Skip to main content
The BMJ logoLink to The BMJ
. 2002 Apr 13;324(7342):916. doi: 10.1136/bmj.324.7342.916

Schizophrenia in ethnic minority groups

Selection bias in prevalence data is difficult to rule out

Dirk Jacobs 1
PMCID: PMC1122853  PMID: 11950753

Editor—To sociologists, Boydell et al's findings are counterintuitive.1 One would expect economic deprivation (at neighbourhood level) to be a decisive factor for an increased incidence of mental illness. But it is surprising to learn that the lower the proportion of non-white ethnic minorities in a local area the higher the incidence of schizophrenia in those minorities (controlled for economic deprivation).

As an explanatory hypothesis the authors point to overt discrimination and institutionalised racism as sources of stress, which can be alleviated by people making use of social capital within the ethnic group. This hypothesis surely necessitates further testing and debate. It is a pity that non-white ethnic minority groups had to be considered as one homogeneous group on an aggregate level. The social networks and levels of social cohesion may be different for different ethnic groups, and follow up research should be able to distinguish these.

Boydell et al assume that all people with schizophrenia will come into contact with psychiatric services, but this requires closer attention. Members of an ethnic minority with a mental disorder who live in predominantly white neighbourhoods may be more likely to come into contact with psychiatric services. Probing for mental disorders might be more likely in predominantly white neighbourhoods than in non-white neighbourhoods. This is not necessarily ruled out by the fact that there is job mobility of clinical staff, since institutional cultures can both consciously and unconsciously shape and influence individual professional practice (and in fact necessitate individual adaptation).

Different processes of self selection in contacting health services or looking for particular types of treatment may operate in different areas. It might be that in mainly non-white neighbourhoods, which often are also the most economically deprived areas, mental health issues among non-white groups are considered to be “luxury” problems compared with other health or social problems. As a result, incidences might be underestimated. The risk of being diagnosed as mentally ill in white (and often better off) neighbourhoods might be higher because of cultural-institutional factors.

The findings of this study are interesting. Sociologists signal different levels of tolerance, or willingness to label someone as deviant (for example, as “ill” or “insane”), according to the social setting.2

References

  • 1.Boydell J, van Os J, McKenzie K, Allardyce J, Goel R, McCreadie R, et al. Incidence of schizophrenia in ethnic minorities in London: ecological study into interactions with environment. BMJ. 2001;323:1336–1338. doi: 10.1136/bmj.323.7325.1336. . (8 December.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Catalano R. Ecological factors in illness and disease. In: Freeman H, Levine S, editors. Handbook of medical sociology. Englewood Cliffs, NJ: Prentice Hall; 1989. pp. 87–101. [Google Scholar]
BMJ. 2002 Apr 13;324(7342):916.

Authors' reply

J Boydell 1, R M Murray 1

Editor—We agree that our findings are preliminary and demand both replication and further investigation. We are currently studying social capital in the area and hope to be able to measure this separately for the larger ethnic minority groups. Regarding the problem of selection bias, we examined incidence (number of new cases) not prevalence (total number of cases). Several studies, including one from the area we studied, have shown that a very high percentage of people with schizophrenia come into contact with psychiatric services.1-11-3 Furthermore in our study most incident cases were admitted via emergency psychiatric services provided centrally for the whole area. We therefore consider it very unlikely that institutional factors and labelling phenomena have influenced our results.

As schizophrenia is still highly stigmatised we do not believe that non-whites in the smaller groups selected themselves for hospitalisation. Similarly, the very high rates of compulsory admission are against the view that schizophrenia is considered a “luxury” problem in our local community. It is, of course, never possible to completely rule out selection bias, as we discussed in the paper, but the magnitude of our findings give us confidence in the conclusion that the smaller the minority, the greater the incidence of schizophrenia.

References

  • 1-1.Cooper JE, Goodhead D, Craig T, Harris M, Howat J, Korer J. The incidence of schizophrenia in Nottingham. Br J Psychiatry. 1987;151:619–629. doi: 10.1192/bjp.151.5.619. [DOI] [PubMed] [Google Scholar]
  • 1-2.Kendell RE, Malcolm DE, Adams W. The problem of detecting changes in the incidence of schizophrenia. Br J Psychiatr. 1993;162:212–218. doi: 10.1192/bjp.162.2.212. [DOI] [PubMed] [Google Scholar]
  • 1-3.Prince M, Phelan M. Invisible schizophrenia: a postal survey of the incidence and management of new cases of schizophrenia in primary care. J Mental Health. 1994;3:91–98. [Google Scholar]

Articles from BMJ : British Medical Journal are provided here courtesy of BMJ Publishing Group

RESOURCES