Poster Presentations: P3

30 faces will be chosen based on its level of difficulty and its ability to differentiate between the 2 groups of subjects. Specificity and sensitivity measures will be used in selecting the faces. A t-test will be used to compare
performance between the 2 groups on the final set of faces to ensure discriminant validity. Conclusions: Following the successful validation of the test,
it will be normed on the local population and then used as a clinical tool in
the assessment of patients with Alzheimers disease as an indicator of their
stage in the disease.

P3-260

THE EUROPEAN CROSS-CULTURAL

NEUROPSYCHOLOGICAL TEST BATTERY (CNTB)

T. Rune Nielsen1, Peter Bekkhus-Wetterberg2, Gunhild Waldemar1,

1
Copenhagen University Hospital, Copenhagen, Denmark; 2Oslo
University Hospital, Oslo, Norway. Contact e-mail: rune.nielsen@rh.
regionh.dk
Background: With the changing demographics of Europes elderly populations, the number of elderly with dementia is expected to increase considerably. Due to the specific history and conditions of some immigrant
populations, this increase may be more pronounced in these populations.
Moreover, validated neuropsychological tests for cross -cultural cognitive
assessment of the older persons belonging to ethnic and linguistic minorities
are generally lacking in Europe. The aim of this study was to examine the
cross-cultural validity of the European Cross-Cultural Neuropsychological
Test Battery (CNTB) in native Danes and immigrants from former Yugoslavia, Poland, Pakistan and Turkey. Methods: Between January 2012 and
January 2014, Turkish, Pakistani, former Yugoslavian and Polish immigrants as well as native Danes were recruited. All subjects underwent an
approximately one and a half hour assessment, including a structured demographic and health interview, measures of acculturation, screening for
depression and the CNTB. The CNTB consists of 12 neuropsychological
tests covering multiple cognitive domains: global cognitive function, memory, language, executive functions and visuospatial functions. All subjects
were assessed in their primary language. Subjects with conditions known
to affect neuropsychological test performance were subsequently excluded.
Results: Of the 200 subjects included in the study, 26 were native Danes and
17 were immigrants from former Yugoslavia, 33 from Poland, 47 from
Pakistan and 77 from Turkey. Significant age (F 6.705, p<0.001) and
educational (F 67.250, p<0.001) differences were present between the
groups. Although significant differences in test performances were found
between the groups on most cognitive measures, regression analyses revealed that this was due to educational and age differences rather than cultural background (explaining <3% of the variance in test performances).
Conclusions: The cross-cultural validity of the CNTB was found to be
excellent, but age and education have to be taken into consideration when
interpreting test scores. Future studies are needed to establish the clinical
utility and diagnostic accuracy of the CNTB.

P3-261

IMPROVING DETECTION OF DEMENTIA IN AN

ARABIC SPEAKING POPULATION WITH LOW
EDUCATION: COMBINING THE ROWLAND
UNIVERSAL DEMENTIA ASSESSMENT SCALE
AND THE INFORMANT QUESTIONNAIRE ON
COGNITIVE DECLINE IN THE ELDERLY

T. Rune Nielsen1, Thien Kieu Thi Phung1, Monique Chaaya2,

Samir Atweh2, Martin Prince3, Gunhild Waldemar1, 1Copenhagen
University Hospital, Copenhagen, Denmark; 2American University of
Beirut, Beirut, Lebanon; 3Kings College London, London, United
Kingdom. Contact e-mail: rune.nielsen@rh.regionh.dk
Background: Screening for dementia in Lebanon is challenging due to the
high rate of illiteracy among older people. Commonly used cognitive
screening tests, such as the Mini Mental State Examination, are rarely suitable as they usually require reading, writing, and arithmetic skills. The Row-

P727

land Universal Dementia Assessment Scale (RUDAS) and the Informant

Questionnaire on Cognitive Decline in the Elderly (IQCODE) have previously been successfully validated in this population. The aim of this study
was to evaluate whether there was any added value of combining the RUDAS and the IQCODE when screening for cognitive impairment among
Arabic speaking older persons with low education. Methods: Data from
225 subjects older than 65 years, 90 with mild to moderate dementia and
135 non-demented controls, were analyzed. The subjects were recruited
from social organizations, community-based primary care clinics, and hospital-based geriatric clinics. The diagnosis of dementia was made according
to the DSM-IV criteria. The RUDAS and the short form of the IQCODE
were used adopting previously established cut-off scores. Area under the
curve (AUC), sensitivity, specificity, positive and negative predictive values
(PPV and NPV) were obtained using three combination methods; the "And"
method, the "Or" method, and the "Weighted Sum" method. Results: The
IQCODE outperformed the RUDAS in all analyses. Logistic regression
demonstrated that the combination of the RUDAS and the IQCODE resulted
in better prediction of dementia than either measure alone. The "Weighted
Sum" method had the best AUC and specificity, while the "Or" rule had
the best sensitivity. A graphical method was developed to facilitate the adoption of the "Weighted Sum" method in everyday clinical practice. Conclusions: We recommend that the IQCODE is used as a supplement to formal
cognitive testing with the RUDAS in Arabic speaking older populations
with low education and that the "Weighted Sum" method is used to combine
the two measures. In order to generalize our results, this approach needs to
be further validated in populations with diverse cultural and educational
characteristics.

P3-262

ADJUSTMENT OF MOCA SCORES IN A SPANISHSPEAKING POPULATION WITH VARIED LEVELS

OF EDUCATION

Yan Zhou1, Freddy Ortiz2, Christopher Nu~nez3, David Elashoff4,

Ellen Woo2, Liana Apostolova4, Sheldon Wolf2, John Ringman4, 1Mary S.
Easton Center for Alzheimers Disease Research at UCLA, Los Angeles,
California, United States; 2Mary S. Easton Center for Alzheimers Disease
Research at UCLA, Los Angeles, California, United States; 3Keck School of
Medicine, University of Southern California Division of Research on
Children, Youth and Families; Department of Pediatrics Childrens Hospital
Los Angeles, Los Angeles, California, United States; 4UCLA, Los Angeles,
California, United States. Contact e-mail: YanZhou@mednet.ucla.edu
Background: Performance on the Montreal Cognitive Assessment (MoCA)
is known to be dependent on educational level. However, there is a dearth of
research on the utility of the MoCA in Spanish speakers with lower levels of
education. In such populations score adjustments may be necessary to increase the accuracy of interpretations. Methods: We studied 47 Latino participants enrolled in a study of cognitive aging who were given the MoCA
test in Spanish and had a cognitive diagnosis assigned by a consensus
team based on all available clinical information and neuropsychological
test performance (Table 1). The number of years of education ranged 020, and 29 participants (62%) had six or less years of education.We examined the pattern of diagnosis-adjusted MoCA residuals in relation to education, and proposed alternative education adjustments for the raw score of
MoCA. To compare different adjustments, we used bootstrap sampling
and examined two criteria: (1) standard deviation of the residuals and (2) association between education and the residuals. Sensitivity and specificity
analysis were also carried out for the raw score and each adjusted score.
Finally, we examined the reliability of the MoCA and performed item analyses to detect items that might be education biased. Results: After adjusting for the diagnosis, participants with fewer than six years of education
needed additional compensation in the MoCA score. We compared four
different adjustments using 1000 bootstrap samples. Results showed that,
with progressive compensation added for those with lower education, unexplained residuals decreased, and the education-residual association moved
to zero. These findings suggest that more compensation was necessary to