268 Chapter 32
clinicians should not forget to assess all the other factors that influence
the life course. Over the longer term, if individuals are followed into
their mid-twenties, attachment patterns are only modestly stable: in-
fant classification typically correlates 0.2 with adult classification. This is
partly explained by different life courses and experiences of the type
indicated above.
Bearing these caveats in mind, a secure attachment nonetheless does
increase the likelihood of the child subsequently forming harmonious
relationships with adults and children. This is most evident for close
relationships with family members and friends. Securely attached children
are more cooperative and responsive with their mothers, more likely to
comfort younger siblings, and more likely to have good friends. They are
less likely to be non-compliant with parents, quarrelsome with siblings,
or controlling with friends. The benefits of secure attachments are also
evident with familiar but less intimate social partners. On average, se-
curely attached children are less emotionally dependent on teachers and
are better able to ask for a teacher’s help when they face a challenge
they cannot manage alone. Typically, they are also more popular with
classmates and less often victimised, perhaps because they show more
empathy for peers and engage in less conflict when playing. Attachment
security has least influence on the quality of social interactions with
unfamiliar adults or children. Such interactions are primarily influenced
by the child’s sociability, which is a moderately heritable temperamental
trait. Conversely, genetic factors seem less important in determining the
quality of close relationships.
Early studies, using the ABC classification of attachment, particularly
emphasised the link between type A (avoidant) attachment and externalis-
ing problems such as aggression. It now seems likely that the more recently
recognised type D (disorganised) attachment is the strongest predictor of
externalising problems. For example, one study showed that disorganised
attachment at 18 months predicted a six-fold increase in serious aggression
towards peers in nursery school. Though the disorganised attachment
often included avoidant elements, it was noteworthy that those chil-
dren who had purely avoidant (and not disorganised) attachments were
not subsequently more likely to be unusually aggressive towards peers.
Perhaps severe family adversity launches children onto a developmental
pathway characterised by disorganised attachment and dysphoria in in-
fancy, oppositional defiant disorder in middle childhood, and more severe
conduct disorder and juvenile delinquency in adolescence.
An increasing number of studies have used the AAI to investigate the
attachment classification of adults with mental illnesses or personality
disorders. In clinical samples, the likelihood of an autonomous (that
is, secure) attachment is only about 10%, as compared with roughly
60% in low-risk samples. The remaining 90% of clinic patients are split
fairly evenly between the three insecure attachment categories: dismiss-
ing, preoccupied and unresolved. So far, there are only hints of links
between particular psychiatric diagnoses and specific types of insecurity,