In my editorial (Reference Woodruff, Murphy and HerjanicTyrer, 2001) I argued that ‘diagnoses that have face validity should at least be tested in the classification arena before being accepted or rejected’. In fact, ‘cothymia’ has more than face validity in its favour. We have just completed a further study (Tyrer et al, 2001) that shows that its long-term outcome (12 years) with regard to clinical symptoms, service contact and social function is nearly 50% worse than that of single mood disorders and it is as powerful a predictor as personality disorder (Reference Tyrer, Seivewright and SimmondsSeivewright et al, 1998) in indicating the prognosis of common neurotic disorders. If we persist in regarding this association as yet another example of comorbidity, we are unlikely to make progress in the treatment of what appears to be a very morbid condition. Grant-giving bodies are very reluctant to provide funding for treatment interventions for conditions that have no formal existence.

Dr Malhi's argument for retaining the separate diagnoses of anxiety and depression, pending further investigation and research into its chronology, is somewhat recondite and would carry more weight if the jury had not been sitting on this case for over 30 years (Woodruff et al, 1967) before coming to a verdict. Leff's early work (Reference Tyrer1978) identifies the heart of the problem: joining anxiety and depression upsets the epidemiologists, pharmacologists and researchers who would like anxiety and depression to stay apart, but for patients and clinicians these emotions are often joined together intimately. Let them at least have a trial marriage before condemning them indefinitely to live in nosological sin.