Where should bipolar disorder appear in the meta-structure?

BACKGROUND: The extant major psychiatric classifications, DSM-IV and ICD-10, are purportedly atheoretical and largely descriptive. Although this achieves good reliability, the validity of a medical diagnosis is greatly enhanced by an understanding of both risk factors and clinical history. In an effort to group mental disorders on the basis of risk factors and clinical manifestations, five clusters have been proposed. The purpose of this paper is to consider the position of bipolar disorder (BPD), which could be either with the psychoses, or with emotional disorders, or in a separate cluster. METHOD: We reviewed the literature on BPD, unipolar depression (UPD) and schizophrenia in relation to 11 validating criteria proposed by the DSM-V Task Force Study Group, and then summarized similarities and differences between BPD and schizophrenia on the one hand, and UPD on the other. RESULTS: There are differences, often substantial and never trivial, for 10 of the 11 validators between BPD and UPD. There are also important differences between BPD and schizophrenia. CONCLUSION: BPD has previously been classified together with UPD, but this is the least justifiable place for it. If it is to be recruited to a 'psychotic cluster', there are several important respects in which it differs from schizophrenia, so the cluster would have a division within it. The alternative would be to allow it to be in an intermediate position in a cluster of its own.