We agree that there is inevitably a limitation in the use of measures developed in a different cultural setting. Our measure of depression, the Composite International Diagnostic Interview (CIDI), lacks sensitivity because of the strict diagnostic rule. This could account for the low prevalence rate in our study and the fact that we may have picked up only those most seriously affected. However, we do not doubt the presence of depression in our society. The impetus for our study came from the ‘lived lives’ of Ethiopian psychiatrists working within Ethiopia who commonly encounter people presenting with symptoms according to a ‘Western’ construct of depression in a tertiary care setting. These people respond to anti-depressants by showing good recovery from symptoms and regaining their original level of functionality. In an ongoing intervention programme, we have found the same for people with depression identified by the CIDI in Butajira (study ongoing).

The CIDI was translated, back-translated and modified by experienced Ethiopian psychiatrists who considered the symptom questions to have face validity and applicability. In addition, convergent validity of CIDI-defined depression was indicated by our finding of strong associations between depression and disability. We believe that the CIDI is unlikely to be merely detecting physical ill health because first, it incorporates specific measures to screen out symptoms that seem to have a physical cause and second, our study participants with persistent depression were most disabled in social domains rather than in those domains of functioning more likely to be influenced by physical impairment (e.g. mobility).

We believe that the difference in mental health across cultures is mainly in the presenting features, not in the nature of the disorder. In low- and middle-income countries it has been said that people tend to present with somatic symptoms (Reference Mumford, Saeed and AhmadMumford et al, 1997; Reference Parker, Gladstone and CheeParker et al, 2001). However, this view of cultural difference between the West and the rest of the world was challenged by the World Health Organization cross-cultural study in primary care (Reference Gureje, Simon and UstunGureje et al, 1997). Although the presentation of depression clearly does vary across cultures, in an African setting depression was found to be better characterised by core depressive symptoms than by somatic complaints (Reference Okulate, Olayinka and JonesOkulate et al, 2004).