Drs Bhui & Bhugra write that categories are confused by comparing Turkish and Moroccan people with Hindustanis. However, it is perfectly clear from the text and the tables that people were classified by country of birth (or country of birth of the parents) and that immigrants from Turkey or Morocco were compared with immigrants from (the country of) Surinam.

We defined Hindustanis as people who migrated in the 19th century from British India to Surinam (not to The Netherlands, as Drs Bhui & Bhugra suggest). In the Netherlands, the designation Hindustani is not a religious category and small proportions are Muslim or Buddhist.

Drs Bhui and Bhugra note that social factors, such as ethnic density, cultural identity, unemployment and separation from parents, are “ entirely unexplored”. Apparently, it escaped their attention that the proportions of children growing up in single-parent families were presented on p. 367 (Selten et al, 2001). This proportion is lowest for the group with the highest schizophrenia risk (Moroccans). Similarly, the pattern of unemployment rates (higher for Turkish and Moroccan people than for Surinamese or Antilleans) does not resemble the pattern of schizophrenia rates (lowest for Turkish people) (Selten & Sijben, 1994). It follows that an explanation in terms of social factors is not as easy as suggested.

Drs Bhui and Bhugra correctly argue that attention should be paid to ethnic differences. Our study reported an increased incidence of schizophrenia among Surinamese citizens of The Hague, mostly Hindustanis. One might ask, therefore, whether the increased incidence also applies to African Surinamese immigrants, who have common origins with African—Caribbeans in the UK. An answer can perhaps be found indirectly. When they migrated to The Netherlands in the 1970s and early 1980s, Hindustanis and African Surinamese differed greatly in their preference for a Dutch residence. As a result, the ethnic composition of the Surinamese community in Amsterdam (70% African and 26% Hindustani) is different from that in The Hague (10% African and 80% Hindustani) (Martens & Verweij, 1997). The Dutch psychiatric registry gave information on the residence of Surinamese-born people who had been discharged with an ICD diagnosis of schizophrenia in the period 1978-1996. The Central Bureau for Statistics provided figures for the Surinamese-born populations in Amsterdam and The Hague on 1 January 1990 (mid-period). The age structures of both populations were very similar. The results can be summarised as follows. In The Hague: 23 815 (Surinamese-born) citizens in 1990; 253 citizens admitted over the 19-year period. In Amsterdam: 38 722 (Surinamese-born) citizens in 1990; 493 admitted. These results although not definitive, suggest that the incidence of schizophrenia is also increased for African Surinamese immigrants. This conclusion is supported by clinical experience and a small incidence study in Amsterdam (Dekker et al, 1996).