Adopted Child Attachment Insecurity
Nola L. Passmore
Candida C. Peterson
Although there is evidence that adopted persons fare worse than non-adoptees in terms of relational adjustment and other psychosocial variables, some studies have produced null results, and others have demonstrated the wide variability in adjustment among adoptees.
Several authors in this area have discussed issues pertaining to loss, abandonment, and rejection. However, researchers have not systematically examined the impact of adoption on adults' attachment security and relationship outcomes, or the moderating role of family and search/reunion experiences. This paper reports on the first phase of a longitudinal study of adults who were adopted as infants, and a comparison sample of adults who were raised by both biological parents. Measures of attachment styles and attachment dimensions indicated less security in the adopted sample than the comparison sample. However, variability was somewhat greater among adoptees, and those who had not searched for birth relatives were generally similar to the comparison sample. Within the adopted sample, attachment security was related to perceptions of childhood experiences and current relationships with adoptive parents and, to a lesser extent, relationships with birth mothers. Ongoing analyses will focus on stability and change in relationship variables, and in-depth exploration of adopted persons' experiences.
Although there is evidence that adopted persons fare worse than non-adoptees in terms of relational adjustment and other psychosocial variables, some studies have produced null results, and others have demonstrated the wide variability in adjustment among adoptees. Several authors in this area have discussed issues pertaining to loss, abandonment, and rejection. However, researchers have not systematically examined the impact of adoption on adults' attachment security and relationship outcomes, or the moderating role of family and search/reunion experiences. This paper reports on the first phase of a longitudinal study of adults who were adopted as infants, and a comparison sample of adults who were raised by both biological parents. Measures of attachment styles and attachment dimensions indicated less security in the adopted sample than the comparison sample. However, variability was somewhat greater among adoptees, and those who had not searched for birth relatives were generally similar to the comparison sample. Within the adopted sample, attachment security was related to perceptions of childhood experiences and current relationships with adoptive parents and, to a lesser extent, relationships with birth mothers. Ongoing analyses will focus on stability and change in relationship variables, and in-depth exploration of adopted persons' experiences.
Both theory and research highlight the importance of personal relationships in meeting needs for comfort and security, and in promoting well being (Baumeister & Leary, 1995; Weiss, 1991). Although issues regarding the quality and stability of relationships are of general importance, there are compelling reasons for studying the relationship experiences and concerns of adult adoptees. Adoptees have lost the major person(s) with whom attachments normally form (i.e., biological parents). Further, with recent changes in legislation, more adoptees are now searching for birth relatives, and this process may entail further loss and rejection. Hence, adoption may be a risk factor for relationship difficulties in adult life.
Adoption, Family Experiences and Adjustment
Many researchers have investigated the extent to which adoption is a risk factor for general adjustment difficulties. There is evidence that adoptees are over-represented in clinical populations, and report higher levels of psychosocial difficulties (e.g., self-esteem, depression) than non-adoptees (Borders, Penny, & Portnoy, 2000; Cubito & Obremski-Brandon, 2000; Levy-Shiff, 2001; Wierzbicki, 1993). However, Collishaw, Maughan, and Pickles (1998) reported that adoptees did not differ from the general population in terms of psychological distress, and Borders et al. (2000) found no difference between adoptees and their friends in terms of life satisfaction. Further, Sharma, McGue, and Benson (1996) reported higher levels of prosocial behaviour among adopted than nonadopted adolescents. Although methodological differences may explain some of these mixed results, another explanation is that the association between adoption status and psychosocial outcomes is moderated by search status and family experiences. On average, adoptees who have searched for birth parents (‘searchers’) have lower self-esteem than non-searchers (Aumend & Barrett, 1984; Borders et al., 2000; Sobol & Cardiff, 1983). Further, Sobol and
If adoption is a risk factor for psychosocial difficulties, at least for some adoptees, then many adoptees may also experience difficulties in interpersonal relationships. Issues concerning loss and betrayal (which are inherently relational) are central to the adoption experience. Not only have adoptees lost their birth parents; they have also lost other birth relatives, knowledge of their genetic heritage, and a sense of being biologically tied to significant others (Jones, 1997; Schechter & Bertocci, 1990). In cases where reunion is unsuccessful, adoptees may feel rejected again, and thus experience a double loss. It is widely accepted that parental loss through death or divorce predicts is linked to insecurity and interpersonal difficulties (Kobak, 1999), but the losses associated with adoption have unique features that may predispose individuals to relationship problems. Specifically, these losses are covert and often unacknowledged or underestimated, and may entail a sense of betrayal, abandonment and mistrust (Brodzinsky, 1990; Jones, 1997; Nickman, 1985). Although several studies have examined the family relationships of adoptees, few have assessed the impact of adoption on the peer relationships of adult adoptees. As argued next, attachment theory is particularly well suited to addressing these issues.
The Adult Attachment Perspective
Some years ago, Hazan and Shaver (1987) argued that adults’ close relationships (especially those with intimate partners) share important emotional, behavioural and functional similarities with the bonds that form between infants and their caregivers. They also argued that the concept of ‘attachment style’ (secure, avoidant, anxious/ambivalent) is relevant to both types of relationships. That is, individual differences in adult security predict key relationship processes and reflect, in part, childhood experiences with attachment figures. Subsequent research has generally supported these propositions (Feeney, 1999). Further, measures of adult attachment have evolved rapidly: Although some studies still rely on categorical measures (e.g., secure, preoccupied, fearful and dismissing), there has been a move toward the use of more reliable multiple-item measures.
Recently, Edens and Cavell (1999) made a strong case for the utility of attachment theory in the study of adoption. They argued that current conceptualisations of adult attachment are directly relevant to relationship phenomena unique to adoptees, including loss of biological ties, and the potential for search and reunion. To date, however, Borders et al. (2000) are the only researchers to have systematically investigated attachment security in adult adoptees. These researchers studied adoptees and their non-adopted friends. Although the two groups were similar in terms of marital satisfaction and sensitivity to rejection, they differed with regard to adult attachment and social support. Adoptees (regardless of search status) were over-represented in the preoccupied and fearful attachment groups, and under-represented in the secure group. Adoptees also reported less social support than their non-adopted friends, although this association was moderated by search status: Searchers reported less support than non-searchers and non-adopted respondents. This study provides an important first step in exploring the link between adoption and adult attachment, but was limited by its cross-sectional nature, its reliance on a categorical measure of attachment, and its failure to fully consider the role of early parenting and ongoing relationship experiences.
In summary, despite the considerable literature on issues of infant attachment, loss, rejection, abandonment and trust in adoptees, no study has comprehensively explored the impact of adoption on dimensions of attachment security and relationship outcomes in adulthood, or the moderating role of family and search/reunion experiences. Such studies are essential in order to develop best practice for adult adoptees who may be at risk of relationship problems. This paper reports on the first phase of a study addressing these issues. We expected that adults who were adopted as infants would report higher levels of attachment insecurity than adults who grew up with both biological parents (Hypothesis 1a). However, the adopted sample was also expected to show greater variability on attachment measures (Hypothesis 1b). Within the adopted sample, insecurity was expected to be higher for those who had searched for birth relatives (Hypothesis 2), and for those who perceived relationships with adoptive parents and birth mothers in a more negative light (Hypothesis 3).
Participants were 140 adults who were adopted as infants, and a comparison sample of 128 adults who grew up with both biological parents. Adopted participants were required to have been adopted in
As part of a larger study, all participants completed measures of attachment security. In addition, adoptees reported on relationships in the adoptive family (parental bonding, discussion of the adoption, current emotional closeness), and on their search and reunion experiences.
Current attachment security was measured in two ways. First, attachment style was assessed by asking participants to endorse one of the four attachment descriptions (secure, preoccupied, dismissing, fearful), developed by Bartholomew and Horowitz (1991). Second, participants completed the 40-item Attachment Style Questionnaire (ASQ), developed by Feeney, Noller, and Hanrahan (1994).
Demographic characteristics of comparison and adopted samples
The ASQ measures five dimensions of adult attachment: confidence in self and others (8 items; e.g., ‘I feel confident about relating to others’), discomfort with closeness (10 items; e.g., ‘I prefer to keep to myself’), need for approval (7 items; e.g., ‘It’s important to me that others like me’), preoccupation with relationships (8 items; e.g., ‘I worry a lot about my relationships’), and relationships as secondary to achievement (7 items; e.g., ‘Achieving things is more important than building relationships’). Each item is rated on a 6-point scale, from 1 (totally disagree) to 6 (totally agree). All five scales were reliable, with alpha coefficients ranging from .74 to .88.
The Parental Bonding Instrument (PBI; Parker, Tupling, & Brown, 1979) contains 25 items describing various parental attitudes and behaviours. This measure requires participants to think back over the first 16 years of life, and to rate each item (separately for mother and father) from 0 (very unlike this parent) to 3 (very like this parent). Adoptees answered these questions with respect to their adoptive parents. The PBI yields scores on the dimensions of care (12 items, e.g., ‘spoke to me with a warm and friendly voice’), and overprotection (13 items, e.g., ‘tried to control everything I did’). Both scales were highly reliable, with alpha coefficients exceeding .90 for each parent.
Additional questions about the adoptive family.
In addition, adoptees were asked about the adoptive family’s attitude toward discussing the adoption (1 = the topic was taboo or the source of lies and misinformation) to 3 = discussion was open and honest). They also rated current emotional closeness to the adoptive mother and adoptive father (1 =extremely distant to 6 = extremely close).
Search and reunion experiences
Adoptees answered a series of questions about their search and reunion experiences, five of which are considered in this paper. First, those who had searched for birth relatives rated how supportive the adoptive mother and adoptive father were of their decision to search (1 = extremely opposed to 6 =very supportive). Finally, those who had met their birth mothers rated their satisfaction with the initial reunion and satisfaction with the current relationship (1 = extremely dissatisfying to 6 = extremely satisfying), and emotional closeness of the current relationship (1 = extremely distant to 6 =extremely close).
For both samples, participants were recruited through the first-year Psychology pools at the
Adoption and Attachment Characteristics
The association between adoption and attachment security was assessed in two ways. First, a frequency comparison was conducted, relating sample (comparison versus adopted) to the four-group (forced-choice) measure of attachment style. The association between sample and attachment category was significant, 2 (3) = 19.71, p < .001. Adopted persons represented only 38% of the secure group; in contrast, they represented 59% of the dismissing group, 64% of the preoccupied group, and 72% of the fearful group.
Second, MANOVA was used to assess differences between the samples on the five scales of the ASQ. This analysis revealed a significant overall difference, multivariate F (5, 262) = 4.10, p < .001; further, univariate tests showed that this difference applied to all five scales. Adopted persons obtained lower scores than comparison persons on confidence, and higher scores on all remaining attachment scales (see Table 2, top rows). The multivariate test of homogeneity variance was marginally significant, indicating greater variability in attachment scores within the adopted sample than the comparison sample. However, this effect applied only to the confidence scale.
To assess the role of search status, a more fine-grained MANOVA was conducted in which searchers (n = 106), nonsearchers (n = 33) and comparison participants (n = 126) were compared on the attachment scales. Significant differences emerged on all scales except for relationships as secondary. Post hoc (Tukey) tests showed a consistent pattern, involving significant differences between searchers and comparison participants (see Table 2): Searchers reported lower levels of confidence, and higher levels of discomfort, preoccupation, and need for approval. Interestingly, non-searchers and comparison participants did not differ on any of the five scales.
Mean scores and standard deviations on attachment scales according to group
Note. Conf. = Confidence, Disc. = Discomfort, Relate second. = Relationship as secondary, Need approv. = Need for approval, Preocc. = Preoccupation
Relationships with Adoptive Parents
To assess the role of relationships with adoptive parents, the attachment scales were correlated with reports of childhood relationships with adoptive parents, openness of discussion, and current emotional closeness (Table 3). All attachment scales were associated with reports of relationships with adoptive parents, although the results were strongest for confidence and discomfort. In terms of parental bonding, confidence was related positively to parental care (from both adoptive mother and adoptive father), and negatively to parental overprotection. Further, confidence was related positively to open discussion of adoption, and to ratings of current emotional closeness. Conversely, discomfort was related negatively to parental care, open discussion and emotional closeness, and positively to parental overprotection.