Introduction

Health policy makers are paying increasing attention to assessing the needs of minority ethnic groups with the aim of basing policy on reliable data.¹ Most epidemiological research concerning mental disorders in different ethnic groups in Britain has concentrated on psychotic disorders and has shown higher treated prevalence and incidence among African Caribbeans. ^{2 3} By contrast, there have been few surveys of common mental disorders (depression and anxiety) in this ethnic group. One recent study found a lower prevalence of anxiety symptoms in Caribbeans (13%) compared with whites (18%) but higher estimated weekly prevalence of depressive neurosis (6.0% in Caribbeans and 3.8% in whites).⁴ The increased unemployment and poverty among British African Caribbeans⁵ together with the effects of racism suggest that anxiety and depression might be more common among this ethnic group than in white Europeans.

Participants and methods

Design and instruments
We used a two phase design in order to screen a large sample but limit the number of lengthy psychiatric interviews.⁶ The whole sample received the first phase screening instrument (12 item general health questionnaire⁷). All those scoring 3 or more were included in the second phase sample together with a 1 in 4 random sample of those scoring 2 or less. We chose a low threshold to ensure that few cases would be missed. The threshold has been validated in previous general population surveys⁷ and used with African Caribbeans in primary care.⁸

Sampling
We obtained a random sample from family health services authority population registers of four participating general practices in central Manchester (Moss Side and Hulme), where over half of Manchester's 10 000 African Caribbean population live. Sampling was done between September 1993 and February 1996 in collaboration with a comparative survey of nutrition, diabetes, and hypertension being run by one of the authors (JKC).¹⁰ General practitioners' records were checked to confirm that subjects were still registered. Those who were dead or had changed address were removed from the final denominator.

Procedure
Each person was contacted by telephone, post, or home visits. Non-respondents were those who refused to participate, consistently failed to keep appointments for interviews, or were unavailable after at least five home visits. Subjects were interviewed by prearranged appointment at home or in their doctor's surgery. Demographic and other health data were collected, and the first phase psychiatric screening instrument administered. Ethnic group was categorised by the subject from the list of 1991 census categories. Black Caribbean and black other (Caribbean) categories were combined to form the category African Caribbean. (As place of birth and that of parents were also recorded Asian or African "black other" subjects were not included.) Respondents selected for the second phase were interviewed at home by the research psychiatrist (CMS). Ethical approval was obtained from the Manchester Health Commission.

Analysis
Data were analysed with SPSS/PC+. We calculated prevalence estimates (and 95% confidence intervals) according to the method of Pickles et al¹¹ for men and women separately, and in total, for each ethnic group.

Results

First phase
We contacted 1467 people; 590 were no longer resident or registered with the index general practices and 13 had died. Of the 864 remaining, 131 refused to participate in the study and 121 were persistently unavailable. Thus 612 people (337 African Caribbeans and 275 white Europeans) completed screening: a 71% response rate.

Second phase
At the time of the second interviews five African Caribbeans and five white Europeans who had scored three or more on the general health questionnaire had died or moved out of the area; interviews were completed with 92 African Caribbeans (response rate 88%) and 62 white Europeans (response rate 79%). A random sample of those scoring below 3 was also identified; 53 African Caribbeans were contacted, of whom 49 were still resident and 35 completed interviews (response rate 71%). Similarly, 55 white Europeans were contacted, 50 were still resident, and 41 interviewed (response rate 82%).

Help seeking behaviour by people with mental disorder
In response to the question "Do you have any problems with your health?" most "cases" reported some kind of psychological illness (24/29 (83%) African Caribbeans and 25/29 (86%) white Europeans). Others denied they were ill or reported exclusively somatic symptoms or physical illnesses despite sufficient symptoms being evident during the psychiatric assessment to diagnose mental disorder.

Discussion

This is one of the first population based surveys in Britain to address the prevalence of common mental disorders in African Caribbeans. The prevalence estimates are similar to those identified in recent general population surveys,¹² and in contrast to findings for psychotic disorders, show that in an inner city population the prevalence of anxiety and depressive disorder is similar in African Caribbeans and white Europeans.

There was no difference in medical help seeking between African Caribbeans and white Europeans with mental disorders. Although over 80% had consulted their doctor in the previous six months, most presented with somatic (rather than psychological) symptoms. Further research is needed to explore the differences in recognition rates by general practitioners as we did not have sufficient numbers to detect significant differences. The prominent belief that medical consultation would not be beneficial has implications for health promotion.

We recruited fewer subjects than we had intended. Lists of registered patients proved to be out of date as many people had changed address. Unfortunately, as the ethnicity of the members of the initial sample was unknown, any possibility of differential migration out of the inner city could not be addressed. Attrition between phases one and two was high but not linked to ethnicity, and this could be due to delay (in some cases) between the two phases. Using the general health questionnaire as a screening instrument with a 2/3 threshold had limitations. Although few true cases were missed, the false positive rate was high. Delay before second phase interviews may also have contributed to this.

The researchers were not blind to the ethnicity of the respondents, and the psychiatric researcher (CMS) was white European. The criticism of eurocentrism could be made, but it is not obvious if this would have led to underestimation or overestimation of psychiatric symptoms among African Caribbeans.

The differences in rates of anxiety (lower in African Caribbeans) and depressive disorders (higher in African Caribbean women) between the ethnic groups agree with previously reported trends.⁴ This difference could be explained by genetic or vulnerability factors or by exposure to different social or environmental experiences.¹³ African Caribbeans may be experiencing more events and difficulties associated with loss, and fewer associated with fear, than white Europeans.¹⁴

Only a minority of people with mental disorders found by the prevalence survey were being treated by their general practitioner. Low recognition of mental disorders remains the greatest barrier to care for both African Caribbean and white European people with depression and anxiety. The training of doctors in the care and management of these common mental disorders therefore needs to continue to focus on the somatic presentation of mental disorder (somatisation) coupled with the development of therapeutic models that include social, psychological, and pharmacological treatments.¹⁵

In conclusion, this study shows that common mental disorders are similarly prevalent in African Caribbeans and white Europeans living in a British inner city. The emphasis on psychotic disorders in both research and service provision should not obscure the fact that depression and anxiety afflict far more people of all ethnic groups. There is no evidence that ethnicity is associated with more or less morbidity, but it may be an important factor in determining what type of disorder is experienced.

Acknowledgments

We thank the researcher workers of the diabetes and hypertension project; project assistants Colin Moore, Jane Rogers and Amanda Lee; and the patients and staff of participating general practices.

Contributors: CMS and FC conceived and designed the study and interpreted the results. FC obtained funding on CMS's behalf, supervised the training, and is the study guarantor. CMS supervised project assistants, collected and analysed data, and wrote the manuscript. JKC and LR conceived and ran the comparative survey of nutrition, diabetes, and hypertension and supervised data collection. BT contributed to study design and, together with CMS, carried out statistical analysis. All authors edited the manuscript.

Funding: Wellcome Clinical Epidemiology Training Fellowship (CMS).

Competing financial interests: None declared.

References

1. Department of Health. Health of the nation key area handbook. Mental illness. 2nd ed. London: HMSO , 1994.
2. Bhugra D, Leff J, Mallet R, Der G, Corridan B, Rudge S. Incidence and outcome of schizophrenia in whites, African Caribbeans and Asians in London. Psychol Med 1997; 26: 791-798.
3. Harrison G, Glazebrook C, Brewin J, Cantwell R, Dalkin T, Fox R, et al. Increased incidence of psychotic disorders in migrants from the Caribbean to the United Kingdom. Psychol Med 1997; 26: 799-806.
4. Nazroo JY. Ethnicity and mental health. London: Policy Studies Institute , 1997.
5. Amin K. Poverty in black and white. London: Child Poverty Action Group , 1992.
6. Vazquez-Barquero JL, Diez-Manrique JF, Pena C, Quintanal RG, Labrador Lopez M. Two stage design in a community survey. Br J Psychiatry 1986; 149: 88-97[Abstract/Free Full Text].
7. Goldberg D, Williams P. A user's guide to the general health questionnaire. Windsor: NEFR-Nelson , 1988.
8. Lloyd KR, Jacob KS, Patel V, St Louis L, Bhugra D, Mann A. The development of the short explanatory model interview (SEMI) and its use among primary-care attenders with common mental disorders. Psychol Med 1998; 28: 1231-1237[Medline].
9. World Health Organisation. SCAN schedules for clinical assessment in neuropsychiatry. Geneva: WHO Division of Mental Health , 1992.
10. Mbanya J-CN, Cruickshank JK, Forrester T, Balkau B, Ngongyang JY, Riste L, et al. Standardized comparison of glucose intolerance in west African-origin populations: rural and urban Cameroon, Jamaica, and Caribbean migrants to Britain. Diabetes Care (in press).
11. Pickles A, Dunn G, Vazquez-Barquero JL. Screening for stratification in two phase ("two stage") epidemiological surveys. Statistical Methods in Medical Research 1995; 4: 73-89[Abstract/Free Full Text].
12. Meltzer H, Gill B, Petticrew M, Hinds K. OPCS surveys of psychiatric morbidity in Great Britain: report 1. London: HMSO , 1995.
13. Kendler KS, Neale MC, Kessler RC, Heath AC, Eaves LJ. Major depression and generalised anxiety disorder. Arch Gen Psychiatry 1992; 49: 716-722[Abstract/Free Full Text].
14. Finlay-Jones R, Brown GW. Types of stressful life events and the onset of anxiety and depressive disorder. Psychol Med 1981; 11: 803-815[Medline].
15. Goldberg D. The management of medical out-patients with non-organic disorders: the reattribution model. In: Creed F, Mayou R, Hopkins A, eds. Medical symptoms not explained by organic disease. London: Royal College of Psychiatrists, Royal College of Physicians , 1992.

Commentary: Counting heads may mask cultural and social factors

Greta Rait, lecturer in primary care. 

Department of Primary Care and Population Sciences, Royal Free and University College London Medical Schools, Archway Resource Centre, Whittington Campus, London N19 5NF

g.rait{at}ucl.ac.uk

This study on the prevalence of common mental health disorders is commendable, but the findings should be interpreted with caution. Cross cultural studies of mental health are controversial and difficult to conduct and interpret.

Ethnicity is a complex variable.¹ For example, African Caribbean, which is used in this study, encompasses people from a multitude of islands with diverse cultures. Presentation and prevalence of illness is also likely to vary. The broad term white European also encompasses many ethnic groups, including Polish, Bosnian, and Irish.² These groups have different rates of mental illness. For example, compared with white English people the Irish have a high rate of diagnosed mental health problems.³ Place of birth is also relevant. The experiences, health beliefs, and patterns of health seeking behaviour of a young African Caribbean born in the United Kingdom are likely to differ from those of an older migrant from Jamaica. In addition, over a fifth of white Europeans in this study were born abroad and may have common experiences with some of the migrants from the Caribbean, particularly of discrimination.

The use of self assigned ethnicity based on census categories is a pragmatic approach but has limitations.⁴ Other data such as language spoken, religion, place of birth, and social and economic factors are necessary to provide an informed picture.⁵ The quoted prevalences of anxiety and depression in African Caribbeans and white Europeans could therefore mask variations between different ethnic groups and the effects of physical factors and social inequalities.

Cross cultural psychiatry aims to provide a scientific basis for the study and comparison of mental health across cultures. Most traditional psychiatric instruments have been devised for a North American or north European population and are not necessarily valid in other groups.

There are two approaches to using psychiatric instruments in different cultural groups. The "emic" approach is a within culture approach. It uses culturally defined terms and an instrument devised for the particular cultural group. The findings are specific to the experience and presentation of illness in that culture.⁶ For example, a recent study with older Caribbean people in London constructed an interview schedule for emotional distress using vignettes and interviews with carers and older Caribbeans with mental health problems.⁷ The approach aims to identify those people who would be recognised as being ill or impaired by those of the same culture and attempts to overcome ethnocentric perspectives. However, instruments developed by this method are not transferable to other cultural or social groups and cannot provide comparative data.

The "etic" approach is an across cultures approach. It uses terms that are similar across different cultures and pre-existing instruments. The approach is based on the assumptions that the underlying features of common mental disorders, such as depression, are similar across cultures⁸ and that no important culture-specific symptoms or presentations would be missed. The charges of ethnocentricity can be reduced by assessing the instrument's performance and acceptability within a new population. This may require adaptations to the standard instrument. The general health questionnaire has been extensively validated in other cultures,⁹ but not specifically with an African Caribbean population in the United Kingdom.

The etic method allows for comparisons between cultural groups. However, if supposedly universal symptoms do not occur in a particular cultural group or do not have the same meaning the result will be invalid and misleading.

Shaw et al used an etic approach to screen for and identify cases and then interviewed these cases with an emic schedule. People whom the initial instruments failed to recognise as distressed would have been missed. By only interviewing identified cases the authors may have lost rich information on cultural aspects of mental health. This is important for those working in primary care, where most people with mental illness are seen and cared for, and where the iceberg of unmet need almost certainly lies.

The authors have attempted a difficult study and provided some insights into mental health problems in different cultural groups. However, prevalences derived from such a heterogeneous sample may conceal important variations among subgroups. Counting heads is important, but further attention must be given to the complexity of cultural and social factors in the experience of mental illness.

References

1. Senior P, Bhopal R. Ethnicity as a variable in epidemiological research. BMJ 1994; 309: 327-330[Free Full Text].
2. Office of Population Censuses and Surveys. 1991 census: local base statistics. London: OPCS , 1992.
3. Bracken P, Greenslade L, Griffin B, Smythe M. Mental health and ethnicity: an Irish dimension. Br J Psychiatry 1998; 172: 103-105[Free Full Text].
4. McKenzie K, Crowcroft NS. Describing race, ethnicity and culture in medical science. BMJ 1996; 312: 1054[Free Full Text].
5. Smaje C. Ethnicity in context. In: Smaje C, ed. Health, race and ethnicity. London: King's Fund Institute , 1995.
6. Kleinman A. Anthropology and psychiatry. The role of culture in cross-cultural research. Br J Psychiatry 1987; 151: 447-454[Abstract/Free Full Text].
7. Abas M, Phillips C, Richards M, Carter J, Levy R. Initial development of a new culture-specific screen for emotional distress in older Caribbean people. Int J Ger Psychiatry 1996; 11: 1097-1103.
8. Marsella A. Thoughts on cross-cultural studies on the epidemiology of depression. Culture, Medicine and Psychiatry 1978; 2: 343-357.
9. Jacob K, Bhugra D, Mann A. The validation of the 12-item general health questionnaire among ethnic Indians living in the United Kingdom. Psychol Med 1997; 27: 1215-1217[Medline].