Attachment, behavioral inhibition, and anxiety in preschool children.

Kagan and colleagues (Kagan, Reznick, & Snidman, 1987) have described a temperament construct termed "behavioral inhibition to the unfamiliar" (BI), which is characterized by the predisposition to be irritable as an infant, unusually shy and fearful as a toddler, and quiet, cautious, and withdrawn in the preschool and early school age years, with marked behavioral restraint and physiological arousal in unfamiliar situations. The opposite temperamental construct has been termed "behaviorally uninhibited" (BUI), and is characterized by a bold, extroverted, sociable, and fearless approach to novel situations, objects, and people. Behavioral inhibition has been widely studied over the past 20 years and has been proposed as a risk factor for anxiety disorders in childhood (Rosenbaum et al., 1991, 1992). However, there are marked differences in the outcomes of inhibited children. Not all behaviorally inhibited children remain inhibited over the early childhood years (Reznick et al., 1986), and of those children who do remain inhibited (i.e., stably inhibited), not all develop anxiety disorders (Biederman et al., 1990), implicating other factors as important in the development of anxiety in childhood. Integrated models propose that environmental factors, such as parent-child attachment, may combine with temperament to increase the risk for the development of childhood anxiety (Manassis & Bradley, 1994; Mills & Rubin, 1993). However, only two empirical attempts have been made to examine both behavioral inhibition and attachment influences on childhood anxiety (Manassis, Bradley, Goldberg, Hood, & Swinson, 1995; Warren, Huston, Egeland, & Sroufe, 1997). Hence, the major aim of this study was to contribute to this literature by examining the relationships between insecure attachment, behavioral inhibition, and child anxiety in an at-risk sample of preschool-aged children.

Behavioral Inhibition and Anxiety

Numerous studies have found that behaviorally inhibited children have increased risk for multiple anxiety and phobic disorders (Biederman et al., 1990; Kagan, Snidman, Zentner, & Peterson, 1999; Rosenbaum et al., 1988), with the association being largely accounted for by children who remain inhibited over time. Using Kagan's longitudinal sample and the original classification system of behavioral inhibition, Hirshfeld and colleagues (1992) found that those children who remained inhibited across four assessments occurring between 21 months and 7.5 years of age had the highest risk for anxiety disorders relative to children who were not persistently inhibited. Prior, Smart, Sanson, and Oberklaid (2000) reported similar results with an Australian sample, using questionnaire ratings of shy temperament. Turner, Beidel, and Wolff (1996) concluded, in their review of studies of behaviorally inhibited children, that the most extremely inhibited children in these studies (approximately 10% of those studied) were the most likely to remain inhibited throughout middle childhood and also to be at the highest risk for developing anxiety disorders.

Kagan and his associates have argued that behavioral inhibition is best examined at its extremes and reflects a categorical rather than continuous dimension of infant temperament (Kagan, Reznick, & Gibbons, 1989). Indeed, in Kagan's research the stability and prediction of inhibition from infancy to early childhood was only evident when extreme groups were considered. Although the majority of temperamental traits are regarded as continuous variables, most agree that it is the individuals at the extremes of the distribution who are most vulnerable (Rothbart & Bates, 1998). As the bulk of the research on the relationship between behavioral inhibition and anxiety disorders has used Kagan's conceptualization of behavioral inhibition (Turner et al., 1996), we followed the Harvard group categorical classification and compared inhibited children with their uninhibited counterparts.

In summary, the research findings taken together indicate that behavioral inhibition represents a proneness to anxiety, placing the child at an increased risk for developing an anxiety disorder. However, behavioral inhibition can be considered neither necessary nor sufficient for the development of anxiety. Family relationship factors, such as parent-child attachment, may increase the likelihood of the transformation from a temperamental proneness to an anxiety disorder.

Attachment and Anxiety

Attachment theorists argue that the development of anxiety may be linked to insecure attachment (Bowlby, 1973; Sroufe & Waters, 1977). From an attachment theory perspective, each infant, however treated, will become attached to its caregiver by the end of the first year. However, individual differences in the secure versus insecure quality of attachments have been observed, resulting from differences in the caregiver's availability and responsiveness to the infant and the degree of reciprocity between the infant and the caregiver (Carlson & Sroufe, 1995). After observing infant-mother interactions in the home every 3 weeks for over a year, Ainsworth and colleagues (Ainsworth, Blehar, Waters, & Wall, 1978; Ainsworth & Wittig, 1969) devised an experimental procedure (the "Strange Situation") to assess differences in quality of attachment that involved two brief separations and reunions between parents and their 12-month-old infants in a laboratory environment. Although separation from an attachment figure in unfamiliar circumstances produces physiological arousal in all infants, they vary in their responses to their caregivers in the face of this arousal. Ainsworth and her collaborators (1978) discovered three types of attachment patterns based on infant behavior displayed in the Strange Situation: secure, insecure-avoidant, and insecure-ambivalent. These patterns of attachment related to the early care the infants had received at home.

Infants classified as secure with their caregiver (B classification) were able to use the caregiver as a secure base for exploration in the unfamiliar situation. They were distressed when separated from their mother but were readily comforted by her on reunion, and eventually returned to play. Secure infant behavior is based on the experience of well-coordinated, sensitive interactions with a mother who is consistently available and appropriately responsive to her infant's needs (Ainsworth et al., 1978). The mother helps the infant to achieve a sense of mastery over threatening or frightening situations and to manage distressed feelings in a well-modulated and effective way (Cassidy, 1994).

Infants classified as insecure-avoidant (A classification) tended to show minimal distress on separation and avoided their mother on reunion. Mothers of these infants were found to express an aversion to physical contact when their infants were upset, frequently rejecting their infants when they sought comfort and reassurance (Ainsworth et al., 1978). These infants tended to minimise or suppress expressions of negative affect (Cassidy & Kobak, 1988) and avoided their mother on reunion, presumably as a defense against painful feelings in relation to their mother's unavailability.

Infants classified as insecure-ambivalent (C classification) were noticeably unable to use the caregiver as a secure base for exploration, tending to seek proximity and contact with the caregiver even prior to the separation. These infants became overtly distressed upon separation, and showed angry, resistant, and ambivalent behavior upon reunion with their mother, displaying proximity-seeking behavior but then resisting contact angrily once it was achieved. Mothers of insecure-ambivalent infants were found to be the most inconsistent in their availability and least competent in comforting their infants, and directly interfered with their infant's exploration (Cassidy & Berlin, 1994).

A fourth group of insecure disorganized/disoriented (D classification) infants was later identified (Main & Solomon, 1986). These infants exhibited unusual, conflicted behaviors, such as behavioral stilling, stereotyped movements, or direct apprehension with regard to the parent that indicated an inability to maintain one coherent pattern of attachment behavior in the face of distress. Mothers of disorganized infants have been described as either frightening to the child or frightened themselves, due to the experience of personal trauma or loss (Main & Hesse, 1990). Hence, for these infants, the caregiver serves as a source of both fear and reassurance, which understandably leads to conflicted behavior in their infants.

Conceptually similar patterns of A, B, and C attachment have subsequently been identified in preschool-aged children (Cassidy & Marvin, 1992) in a comparable separation and reunion context. The major change in attachment behavior in the preschool age group is the child's greater reliance on verbal communication rather than physical contact for expressing attachment needs, for negotiating issues like separation from the mother, and for re-establishing the relationship following the mother's return. In addition, while some insecure disorganized/disoriented children continue to demonstrate a lack of a coherent attachment strategy at preschool age, other children show a new profile of controlling behavior. This new insecure pattern involves the child taking some control of the mother--child relationship in either a care-giving or punitive way when the mother herself fails to take appropriate responsibility in the caregiving role.

According to Bowlby, anxiety originates in an infant's uncertainty about caregiver availability, which is the fundamental condition underlying insecure attachment. It has been suggested that the type of insecurity that is most likely to be associated with later anxiety disorders is the insecure-ambivalent classification (Cassidy & Berlin, 1994). As a consequence of the unpredictable and inconsistent availability of their caregiver, children with insecure-ambivalent attachments are chronically anxious, worrying about whether their needs will be met (Warren et al., 1997), and constantly fearful of being left vulnerable and alone. These children are characterized by heightened monitoring of their mother and inhibited exploratory behavior. Fearfulness may result from limited familiarity with the environment. In …