Most attachment research has focused exclusively on environmental predictors of attachment security like caregiving, which have been supported by several twin studies. Bokhorst and colleagues (2003) found that more than half of the variance in attachment security was attributed to twins’ shared environment, with 48% being attributed to the non‐shared environment. However, Fearon and colleagues (2014) documented a high rate of heritability among adolescents! The authors suggest that adolescence is a phase of reorganization during which attachment changes from predominantly behavioral to predominantly cognitive, and that genetic factors are particularly influential during this time. Similarly, research examining genetic links to dimensions of attachment anxiety and avoidance suggest that about 40% of the variability in attachment security is attributable to genes. Critically, several studies have noted that this genetic influence is strongest in relation to attachment anxiety (Picardi, Fagnani, Nistico, & Stazi, 2011).
Molecular genetic studies
There does not appear to be any one attachment gene (Fearon et al., 2014). Some evidence suggests a link between OXTR, the receptor that controls oxytocin—a neuropeptide that has been linked to social and attachment behaviors (Feldman, Weller, Zagoory‐Sharon, & Levine, 2007)—and attachment styles and psychopathology (Costa et al., 2009). However, genes associated with dopamine receptors (i.e., DRD4 and DRD2), which are implicated in social behavior, have also been linked to attachment (disorganization, Lakatos et al., 2002; and anxiety, Gillath et al., 2008). Research has also suggested a genetic link between the serotonin receptor HTR2A and attachment (Gillath et al., 2008).
FIGURE 4.4 A Developmental Psychopathology Model of Insecure Attachment
Neurobiological Factors (Proximal Vulnerability Factors)
Structural and functional imaging findings
Like with genes, there is no “attachment circuit” explicitly defined in the brain. Attachment behaviors have complex relations to neuroanatomy, and, thus, many brain regions are implicated. Nonetheless, a few brain structures and circuits implicated in threat detection and regulating experienced emotions are particularly important, such as the amygdala, the hypothalamus, and other structures included in the hypothalamic–pituitary–adrenocortical (HPA) axis. Indeed, studies have linked greater activation in the amygdala to the telling of attachment stories that contain themes of loss, death, abuse, or abandonment (Buchheum et al., 2006) and anxious attachment (Vrticˇka, Anderson, Grandjean, Sander, & Vuilleumier, 2008). Further, the hypothalamus has been implicated in the soothing effect of attachment figures (see Coan, Schaefer, & Davidson, 2006). These studies suggest that neurobiological vulnerabilities underlie insecurely attached individuals’ tendency to be over‐reactive to threat and less equipped to regulate that threat response.
Neurotransmitters
As you may have guessed based on the genetic research linking attachment to DRD4 and DRD2, there is a consensus that the neurotransmitter dopamine, which regulates the brain’s reward systems, plays a role in bonding between parents and children as well as between adults (Insel, 2003). For example, low maternal care has been linked to an elevated stress dopamine response (Pruessner, Champagne, Meaney, & Dagher, 2004). Strathearn and colleagues (2008) found that brain areas associated with dopamine‐related reward processing displayed increased activation in response to the mother’s own infant, when compared with an unknown infant. Further, Strathearn and colleagues (2009) found that secure mothers showed greater activation in areas of the brain associated with the dopaminergic