Age and Ageing 2009; 38: 669–675 C© The Author 2009. Published by Oxford University Press [on behalf of the British Geriatrics Society].All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.orgdoi: 10.1093/ageing/afp127

Published electronically 3 August 2009

Cognitive performance in community-dwellingEnglish- and Spanish-speaking seniors

ALEX D. FEDERMAN1, HELEN COLE1, MARY SANO2,3

1Division of General Internal Medicine, Mount Sinai School of Medicine, New York, NY, USA2Department of Psychiatry, Mount Sinai School of Medicine, New York, NY, USA3Geriatric Research, Education, and Clinical Center, Bronx Veterans Affairs Medical Center, Bronx, NY, USA

Address correspondence to: A. D. Federman. Division of General Internal Medicine, Mount Sinai School of Medicine, OneGustave L. Levy Place, Box 1087, New York, NY 10029, USA. Tel: (+1) 212 824 7565; Fax: (+1) 212 824 2317.Email: alex.federman@mssm.edu

Abstract

Objectives: to examine the association of language (English vs Spanish), and commonly used measures of memory and wordfluency among older adults.Design: cross-sectional.Setting: community-based settings in New York City, including senior centres and residential complexes.Subjects: four hundred and twenty independently living adults aged 60 or older (mean 73.8 years).Methods: participants completed the Mini-Mental State Exam (MMSE), animal naming test (ANT) and Wechsler MemoryScale III (WMS) Story A immediate and delayed subtests. Scores were examined by strata of language, age or education and fordifferent thresholds of the MMSE. We tested the association of language and cognitive test performance using multivariablelinear regression.Results: twenty-one per cent of subjects were interviewed in Spanish and 16.2% reported poor-fair English proficiency. Themean WMS scores were not statistically different between English and Spanish groups (immediate recall, 9.9 vs 9.5, P = 0.44;delayed recall, 8.0 vs 7.6, P = 0.36, respectively), whereas ANT scores did differ (16.6 vs 14.3, P < 0.0001). These associationswere consistent across MMSE thresholds. The association of language and ANT score was not significant after accountingfor education.Conclusions: we found little difference in performance on the Story A subtests from the WMS suggesting that this test maybe used for both English- and Spanish-speaking populations. Results suggest that variations in ANT performance may beaccounted for by adjusting for the level of education. These results have important implications for the generalisability of testscores among diverse older populations.

Keywords: elderly, cognition, language, Spanish

Introduction

Community-based research often seeks to examine the asso-ciation between cognition and health outcomes. However,this is a major challenge when multiple languages are spokenwithin a given cohort. Test selection must account for demo-graphic characteristics of the sample, such as age and educa-tion that are known to affect test performance [1–3]. Whilethe effects of age and education are well established, lessis known about language of assessment, although evidencesuggests that it influences test performance [3–5]. Spanish isthe second most commonly spoken language in the UnitedStates with prevalence estimates of 12% of the total popu-lation [6] and one-third of the older population. Yet there

is little normative data for US Spanish-speaking cohorts thatmight inform the selection of cognitive tests.

There are many variables to consider when assessingSpanish speakers in the United States such as literacy inSpanish, country of origin, education, length of time in theUnited States and proficiency of English [4, 5]. In studies ofSpanish-speaking adults, previous investigators have founda nominal impact of language on test performance whenthe research subjects were equally skilled in English andSpanish language, whereas Spanish-dominant speakers havetended to perform less well on tests of memory and wordfluency, suggesting that acculturation may contribute to dif-ferences in performance [4, 5]. Since the association betweencognition and health outcomes may be used to improve

669

at Universidade Federal do R

io Grande do N

orte on April 14, 2015

http://ageing.oxfordjournals.org/D

ownloaded from

A. D. Federman et al.

health care information and dissemination, it is importantto select assessment tools that can be used among Spanishspeakers.

In this report, we took a pragmatic approach to charac-terise performance of Spanish speakers on cognitive tests ofmemory and word fluency. Spanish speakers were defined asthose who preferred to be tested in Spanish. We also assessedself-reported English proficiency. While this approach maynot permit the assessment of the impact of all nuances ofculture and language, it reflects a practical way to includeSpanish speakers in community-based research and may pro-vide the basis for developing additional sensitivity for clinicalassessments of cognition. We conducted an analysis of datafrom a subset of tests from established neuropsychologicalbatteries [7], specifically the first paragraph from the Wech-sler Memory Scale III (WMS) logical memory I and II subtest[8] and the animal naming test (ANT) [9], from a diversepopulation of urban, independently living older adults. Weexamined performance on these assessments stratified by ageand education, and present data on the effect of both test-ing language (English or Spanish) and self-reported Englishproficiency.

Methods

Subjects and setting

Data for these analyses were collected in a larger studythat addressed health insurance decision-making amongolder adults. We recruited independently living adults aged60 years and older from 30 community-based settings inNew York City, including senior centres and residential com-plexes. Senior centres were identified through listings fromthe New York City Department for the Aging and resi-dential complexes through a listing of federal Housing andUrban Development-supported low-income housing facili-ties. We selected sites in zip code areas with median house-hold incomes below $50,000, and we over-sampled menbecause they are outnumbered by women in these communi-ties and in this age group and in our experience are less likelyto participate in survey interviews [6].

Individuals were recruited during site-sponsored meals orspecial events for a longitudinal study about ‘health, healthcare use and health insurance’ that provided $20 for thebaseline interview and $10 for a follow-up interview. Inter-views were conducted with only one member of a house-hold and were conducted on-site in English and Spanish bytrained bilingual interviewers. The interviewers were nativeSpanish speakers from Puerto Rico, the Dominican Repub-lic, and Central America, congruent with the cultural back-grounds of much of the study population. To ensure theability to complete the visual tasks of the Mini-Mental StateExam (MMSE), we only included individuals whose visualacuity was 20/50 or better as determined with a handheldSnellen chart, with or without use of corrective lenses. Writ-ten informed consent was obtained from all participants priorto screening and interviews. The study was approved by

the Mount Sinai School of Medicine Institutional ReviewBoard.

Measures of cognition

Our cognitive assessments focused on memory and cate-gory fluency. We assessed memory using the immediate anddelayed recall for Story A of the Logical Memory test from theWMS III [8]. Story A is a brief passage read by an interviewerwho asks the subject to recall as much of the story as possibleimmediately after it is read. The interviewer then informs theparticipant that they will be asked to recall it again ‘later’. Theparticipant is asked to recall as much of the story as possibleagain 25–30 min later. Subjects are given credit for recallingcore elements of the story. This method is comparable tothose used in other large multicentre studies [7].

We assessed category fluency using the ANT [10]. TheANT measures semantic verbal fluency, an ability thatinvolves language and executive functions [9, 11–13]. Thetest requires subjects to name as many animals as they canin 1 min. Scores equal the total number of unique animalsnamed within 1 min.

We also assessed levels of global cognitive function usingthe MMSE [14], a widely used dementia screening instru-ment. Interviewers were formally trained by a neuropsychol-ogist in the Mount Sinai School of Medicine, Alzheimer’sDisease Research Center, in the administration of all cogni-tive assessments. Their proficiency in test administration wasestablished prior to data collection.

Language

We used two variables to represent language. The first vari-able indicated whether the subject was assessed in Spanish.The study participants were given this option irrespective oftheir English-speaking ability. The second variable was a self-reported measure of English language proficiency, assessedusing the question: ‘How would you describe your ability tospeak and understand English?’ with six responses rangingfrom very poor to excellent.

Other variables

We collected data on variables having established associa-tions with cognitive function including age, education level,income and health and functional status [2, 15–19]. Healthstatus measures included self-reported general health fromthe Short Form-36 [20], total number of chronic diseases andbasic activities of daily living (ADL) and instrumental activ-ities of daily living (IADL). We documented self-reportedperformance on six ADLs (bathing or showering, walking,getting in or out of bed or chairs, eating, dressing and usingthe toilet) [21] and five IADLs (managing money, preparingmeals, doing light or heavy housework and shopping) [21].The study participants were considered to have impairmentif they experienced ‘a lot of’ difficulty with or were unableto do one or more tasks. Additional variables included sex,race and self-identified ethnicity (non-Hispanic black, non-Hispanic white, Hispanic and other).

670

at Universidade Federal do R

io Grande do N

orte on April 14, 2015

http://ageing.oxfordjournals.org/D

ownloaded from

Language and cognitive assessment scores

Analysis

We tabulated means and standard deviations for scores onthe immediate and delayed recall tasks and the ANT, strat-ified by language. In comparing test performance betweenEnglish and Spanish speakers, we aimed to provide datathat would inform future studies involving cognitive assess-ments in older, community-dwelling adults. Hence, we usedthree commonly applied MMSE score thresholds to iden-tify clinically relevant subgroups often used in community-based research: MMSE of 22 or higher to ensure that wecaptured individuals with moderate to no cognitive impair-ments; MMSE of 26 or higher to limit the subgroup to indi-viduals with possible mild cognitive impairment (MCI) to nocognitive impairments; and MMSE of 28 or higher to have asubgroup minimal or no cognitive impairments [22, 23]. Testscores were compared using t-tests. Because of the prob-lem of small cell sizes with stratification, age strata werelimited to 60–69 years, 70–79 and 80 and older, and educa-tion to high school or higher versus did not graduate highschool.

We also conducted a series of linear regression analysesto examine the association of language (language of interviewadministration or self-reported English proficiency) withtest performance. We first modelled test performance as afunction of language alone. We then sequentially added age,education and then other covariates that had a statisticallysignificant bivariate association (P < 0.05) with the outcomemeasure. The participants reporting poor or fair English-speaking ability were also likely to request interview admin-istration in Spanish (kappa 0.73), and results for languageof interview administration and self-reported language pro-ficiency were qualitatively the same. We therefore only pre-sented results for interview language (results for analysesusing the language proficiency variable are available from theauthors). The test results were considered statistically signif-icant at the P < 0.05 level.

Most questions were missing data for fewer than 5%of study participants. However, 13.0% were missing dataon income, consistent with self-reported income data fromother studies [24, 25]. We used multiple hot-deck imputa-tions to replace missing observations for income in logisticregression analyses [26]. Imputed data sets were created withSTATA version 10 using the hotdeck command (Stata Cor-poration, College Station, TX, USA). All analyses were con-ducted with SAS version 9.1 (SAS Institute, Inc., Cary, NC,USA).

Results

Characteristics of the study participants

Of the 453 individuals recruited, 420 (92.7%) had MMSEscores of 22 or higher and were included in our analyses.The study subjects who were administered the interview inSpanish were more likely to be among the 31 excluded indi-viduals than English speakers (58.1% vs 41.9%, P < 0.0001).

Sixty-eight subjects (16.2%) reported having very poorto fair English language skills, and the interview (includingcognitive tests) was administered in Spanish to 86 (20.5%)(Table 1). Among those reporting very poor to fair Englishlanguage skills, 61 (89.7%) completed the Spanish-languageinterview. Those who completed the interview in Spanish hadlower levels of education and income and had worse generalhealth and greater use of emergency department care thanthose who completed it in English (Table 1). The mean ageof the cohort was 73.8 years (Table 1). Two-thirds (63.1%)were women, one-quarter (27.2%) had not graduated highschool and more than half (54.3%) had household incomesat or below $1,350 per month. The sample was approximatelyone-third black (30.5%), one-third white (31.9%) and one-third Hispanic (30.7%).

Cognitive assessments

Table 2 shows the mean and standard deviations for immedi-ate recall scores, delayed recall scoresand ANT scores, strati-fied by age and education, for English- and Spanish-speakingsubjects with MMSE scores of 22 or higher. As expected,performance on the assessments declined with increasingage and lower educational attainment, for both English andSpanish speakers.

Comparisons of test administered in English andSpanish

Subjects interviewed in English and Spanish had similarscores on tests of memory across all three MMSE thresholds,whereas those who were administered the ANT in Spanishhad poorer performance (see Appendix Table 1 in the sup-plementary data available at Age and Ageing online). Similarly,individuals who reported poor English-speaking ability hadpoorer performance on the ANT than better English speak-ers (data not shown). When test scores were stratified by thelevel of education, there were no significant differences bylanguage for any of the tests, although the difference in WMSStory A immediate recall scores between Spanish and Englishspeakers at or above the 12th grade level of education was ofborderline statistical significance, with Spanish speakers hav-ing slightly higher performance on the assessment (11.7 ±3.5 vs 10.3 ± 3.7, P = 0.05) (see Appendix Table 2 in thesupplementary data available at Age and Ageing online).

Although the differences between Spanish and Englishlanguage performance on memory assessment scores werenot statistically significant on bivariate analyses, we con-ducted regression analyses adjusting for age and educationsince these are potent correlates of test performance, andSpanish- and English-speaking study participants differedmarkedly by these characteristics. Indeed, the associationbetween Spanish language interview and immediate recallscores was statistically significant after adjusting for educa-tion and closely approached significance for delayed recall(Table 3). For immediate recall, language narrowly lost sta-tistical significance when age was added to the model, incontrast to that for delayed recall.

671

at Universidade Federal do R

io Grande do N

orte on April 14, 2015

http://ageing.oxfordjournals.org/D

ownloaded from

A. D. Federman et al.

Table 1. Subject characteristics

Total sample English Spanisha P-value. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

N = 420 N = 334 (79.5%) N = 86 (20.5%)Age (mean, SD) 73.8 (8.7) 73.9 (9.0) 73.4 (7.4) 0.62b

Age (years) 0.00260–69 153 (36.5) 128 (38.4) 25 (29.1)70–79 160 (38.2) 113 (33.9) 47 (54.6)80+ 106 (25.3) 92 (27.6) 14 (16.3)

Female 265 (63.1) 208 (62.3) 57 (66.3) 0.49Race <0.0001

White, non-Hispanic 133 (31.9) 133 (40.2) 0 (0)Black, non-Hispanic 127 (30.5) 127 (38.4) 0 (0)Hispanic 128 (30.7) 42 (12.7) 86 (100)Other 29 (7.0) 29 (8.8) 0 (0)

Education <0.0001Less than high school 66 (15.8) 16 (4.8) 50 (58.1)Some high school 48 (11.5) 42 (12.6) 6 (7.0)High school graduate 95 (22.7) 86 (25.8) 9 (10.5)Some college 92 (22.0) 78 (23.4) 14 (16.3)College graduate or higher 118 (28.2) 111 (33.3) 7 (8.1)

Married 79 (18.8) 59 (17.7) 20 (23.3) 0.24Income <0.0001

<$750 104 (25.2) 65 (19.9) 39 (45.4)<$751–$1,350 120 (29.1) 89 (27.3) 31 (36.1)<$1,351–$2,000 74 (18.0) 62 (19.0) 12 (14.0)<$2,001–$3,000 52 (12.6) 51 (15.6) 1 (1.2)>$3,001 62 (15.1) 59 (18.1) 3 (3.5)

General health <0.0001Excellent 39 (9.3) 31 (9.3) 8 (9.3)Very good 80 (19.1) 71 (21.3) 9 (10.5)Good 137 (32.7) 124 (37.2) 13 (15.1)Fair 134 (32.0) 86 (25.8) 48 (55.8)Poor 29 (6.9) 21 (6.3) 8 (9.3)

Chronic diseases 0.24None 34 (8.1) 28 (8.4) 6 (7.0)1 64 (15.2) 55 (16.5) 9 (10.5)2 87 (20.7) 71 (21.3) 16 (18.6)3 92 (21.9) 75 (22.5) 17 (19.8)4 or more 143 (34.1) 105 (31.4) 38 (44.2)

ADL deficiency 71 (16.9) 52 (15.6) 19 (22.1) 0.15IADL deficiency 180 (42.9) 136 (40.7) 44 (51.2) 0.08Any hospital stay 98 (23.3) 80 (24.0) 18 (20.9) 0.56Any ED visit 165 (39.3) 121 (36.2) 44 (51.2) 0.01

aCompleted the interview, including all neurocognitive assessments, in Spanish.bt-test, all other tests χ2.ADL denotes activities of daily living; IADL, instrumental activities of daily living; ED, emergency department.

Table 2. Means and standard deviations for scores on memory and word fluency assessments

English Spanish

WMS immediate WMS delayed Animal naming WMS immediate WMS delayed Animal namingrecall recall test recall recall test

n Mean (SD) Mean (SD) Mean (SD) n Mean (SD) Mean (SD) Mean (SD). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Overall 334 9.9 (3.8) 8.0 (4.1) 16.6 (5.3) 86 9.5 (4.1) 7.6 (3.8) 14.3 (3.9)Age (years)

60–69 128 10.2 (3.6) 8.4 (4.0) 17.7 (5.4) 25 10.5 (2.9) 8.6 (3.0) 15.4 (3.9)70–79 114 10.1 (3.9) 8.4 (4.1) 16.4 (5.4) 47 9.7 (4.4) 7.7 (4.1) 14.0 (4.0)80+ 92 9.1 (4.0) 7.1 (4.2) 15.4 (4.9) 14 7.2 (4.3) 5.5 (3.5) 13.0 (3.5)

Education (years)12+ 276 10.2 (3.8) 8.3 (4.0) 17.1 (5.4) 30 11.7 (3.5) 9.8 (3.1) 15.4 (4.6)<12 58 7.4 (3.6) 5.6 (3.7) 13.4 (3.9) 56 8.3 (4.0) 6.4 (3.7) 13.6 (3.4)

672

at Universidade Federal do R

io Grande do N

orte on April 14, 2015

http://ageing.oxfordjournals.org/D

ownloaded from

Language and cognitive assessment scores

Table 3. Multivariable analysis of neurocognitive test scores: effect of Spanish languagea

Immediate recall Delayed recall Animal naming test

β (SD) P-value β (SD) P-value β (SD) P-value. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Model 1

Spanish interview −0.36 (0.47) 0.44 −0.44 (0.49) 0.36 −2.35 (0.62) 0.0002Model 2

Spanish interview −0.47 (0.47) 0.32 −0.58 (0.49) 0.24 −2.34 (0.62) 0.0002Age (years)

60–69 Ref. – Ref. – Ref. –70–79 −0.23 (0.44) 0.60 −0.14 (0.46) 0.76 −1.29 (0.57) 0.0280+ −1.44 (0.49) 0.003 −1.51 (0.51) 0.003 −2.28 (0.63) 0.0004

Model 3Spanish interview 1.10 (0.50) 0.03 0.96 (0.52) 0.07 −0.81 (0.66) 0.22Education <12 years −3.00 (0.45) <0.0001 −2.87 (0.47) <0.0001 −3.20 (0.60) <0.0001

Model 4Spanish interview 0.97 (0.50) 0.053 0.80 (0.52) 0.13 −0.81 (0.66) 0.22Age

60–69 Ref. – Ref. – Ref. –70–79 −0.28 (0.42) 0.51 −0.19 (0.44) 0.67 −1.33 (0.56) 0.0280+ −1.35 (0.47) 0.004 −1.44 (0.49) 0.003 −2.16 (0.62) 0.0005Education <12 years −3.11 (0.59) <0.0001 −2.78 (0.47) <0.0001 −3.12 (0.60) <0.0001

Model 5b

Spanish interview 1.16 (0.51) 0.02 1.05 (0.53) 0.05 −0.33 (0.67) 0.62Age

60–69 Ref. – Ref. – Ref. –70–79 −0.24 (0.42) 0.57 −0.25 (0.44) 0.56 −1.28 (0.56) 0.0280+ −1.45 (0.47) 0.002 −1.57 (0.49) 0.001 −2.36 (0.62) 0.0002Education <12 years −2.35 (0.48) <0.0001 −1.96 (0.50) <0.0001 −2.39 (0.63) 0.0002

Linear regression for individuals with Mini-Mental State Exam scores ≥22.aThe primary independent variable for Spanish language is administration of the interview, including all neurocognitive assessments.bModels additionally adjusted for income, general health and functional status (any deficiency of in activities of daily living).

In multivariable linear regression analysis (Table 3),Spanish language remained significantly associated with ANTscores after adjusting for age. However, the magnitude of theassociation between Spanish language and ANT scores wasreduced by 62% when education was added to the model(Table 3, model 3) and the association lost its statistical sig-nificance. The magnitude of the association was reduced fur-ther by the addition of income, general health and functionallimitations to the model (model 5). Analyses demonstratedthat among these latter variables, income was responsible forthe additional confounding (data not shown).

Discussion

Assessing the cognitive functioning of older individuals canbe particularly challenging in diverse urban areas where dif-ferent languages are represented. Fortunately, for two com-monly used neurocognitive assessments, we have foundlittle difference in performance between English and Span-ish speakers among a community-dwelling urban popula-tion. This population is notable for a high percentage ofnon-Hispanic black English speakers and is typical of otherurban older communities. We administered the immediateand delayed recall assessments of the Story A subtest of theWMS III and a measure of word fluency, the ANT, to this

diverse sample of urban, independently living older adults.We used easily defined and practical definitions of language—self-reported English language proficiency and preferencefor Spanish language assessments—and found similar pat-terns of test performance for each. Only with the ANT, ameasure of word fluency, did we observe significantly lowerscores by Spanish speakers. Stratified and multivariable anal-yses demonstrated that the poorer performance by Spanishspeakers on the ANT were largely attributable to lower lev-els of education. Previous research has suggested that theANT is relatively insensitive to variations in English lan-guage ability [11, 27]. Our findings somewhat contradict theseprior observations, and suggest the need for education-basedadjustments to ANT scores in multilingual samples. On theother hand, our findings indicate that education adjustmentsby language may not be needed when the WMS Story Amemory assessment is administered in multilingual samples.

Independent of language effects, our data on the ANT areconsistent with previously published data from other popu-lations, in that age and education had an impact on perfor-mance [8, 11, 28, 29]. Scores on the WMS Story A delayedrecall task in our study were higher than those observed inthe Framingham Heart Study cohort [30], but they were nor-mally distributed and followed the expected patterns of ageand education associations [27, 28]. Specifically, we foundthat cognitive test performance declined with increasing age

673

at Universidade Federal do R

io Grande do N

orte on April 14, 2015

http://ageing.oxfordjournals.org/D

ownloaded from

A. D. Federman et al.

and lower levels of education, consistent with findings fromstudies across a host of populations.

Our study provides data for two commonly used neu-ropsychiatric assessments for independently living, olderadults from demographically diverse urban communities.However, a number of factors should be considered beforeextrapolating these data to other populations. First, the sam-ple size of 420 individuals restricted our analyses to largestrata of age and education. Second, we employed conve-nience sampling rather than true random sampling whichcould result in selection bias. In addition, a greater proportionof Spanish speakers were excluded from our analyses becauseof low MMSE scores. Nonetheless, the impact of selectionbias is probably nominal in the range of MMSE scores of 22and greater as indicated by the balanced distribution of partic-ipant characteristics and the normal distribution of scores onall neurocognitive assessments. Third, we administered onlyone component of the logical memory subtests of the WMSIII during subject interviews. Hence, we have not obtainedcomprehensive measures of memory.

In conclusion, our study provides descriptive data forStory A of the WMS III memory assessments and the ANTwhich suggests that these tests have relatively comparableranges permitting research to be extended to Spanish lan-guage speakers.

Key points� Spanish- and English-speaking older adults perform sim-

ilarly on Story A of the WMS III memory assessment andthe ANT when controlling for education.

� Poorer performance by Spanish speakers on some assess-ments of cognition is attributable to lower education.

� Education adjustments by language may not be neededwhen the WMS Story A memory assessment is adminis-tered in multilingual samples.

Conflict of interest

None.

Funding

This study was supported by a Paul B. Beeson Career Devel-opment Award in Aging from the National Institute on Aging(Dr. Federman, 1K23AG028955-01). Additional support wasprovided by the Mount Sinai School of Medicine Alzheimer’sDisease Research Center (NIH AG0051318).

Supplementary data

Supplementary data are available at Age and Ageing online.

References

1. Crum RM, Anthony JC, Bassett SS, Folstein MF. Population-based norms for the Mini-Mental State Examination by age andeducational level. JAMA 1993; 269: 2386–91.

2. Fisk JE, Sharp CA. Age-related impairment in executive func-tioning: updating, inhibition, shifting, and access. J Clin ExpNeuropsychol 2004; 26: 874–90.

3. Johnson-Selfridge MT, Zalewski C, Aboudarham JF. The rela-tionship between ethnicity and word fluency. Arch Clin Neu-ropsychol 1998; 13: 319–25.

4. Gasquoine PG, Croyle KL, Cavazos-Gonzalez C, SandovalO. Language of administration and neuropsychological testperformance in neurologically intact Hispanic Americanbilingual adults. Arch Clin Neuropsychol 2007; 22: 991–1001.

5. Rosselli M, Ardila A, Araujo K et al. Verbal fluency and repe-tition skills in healthy older Spanish–English bilinguals. ApplNeuropsychol 2000; 7: 17–24.

6. US Census Gateway, 2000. Available at http://www.census.gov/main/www/cen2000.html. Accessed March 12, 2007.

7. Morris JC, Weintraub S, Chui HC et al. The Uniform DataSet (UDS): clinical and cognitive variables and descriptive datafrom Alzheimer Disease Centers. Alzheimer Dis Assoc Disord2006; 20: 210–6.

8. Wechsler D. Manual: Wechsler Memory Scale-III Abbreviated,3rd edition. San Antonio: Harcourt Assessment, 2002.

9. Lezak M, ed. Neuropsychological Assessment, 3rd edition.Oxford: Oxford University Press, 1995.

10. Rosen WG. Verbal fluency in aging and dementia. J Clin ExpNeuropsychol 1980; 2: 135–46.

11. Ardila A, Ostrosky-Solis F, Bernal B. Cognitive testing towardthe future: the example of semantic verbal fluency (ANIMALS).Int J Psychol 2006; 41: 324–32.

12. Abrahams S, Leigh PN, Harvey A, Vythelingum GN, GriseD, Goldstein LH. Verbal fluency and executive dysfunction inamyotrophic lateral sclerosis (ALS). Neuropsychologia 2000;38: 734–47.

13. Henry JD, Crawford JR, Phillips LH. Verbal fluency perfor-mance in dementia of the Alzheimer’s type: a meta-analysis.Neuropsychologia 2004; 42: 1212–22.

14. Folstein MF, Folstein SE, McHugh PR. ‘Mini-Mental State’. Apractical method for grading the cognitive state of patients forthe clinician. J Psychiatr Res 1975; 12: 189–98.

15. Manly JJ, Schupf N, Tang MX, Stern Y. Cognitive decline andliteracy among ethnically diverse elders. J Geriatr PsychiatryNeurol 2005; 18: 213–7.

16. Manly JJ, Jacobs DM, Touradji P, Small SA, Stern Y. Read-ing level attenuates differences in neuropsychological test per-formance between African American and White elders. J IntNeuropsychol Soc 2002; 8: 341–8.

17. Schmand B, Smit JH, Geerlings MI, Lindeboom J. The effectsof intelligence and education on the development of dementia.A test of the brain reserve hypothesis. Psychol Med 1997; 27:1337–44.

18. Barnes DE, Tager IB, Satariano WA, Yaffe K. The relation-ship between literacy and cognition in well-educated elders.J Gerontol A Biol Sci Med Sci 2004; 59: 390–5.

19. Wray LA, Lynch JW. Does cognitive functioning mediate thewell-documented link between education and functional dis-ability in middle-aged adults? Ann N Y Acad Sci 1999; 896:480–3.

674

at Universidade Federal do R

io Grande do N

orte on April 14, 2015

http://ageing.oxfordjournals.org/D

ownloaded from

Bone and mineral metabolism in older adults with Parkinson’s disease

20. DeSalvo KB, Fan VS, McDonell MB, Fihn SD. Predicting mor-tality and healthcare utilization with a single question. HealthServ Res 2005; 40: 1234–46.

21. Linn MW, Linn BS. The rapid disability rating scale-2. J AmGeriatr Soc 1982; 30: 378–82.

22. Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG,Kokmen E. Mild cognitive impairment: clinical characteriza-tion and outcome. Arch Neurol 1999; 56: 303–8.

23. Mitchell AJ. A meta-analysis of the accuracy of the Mini-MentalState Examination in the detection of dementia and mild cog-nitive impairment. J Psychiatr Res 2008; 43: 411–31.

24. Kim S, Egerter S, Cubbin C, Takahashi ER, Braveman P. Poten-tial implications of missing income data in population-basedsurveys: an example from a postpartum survey in California.Public Health Rep 2007; 122: 753–63.

25. Turrell G. Income non-reporting: implications for healthinequalities research. J Epidemiol Community Health 2000;54: 207–14.

26. Barzi F, Woodward M. Imputations of missing values in prac-tice: results from imputations of serum cholesterol in 28 cohortstudies. Am J Epidemiol 2004; 160: 34–45.

27. Acevedo A, Loewenstein DA, Barker WW et al. Category flu-ency test: normative data for English- and Spanish-speakingelderly. J Int Neuropsychol Soc 2000; 6: 760–9.

28. Tombaugh TN, Kozak J, Rees L. Normative data stratified byage and education for two measures of verbal fluency: FASand animal naming. Arch Clin Neuropsychol 1999; 14: 167–77.

29. Stricks L, Pittman J, Jacobs DM, Sano M, Stern Y. Norma-tive data for a brief neuropsychological battery administeredto English- and Spanish-speaking community-dwelling elders.J Int Neuropsychol Soc 1998; 4: 311–8.

30. Elias MF, Elias PK, D’Agostino RB, Silbershatz H, Wolf PA.Role of age, education, and gender on cognitive performance inthe Framingham Heart Study: community-based norms. ExpAging Res 1997; 23: 201–35.

Received 22 January 2009; accepted in revised form 14 May 2009

Age and Ageing 2009; 38: 675–680 C© The Author 2009. Published by Oxford University Press [on behalf of the British Geriatrics Society].All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.orgdoi: 10.1093/ageing/afp137

Published electronically 15 August 2009

Bone and mineral metabolism in older adultswith Parkinson’s disease

SUZAN ABOU-RAYA1, MADIHAH HELMII2, ANNA ABOU-RAYA3

1Geriatric Division, Department of Internal Medicine, Faculty of Medicine, University of Alexandria, Alexandria, Egypt2Department of Biochemistry, Medical Research Institute, Alexandria, Egypt3Rheumatology Division, Department of Internal Medicine, Faculty of Medicine, University of Alexandria, Alexandria, Egypt

Address correspondence to: S. Abou-Raya. Tel: (203) 5924601; Fax: (203) 5457611. Email: suzanraya@yahoo.com

Abstract

Objectives: Parkinson’s disease (PD) and osteoporosis are two common chronic disabling conditions in older adults thatadversely affect quality of life. The aim of the present work was to study the relationship between bone changes and PD.Methods: eighty-two patients with established PD aged 65 years or older and 68 age-, sex- and body mass index (BMI)-matchedhealthy control subjects were recruited. Exclusion criteria included other known causes of osteoporosis. Data including BMI,sunlight exposure, Hoehn and Yahr stage, disease duration and history of previous falls and/or fractures were collected. Bonemineral density was measured using dual energy x-ray absorptiometry. Sera were analysed for ionised calcium, vitamin D, bonealkaline phosphatase (BALP) and urinary N-terminal telopeptide of type I collagen (NTx). Physical and mental performancewas also assessed.Results: the findings show that the bone mineral density (BMD) of all PD patients was significantly lower compared tocontrols. PD patients had significantly decreased vitamin D levels, significantly increased BALP and NTx levels, reducedphysical and mental performance and more falls and/or fractures in comparison to healthy controls.Conclusion: PD is associated with an increased incidence of osteoporosis, falls and fractures. PD is thus a risk factor forosteoporosis and appropriate therapeutic interventions should be initiated to slow or prevent disability.

Keywords: Parkinson’s disease, osteoporosis, older adults, elderly

675

at Universidade Federal do R

io Grande do N

orte on April 14, 2015

http://ageing.oxfordjournals.org/D

ownloaded from