Jones et al (Reference Jones, Scott and Haque2005) have focused on the important although relatively neglected area of psychosocial aspects/intervention in bipolar affective disorder. Although there are several previous reports on the subject by the same group, this study has a better design and a much larger sample size. However, some central issues remain unresolved.

The authors were unable to find dysfunctional beliefs specific to bipolar disorder. Cognitive therapy as practised in depressive or panic disorders attempts to correct characteristic dysfunctional beliefs (Reference Beck, Rush, Kaplan and SaddockBeck & Rush, 2000). In the absence of a specific pattern of dysfunctional beliefs, devising effective and specific cognitive strategies to treat bipolar disorder may be difficult. This is illustrated by the pilot study of cognitive therapy in bipolar disorders by the same group (Reference Scott, Garland and MoorheadScott et al, 2001) in which relatively non-specific strategies such as self-management of symptoms, dealing with non-adherence, anti-relapse techniques, etc. were employed. The lack of precise techniques could also have resulted in the differential efficacy of cognitive therapy, with effects mainly on depressive, rather than manic symptoms.

In the current study Jones et al used a 24-item sub-scale version of the Dysfunctional Attitude Scale, whereas in earlier studies (Reference Scott, Stanton and GarlandScott et al, 2000; Reference Scott and PopeScott & Pope, 2003) a 40-item scale was used. It is not clear whether the use of different versions of this scale contributed to the ambiguous nature of the dysfunctional beliefs found in bipolar disorder, especially since the two different versions appear to have different sub-scales. Finally, although some potential confounding variables, such as current mental state, were controlled for, others, such as duration of illness, severity, chronicity and possible effects of pharmacophrophylaxis, were not. Cognitive style may vary according to these factors (Reference Scott and PopeScott & Pope, 2003) making it necessary to control for them.

It is possible that these concerns will be addressed by future research. This study paves the way for examination of psychosocial factors in bipolar disorder.