Using the Barnes Language Assessment with older ethnic minority groups

6
Using the Barnes Language Assessment with older ethnic minority groups Victoria Ramsey 1 * , Susan Stevens 2 , Karen Bryan 2 , Julia Binder 3 and Jane Cockle-Hearne 2 1 Kingston Primary Care Trust, Barnes Hospital, London, UK 2 University of Surrey, Guildford, Surrey, UK 3 Hammersmith and Fulham Primary Care Trust, London, UK SUMMARY Objective There are many issues concerning the assessment of older people from ethnic minority groups, the most significant being the language barrier experienced by those whose English is an additional language (EAL). This study aimed to test the hypothesis that EAL participants would score less well than those with English as a first language (EFL) on the sub-tests of the Barnes Language Assessment (BLA), elucidate the reasons for any such differences and discuss the implications. Methods The Barnes Language Assessment (BLA) is an accurate tool providing information about expected patterns of language in different dementia syndromes. This study compares the performance of EAL participants with EFL participants. The BLA was administered to 144 participants, divided into sub-groups with respect to age, gender and educational background, none of whom had a working diagnosis of dementia. Results Results suggest that EAL speakers performed less well compared to EFL speakers when other variables were matched. Significantly better BLA scores, at the one percent level, were found in both EAL and EFL groups with higher educational achievement for eight of the 15 sub-tests. Conclusion Differences were found in performance on the BLA between EAL and EFL participants. The degree of difference between EAL and EFL speakers decreased as educational achievement rose. The consequences of these findings for service delivery and the problems of recruitment of older EAL participants are discussed. Copyright # 2009 John Wiley & Sons, Ltd. key words — older ethnic minority groups; language testing; dementia INTRODUCTION The number of older people from minority ethnic communities in the United Kingdom has risen sharply over recent decades (Odutoye and Shah, 1999) and are predicted to continue to do so. It is recognised that social research has not yet fully explored the issues facing black and minority ethnic elders (Lindesay et al., 1997; Richards et al., 2000). The results of such research may have implications for policy develop- ment and service provision. Studies suggest that ethnic minority populations have a greater incidence of dementia, the cause attributed to unidentified genes or other risk factors (Tang et al., 1998). Dementia is diagnosed later in these groups (Ayalon and Area ´n, 2004), so ethnic elders are less likely to access important early interventions and support. A number of explanations are given for this delay in diagnosis, such as a reluctance to access services (Eolas, 1999), difficulty getting appropriate information, and lack of know- ledge about services, together with a lack of appro- priate test materials (Richards et al., 2000). Even when interpreters or bilingual interviewers are used, there seem to be particular problems around the reliability of the interview process in ethnic minority groups (Shah, 1999; Rait et al., 2000). Finally, there are differences in culture and belief, with some traditional approaches to health and well- being conflicting with western values (Butt and INTERNATIONAL JOURNAL OF GERIATRIC PSYCHIATRY Int J Geriatr Psychiatry 2009; 24: 426–431. Published online 10 February 2009 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/gps.2158 *Correspondence to: V. Ramsey, c/o Karen Bryan, Facultly of Health and Medical Science, Duke of Kent Building, University of Surrey, Guildford, GU27TE, UK. E-mail: [email protected] Copyright # 2009 John Wiley & Sons, Ltd. Received 21 December 2007 Accepted 1 October 2008

Transcript of Using the Barnes Language Assessment with older ethnic minority groups

Page 1: Using the Barnes Language Assessment with older ethnic minority groups

INTERNATIONAL JOURNAL OF GERIATRIC PSYCHIATRY

Int J Geriatr Psychiatry 2009; 24: 426–431.

Published online 10 February 2009 in Wiley InterScience

(www.interscience.wiley.com) DOI: 10.1002/gps.2158

Using the Barnes Language Assessment with older ethnicminority groups

Victoria Ramsey1*, Susan Stevens2, Karen Bryan2, Julia Binder3 and Jane Cockle-Hearne2

1Kingston Primary Care Trust, Barnes Hospital, London, UK2University of Surrey, Guildford, Surrey, UK3Hammersmith and Fulham Primary Care Trust, London, UK

SUMMARY

Objective There are many issues concerning the assessment of older people from ethnic minority groups, the mostsignificant being the language barrier experienced by those whose English is an additional language (EAL). This study aimedto test the hypothesis that EAL participants would score less well than those with English as a first language (EFL) on thesub-tests of the Barnes Language Assessment (BLA), elucidate the reasons for any such differences and discuss the implications.Methods The Barnes Language Assessment (BLA) is an accurate tool providing information about expected patterns oflanguage in different dementia syndromes. This study compares the performance of EAL participants with EFL participants.The BLA was administered to 144 participants, divided into sub-groups with respect to age, gender and educationalbackground, none of whom had a working diagnosis of dementia.Results Results suggest that EAL speakers performed less well compared to EFL speakers when other variables werematched. Significantly better BLA scores, at the one percent level, were found in both EAL and EFL groups with highereducational achievement for eight of the 15 sub-tests.Conclusion Differences were found in performance on the BLA between EAL and EFL participants. The degree ofdifference between EAL and EFL speakers decreased as educational achievement rose. The consequences of these findingsfor service delivery and the problems of recruitment of older EAL participants are discussed. Copyright# 2009 John Wiley& Sons, Ltd.

key words—older ethnic minority groups; language testing; dementia

INTRODUCTION

The number of older people from minority ethniccommunities in the United Kingdom has risen sharplyover recent decades (Odutoye and Shah, 1999) and arepredicted to continue to do so. It is recognised thatsocial research has not yet fully explored the issuesfacing black and minority ethnic elders (Lindesayet al., 1997; Richards et al., 2000). The results of suchresearch may have implications for policy develop-ment and service provision.Studies suggest that ethnic minority populations

have a greater incidence of dementia, the cause

*Correspondence to: V. Ramsey, c/o Karen Bryan, Facultly of Healthand Medical Science, Duke of Kent Building, University of Surrey,Guildford, GU27TE, UK. E-mail: [email protected]

Copyright # 2009 John Wiley & Sons, Ltd.

attributed to unidentified genes or other risk factors(Tang et al., 1998). Dementia is diagnosed later inthese groups (Ayalon and Arean, 2004), so ethnicelders are less likely to access important earlyinterventions and support. A number of explanationsare given for this delay in diagnosis, such as areluctance to access services (Eolas, 1999), difficultygetting appropriate information, and lack of know-ledge about services, together with a lack of appro-priate test materials (Richards et al., 2000). Even wheninterpreters or bilingual interviewers are used, thereseem to be particular problems around the reliabilityof the interview process in ethnic minority groups(Shah, 1999; Rait et al., 2000).

Finally, there are differences in culture and belief,with some traditional approaches to health and well-being conflicting with western values (Butt and

Received 21 December 2007Accepted 1 October 2008

Page 2: Using the Barnes Language Assessment with older ethnic minority groups

Table 1. Breakdown of education level vs language groups

Education groups English European Asian Other Totals

No qualifications 13 1 2 1 17O/A level 46 8 6 2 62Degree/diploma 30 19 12 4 65Totals language 89 28 20 7 144

language assessment in ethnic elders 427

O’Neil, 2004). The effects of ageing may beexperienced and treated in a different way in minoritycultures.

Language, age and education

Ethnicity and bilingualism are not the same, althoughmany people in ethnic minority groups are bilingual.However, language barriers are often cited as the causeof difficulties when assessing ethnic elders.

Age is significant when considering languagebarriers in ethnic communities. At present ethnicindividuals of over 85 years are more likely tobe monolingual than younger people. In the future,although this ‘older old’ group will expand, theproportion with some knowledge of English mayincrease. However, barriers in the use of Englishlanguage assessments may continue to exist, as maythe effect of lack of formal education.

Education level is known to be a significant influenceon ethnic elders’ performance on cognitive testing(Lindesay et al., 1997; Richards et al., 2000). However,education level alone is an inadequate indicator ofperformance when considering a group with suchvaried ‘education experience’. This is due to an absenceof internationally standardised educational levels.

The Barnes Language Assessment

The initial research leading to the development of theBLA was carried out by a group of speech andlanguage therapists, specialised in working with olderpeople. The test provides a valuable screen andbaseline measurement of language skills (Bryan et al.,2001). Feedback from participants and clinicians wasthat the project highlighted the need to distinguishbetween EAL and EFL speakers, as well as to controlfor variables other than age and gender (Bryan et al.,2006). These were social class, occupation andeducation levels. There are indications, based on theliterature, that EAL speakers will do significantlyworse on language tests when compared to EFLspeakers. The second project, carried out between2002–2005, established robust normative data for EFLelders, and norms for EAL elders. This study aimed totest the hypothesis that EAL participants would scoreless well than EFL participants, elucidate the reasonsfor any such differences and discuss the implications.

METHOD

Sampling

One hundred and forty-four healthy community livingparticipants over the age of 50 were recruited from a

Copyright # 2009 John Wiley & Sons, Ltd.

variety of settings including day centres, communitygroups, work-places and via word of mouth. Theproject was advertised by A4 posters, letters andpersonal contact. Interviews were carried out inparticipants’ homes, central locations (e.g. daycentres) and at the University of Surrey. Criteria forinclusion in the study included no evidence of adementia, no significant current depression, nosignificant neurological history, no reading or writingimpairment. The criteria were applied using screeningtools described below. The 144 participants weredivided into EFL (89) and EAL (55).During the analysis stage, the EAL participants

were grouped according to ethnic background;European, Asian (including Chinese and Japanese),Other, e.g. Creole and Guyanese, and level ofeducation (see Table 1)The aim was to identify any patterns emerging in

the data, particularly error patterns, although it isacknowledged that such broad groupings will need tobe refined. Discussion of emerging error patternsparticular to ethnic groups is the subject of a futurepaper.Participants were reassured throughout the process

that the project was to collect data about ‘normal’functioning and not in order to identify anypathological process. However, if participantsexpressed concern about their performance they wereadvised to contact their GP. Only one individual had asignificantly impaired performance and it was foundthat she was being monitored by her GP; her data wereeliminated.

Tools

Before the BLA was administered, a pre-assessmentscreen was carried out to eliminate any significantconditions, e.g. dyslexia or depression. The pre-assessment contained the following screens; MiniMental State Examination (Folstein et al., 1975) toexclude cognitive impairment; BASDEC (BriefAssessment Schedule Depression Cards) (Adsheadet al., 1992) to exclude depression; WRAT (WideRange Achievement Test) (Jastak and Jastak, 1978) to

Int J Geriatr Psychiatry 2009; 24: 426–431.

DOI: 10.1002/gps

Page 3: Using the Barnes Language Assessment with older ethnic minority groups

428 v. ramsey ET AL.

exclude reading and writing problems; Health Screento establish the presence of any neurological conditions.The BLA is made up of 15 sub-tests. The sub-tests

are grouped into expression (five sub-tests), compre-hension (three sub-tests), reading and writing (threesub-tests), memory (two sub-tests) and executivefunction (one sub-test). Key areas of languagefunctioning are included in each modality, e.g. wordfluency, naming, word and sentence comprehension,word and sentence reading and writing.

Data analysis

Analysis was undertaken on two levels. Using SPSSversion 15, first summary statistics were extractedand secondly bivariate analysis was conducted on theBLA sub-tests with respect to EFL/EAL. Analysisof education levels, age and gender were alsoconducted. Where necessary, the variables were testedfor normality (Shapiro-Wilk) and appropriate tests(t-tests, Mann–Whitney U, Spearman’s Rank) wereperformed.

RESULTS

The results confirmed the hypothesis that EALspeakers would perform less well on the BLAcompared to EFL speakers. Gender was not foundto be a significant variable at the 1% level in relation toperformance on the BLA, and age was only significantat the 1% level for two sub-tests (verbal fluencyanimals and trail time). Both of these are timed tests.However, significantly better BLA scores at the 1%level were found in both EAL and EFL groups withhigher educational achievement for eight out of the15 sub-tests (spoken word to picture matching, verbalfluency S and animals, word definition, TROG,forward digit span, sentence writing errors, storyre-telling). Two further sub-tests were significant atthe 5% level (spelling to dictation and followingcommands).Table 2 presents summary statistics across language

background and education groups. Although EALspeakers’ mean scores were usually lower than scoresof the EFL participants (see Table 2), there were somesub-tests where EFL/EAL scores in the highereducation sub-groups (O/A levels and diploma/degree) were similar, e.g. following commands andsentence writing, picture description and word-picturematching.Generally, scores increased for both EFL and EAL

when the subject’s education levels were higher.Differences between EFL and EAL tended to reduce

Copyright # 2009 John Wiley & Sons, Ltd.

as the education level rose, i.e. there was a greaterdifference between EFL and EAL in the no qualifica-tions sub-group.

The no qualification EAL sub-group did poorly onverbal production sub-tests (picture naming, wordfluency, word definition) when compared with theirEFL counterparts and other EAL speakers in highereducation groups. On a test of picture description, thissub-group’s score differed from all the other groups(2.5, compared to 4 for other groups) and single wordspelling was significantly lower than all other groups.Also, in the no qualification EAL sub-group, Trail Testtime was much higher than other groups (a higherscore indicating less ability).

DISCUSSION

The findings show that in order to effectively assess awide population of older people that includes ethnicelders, adapted norms on formal language tests areneeded. This finding is supported by the differencesfound between the EAL and EFL groups tested on theBLA. The EAL group did less well on a language testthan their contemporaries in the EFL group. Theresults show that education is a significant factor inEAL performance, and that the higher the level ofeducation, the less the difference between EAL andEFL performance.

The sub-group that did consistently worse was theEAL group with no formal qualifications. This couldbe explained by a number of factors. The first may beunfamiliarity with the test format, so poor Trail Testscores (time and error) could be due to unfamiliaritywith the test concept (a type of puzzle) that may havestrong cultural bias. The fact that tests of verbal output(expression) were more poorly performed by EALsub-groups lends support to the view that some ethnicolder people will have limited functional use ofEnglish, poor vocabulary and reduced use and under-standing of more complicated syntax. Those withhigher levels of education are more likely to haveworked in the UK, have required English as part oftheir job specification, and/or have been exposed tosome formal English teaching.

There are, therefore, two types of EAL individuals,for whom different approaches to assessment andmanagement are required, one of whom has somefunctional English, the other, very limited or nofunctional English.

For those with functional English, consideration oferror profiles is also an important finding from thestudy (to be expanded in a future paper). It is likelythat some ‘errors’ found on sub-tests such as reading

Int J Geriatr Psychiatry 2009; 24: 426–431.

DOI: 10.1002/gps

Page 4: Using the Barnes Language Assessment with older ethnic minority groups

Table 2. Summary statistics across education groups and language background

Sub test No qualifications O or A levels Diploma or degree

EFL EAL EFL EAL EFL EAL

Spoken-word picMatching NR* 13–15 11–14 14–15 11–15 15 14–15Mean 14.75 12.25 14.84 13.80 15.00 14.71(SD) (0.866) (1.708) (0.367) (1.612) (0.000) (0.667)Fluency ‘s’ NR* 9 þ 4 þ 9 þ 7 þ 11 þ 8 þMean 13.33 6.50 15.73 12.93 18.55 16.34(SD) (3.257) (2.887) (5.370) (4.026) (5.422) (5.573)Picture naming NR* 13 þ 7 þ 13 þ 9 þ 14 þ 11 þMean 14.42 9.00 14.60 12.87 14.62 13.97(SD) (0.793) (2.160) (0.688) (2.326) (0.622) (1.175)Spelling NR* 15 þ 5 þ 16 þ 5 þ 16 þ 12 þMean 17.00 5.50 17.49 12.27 17.38 16.49(SD) (1.128) (3.873) (0.895) (4.862) (1.293) (2.077)Trail test time NR* Up to 38 Up to 27 Up to 26 Up to 27 Up to 18 Up to 20Mean 17.47 23.98 10.54 16.14 9.95 11.44(SD) (15.799) (6.667) (6.254) (14.018) (6.620) (6.010)Trail test errors NR* Up to 9 Up to 8 Up to 5 Up to 8 Up to 5 Up to 8Mean 1.92 1.75 0.44 2.80 0.55 1.77(SD) (3.895) (2.062) (1.560) (4.814) (1.478) (3.126)Fluency animals NR* 10 þ 8 þ 16 þ 13 þ 16 þ 14 þMean 19.58 10.75 23.91 19.47 26.52 21.63(SD) (6.515) (4.272) (5.977) (7.100) (5.462) (6.567)Word definition NR* 14 þ 9 þ 15 þ 9 þ 15 þ 14 þMean 17.42 8.75 20.16 15.87 21.55 19.06(SD) (3.288) (2.062) (3.960) (4.565) (3.915) (3.757)Trog NR* 4 þ 3 þ 6 þ 4 þ 7 þ 5 þMean 6.83 3.50 7.76 6.07 8.34 7.17(SD) (1.193) (1.291) (1.048) (1.710) (0.769) (1.317)Forward digit span NR* 5 þ 5 þ 5 þ 5 þ 5 þ 5 þMean 5.58 5.25 6.91 5.47 7.17 6.34(SD) (0.996) (0.500) (1.164) (0.834) (1.227) (1.349)Following commands NR* 5 2 þ 5 2 þ 5 4 þMean 4.92 3.00 5.00 4.53 4.97 4.89(SD) (0.289) (1.155) (0.000) (0.834) (0.186) (0.404)Sentence writing NR* 1 2 1 Up to 1 0 0Mean 0.17 2.25 0.11 0.20 0.03 0.03(SD) (0.389) (2.630) (0.383) (0.561) (0.186) (0.169)Story retell NR* 9 þ 7 þ 9 þ 8 þ 12 þ 8 þMean 11.42 10.25 13.00 11.93 14.17 12.80(SD) (1.379) (3.304) (2.646) (3.011) (1.794) (2.495)Reading NR* 38 þ 21 þ 39 þ 27 þ 39 þ 31 þMean 39.17 23.50 39.84 33.27 39.79 36.26(SD) (1.030) (4.933) (0.367) (5.800) (0.559) (2.559)Pic description NR* 4 þ 2 þ 4 þ 2 þ 4 þ 2 þMean 4.25 2.50 4.38 4.13 4.45 4.14(SD) (0.622) (1.000) (0.490) (1.060) (0.572) (1.004)

NR*¼Normal range.

language assessment in ethnic elders 429

aloud may be based on regular patterns of pronuncia-tion within certain language cultures. Understandingthis would reduce the possibilities of ‘false positive’diagnoses, but further analysis is needed to provide aframework within which to understand and measuresuch error patterns. This work could be the subject offuture research.

Copyright # 2009 John Wiley & Sons, Ltd.

Surprisingly, age alone was not a significantvariable. The ‘older old’ in our study tended to fallinto the EAL no qualifications group and therefore, asdescribed above, were the worst performers. However,data analysis suggested this was due to lack ofknowledge of the English language rather than ageper se. This does not negate the fact that there is a

Int J Geriatr Psychiatry 2009; 24: 426–431.

DOI: 10.1002/gps

Page 5: Using the Barnes Language Assessment with older ethnic minority groups

KEY POINTS

� Subjects with English as a first language (EFL)and English as an additional language (EAL)perform differently on the Barnes languageAssessment

� The differences are mainly related to levels ofeducation

� Difficulties were found in recruiting subjectsfrom ethnic minority groups

430 v. ramsey ET AL.

group of older ethnic people for whom accessingservices remains problematic due to lack of Englishand for whom using a tool like the BLAwould not beappropriate.Another finding was the difficulty in recruiting

participants fromminority ethnic groups. The problemincreased with age, the 80þ group being the mostdifficult to recruit. The method of recruitment variedfrom contacting agencies such as local day centres,facilities for older people and exercise groups, to wordof mouth via friends and work contacts. It wasassumed at the beginning of the research that older agerecruitment would be easier than it was. The result wasthat the target number of participants (five) in someEAL sub-groups was not achieved, notably the80þ group.One factor that could have contributed is a suspicion

amongst older ethnic elders about participating inresearch and answering questions. This may be due todifferent cultural attitudes to ageing, and resistance topressure to conform to a different culture not fullyunderstood. Some participants may have felt beingtested meant being ‘checked on’, while others wereunwilling to disclose frailties. Some may have felt thatbecause you are old there is no value in you beingtested. Elsewhere, a lack of respect or understandingof culture has been attributed to the poor developmentof relationship between researchers and minority elderparticipants (Brangman, 1995; Bedolla, 1995).A functional level of English is required to

complete the BLA. Many older people from anethnic background may have limited English, relyingon family to translate where necessary, and mixingonly with those from the same ethnic, and thereforelanguage, background. This was particularly true ofthe Chinese community, and is supported by evidence(Yu, 2000).Also, as noted earlier, there may be cultural

differences amongst ethnic groups, not only withregard to attitude to research but also in terms of thetype of tasks that they were asked to do. The

Copyright # 2009 John Wiley & Sons, Ltd.

differences applied not only to linguistic tasks such asreading and writing, but also to other tasks, e.g. picturerecognition or the Trail Test (Carter et al., 2005).Therefore, the very act of being involved in a formaltest may be a factor that prevented people from takingpart in the study. A lack of familiarity with the testsituation per se may cause difficulties in participatingin structured tests or interviews and has been cited asone cause of misdiagnosis in minority elders (Espinoand Lewis, 1998).

It is also possible that some people have no faith inmainstream services and would prefer to gain supportand advice from their own community voluntaryorganisations (Butt and O’Neil, 2004).

CONCLUSION

The findings of this research support the suggestionthat there are differences in performance on formallanguage tests between those who are EAL and EFLspeakers. As the results indicate, in order to accuratelytest language, adapted norms are needed to allow forthose differences. The BLA has adapted norms.Distinct groups were found within the EAL group, i.e.those with useable English and those with very limitedor no functional English. The use of formal tests suchas the BLA may not be appropriate for the secondgroup. The project has raised issues about the access ofEAL individuals to assessment services, together witha number of potential research questions:

� A

re there a group of people for whom alternativemethods of assessment and service provision arenecessary in order to meet their needs successfully?

� I

f that is the case, then in what form would languageassessment take?

� H

ow can information be provided to those whosepre-morbid English language use is limited?

� A

re there error patterns that can be predicted andattributed to specific ethnic groups?

� A

re there cultural factors that give rise to difficultieson particular test formats, e.g. animal fluency orTrail Test?

CONFLICT OF INTEREST

None known.

ACKNOWLEDGEMENTS

Sponsored with a grant from the Royal College ofSpeech and Language Therapists.

Int J Geriatr Psychiatry 2009; 24: 426–431.

DOI: 10.1002/gps

Page 6: Using the Barnes Language Assessment with older ethnic minority groups

language assessment in ethnic elders 431

The work was approved by the University of SurreyEthics Committee and the Hammersmith HospitalResearch Ethics Committee.

REFERENCES

Adshead F, Day Cody D, Pitt B. 1992. BASDEC: a novel screeninginstrument for depression in elderly medical inpatients. BMJ 305:397.

Ayalon L, Arean P. 2004. Knowledge of Alzheimer’s disease in fourethnic groups of older adults. Int J Geriatr Psychiatry 19:51–57.

Bedolla MA. 1995. The principles of medical ethics and theirapplication to Mexican-American elderly patients. Clin GeriatrMed 11: 131–137.

Butt J, O’Neil A. 2004. ‘Lets Move On’: Black and Minority EthnicOlder People’s Views on Research Findings. Joseph RowntreeFoundation ISBN 1 85935 175 1. www.jrf.org.uk/knowledge/findings/socialcare/564.asp

Brangman SA. 1995. African-American Elders: implications forhealth care providers. Clin Geriatr Med 11: 15–23.

Bryan K, Stevens S, Binder J, et al. 2006. The Barnes LanguageAssessment: Cultural Dimensions in the Assessment of Langugaein Older People. Australian Speech Pathology Association Con-ference, Perth, November.

Bryan K, Binder J, Dann C, et al. 2001. Development of a screeningassessment for language in older people (Barnes LanguageAssessment). Aging Ment Health 5: 371–378.

Carter JA, Lees JA, Murira GM, et al. 2005. Issues in the devel-opment of cross-cultural assessments of speech and language forchildren. Int J Lang Comm Dis 40: 385–401.

Copyright # 2009 John Wiley & Sons, Ltd.

Eolas. 1999. Health and Social Care for Older Black and EthnicMinority Residents of Sefton. Commissioned by Sefton HealthAuthority and Sefton Council.

Espino DV, Lewis RL. 1998. Dementia in Older Minority Popu-lations. Issues of prevalence, diagnosis and treatment. Am JGeriatr Psychiatry 6: S19–S25.

Folstein MF, Folstein SE, McHugh JPR. 1975. ‘Mini-Mental State.’A practical method for grading the cognitive state of patients forthe clinician. Psychiat Res 12(3): 189–198.

Jastak J, Jastak S. 1978. The Wide Range Achievement Test. JastakAssociates: Wilmington, DE.

Lindesay J, Jagger C, Mlynik-Szmid A, et al. 1997. The Mini-Mental State Examination (MMSE) in an elderly population in theUnited Kingdom. Int J Lang Comm Dis 12: 1155–1167.

Odutoye K, Shah A. 1999. The characteristics of Indian Subconti-nent Origin Elders newly referred to a Psychogeriatric Service. IntJ Geriatr Psychiatry 14: 446–453. DOI: 10.1002/(SICI)1099-1166(199906)14:6<446::AID.-GPS950>3.0.CO;2-L

Rait G, Burns A, Baldwin R, et al. 2000. Validating screeninginstruments for cognitive impairment in older South Asians in theUnited Kingdom. Int J Geriatr Psychiatry 15: 54–62.

Richards M, Brayne C, Dening T, et al. 2000. Cognitive function inUK community-dwelling African Caribbean and White Elders: apilot study. Int J Geriatr Psychiatry 15: 621–630.

Shah A. 1999. Difficulties experienced by a Gujarati GeriatricPsychiatrist in interviewing Gujarati Elders in Gujarati. Int JGeriatr Psychiatry 14: 1072–1074. DOI: 10.1002/(SICI)1099-1166(199912)14:12<1072::AID.-GPS93>3.0.CO;2-W

TangM, Stern Y, Marder K, et al. 1998. The APOE-e4 Allele and therisk of Alzheimer Disease among African Americans, Whites andHispanics. J Am Med Assoc 279: 751–755.

YuWK. 2000. Chinese Older People: A Need for Social Inclusion inTwo Communities. The Policy Press for the Joseph RowntreeFoundation: Bristol.

Int J Geriatr Psychiatry 2009; 24: 426–431.

DOI: 10.1002/gps