The October 2001 issue of the Journal reports two prevalence studies of depressive disorders (Reference Ayuso-Mateos, Vázquez-Barquero and DowrickAyuso-Mateos et al, 2001; Reference Thompson, Ostler and PevelerThompson et al, 2001). Both studies used a self-report questionnaire as an initial screening device although both avoided the sometimes reported but unjustified assertion of diagnosis based upon scores of the scales. Such scales are widely used in the manner reported by these studies and a cautionary comment is in order. There is a widespread view that the selection of instrument is unimportant so long as it is designated as a ‘depression’ scale; this is not true. For instance, the scales used in the above-mentioned studies were the Beck Depression Inventory (BDI) (in Reference Ayuso-Mateos, Vázquez-Barquero and DowrickAyuso-Mateos et al, 2001) and the depression sub-scale of the Hospital Anxiety and Depression scale (HAD—D) (in Reference Thompson, Ostler and PevelerThompsonet al, 2001). These two instruments high-light very different aspects of depressive disorders (Reference SnaithSnaith, 1993). The HAD—D has 86% of its variance directed to mood symptoms (depressed mood and anhedonia) but an absence of cognitive symptoms (hopelessness, low self-esteem and guilt ideation). With the BDI the reverse is the case, with 14% directed to mood and/or anhedonia but 33% focusing on the cognitive symptoms.

There is an unfortunate tendency to refute the importance of difference based upon predominant psychopathology and even, within the realm of depressive disorders, to deny the importance of diagnosis. Indeed, the first study uses the term ‘prejudice’ when referring to the separation of disorders and frankly advocates the conflation of disorders of major depression (for which one or other of the mood symptoms is prerequisite for diagnosis) and the other group of ‘dysthymia and adjustment disorders’, which are characterised by the cognitive distortion. Until diagnostic practice is based on exact psychopathology, research will remain in its present state of confusion. For instance, the oft-repeated statement that cognitive therapy and biological treatments are of equal worth in the treatment of ‘depression’ will continue to be made. The statement may be true if no distinction is made between different depressive disorders but non-responders to the one or other treatment will have different characteristics: the psychotherapeutic approach will be more successful in the disorders based on cognitive distortion whereas the biological treatments are likely to be more effective when major depressive disorder is present.