It was encouraging to see the study by Westen et al (Reference Westen, Dutra and Shelder2005) published in a mainstream journal such as the British Journal of Psychiatry. For a variety of reasons, there is a reluctance among many British adolescent mental health clinicians to diagnose personality disorders in their patients, despite the clear presence often of the requisite diagnostic features. This study shows that personality disorders in adolescents can be validly diagnosed, whether using an established framework such as the DSM-IV or a new, empirically derived taxonomy.

I was struck by the extent to which certain factors derived from the Q-factor analysis appear to map onto current conceptualisation of attachment categories, a point borne out partially by the same research group (Reference Nakash-Eisikovits, Dutra and WestenNakash-Eisikovits et al, 2002) using Bartholomew's attachment classification (Reference Bartholomew and HorowitzBartholomew & Horowitz, 1991). Specifically, the ‘psychological health’ factor shows strong correspondence to features of a secure internal working model, while the ‘histrionic sexualisation’ and ‘emotional dysregulation’ factors contain items integral to the conceptualisation of ambivalent/preoccupied attachment. It is interesting that in their 2002 study, the group found that attachment avoidance was correlated with their ‘avoidant’ Q-factor but not with DSM-IV avoidant personality disorder; on this basis, they questioned the prevailing conceptualisation of avoidant personality disorder. It is unclear how attachment disorganisation is related to the SWAP-200-A factors, as it is still uncertain the extent to which Bartholomew's ‘fearful’ category corresponds to disorganised/unresolved attachment.

Therefore, it is perhaps logical to hypothesise that some personality trait constellations (the most maladaptive of which may constitute personality disorders) are indeed disorders of attachment. This hypothesis, which is supported theoretically (Reference Nakash-Eisikovits, Dutra and WestenNakash-Eisikovits et al, 2002) and which makes intuitive sense to many adolescent mental health professionals, needs to be tested with longitudinal research. In addition to other empirical work, the above research shows the continuing clinical importance of attachment theory. However, there is still no easily administered validated measure of adolescent attachment in widespread clinical use currently in the UK. Surely, this is a deficit that needs to be addressed.