Incidence of schizophrenia and other psychoses in
ethnic minority groups: results from the MRC
Paul Fearon 1, James B. Kirkbride2, Craig Morgan1, Paola Dazzan1,
Kevin Morgan* 1, Tuhina Lloyd3, Gerard Hutchinson4, Jane Tarrant3, Wai
Lun Alan Fung2, John Holloway5, Rosemarie Mallett1, Glynn Harrison6,
Julian Leff1, Peter B. Jones2 and Robin M. Murray1, on behalf of the
AESOP Study Group
* Kevin Morgan now works within the Department of Psychology, University of Westminster
1 Division of Psychological Medicine, Institute of Psychiatry, King’s College, London, UK
2 Department of Psychiatry, University of Cambridge, Cambridge, UK
3 Division of Psychiatry, University of Nottingham, Nottingham, UK
4 Faculty of Medical Sciences, University of the West Indies, St Augustine, Trinidad
5 Department of Mental Health, University of Bristol, Bristol, UK
6 Division of Psychiatry, University of Bristol, Bristol, UK
This paper has been published in Psychological Medicine, 36 (11). pp. 1541-
1550, November 2006.
© Cambridge University Press 
Psychological Medicine is available online at:
The WestminsterResearch online digital archive at the University of Westminster
aims to make the research output of the University available to a wider audience.
Copyright and Moral Rights remain with the authors and/or copyright owners.
Users are permitted to download and/or print one copy for non-commercial private
study or research. Further distribution and any use of material from within this
archive for profit-making enterprises or for commercial gain is strictly forbidden.
Whilst further distribution of specific materials from within this archive is forbidden,
you may freely distribute the URL of WestminsterResearch.
In case of abuse or copyright appearing without permission e-mail firstname.lastname@example.org.
Incidence of schizophrenia and other psychoses in
ethnic minority groups: results from the MRC
PAUL FEARON1*, JAMES B. KIRKBRIDE2, CRAIG MORGAN1, PAOLA DAZZAN1,
KEVIN MORGAN1, TUHINA LLOYD3, GERARD HUTCHINSON4,
JANE TARRANT3, WAI LUN ALAN FUNG2, JOHN HOLLOWAY5,
ROSEMARIE MALLETT1, GLYNN HARRISON6, JULIAN LEFF1, PETER B. JONES2
AND ROBIN M. MURRAY1, on behalf of the AESOP Study Group#
1Division of Psychological Medicine, Institute of Psychiatry, King’s College, London, UK;2Department of
Psychiatry, University of Cambridge, Cambridge, UK;3Division of Psychiatry, University of Nottingham,
Nottingham, UK;4Faculty of Medical Sciences, University of the West Indies, St Augustine, Trinidad;
5Department of Mental Health, University of Bristol, Bristol, UK;6Division of Psychiatry, University of
Bristol, Bristol, UK
Background. The incidence of schizophrenia in the African-Caribbean population in England is
reported to be raised. We sought to clarify whether (a) the rates of other psychotic disorders
are increased, (b) whether psychosis is increased in other ethnic minority groups, and (c) whether
particular age or gender groups are especially at risk.
Method. We identified all people (n=568) aged 16–64 years presenting to secondary services with
their first psychotic symptoms in three well-defined English areas (over a 2-year period in Southeast
London and Nottingham and a 9-month period in Bristol). Standardized incidence rates and
incidence rate ratios (IRR) for all major psychosis syndromes for all main ethnic groups were
Results. We found remarkably high IRRs for both schizophrenia and manic psychosis in both
African-Caribbeans (schizophrenia 9.1, manic psychosis 8.0) and Black Africans (schizophrenia
5.8, manic psychosis 6.2) in men and women. IRRs in other ethnic minority groups were modestly
increased as were rates for depressive psychosis and other psychoses in all minority groups. These
raised rates were evident in all age groups in our study.
Conclusions. Ethnic minority groups are at increased risk for all psychotic illnesses but African-
Caribbeans and Black Africans appear to be at especially high risk for both schizophrenia and
mania. These findings suggest that (a) either additional risk factors are operating in African-
Caribbeans and Black Africans or that these factors are particularly prevalent in these groups, and
that (b) such factors increase risk for schizophrenia and mania in these groups.
An elevated incidence of schizophrenia in
African-Caribbean populations living in Eng-
land is a well replicated, yet contentious find-
ing (Fernando, 1998). Less is known about
whether these increased rates exist to the same
extent for other psychotic disorders and in other
ethnic minority groups, including non-British
Whites. If the raised incidence is a genuine
phenomenon, it presents an urgent need in terms
of individual and public health, as well as a
* Address for correspondence: Dr Paul Fearon, Box 63, Institute
of Psychiatry, De Crespigny Park, London SE5 8AF, UK.
# Members of the AESOP Study Group are listed in the
Psychological Medicine, 2006, 36, 1541–1550.
f 2006 Cambridge University Press
Printed in the United Kingdom
remarkable opportunity to explore some of the
causes of the schizophrenia syndrome.
The AESOP (Aetiology and Ethnicity of
Schizophrenia and Other Psychoses) study was
designed to investigate these phenomena, using
a large-scale, multi-centre epidemiological de-
sign. Our three primary objectives in the first
stage of the study were:
(1) to define the degree of any increased inci-
dence of schizophrenia in the African-
Caribbean population in England;
(2) to explore the specificity of this phenom-
enon in terms of other psychotic syndromes;
(3) to determine whether the incidence of
schizophrenia and other psychoses is raised
in other ethnic minority groups.
The AESOP study was a population-based
incidence survey of all people with any psychosis
contacting health services from a defined popu-
lation during a 2-year period. The study
was based upon the methodology of the
World Health Organisation (WHO) ten-country
study (Jablensky et al. 1992), with additional
features to reduce biases associated with eth-
nicity in both population at risk and in case
Population at risk
The study took place in three areas in England:
Southeast London, Nottingham and Bristol.
Each has long-established African-Caribbean
populations as well as other ethnic minority
groups. Populations were estimated according
to the 2001 Census, with tables commissioned
from the Office of National Statistics stratified
by age (5-year age-bands), sex and ethnicity.
Self-ascription of ethnicity in this Census was
according to the categories of: White British,
White Irish, White Other, Mixed (four groups),
Black Caribbean, Black African, Black Other,
Indian, Pakistani, Bangladeshi, Chinese, Asian
Other, and Other. We collapsed these groups
into the following seven groups: White British,
African-Caribbean (Black Caribbean & Black
Other), Black African, Asian (Indian, Pakistani,
Bangladeshi), Other (Chinese, Asian Other,
Other), Mixed and White Other. The 2001
Census defined an African-Caribbean as a
person who was born in the Caribbean, or
whose family originated there. Similarly, Black
African was defined as a person either born in
sub-Saharan Africa or whose family hailed from
that region. We adhered to these definitions in
our ethnicity ascription in the AESOP study.
In Southeast London 33 adjacent electoral
wards served by the South London & Maudsley
Trust contained 282788 people between the ages
of 16 and 64 years; in Nottingham, 95 wards
containing 404208 people aged 16–64 were also
served by a single mental health provider,
Nottingham Healthcare Trust; in Bristol, 52
wards included 342806 people between the ages
of 16 and 64 years, served by the Avon Mental
Health Trust. These populations were con-
sidered to be at risk for 24 months in Southeast
London & Nottingham from September 1997
until the end of August 1999, and for the first 9
months of this period in Bristol. Thus, the study
was based upon an estimated total of 1.6 million
person-years at-risk between the ages of 16 and
Initial screening criteria
The initial target group for the study was people
in the populations at-risk who presented to
psychiatric services for the first time during the
study period with evidence of the following:
delusions, hallucinations, thought disorder or
negative symptoms of schizophrenia, irrespec-
tive of cause.
Subjects with psychosis
All mental health services serving the relevant
areas took part in the study, both in-patient and
community-based facilities. Cases were ascer-
tained through two routes. First, routine case
ascertainment was conducted through on-going
liaison between the study teams in each centre
and the mental health services. Clinical staff
were encouraged to refer all people who met the
initial screening criteria to the study offices using
a variety of agreed routes including telephone,
24-hour answering services, postal pro-forma
and dedicated fax returns. There was regular
phone or face-to-face contact by study teams
with both the in-patient and community mental
health teams (including child and adolescent
services) serving the
Regular training events for clinical teams
ensured that all staff knew about AESOP,
1542 P. Fearon et al.
regardless of staff turnover. Advertising ma-
terials were made available in all clinical areas to
ensure awareness and continuation of referrals,
and presentations were made to user and carer
groups within the relevant areas. Secondly, a
‘leakage study’ was undertaken following the
case ascertainment period in Southeast London
and Nottingham to identify any cases missed
through the routine procedures. All electronic
and paper information systems were scrutinized
for any cases aged 16–64 years presenting to the
services for the first time with the following di-
agnostic codes, according to ICD-10 (WHO,
1993), routinely used in the UK National
Health Service: F1x.04; F1x.5; F1x.7; F20–29;
F30–31.9; F32–33. These data were corrob-
orated with case records to confirm eligibility.
Data collection from eligible subjects
Those subjects identified were approached by
the research team. Following explanation and
consent, data were collected through interview,
in as many interviews as were required.
Psychopathology was assessed using the PSE
SCAN interview version 2.0 (WHO, 1992).
Psychopathology information on those subjects
who could not be interviewed was compiled
following thorough case-note review and em-
ploying the Item Group Checklist (IGC) of the
SCAN. Sociodemographic information was ob-
tained using a specially designed questionnaire,
particularly with respect to accurate recording
of ethnicity to closely mirror the 2001 Census
categories. Age at onset was defined as the age,
to the nearest year, at which someone passed the
initial screen. Ethnicity was ascribed indepen-
dently by three researchers (P.F., J.B.K., C.M.),
with discrepant cases (n=34) resolved by
consensus with a principal investigator (PBJ).
Inter-rater reliability was high (kappa 0.91). We
used all available information, including self-
ascription, place of birth and place(s) of par-
ental birth. The principal source of information
was self-ascription of ethnicity; when this was
unavailable, other information sources, such as
other informants and case-notes were used.
The clinical assessor prepared clinical vignettes
which summarized the presenting complaints
and initial history. These vignettes, together
with all available clinical information including
SCAN interview or IGC for those who had not
consented to interview, were presented, blind to
ethnicity, to a group of clinicians who made
criteria. These consensus diagnostic groups met
at each site throughout the duration of the
study, and included at least one Principal
Investigator and experienced diagnostician, and
the clinical assessor. Inter-rater and inter-centre
reliability results for this procedure based on
20 cases were satisfactory with kappa scores
ranging between 0.63 and 0.75.
Ethical approval was granted from the local re-
search ethical committees at each site.
The basic characteristics of the datasets – such
as age and sex distribution of both the numer-
ator and denominator – were examined, to
check for internal consistency of the data.
Incidence rates (age-standardized using the
population of England and Wales) of overall
psychotic illness, schizophrenia (F20), manic
(F31–32), and ‘Other Psychoses’ (F10–19 and
F21–29) were calculated for each ethnic group
using direct standardization. Age- and sex-
specific incidence rates were also calculated.
Unadjusted incidence rate ratios were calcu-
lated and then adjusted for the potential con-
founders of age at first contact of service (as a
categorical 10-item ‘ageband’ variable) and sex,
using Poisson regression. The analysis was
carried out using the ‘xi:poisson’ command in
STATA 6 (StataCorp, 1999). Age and sex-specific
incidence rate ratios (IRR) were also calculated.
Finally, we tested for an interaction between
study area and ethnicity by comparing a model
fitted first with, then without, a specific interac-
tion term using the likelihood ratio test.
Across the three centres, 568 cases of psychosis
met our inclusion criteria and were given a
consensus diagnosis of psychotic illness during
over 1.6 million person-years of follow-up
of more than 5% of the relevant English
Psychosis in UK ethnic minority groups1543
population (Kirkbride et al. 2006). Three
in Southeast London (over 2 years), 203 in
Nottingham (over 2 years), and 57 in Bristol
(over 9 months).
standardized incidence rates for each group
for each diagnostic category in both males and
females and both sexes combined. African-
Caribbeans [140.8 per 100000 persons/year
(PPY)] and Black Africans (80.6 PPY) have
markedly raised incidence
with the White British group (20.2 PPY) for all
psychosis. By diagnosis, rates are particularly
raised for both schizophrenia and mania in
African-Caribbeans and Black Africans. Both
the White Other and Mixed groups have mod-
estly increased rates for all diagnoses. The
Asian group have a more modest increased rate
for all psychosis (31.6 PPY), which holds true
for schizophrenia and depressive psychosis. The
rates for the ‘Other’ group are also raised,
both for all psychosis (55.0 PPY) and for all
diagnostic categories except ‘other’ psychosis.
When examined by gender, Table 1 demon-
strates that these raised rates are present for
both males and females. Male rates are generally
higher than the corresponding female rates for
schizophrenia and other psychosis.
Table 2 shows the IRR for each ethnic min-
ority group, using the White British group as
baseline. African-Caribbeans and Black Afri-
cans have markedly raised IRRs compared with
the White British group. For all psychosis,
African-Caribbeans have an IRR of 6.7 [95%
confidence interval (CI) 5.4–8.3], and Black
Africans, an IRR of 4.1 (3.2–5.3) after adjust-
ment for age and sex. These adjusted IRRs are
very similar to the crude IRRs (not shown). It is
worth noting that African-Caribbeans have a
significantly higher IRR than Black Africans.
All other groups have more modestly increased
IRRs, ranging from 1.6 (1.1–2.2) in the White
Other group to 2.7 (1.8–4.2) in the Mixed group.
Although Asians have an increased IRR of
1.5 (0.9–2.4), this was just beyond statistical
Incidence rate ratios for schizophrenia and
for mania in both African-Caribbeans and
Black Africans are markedly elevated. Thus, the
IRR for schizophrenia in African-Caribbeans
is 9.1 (6.6–12.6) and for Black Africans is 5.8
Table 1. Age-standardized incidence rates (95% confidence intervals) for each diagnostic
category in all ethnic groups
1544 P. Fearon et al.
(3.9–8.4). The IRR for mania in African-
Caribbeans is 8.0 (4.3–14.8) and for Black
Africans is 6.2 (3.1–12.1); these raised IRRs are
present for both men and women. Rate ratios
for the Asian, Other, Mixed and White Other
groups are also elevated for both diagnoses,
with for example, White Others having a 2.5-
fold increased rate of schizophrenia compared
with White British.
A somewhat different pattern is evident for
depressive psychosis, with IRRs for all groups
except the White Other group being raised to
comparable degrees. Thus, African-Caribbeans
have an IRR of 3.1 (1.5–6.1), Black Africans 2.1
(0.9–5.0), Asians 3.0 (1.3–7.1), Others 5.6
(2.5–12.4), and Mixed 4.0 (1.6–10.2).
Table 3 shows that the rate ratios observed
for all psychosis in African-Caribbeans and
Black Africans are raised across all age bands in
this study and follow a similar pattern both for
men and for women (data not shown).
There was no evidence of any interaction be-
tween study area and ethnicity for any of the
diagnostic categories included in the study, im-
plying that the differences in rates between
groups did not differ significantly by study area.
Table 2. Combined and sex-specific age-adjusted incidence rate ratios (IRRs) with
corresponding 95% confidence intervals in ethnic minority groups
Table 3. Age-specific incidence rate ratios (IRRs) with corresponding 95% confidence intervals
in ethnic minority groups for all psychosis
AfricanAsian OtherMixedWhite Other
Psychosis in UK ethnic minority groups1545
In this, the largest incidence study of psychosis
in England and the first to utilise the 2001
Census data, we have confirmed the previously
reported raised rates of schizophrenia in the
African-Caribbean population. Our findings
extend those of previous studies. We found ro-
bust evidence that these markedly raised rates
exist in both men and women, in both African-
Caribbeans and Black Africans, in all categories
of psychosis, in all age groups and across three
different areas in England contemporaneously.
Furthermore, we were able to calculate rates
for Asians, White Other and Mixed groups and
to use White British (rather than All White)
incidence rates as a baseline for calculating
In the UK, research in this area was hampered
prior to the 1991 Census because denominator
data was not collected at the individual level but
only on the head of household. Research using
1991 data (King et al. 1994; van Os et al. 1996;
Bhugra et al. 1997; Harrison et al. 1997) in-
dicated a high incidence of schizophrenia in
ethnic minority groups. However, until recently,
it was unclear whether these raised rates were
specific to schizophrenia or whether they ex-
tended to all psychotic disorders, though several
studies (Leff et al. 1976; Bebbington et al. 1981;
Hunt et al. 1993) had suggested rates of mania
were also raised. However, in a recent paper
from our group reporting rates of mania in our
three study areas, we reported incidence rates of
mania in ethnic minority groups, particularly
African-Caribbeans and Black Africans, were
significantly raised when compared with an all-
White group (Lloyd et al. 2005). It remained
uncertain whether these increased rates ex-
tended to all psychotic disorders and to all eth-
nic minority groups in the UK. Some studies
have suggested increased rates of schizophrenia
in Africans (van Os et al. 1996), but findings for
Asians have been contradictory (King et al.
1994; van Os et al. 1996; Bhugra et al. 1997;
Harrison et al. 1997). The EMPIRIC study
(King et al. 2005), a community-based preva-
lence study, found a more modest association
between ethnicity and the likelihood of report-
ing psychotic symptoms. African-Caribbeans
were only about twice as likely to report such
symptoms compared with Whites. However, it
should be noted that findings from prevalence
studies, which measure the number of prevalent
cases in a population at any given time, differs
from incidence, and that one should not make
direct comparisons between these two measures.
Studies from continental Europe report that
the rate of schizophrenia for migrants to other
countries is also raised. Two studies of migrants
to The Netherlands from Surinam, the Dutch
Antilles and Morocco (Selten et al. 1997, 2001),
and one study of immigrants to Sweden from
East Africa (Zolkowska et al. 2001) indicate a
higher incidence of schizophrenia and other
psychotic illnesses in these groups. The popu-
lation in Surinam and the Surinamese com-
munity in The Netherlands are ethnically
diverse, with around 40% each from Asian and
Caribbean does not appear to be markedly
raised. Three major incidence studies have been
conducted, covering Jamaica (Hickling et al.
1995), Trinidad (Bhugra et al. 1996) and
Barbados (Mahy et al. 1999), the three islands
from where the majority of UK migrants orig-
inated. The incidence of schizophrenia in each
study was comparable to the rate for the UK
White population, and significantly lower than
the comparable rate for the UK African-
The raised rates of psychosis in African-
Caribbeans do not appear to be due to mis-
diagnosis (Lewis et al. 1990; Hickling et al.
1999) or disproportionate referral to services
(Wessely et al. 1991; Mortensen et al. 1997).
These factors, together with our confirmation of
findings in previous studies (particularly those
using 1991 Census data and employing strict
operational diagnostic criteria) suggest that
these raised rates cannot be merely explained by
Interpretation of current findings
We observed only modestly raised rates for our
Asian group, with the relatively small numbers
of cases involved reducing our ability to detect
a true difference unequivocally. Two previous
Bhugra and colleagues (1997) found no overall
increase in the incidence of schizophrenia in the
1546P. Fearon et al.
Asian population in Ealing, but noted elevated
rates in those over 30 years old. King and co-
workers (1994) found elevated rates in the Asian
population in their North London study. What
can be stated with some confidence is that the
rates for schizophrenia and mania in Asians do
not appear to be raised to the same extent as
for African-Caribbeans and Black Africans.
Understanding this difference between these
groups, who migrated to the UK during the
same era, and who appear to experience similar
levels of racial discrimination, may reveal im-
portant factors which play a part in the mark-
edly raised rates of psychosis in the Black British
populations. Brugha and colleagues (2004), in
their survey of householders in Britain found
that African-Caribbeans and Black Africans
were more likely than other ethnic minority
groups to suffer from indicators of social dis-
advantage, such as: unemployment; lone parent
status; lower social class; low perceived social
support; poverty (indicated by lack of car own-
ership) and having a primary social support
group of fewer than three close others. Further,
they found that adjusting for these factors
modestly attenuated the risk of psychosis in
The rates of psychosis among the ‘Other’
group were also raised. This group consisted of
members of other smaller minority groups and
was composed of people mainly from the
Middle East, North Africa, China, Vietnam and
Japan. Although this is a rather disparate
group, these findings add weight to the concept
that all migrant groups are at some degree of
increased risk of psychotic illness. A recent
meta-analysis of population-based incidence
studies of schizophrenia in migrant populations
(Cantor-Graae & Selten, 2005) demonstrated
ethnic minorities had, overall, an increased
relative risk of 2.9 compared with the indigen-
ous populations. This effect was more marked in
migrants from either developing countries or
from countries where the majority population is
Black. A systematic review by McGrath and
colleagues (2004) found a similar effect for
schizophrenia in migrant populations, with an
overall median rate ratio of 4.6. Furthermore,
our White Other group (comprised mainly of
White Irish and White Europeans) had over a
twofold increased IRR for schizophrenia. Many
of these groups have migrated more recently
counterparts, which suggests that the rates in
these groups may represent the degree of
increased risk conferred by relatively recent
migration. Further studies are required to
specifically test this hypothesis. This study also
provides the first evidence of raised rates of
psychosis in people of mixed race.
Although we have previously reported IRRs
for mania (Lloyd et al. 2005), the present
analysis used the White British group, rather
than an all-White group as a baseline. Thus, the
IRRs for manic psychosis reported here are
more elevated than in our previous report, but
follow a remarkably similar pattern to those of
schizophrenia. These findings are included in
this paper in order to provide a comprehensive
picture of the rates of all psychotic disorders,
and to allow direct comparison of IRRs for each
diagnosis in each ethnic minority group.
The finding of raised IRRs for depressive
psychosis is, to our knowledge, the first such
report from the UK. The pattern of these raised
rates differed from that for both schizophrenia
and mania in that rates were more hom-
ogeneously raised for all minority groups. There
is little literature on rates of this disorder among
ethnic minority groups. The closest comparable
study is that of Selten and colleagues (2003),
who examined the admission rates for both first-
episode bipolar disorder of manic and depressed
types in migrant groups in The Netherlands,
and found only small increases in the admission
rates for manic illness in Surinamese, but not
Turkish migrants. However, they found more
substantial increases in admission rates for both
Surinamese and Turkish migrants for bipolar
disorder depressed type, particularly for men.
These findings are in partial agreement with the
findings for depressive psychosis reported here,
in that both groups had modestly raised rate
There does not appear to be a major effect of
generation, with raised rates of schizophrenia-
like psychosis being found in the elderly first
generation in London (Reeves et al. 2001). Our
study provides further evidence of a lack of an
age or ‘generational’ effect, with IRRs raised
for all psychoses across all age groups.
We found no evidence of an interaction be-
tween study area and ethnicity for any psychotic
disorders examined in our study. While this may
their African-Caribbeanand Asian
Psychosis in UK ethnic minority groups1547
suggest that factors related to urbanicity are not
associated with the findings of increased rates of
psychotic disorder in ethnic minority groups
(a fact in part supported by the consistent find-
ings of increased rates of psychosis among
African-Caribbeans in several English areas
over the last few decades), it is worth bearing in
mind that this is a somewhat crude method of
investigating the role of urban factors in these
groups. For example, although the Nottingham
area is overall less densely urban that the
Southeast London area, this does not take
into account variation in urbanicity within
our centres; our Nottingham centre ranges
from highly urban to rural. Thus, in order to
definitively explore the role of urban factors, it
may prove necessary to do such analyses at a
smaller, more tightly defined geographical scale.
Our study has a number of strengths. We em-
ployed a prospective case ascertainment design
in three well-defined geographical areas; we
used operational consensus diagnoses per-
formed blind to ethnic group status; we used
White British (rather than all-White groups) as
our baseline group; and we performed leakage
studies in Southeast London and Nottingham to
minimize the chance of under-ascertainment.
We collapsed the 16 ethnic group categories
of the 2001 Census into seven categories for this
study. We chose this option for clarity of pres-
entation and also to maximize our available
power. However, this seven-item classification
still represents a more fine-grain categorization
than previous studies in this field. When exam-
ined individually, each ethnic minority sub-
group within our seven groups had comparably
raised rates. For example, Indians, Pakistanis
and Bangladeshis had similar rates and rate
ratios, an interesting finding given the differing
cultures and traditions across these ethnic
groups. Similarly, the Black Caribbean and
Black Other groups were merged, as they had
similar rates. We acknowledge, however, that
while the majority of those who ascribe them-
selves as Black Other are of African-Caribbean
origin, a minority will be of Black African, or
even African-American origin. We also ac-
knowledge that, despite the consistency of our
findings for each ethnic group in all centres,
these groupings remain crude and do not reflect
different contexts. Further work at a more
detailed level is required to elucidate the under-
lying causes of the different rates found in
different ethnic groups.
Although the Bristol centre recruited over a
shorter time period than Southeast London or
Nottingham and a formal leakage study was
not undertaken in this centre, the overall rates
of psychosis in Bristol were remarkably com-
parable to those of Nottingham, suggesting that
there is unlikely to have been a significant error
in our estimates for this centre. Furthermore,
analysis of the data excluding Bristol cases did
not alter our findings.
The majority of cases in this study were in-
terviewed directly, and the remainder had their
ethnicity classified using their and their parents’
places of birth, rather than by self-ascription.
Although this might lead to misclassification
of a small proportion of cases, the likelihood of
this happening was minimized by fact that we
collapsed our ethnic group classification into
seven broad categories.
Case ascertainment was achieved by service
contact. Selection biases may have been present
if psychotic patients from ethnic minority
groups were more likely to contact services than
their White counterparts, thus contributing to
the increased incidence of psychoses based on
first contact of service. However, since a five- to
tenfold difference in incidence was observed in
African-Caribbeans and Black Africans, it is
highly unlikely that psychotic disorders of
comparable severity would remain undetected
in the White population on a scale large enough
to account for such incidence rate ratios.
Population studies have also failed to support
the notion that a large number of psychotic in-
dividuals remain undetected by the services
(Harrison, 1990). However, it should be noted
that we cannot exclude the possibility that
incidence rates calculated on the basis of first
presentation to services may not correspond
exactly to the true incidence of psychosis in any
community at any given time.
A well recognized limitation of the 1991 UK
Census was underenumeration of certain ethnic
minority groups. By using data obtained from
the 2001 Census, which was closer in time to the
AESOP survey period than its 1991 predecessor,
and which accounted for underenumeration of
1548P. Fearon et al.
certain groups (particularly the young and cer-
tain ethnic minorities) through its one-number
methodology (Pereira, 2002), we believe that
we have used the most accurate denominator
estimates available. Furthermore, analysis of
our data using the 1991 Census data (corrected
for underenumeration) does not alter our main
findings (data not shown).
Although our study is the largest first-
presentation study in the UK to date, with
an effective denominator of over 1.6 million
person-years, our power to detect signifi-
cant differences in smaller groups was still lim-
ited. Thus, the Asian group, despite having
rate ratios consistently above 1.0 often had
confidence intervals that did not exclude the
possibility of our findings being due to chance.
We feel that our findings, taken as a whole,
provide evidence of a modest increase in rates in
this group which requires exploration in future
Thus, although some methodological issues
are present in our study, we believe their com-
bined effect is insufficient to severely undermine
the impact of our primary findings.
We have demonstrated remarkably high in-
cidence rates ofschizophrenia andmania inboth
African-Caribbeans and Black Africans, in men
and women, in three urban UK settings. Rates
in other ethnic minority groups are more mod-
estly increased. We have shown, for the first
time, that people of mixed race have an elevated
risk of psychosis and that non-British Whites
have an elevated risk of schizophrenia. We have
also demonstrated increased rates for depressive
psychosis, and that all minority populations
show a similar degree of increase for this diag-
nosis. Raised rates of psychosis were found in
all age groups in our study.
If we can explain these findings, we stand
to learn much about schizophrenia and other
psychoses, as well as providing improved psy-
chiatric services for these particular population
APPENDIX. The AESOP Study Group
Bristol: Glynn Harrison, Florence Muga, John
Holloway. Cambridge: Peter B. Jones, Wai Lun
Alan Fung, Jouko Mietunen, Maureen Ashby,
Hazel Hayhurst. London: Robin M. Murray,
Julian Leff, Tom Craig, Rosemarie Mallett, Paul
Fearon, Craig Morgan, Kevin Morgan, Paola
Gwenzi, Mandy Sharpley, Simon Vearnals,
Gerard Hutchinson, Rachel Burnett, Jane Kelly,
Kenneth Orr, Jeza Salvo, Kathy Greenwood,
David Raune, Maria Lambri, Samantha Jones,
Stefan Auer, Per Rohebak, Laura McIntosh.
Nottingham: Gillian Doody, Jane Tarrant, Sue
Window, Pat Williams, Tuhina Lloyd, Hemant
Bagalkote, Becci Dow, Daphne Boot, Annette
Farrant, Steve Jones, Jayne Simpson, Ramona
Moanette, Philp Z. Sirip Suranim, Mark
Ruddell, John Brewin, Ian Medley.
We are grateful to mental health services and
patients in Bristol, Nottingham and Southeast
London for their co-operation and support with
this study. This study was funded by grants
from the UK Medical Research Council and the
Stanley Medical Research Institute.
DECLARATION OF INTEREST
Bebbington, P. E., Hurry, J. & Tennant, C. (1981). Psychotic
disorders in selected immigrant groups in Camberwell. Social
Psychiatry 16, 43–51.
Bhugra, D., Hilwig, M., Hossein, B., Marceau, H., Neehall, J., Leff,
J., Mallett, R. & Der, G. (1996). First-contact incidence rates of
schizophrenia in Trinidad and one-year follow-up. British Journal
of Psychiatry 169, 587–592.
Bhugra, D., Leff, J., Mallet, R., Der, G., Corridan, B. & Rudge, S.
(1997). Incidence and outcome of schizophrenia in whites,
African-Caribbeans and Asians in London. Psychological Medi-
cine 27, 791–798.
Brugha, T., Jenkins, R., Bebbington, P., Meltzer, H., Lewis, G.
& Farrell, M. (2004). Risk factors and the prevalence of neurosis
and psychosis in ethnic groups in Great Britain. Social Psychiatry
& Psychiatric Epidemiology 39, 939–946.
Cantor-Graae, E. & Selten, J. P. (2005). Schizophrenia and migra-
tion: a meta-analysis and review. American Journal of Psychiatry
Fernando, S. (1998). Studies into issues of ‘race’ and culture in
psychiatry. Psychological Medicine 28, 496–497.
Harrison, G. (1990). Searching for the causes of schizophrenia: the
role of migrant studies. Schizophrenia Bulletin 16, 663–671.
Harrison, G., Glazebrook, C., Brewin, J., Cantwell, R., Dalkin, T.,
Fox, R., Jones, P. & Medley, I. (1997). Increased incidence of
psychotic disorders in migrants from the Caribbean to the United
Kingdom. Psychological Medicine 27, 799–806.
Psychosis in UK ethnic minority groups1549
Hickling, F. W., McKenzie, K., Mullen, R. & Murray, R. M. Download full-text
(1999). A Jamaican psychiatrist evaluates diagnosis at a
London psychiatric hospital. British Journal of Psychiatry 175,
Hickling, F. W. & Rodgers-Johnson, P. (1995). The incidence of first-
contact schizophrenia in Jamaica. British Journal of Psychiatry
Hunt, N., Adams, S., Coxhead, N., Sayer, H., Murray, C. &
Silverstone, T. (1993). The incidence of mania in two areas in the
United Kingdom. Social Psychiatry and Psychiatric Epidemiology
Jablensky, A., Sartorius, N., Ernberg, G., Anker, M., Korten, A.,
Cooper, J. E., Day, R. & Bertelsen, A. (1992). Schizophrenia:
manifestations, incidence and course in different cultures. A World
Health Organization ten-country study. Psychological Medicine
Monograph Supplement 20, 1–97.
King, M., Coker, E., Leavey, G., Hoare, A. & Johnson-Sabine, E.
(1994). Incidence of psychotic illness in London: compari-
son of ethnic minority groups. British Medical Journal 309, 1115–
King, M., Nazroo, J., Weich, S., McKenzie, K., Bhui, K., Karlsen, S.,
Stansfeld, S., Tyrer, P., Blanchard, M., Lloyd, K., McManus, S.,
Sproston, K. & Erens, B. (2005). Psychotic symptoms in the general
population of England – a comparison of ethnic groups (the
EMPIRIC study). Social Psychiatry & Psychiatric Epidemiology
Kirkbride, J. B., Fearon, P., Morgan, C., Dazzan, P., Morgan, K.,
Tarrant, J., Lloyd, T., Holloway, J., Hutchinson, G., Leff, J.,
Mallett, R. M., Harrison, G. L., Murray, R. M. & Jones, P. B.
(2006). Heterogeneity in incidence rates of schizophrenia and
other psychotic syndromes. Archives of General Psychiatry 63,
Leff, J., Fisher, M. & Bertelsen, A. (1976). A cross-national epi-
demiological study of mania. British Journal of Psychiatry 129,
Lewis, G., Croft-Jeffreys, C. & David, A. (1990). Are British
psychiatrists racist? British Journal of Psychiatry 157, 410–415.
Lloyd, T., Kennedy, N., Fearon, P., Kirkbride, J., Mallett, R., Leff, J.,
Holloway, J., Harrison, G., Dazzan, P., Morgan, K., Murray,
R. M. & Jones, P. B. (2005). Incidence of bipolar affective disorder
in three UK cities: results from the AESOP study. British Journal
of Psychiatry 186, 126–131.
Mahy, G. E., Mallett, R., Leff, J. & Bhugra, D. (1999). First-contact
incidence-rate of schizophrenia on Barbados. British Journal of
Psychiatry 175, 28–33.
McGrath, J., Saha, S., Welham, J., El Saadi, O., MacCauley, C.
& Chant, D. (2004). A systematic review of the incidence of
schizophrenia: the distribution of rates and the influence of sex,
urbanicity, migrant status and methodology. BMC Medicine 2
Mortensen, P. B., Cantor-Graae, E. & McNeil, T. F. (1997).
Increased rates of schizophrenia among immigrants: some meth-
odological concerns raised by Danish findings. Psychological
Medicine 27, 813–820.
Pereira,R. (2002).The Census
Element of a One-number Census. Office for National Statistics:
Reeves, S. J., Saurer, J., Stewart, R., Granger, A. & Howard, R. J.
(2001). Increased first-contact rates for very late-onset schizo-
phrenia-like psychosis in African- and Caribbean-born elders.
British Journal of Psychiatry 179, 172–174.
Selten, J. P., Slaets, J. P. J. & Kahn, R. S. (1997). Schizophrenia in
Surinamese and Dutch Antillean immigrants to The Netherlands:
evidence of an increased incidence. Psychological Medicine 27,
Selten, J. P., van Os, J. & Nolen, W. A. (2003). First admissions for
mood disorders in immigrants to the Netherlands. Social
Psychiatry & Psychiatric Epidemiology 38, 547–550.
Selten, J. P., Veen, N. N., Feller, W., Blom, J. D., Schols, D.,
Camoenie ¨ , W., Oolders, J., Van der Velden, M., Hoek, H. W.,
Vlada ´ r Rivero, V. M., Van der Graaf, Y. & Kahn, R. (2001).
Incidence of psychotic disorders in immigrant groups to The
Netherlands. British Journal of Psychiatry 178, 367–372.
STATA statistical software: release 6.0. Stata
Corporation: College Station, TX.
van Os, J., Castle, D. J., Takei, N., Der, G. & Murray, R. M. (1996).
Psychotic illness in ethnic minorities: clarification from the 1991
Census. Psychological Medicine 26, 203–208.
Wessely, S., Castle, D., Der, G. & Murray, R. M. (1991).
Schizophrenia and Afro-Caribbeans. A case-control study. British
Journal of Psychiatry 159, 795–801.
WHO (1992). Schedule for Clinical Assessment in Neuropsychiatry.
World Health Organisation: Geneva.
WHO (1993). The ICD-10 Classification of Mental and Behavioural
Disorders: Clinical Descriptions and Diagnostic Guidelines. World
Health Organisation: Geneva.
Zolkowska, K., Cantor-Graae, E. & McNeil, T. F. (2001) In-
creased rates of psychosis amongst immigrants to Sweden: is
migration a risk factor for psychosis? Psychological Medicine
CoverageSurvey – the Key
1550 P. Fearon et al.