Asia will experience a surge in dementia prevalence within the next 20–40 years, but there is a dearth of well-normed neuropsychological tests that could assist with dementia diagnosis. Here, we report normative data for the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) in Elderly ethnic Chinese Singaporeans aged 55–91 years of age. A total of 1,165 male and female community-dwelling, cognitively normal elderly Chinese persons in Singapore, with varying levels of education and range of languages, were tested with the RBANS version A. The effects of age, education, language and gender on RBANS performance were examined. Negative effects of increased age and positive effects of education on the RBANS subtests, Index and Total Scale scores were found suggesting differential associations between age-related cognitive decline and education that vary according to the specific cognitive ability measured. The findings indicate that unique cultural and educational profile of elderly Chinese should be considered when applying the RBANS in this population.

Introduction

It is well known that individual differences, particularly those relating to age and education but also sociocultural status and gender, exert significant influences on performance on neuropsychological tests. Although both age and education are well-established correlates of performance, the relationship and the degree of their association vary depending on the sample characteristics and the test itself. The elderly, who are most at risk of degenerative neurological change, generally demonstrates the largest variation on neuropsychological tests due to the effects of age or disease-related neurological decline, and the independent influence of such individual differences.

Typically, advanced age is associated with decreased overall performance (Ardilla, 1998; Ardila & Roselli, 1989; Long & Klein, 1990; Mangione et al., 1993; Vanderploeg, Axelrod, Sherer, Scott, & Adams, 1997) particularly in mental flexibility and psychomotor speed (Sewell, Vigario, & Sano, 2010). A range of age-related processes may account for these impairments including reduced brain volume, loss of myelin integrity, cortical thinning, impairment in neurotransmitter binding and signaling, accumulation of neurofibrillary tangles, and altered concentrations of brain metabolites (Allen, Bruss, Brown, & Damasio, 2005; Del Arco et al., 2011; Erixon-Lindroth et al., 2005; Fotenos, Snyder, Girton, Morris, & Buckner, 2005; Hsu et al., 2008; Kadota, Horinouchi, & Kuroda, 2001; Kruggel, 2006; Magnotta et al., 1999; Pieperhoff et al., 2008).

In addition to advanced age, fewer years of formal education are also associated with poorer test performance (Ardila, Ostrosky-Solis, Rosselli, & Gomez, 2000; Belle et al., 1996; Brucki & Nitrini, 2007; Lee, Collinson, Feng, & Ng, 2012; Lim, Collinson, Feng, & Ng, 2010; Marcopulos, McLain, & Guiliano, 1997; Rosselli & Ardila, 1993) with language tests generally more sensitive to education variables. Low scores in neuropsychological tests may vary due to population and cohort differences (Manly, 2008), education techniques or policies (Ardila et al., 2000; Lam et al., 2013), lack of test familiarity or that testing itself represents a personally irrelevant situation (Ardila, 1998). Conversely, higher education is a recognized protective factor for age-related neurological disease. In a review paper of 22 cohort studies of the effects of education, occupation, premorbid IQ, and mental activities in incident dementia, 10 out of 15 studies demonstrated a significant protective effect of education (Valenzuela & Sachdev, 2005). Studies in patients with established dementia have also reported the protective effects of education in delaying cognitive deterioration (Hall et al., 2007; McDowell, Xi, Lindsay, & Tierney, 2007; Roe, Xiong, Miller, & Morris, 2007) as higher education early in life appears to make individuals more neurologically resistant to the pathological burden of dementia (Brayne et al., 2010; Wharton et al., 2009).

Asia will experience a surge in dementia prevalence in the next 20–40 years (Feng, Chiu, Chong, Yu, & Kua, 2011; Parry & Cui 2011) but presently there are relatively few reliable, internationally recognized and appropriately normed neuropsychological tests that can be applied in the aging populations of the region (Collinson & Yeo, 2010). In Asia and most developing countries, the prevalence of illiteracy is high and educational systems differ markedly. In Singapore, for example, a wealthy and economically advanced south east Asian country of 5 million people, 65.5% of residents aged 55 and above have less than 6 years of education, whereas only 17.2% have greater than 10 years of education (Department of Statistics, 2010). Even as the general literacy rate increased from 92.5% in 2000 to 95.9% in 2010, there remains a significant proportion of older individuals with very little or no formal education. This demographic is repeated across the expanse of Asia. Asians have been shown to perform more poorly than Caucasians on certain neuropsychological tests and better on others (Boone, Victor, Wen, Razani, & Pontón, 2007; Fuji, 2010; Hedden et al., 2002; Shan, Chen, Lee, & Su, 2008). As such, accounting for the influences of individual and sociocultural differences such as age and education is particularly important to Asian norming studies, particularly in those tests which relate to tests used in the elderly to detect early dementia.

The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) (Randolph, 1998) was developed for the specific purpose of identifying and characterizing abnormal cognitive decline in the elderly (Randolph, 1998) and its efficacy at detecting and characterizing dementia of different etiologies has been demonstrated (Randolph, Tierney, Mohr, & Chase, 1998). Studies have reported the influence of individual differences on RBANS (Beatty, Mold, & Gontkovsky, 2003; Gontkovsky, Beatty, & Mold, 2002). Of these, education was found to have the greatest influence, accounting for a statistically significant proportion of variance across the RBANS indices (Visuospatial/Construction and Delayed Memory indexes; all RBANS indices, Gontkovsky et al., 2002). Others have provided corrections for age and education (Duff et al., 2003) but few studies have examined these effects outside of the United States or in non-Western countries.

In our previous study (Lim et al., 2010), we found advanced age and lower education levels to be significantly associated with poorer subtest scores. Education was also a significant predictor of performance of four out of five RBANS indices as well as the Total Scale score. In this study, we extend this normative dataset in a larger sample of 1,165 elderly Chinese participants and further investigate the effects of age, education, language and gender on the performance of the RBANS.

Methods

Subjects in the present study participated in the Singapore Longitudinal Ageing Study (SLAS-II), an on-going population-based study of ageing and health of community-dwelling older adults living in the geographically defined areas in the Central and South-Western Regions of Singapore. Trained research nurses conducted recruitment via door-to-door visits and performed the RBANS overseen by research psychologists and the study administrators including SC, a qualified clinical neuropsychologist. The study was approved by the National University of Singapore Institutional Review Board. All participants provided written informed consent.

Participants were excluded if they had known or suspected history of neurological illness, psychiatric disturbance, substance abuse or any other condition that could compromise neurocognitive ability. Those with a Mini-Mental State Examination (MMSE) score below the cutoff of 26 and/or a Geriatric Depression Scale (GDS) score of >5 were also excluded. The mean MMSE score and GDS score for the current sample was 28.8 and 0.49, respectively. The small proportion of non-Chinese (Malays, Indians, and Others) was also excluded from analysis.

Procedure

The procedure for the present study is described in Lim and colleagues (2010). The RBANS (Form A) was administered to the participants as part of the battery of neuropsychological and clinical assessments of the SLAS. The RBANS was conducted in an individual setting and was administered in the language–dialect according to the participant's dominant or habitual preference by trained research nurses who are fluent in that particular language–dialect.

Using the forward–backward translation procedure, the RBANS (Repeatable Battery for the Assessment of Neuropsychological Status. Copyright © 1998. NCS Pearson, Inc. Reproduced with permission. All rights reserved. Mandarin Chinese translation copyright © 2012. NCS Pearson, Inc. Translated and reproduced with permission.) was translated into Mandarin and dialect by a committee of trained multilingual research psychologists. The nature of the translation procedure ensured that the phrasing of the test items was accurate and appropriate. As certain words were not translatable or meaningful in some dialects, appropriate alternatives that were comprehensible in all languages and dialects were chosen. Wherever possible the words chosen were as semantically and phonemically as close as possible to the original English word. In some cases, culturally foreign items were replaced with a local equivalent. Individual test performance was scored according to standardized instructions (Randolph, 1998) except for the Figure Copy and Figure Recall subtests, which were scored according to the modified criteria suggested by Duff and colleagues (2003). Quality control and inter-rater reliability was assessed for a portion of the test booklets and subtests that required interpretive scoring.

Statistical Analyses

Descriptive analyses were carried out using SPSS version 20.0 to derive the frequencies and cumulative percentages for the demographic characteristics of the sample. Participants were categorized according to their age (54–59 years old, 60–64 years old, 65–69 years old, 70–74 years old, and 75 years and above) and their education level (0 years, 1–3 years, 4–6 years, 7–10 years, and more than 10 years). The education level categories were based on the Singaporean educational system whereby primary school education typically spans 6 years. Pearson's correlation coefficients were calculated to determine the relationship between age, education and RBANS subtests, Index and Total Scale scores. Hierarchical regression analyses were conducted to examine the relative contributions of age and education to the prediction of the subtests, Index and Total Scale scores.

A three-way MANOVA was conducted to investigate the effects of age, education, and language administration. Univariate ANOVAs for each RBANS subtest and index were conducted as follow-up tests to the MANOVA with the Bonferroni method to control for Type 1 error rates for multiple comparisons. Post hoc analyses with Hochberg's GT2 or Games-Howell comparison procedures were then conducted to ascertain specific differences in RBANS performance between groups. Hochberg's GT2 procedure was selected for equal group variances but unequal sample sizes and Games-Howell procedure was selected for unequal group variances (Toothaker, 1993).

To establish norms, group means, and standard deviations for all 12 subtests raw scores were derived using descriptive analysis. These were stratified first by age and subsequently by education. Raw scores were then standardized to T scores with a mean of 100 and standard deviation of 15. Individual index scores were calculated by summing the T scores of the subtests that contribute to the Index. The Total Scale was derived by summing all the Index scores. The means and standard deviations, stratified by age and education, were also calculated for the Index and Total Scale scores.

Missing data were present in <5% of the sample. Missing value analysis was conducted to impute the missing data but it was found that the missing data were unsystematic, thus rendering the analysis inappropriate. Replacing the missing data with its mean did not result in any changes to the distribution of raw scores or group variances; hence, the missing data were retained for its contribution to the remaining subtests.

A number of the indices were significantly non-normal. Certain subtests were more normally distributed, such as List Learning, Story Memory, Semantic Fluency, and Coding. Subtests that were negatively skewed included Figure Copy, Picture Naming, List Recall, List Recognition, and Story Recall. However, subtests including Figure Copy, Picture Naming, and List Recognition are vulnerable to low ceiling effects; hence, these scores were not expected to be normally distributed. Transforming the subtest raw scores into T scores did not make the RBANS indices more normally distributed, statistically. However, some of the indices were more normally distributed and others, such as the Visuospatial/Constructional Index, whereas the Language and Delayed Memory indexes were more negatively skewed.

Results

The final sample consisted of 1,165 participants, who ranged in age from 55 to 91, with a mean of 65.6 years of which 443 were male and 722 were female. The mean number of years of education was 6.0 and 61.7% of the sample had 6 or fewer years of education, 15.1% had no formal education and 9.4% had >10 years of education. Table 1 presents the demographic characteristics stratified by the language of administration, either by their dominant language (English or Mandarin) or dialect (Hokkien, Teochew, and Cantonese) for the subtests. These sub-samples were not matched in terms of age or education.

Table 1.

Demographic characteristics of sample by dominant language (N = 1,160)

Dominant language English (n = 195) (%) Mandarin (n = 350) (%) Hokkien (n = 246) (%) Teochew (n = 93) (%) Cantonese (n = 276) (%) 
Gender 
 Male 48.2 33.7 35.8 28.0 41.3 
 Female 51.8 66.3 64.2 72.0 58.7 
Age 
 55–59 years old 32.8 25.7 15.4 18.3 14.5 
 60–64 years old 29.2 29.7 25.6 34.4 23.6 
 65–69 years old 21.0 23.7 26.0 19.4 23.2 
 70–74 years old 13.3 14.0 20.7 15.1 19.9 
 75 years old and above 3.6 6.9 12.2 12.9 18.8 
Years of education 
 0 year 1.0 7.7 31.3 21.5 17.8 
 1–3 years 3.6 10.6 20.3 20.4 20.7 
 4–6 years 9.2 40.0 31.3 39.8 35.9 
 7–10 years 60.5 30.6 11.8 12.9 19.2 
 >10 years 24.6 8.9 4.5 4.3 4.7 
Dominant language English (n = 195) (%) Mandarin (n = 350) (%) Hokkien (n = 246) (%) Teochew (n = 93) (%) Cantonese (n = 276) (%) 
Gender 
 Male 48.2 33.7 35.8 28.0 41.3 
 Female 51.8 66.3 64.2 72.0 58.7 
Age 
 55–59 years old 32.8 25.7 15.4 18.3 14.5 
 60–64 years old 29.2 29.7 25.6 34.4 23.6 
 65–69 years old 21.0 23.7 26.0 19.4 23.2 
 70–74 years old 13.3 14.0 20.7 15.1 19.9 
 75 years old and above 3.6 6.9 12.2 12.9 18.8 
Years of education 
 0 year 1.0 7.7 31.3 21.5 17.8 
 1–3 years 3.6 10.6 20.3 20.4 20.7 
 4–6 years 9.2 40.0 31.3 39.8 35.9 
 7–10 years 60.5 30.6 11.8 12.9 19.2 
 >10 years 24.6 8.9 4.5 4.3 4.7 

Note: Missing information for language of administration (n = 5).

Age and Education

Hierarchical regression analyses are presented in Table 2. Education was entered as the primary predictor and age as the secondary predictor. The results indicated that age and education significantly predicted RBANS performance but had differential contributions in its predictions to performance on the subtests and indices. Education was a significant predictor of 11 out of 12 RBANS subtests and all RBANS indices and the Total Scale and the primary predictor for the Story Memory, Figure Copy, Line Orientation, Picture Naming, Coding, Story Recall and Figure Recall subtests and the Immediate Memory, Visuospatial/Constructional and Attention indices, and Total Scale. Neither age nor education significantly predicted performance on the Digit Span subtest.

Table 2.

Hierarchical regression of age and education on RBANS subtests, index, and Total Scale scores

 Primary Predictor β p r2 Secondary predictor β p r2 change r2 
Subtests 
 List Learning Education .15 <.01 .022*** Age −.26 <.01 .063*** .084*** 
 Story Memory Education .31 <.01 .096*** Age −.14 <.01 .017*** .114*** 
 Figure Copy Education .30 <.01 .088*** Age −.11 <.01 .012*** .100*** 
 Line Orientation Education .39 <.01 .153*** Age −.17 <.01 .028*** .181*** 
 Picture Naming Education .09 .002  .008** Age −.04 .180  .002  .010** 
 Semantic Fluency Education .12 <.01 .015*** Age −.19 <.01 .035*** .049*** 
 Digit Span Education .02 .567  .000 Age −.02 .600  .000  .001 
 Coding Education .61 <.01 .366*** Age −.34 <.01 .107*** .474*** 
 List Recall Education .13 <.01 .016*** Age −.30 <.01 .079*** .095*** 
 List Recognition Education .08 <.01  .007** Age −.20 <.01 .036*** .043*** 
 Story Recall Education .33 <.01 .107*** Age −.19 <.01 .033*** .140*** 
 Figure Recall Education .31 <.01 .096*** Age −.19 <.01 .032*** .128*** 
Indexes 
 Immediate Memory Education .28 <.01 .075*** Age −.24 <.01 .054*** .128*** 
 Visuospatial/Constructional Education .42 <.01 .176*** Age −.18 <.01 .028*** .204*** 
 Language Education .15 <.01 .022*** Age −.16 <.01 .024*** .045*** 
 Attention Education .43 <.01 .183*** Age −.24 <.01 .053*** .235*** 
 Delayed Memory Education .25 <.01 .064*** Age −.29 <.01 .078*** .142*** 
 Total Scale Education .43 <.01  .188** Age −.32 <.01 .093*** .280*** 
 Primary Predictor β p r2 Secondary predictor β p r2 change r2 
Subtests 
 List Learning Education .15 <.01 .022*** Age −.26 <.01 .063*** .084*** 
 Story Memory Education .31 <.01 .096*** Age −.14 <.01 .017*** .114*** 
 Figure Copy Education .30 <.01 .088*** Age −.11 <.01 .012*** .100*** 
 Line Orientation Education .39 <.01 .153*** Age −.17 <.01 .028*** .181*** 
 Picture Naming Education .09 .002  .008** Age −.04 .180  .002  .010** 
 Semantic Fluency Education .12 <.01 .015*** Age −.19 <.01 .035*** .049*** 
 Digit Span Education .02 .567  .000 Age −.02 .600  .000  .001 
 Coding Education .61 <.01 .366*** Age −.34 <.01 .107*** .474*** 
 List Recall Education .13 <.01 .016*** Age −.30 <.01 .079*** .095*** 
 List Recognition Education .08 <.01  .007** Age −.20 <.01 .036*** .043*** 
 Story Recall Education .33 <.01 .107*** Age −.19 <.01 .033*** .140*** 
 Figure Recall Education .31 <.01 .096*** Age −.19 <.01 .032*** .128*** 
Indexes 
 Immediate Memory Education .28 <.01 .075*** Age −.24 <.01 .054*** .128*** 
 Visuospatial/Constructional Education .42 <.01 .176*** Age −.18 <.01 .028*** .204*** 
 Language Education .15 <.01 .022*** Age −.16 <.01 .024*** .045*** 
 Attention Education .43 <.01 .183*** Age −.24 <.01 .053*** .235*** 
 Delayed Memory Education .25 <.01 .064*** Age −.29 <.01 .078*** .142*** 
 Total Scale Education .43 <.01  .188** Age −.32 <.01 .093*** .280*** 

Notes: **p < .01. ***p < .001.

MANOVA analysis of group differences using Pillai's trace, indicated significant main effects of age, V = 0.13, F(48, 3936) = 2.71, p < .001, partial η2 = 0.03 and education, V = 0.25, F(60, 4925) = 4.29, p < .001, partial η2 = 0.05 for the combined RBANS subtests. Significant main effects of age, V = 0.06, F(20, 3964) = 2.75, p < .001, partial η2 = 0.01 and education, V = 0.16, F(25, 4960) = 6.41, p < .001, partial η2 = .03 were also found for combined RBANS indices. The age × education interaction term for the RBANS subtests, and for the RBANS indices were non-significant.

Univariate ANOVAs showed that the main effect of age was significant for the List Learning, Coding, List Recall and Figure Recall subtests, and all the RBANS indices, with the exception of the Visuospatial/Constructional and Language Indices (see Table 3). Bonferroni and Hochberg's GT2 post hoc comparisons showed that RBANS performance declined significantly with increasing age in this order: 54–59 years > 60–64 years > 65–69 years > 70–74 years > 75 years and above (all ps < .001). In addition, significant differences between participants in the 70- to 74-age group and the 75 years and above age group were found only for the Coding (p < .001) subtest. The main effect of education was significant for the Story Memory, Figure Copy, Line Orientation, Coding, Story Recall and Figure Recall subtests, and all the RBANS indices, except for the Language index. Post hoc Bonferroni and Hochberg's GT2 comparisons showed that RBANS performance improved with an increase in the number of years of education. Specifically, RBANS performance increase with an increase in the number of years of education in this order: 0 < 1–3 years < 4–6 years < 7–10 years < 10 years and above (all ps < .001). There were no significant age × education effects.

Table 3.

Analysis of variance—effects of age and education on performance of the RBANS subtests, index, and Total Scale scores

 Age
 
Education
 
Age × education
 
 F p Partial η2 F p Partial η2 F p Partial η2 
Subtests 
 List Learning 6.11 <.01 .024 3.05 .010 .015 0.81 .696 .015 
 Story Memory 1.95 .100 .008 8.10 <.01 .039 1.05 .400 .020 
 Figure Copy 1.86 .115 .007 6.00 <.01 .029 1.15 .296 .022 
 Line Orientation 2.64 .033 .011 16.02 <.01 .075 0.33 .997 .006 
 Picture Naming 0.44 .783 .002 1.60 .157 .008 0.78 .734 .015 
 Semantic Fluency 2.37 .051 .009 1.63 .150 .008 0.85 .647 .016 
 Digit Span 2.20 .067 .009 2.90 .013 .014 0.81 .693 .015 
 Coding 17.80 <.01 .067 37.7 <.01 .160 0.80 .714 .015 
 List Recall 8.49 <.01 .033 2.07 .067 .010 1.00 .499 .018 
 List Recognition 3.83 .004 .015 1.17 .323 .006 0.89 .604 .017 
 Story Recall 3.32 .010 .013 10.82 <.01 .052 1.00 .488 .018 
 Figure Recall 5.15 <.01 .020 5.72 <.01 .028 0.68 .844 .013 
Indexes 
 Immediate Memory 5.34 <.01 .021 7.68 <.01 .037 1.11 .333 .021 
 Visuospatial/Constructional 2.86 .023 .011 16.01 <.01 .075 0.68 .838 .013 
 Language 2.00 .093 .008 2.41 .035 .012 0.92 .557 .017 
 Attention 4.73 <.01 .019 21.26 <.01 .097 0.67 .851 .013 
 Delayed Memory 8.84 <.01 .034 4.81 <.01 .024 0.92 .559 .017 
 Total Scale 10.18 <.01 .039 19.23 <.01 .088 0.65 .876 .012 
 Age
 
Education
 
Age × education
 
 F p Partial η2 F p Partial η2 F p Partial η2 
Subtests 
 List Learning 6.11 <.01 .024 3.05 .010 .015 0.81 .696 .015 
 Story Memory 1.95 .100 .008 8.10 <.01 .039 1.05 .400 .020 
 Figure Copy 1.86 .115 .007 6.00 <.01 .029 1.15 .296 .022 
 Line Orientation 2.64 .033 .011 16.02 <.01 .075 0.33 .997 .006 
 Picture Naming 0.44 .783 .002 1.60 .157 .008 0.78 .734 .015 
 Semantic Fluency 2.37 .051 .009 1.63 .150 .008 0.85 .647 .016 
 Digit Span 2.20 .067 .009 2.90 .013 .014 0.81 .693 .015 
 Coding 17.80 <.01 .067 37.7 <.01 .160 0.80 .714 .015 
 List Recall 8.49 <.01 .033 2.07 .067 .010 1.00 .499 .018 
 List Recognition 3.83 .004 .015 1.17 .323 .006 0.89 .604 .017 
 Story Recall 3.32 .010 .013 10.82 <.01 .052 1.00 .488 .018 
 Figure Recall 5.15 <.01 .020 5.72 <.01 .028 0.68 .844 .013 
Indexes 
 Immediate Memory 5.34 <.01 .021 7.68 <.01 .037 1.11 .333 .021 
 Visuospatial/Constructional 2.86 .023 .011 16.01 <.01 .075 0.68 .838 .013 
 Language 2.00 .093 .008 2.41 .035 .012 0.92 .557 .017 
 Attention 4.73 <.01 .019 21.26 <.01 .097 0.67 .851 .013 
 Delayed Memory 8.84 <.01 .034 4.81 <.01 .024 0.92 .559 .017 
 Total Scale 10.18 <.01 .039 19.23 <.01 .088 0.65 .876 .012 

Post hoc comparisons across the RBANS subtests and Index and Total Scale scores among the three lowest education strata (0, 1–3 years, and 4–6 years of education) showed significant differences for education on the Figure Copy, Line Orientation and Coding subtests as well as the Visuospatial/Constructional, Attention and Total Scale indices (all ps < .001). There were also significant differences between participants with 0 years and 4–6 years of education for the Story Memory, Figure Copy, Line Orientation, Coding, Story Recall and Figure Recall subtests, and all RBANS indices, with the exception of the Language index (all ps < .001). Significant differences between participants with 1–3 and 4–6 years of education were observed for the Coding (p < .001) and Attention (p < .001) index. Among the higher education stratum, there were also significant differences between participants with 7–10 years and >10 years of education for the Coding (p < .001) subtest.

Decline in performance was analyzed in the different educational groups. Based on Bonferroni and Hochberg's GT2 post hoc comparisons, two distinct groups were identified: low education (0 years, 1–3 years and 4–6 years) and high education (7–10 years and > 10 years). Mean levels of education for the low- and high-educated groups were ∼3 and 10 years, respectively. Performance between these two groups was compared across the different age groups to determine the patterns of decline in RBANS performance. Adopting a similar method as described by Ardila and colleagues (2000), performance of each RBANS subtest at the youngest age group (54–59 years) was taken as 100%. Percentages of performance based on this reference group were calculated for each age group and education group for each RBANS subtest. Subsequently, the decline in performance across age groups was analyzed. Table 4 presents decline for each RBANS subtest. Two main patterns (parallelism and protection) were identified, while confluence was observed for two subtests.

Table 4.

Different patterns of age-related cognitive decline across different educational groups

Subtests Age bands
 
 54–59 years 60–64 years 65–69 years 70–74 years 75 years and above 
Parallelism profiles 
List Learning      
 Low education 28.66 (100%) 27.12 (95%) 27.09 (95%) 26.07 (91%) 24.52 (86%) 
 High education 29.16 (100%) 28.64 (98%) 27.52 (94%) 25.87 (89%) 25.87 (89%) 
Picture Naming 
 Low education 9.92 (100%) 9.90 (100%) 9.92 (100%) 9.84 (99%) 9.85 (99%) 
 High education 9.91 (100%) 9.96 (101%) 9.92 (100%) 9.81 (99%) 10.00 (101%) 
Semantic Fluency 
 No or low education 17.35 (100%) 16.44 (95%) 16.25 (94%) 15.66 (90%) 15.22 (88%) 
 High education 17.65 (100%) 17.33 (98%) 16.22 (92%) 15.50 (88%) 16.20 (92%) 
Coding 
 Low education 34.20 (100%) 28.76 (84%) 23.89 (70%) 20.71 (61%) 18.67 (55%) 
 High education 44.93 (100%) 41.40 (92%) 36.38 (81%) 34.06 (76%) 25.93 (58%) 
List Recall 
 Low education 7.67 (100%) 6.97 (91%) 6.55 (85%) 5.87 (77%) 5.64 (74%) 
 High education 7.62 (100%) 7.45 (98%) 6.70 (88%) 6.28 (82%) 5.80 (76%) 
List Recognition 
 Low education 19.69 (100%) 19.63 (100%) 19.49 (99%) 19.29 (98%) 19.14 (97%) 
 High education 19.71 (100%) 19.77 (100%) 19.45 (99%) 19.28 (98%) 19.47 (99%) 
Protection profiles 
Line Orientation 
 Low education 15.65 (100%) 14.58 (93%) 14.13 (90%) 13.30 (85%) 13.18 (84%) 
 High education 17.29 (100%) 16.71 (97%) 16.23 (94%) 15.64 (90%) 16.73 (97%) 
Figure Copy 
 Low education 17.32 (100%) 16.91 (98%) 16.69 (96%) 16.37 (95%) 15.85 (92%) 
 High education 18.26 (100%) 18.15 (99%) 17.78 (97%) 18.00 (99%) 17.60 (96%) 
Story Recall 
 Low education 9.44 (100%) 9.04 (96%) 8.62 (91%) 8.60 (91%) 7.81 (83%) 
 High education 10.07 (100%) 9.93 (99%) 9.90 (98%) 9.39 (93%) 7.93 (79%) 
Figure Recall 
 Low education 14.75 (100%) 13.91 (94%) 13.45 (91%) 12.70 (86%) 12.22 (82%) 
 High education 15.96 (100%) 15.66 (98%) 14.97 (94%) 15.19 (95%) 14.67 (92%) 
Confluence Downwards 
Story Memory 
 Low education 16.47 (100%) 16.59 (101%) 15.91 (97%) 15.66 (95%) 14.83 (90%) 
 High education 18.22 (100%) 18.17 (100%) 18.35 (101%) 16.51 (91%) 15.47 (85%) 
Confluence Upwards 
Digit Span 
 Low education 13.87 (100%) 13.77 (99%) 13.86 (100%) 13.46 (97%) 13.84 (100%) 
 High education 13.74 (100%) 13.47 (98%) 13.37 (97%) 12.94 (94%) 13.93 (101%) 
Subtests Age bands
 
 54–59 years 60–64 years 65–69 years 70–74 years 75 years and above 
Parallelism profiles 
List Learning      
 Low education 28.66 (100%) 27.12 (95%) 27.09 (95%) 26.07 (91%) 24.52 (86%) 
 High education 29.16 (100%) 28.64 (98%) 27.52 (94%) 25.87 (89%) 25.87 (89%) 
Picture Naming 
 Low education 9.92 (100%) 9.90 (100%) 9.92 (100%) 9.84 (99%) 9.85 (99%) 
 High education 9.91 (100%) 9.96 (101%) 9.92 (100%) 9.81 (99%) 10.00 (101%) 
Semantic Fluency 
 No or low education 17.35 (100%) 16.44 (95%) 16.25 (94%) 15.66 (90%) 15.22 (88%) 
 High education 17.65 (100%) 17.33 (98%) 16.22 (92%) 15.50 (88%) 16.20 (92%) 
Coding 
 Low education 34.20 (100%) 28.76 (84%) 23.89 (70%) 20.71 (61%) 18.67 (55%) 
 High education 44.93 (100%) 41.40 (92%) 36.38 (81%) 34.06 (76%) 25.93 (58%) 
List Recall 
 Low education 7.67 (100%) 6.97 (91%) 6.55 (85%) 5.87 (77%) 5.64 (74%) 
 High education 7.62 (100%) 7.45 (98%) 6.70 (88%) 6.28 (82%) 5.80 (76%) 
List Recognition 
 Low education 19.69 (100%) 19.63 (100%) 19.49 (99%) 19.29 (98%) 19.14 (97%) 
 High education 19.71 (100%) 19.77 (100%) 19.45 (99%) 19.28 (98%) 19.47 (99%) 
Protection profiles 
Line Orientation 
 Low education 15.65 (100%) 14.58 (93%) 14.13 (90%) 13.30 (85%) 13.18 (84%) 
 High education 17.29 (100%) 16.71 (97%) 16.23 (94%) 15.64 (90%) 16.73 (97%) 
Figure Copy 
 Low education 17.32 (100%) 16.91 (98%) 16.69 (96%) 16.37 (95%) 15.85 (92%) 
 High education 18.26 (100%) 18.15 (99%) 17.78 (97%) 18.00 (99%) 17.60 (96%) 
Story Recall 
 Low education 9.44 (100%) 9.04 (96%) 8.62 (91%) 8.60 (91%) 7.81 (83%) 
 High education 10.07 (100%) 9.93 (99%) 9.90 (98%) 9.39 (93%) 7.93 (79%) 
Figure Recall 
 Low education 14.75 (100%) 13.91 (94%) 13.45 (91%) 12.70 (86%) 12.22 (82%) 
 High education 15.96 (100%) 15.66 (98%) 14.97 (94%) 15.19 (95%) 14.67 (92%) 
Confluence Downwards 
Story Memory 
 Low education 16.47 (100%) 16.59 (101%) 15.91 (97%) 15.66 (95%) 14.83 (90%) 
 High education 18.22 (100%) 18.17 (100%) 18.35 (101%) 16.51 (91%) 15.47 (85%) 
Confluence Upwards 
Digit Span 
 Low education 13.87 (100%) 13.77 (99%) 13.86 (100%) 13.46 (97%) 13.84 (100%) 
 High education 13.74 (100%) 13.47 (98%) 13.37 (97%) 12.94 (94%) 13.93 (101%) 

Notes: In each case, performance at 54–59 years is taken as 100%.

Parallelism

Age-related cognitive decline occurring in a parallel fashion for the different education groups. In some instances, scores remain relatively stable across age ranges and in others they decrease in a parallel manner between the high- and low-education groups. Examples of parallelism are observed in the List Learning, Picture Naming, Semantic Fluency, Coding, List Recall, and List Recognition subtests.

Protection

Age-related cognitive decline is attenuated in well-educated participants. Examples of protection are observed in the Line Orientation, Figure Copy, Story Recall, and Figure Recall subtests. For the Line Orientation subtest, performance at the age of 70–74 years was 95% of performance observed at 54–59 years in the high-education group, whereas in the low-education group, this was only 86%.

Confluence downwards

Scores in the groups converge across ages, following a decrease in scores in the high-education group. Confluence downwards was observed in the Story Memory subtest. Scores in the low-education group were lower and remained low across age groups. In the high-education group, scores were initially high but they decrease across the age groups.

Confluence upwards

Scores in the two groups converge across ages, due to an increase in scores in the low-education group. Confluence upwards was observed on the Digit Span subtest. Scores for the low-education group increased across ages, while for the high-education group, scores mildly decreased.

Language

There was a significant main effect of language administration on the combined subtests, V = 0.09, F(48, 3936) = 1.85, p < .001, partial η2 = 0.02 and indices, V = 0.07, F(20, 3936) = 3.33, p < .001, partial η2 = 0.02. The language × age interaction effect was not significant for the subtests, V = 0.02, F(192, 11904) = 1.10, p = .157, partial η2 = 0.02 but was significant for the indexes, V = 0.11, F(80, 4960) = 1.33, p < .05, partial η2 = .02. There was also a significant language × education interaction effect for the subtests, V = 0.28, F(240, 11904) = 1.19, p < .05, partial η2 = 0.02 but was non-significant for the indexes, V = 0.12, F(100, 4960) = 1.20, p = .086, partial η2 = 0.02. The three-way interaction effect, language × age × education was not significant for both the RBANS subtests, V = 0.59, F(636, 11904) = 0.97, p = .141, partial η2 = 0.05 and the indices, V = 0.28, F(265, 4960) = 1.10, p = .157, partial η2 = 0.06. Separate univariate ANOVAs for the main and interaction effects of age, education, and language are presented in Table 5.

Table 5.

Means and standard deviations for RBANS subtests raw scores and index scores stratified by age (N = 1,165)

Subtests 54–59 years 60–64 years 65–69 years 70–74 years 75 years and above 
 (n = 250) (n = 323) (n = 270) (n = 197) (n = 125) 
List Learning 29.0 (3.9) 27.8 (4.3) 27.2 (4.4) 26.0 (4.3) 24.7 (4.7) 
Story Memory 17.4 (3.4) 17.3 (3.6) 16.8 (3.5) 15.9 (3.6) 14.9 (3.2) 
Figure Copy 17.9 (2.2) 17.5 (2.6) 17.1 (2.7) 16.8 (2.6) 16.1 (3.2) 
Line Orientation 16.5 (2.9) 15.5 (3.5) 14.9 (3.3) 14.0 (3.4) 13.6 (3.9) 
Picture Naming 9.9 (0.3) 9.9 (0.3) 9.9 (0.3) 9.8 (0.4) 9.9 (0.4) 
Semantic Fluency 17.5 (3.1) 16.8 (3.3) 16.2 (3.6) 15.6 (3.0) 15.3 (3.3) 
Digit Span 13.8 (2.3) 13.7 (2.1) 13.7 (2.2) 13.3 (2.1) 13.9 (2.1) 
Coding 39.8 (12.0) 34.2 (12.1) 28.3 (11.5) 24.5 (11.3) 19.5 (10.2) 
List Recall 7.7 (1.8) 7.2 (2.1) 6.6 (2.2) 6.0 (2.5) 5.7 (2.3) 
List Recognition 19.7 (0.6) 19.7 (0.7) 19.5 (0.9) 19.3 (0.9) 19.2 (1.1) 
Story Recall 9.8 (1.9) 9.4 (1.8) 9.1 (2.1) 8.8 (2.0) 7.8 (2.1) 
Figure Recall 15.4 (2.9) 14.7 (3.3) 14.0 (3.2) 13.4 (3.5) 12.5 (3.3) 
Index Scores 
 Immediate Memory 104.3 (11.4) 102.1 (11.9) 100.1 (12.2) 96.3 (12.2) 91.8 (11.7) 
 Visuospatial/Constructional 104.8 (9.9) 101.6 (11.9) 99.2 (11.6) 96.4 (12.0) 93.5 (14.9) 
 Language 102.7 (10.2) 101.3 (10.4) 100.1 (10.5) 96.7 (12.0) 96.6 (11.8) 
 Attention 105.6 (10.7) 101.9 (10.5) 98.7 (9.9) 95.3 (9.8) 94.5 (10.3) 
 Delayed Memory 105.1 (9.6) 102.9 (10.8) 98.9 (12.2) 95.6 (13.1) 91.5 (14.4) 
 Total Scale 104.6 (6.9) 102.0 (7.5) 99.3 (7.4) 96.1 (7.9) 93.8 (8.2) 
Subtests 54–59 years 60–64 years 65–69 years 70–74 years 75 years and above 
 (n = 250) (n = 323) (n = 270) (n = 197) (n = 125) 
List Learning 29.0 (3.9) 27.8 (4.3) 27.2 (4.4) 26.0 (4.3) 24.7 (4.7) 
Story Memory 17.4 (3.4) 17.3 (3.6) 16.8 (3.5) 15.9 (3.6) 14.9 (3.2) 
Figure Copy 17.9 (2.2) 17.5 (2.6) 17.1 (2.7) 16.8 (2.6) 16.1 (3.2) 
Line Orientation 16.5 (2.9) 15.5 (3.5) 14.9 (3.3) 14.0 (3.4) 13.6 (3.9) 
Picture Naming 9.9 (0.3) 9.9 (0.3) 9.9 (0.3) 9.8 (0.4) 9.9 (0.4) 
Semantic Fluency 17.5 (3.1) 16.8 (3.3) 16.2 (3.6) 15.6 (3.0) 15.3 (3.3) 
Digit Span 13.8 (2.3) 13.7 (2.1) 13.7 (2.2) 13.3 (2.1) 13.9 (2.1) 
Coding 39.8 (12.0) 34.2 (12.1) 28.3 (11.5) 24.5 (11.3) 19.5 (10.2) 
List Recall 7.7 (1.8) 7.2 (2.1) 6.6 (2.2) 6.0 (2.5) 5.7 (2.3) 
List Recognition 19.7 (0.6) 19.7 (0.7) 19.5 (0.9) 19.3 (0.9) 19.2 (1.1) 
Story Recall 9.8 (1.9) 9.4 (1.8) 9.1 (2.1) 8.8 (2.0) 7.8 (2.1) 
Figure Recall 15.4 (2.9) 14.7 (3.3) 14.0 (3.2) 13.4 (3.5) 12.5 (3.3) 
Index Scores 
 Immediate Memory 104.3 (11.4) 102.1 (11.9) 100.1 (12.2) 96.3 (12.2) 91.8 (11.7) 
 Visuospatial/Constructional 104.8 (9.9) 101.6 (11.9) 99.2 (11.6) 96.4 (12.0) 93.5 (14.9) 
 Language 102.7 (10.2) 101.3 (10.4) 100.1 (10.5) 96.7 (12.0) 96.6 (11.8) 
 Attention 105.6 (10.7) 101.9 (10.5) 98.7 (9.9) 95.3 (9.8) 94.5 (10.3) 
 Delayed Memory 105.1 (9.6) 102.9 (10.8) 98.9 (12.2) 95.6 (13.1) 91.5 (14.4) 
 Total Scale 104.6 (6.9) 102.0 (7.5) 99.3 (7.4) 96.1 (7.9) 93.8 (8.2) 

Univariate ANOVAs showed that the main effect of language was significant only for the Digit Span subtest and the Visuospatial/Constructional, Attention and Total Scale indexes. Post hoc Hochberg's GT2 and Games-Howell comparisons revealed that the English-speaking group performed significantly better than the dialects (Hokkien, Teochew, and Cantonese)-speaking groups on the RBANS indices identified to be significant by the univariate ANOVAs (all ps < .001). The Mandarin-speaking group performed significantly better than the dialect-speaking groups (all ps < .001). Among the dialect-speaking groups, the Cantonese-speaking group outperformed the Hokkien- and Teochew-speaking groups (all ps < .001). RBANS performance was not significantly different between the Hokkien- and Teochew-speaking groups. The reverse of this general trend was observed to be true for the Digit Span subtest.

Based on the univariate ANOVAS, the language × education interaction effect was significant for List Learning, F(20, 992) = 2.26, p = .001, partial η2 = 0.04 subtest. Post hoc comparisons showed that among participants with 4–6 years of education, the Mandarin-speaking group performed better than the Hokkien-speaking group (p < .01).

Gender

MANCOVA was conducted to investigate gender effects after controlling for the effects of age and education. A significant effect of gender on the performance for the RBANS subtests, V = 0.16, F(12, 1104) = 17.76, p < .001, partial η2 = 0.16 and for the RBANS indices, V = 0.08, F(5, 1111) = 19.63, p < .001, partial η2 = 0.08 was found.

Separate univariate ANOVAs showed gender was significant for most subtests, except for Story Memory, Figure Copy and Story Recall and for all RBANS indexes, except for the Total Scale. Bonferroni comparisons showed that the males performed significantly better on Story Memory, Figure Copy, Line Orientation, Picture Naming, Digit Span, Coding, Story Recall, and Figure Recall. Consistent with their performance on the subtests, males significantly outperformed females on the Visuospatial/Constructional and Attention indices. On the other hand, the females were better on the List Learning, Semantic Fluency, List Recall, and List Recognition subtests. Similarly, females outperformed males on the Immediate Memory, Language, and Delayed Memory indices.

Means and Standard Deviations for RBANS Subtests, Index, and Total Scale

Table 5 presents the means and standard deviations for the 12 RBANS subtest raw scores stratified by age bands. Table 6 presents these raw scores stratified by age and education. As illustrated in Table 5, mean performances decreased with increasing age across all the RBANS subtests, however, when the means and standard deviations were further stratified by education, the underlying protection effect was evident. Highly educated elderly performed better than their younger counterparts who had less education. The means and standard deviations for the Index and Total Scale scores, by age and education are displayed in Table 7. Education was stratified into bands based on the post hoc comparisons. As significant differences in mean performances across the RBANS subtests were found among the three lowest education strata (0 years, 1–3 years and 4–6 years), these education bands were retained.

Table 6.

Means and standard deviations for RBANS subtests raw scores stratified by age and education (N = 1,147)

Subtests 54–59 years
 
60–64 years
 
 0 Edu 1–3 Edu 4–6 Edu 7–10 Edu >10 Edu 0 Edu 1–3 Edu 4–6 Edu 7–10 Edu >10 Edu 
List Learning 29.2 (3.7) 28.3 (3.9) 28.7 (3.7) 29.2 (3.8) 29.2 (4.7) 26.7 (4.2) 27.2 (3.6) 27.2 (4.6) 28.6 (4.1) 28.9 (3.9) 
Story Memory 16.8 (1.8) 15.4 (3.2) 16.7 (3.3) 18.0 (3.4) 18.9 (3.2) 15.6 (3.9) 16.7 (3.8) 16.8 (3.6) 17.8 (3.0) 19.5 (3.1) 
Figure Copy 15.6 (3.7) 17.2 (2.5) 17.6 (1.9) 18.2 (1.9) 18.4 (2.2) 15.9 (3.3) 16.7 (2.6) 17.3 (2.8) 18.1 (1.9) 18.4 (1.6) 
Line Orientation 14.4 (4.3) 14.8 (3.2) 16.1 (2.7) 17.1 (2.4) 17.9 (2.3) 12.8 (2.9) 14.1 (4.1) 15.3 (3.2) 16.5 (3.2) 17.6 (2.6) 
Picture Naming 10.0 (0.0) 9.8 (0.5) 9.9 (0.2) 9.9 (0.3) 9.9 (0.2) 9.9 (0.3) 9.9 (0.4) 9.9 (0.4) 10.0 (0.2) 10.0 (0.2) 
Semantic Fluency 16.6 (2.8) 16.8 (3.2) 17.7 (3.0) 17.4 (3.1) 18.4 (3.7) 15.9 (3.2) 16.8 (2.4) 16.5 (3.4) 17.1 (3.3) 18.7 (3.4) 
Digit Span 12.9 (2.3) 13.7 (2.6) 14.1 (1.9) 13.8 (2.3) 13.5 (2.9) 13.4 (2.1) 13.5 (2.0) 43.0 (1.9) 13.4 (2.3) 13.7 (2.4) 
Coding 25.5 (9.9) 31.2 (13.4) 36.4 (8.4) 44.1 (11.0) 47.3 (11.8) 17.6 (8.6) 27.2 (13.0) 32.3 (8.1) 39.5 (8.8) 48.0 (10.7) 
List Recall 8.8 (1.5) 7.2 (2.1) 7.7 (1.8) 7.6 (1.7) 7.6 (1.7) 7.2 (1.6) 7.1 (2.0) 6.9 (2.2) 7.4 (2.1) 7.7 (1.8) 
List Recognition 19.8 (0.6) 19.7 (0.8) 19.7 (0.7) 19.7 (0.7) 19.7 (0.6) 19.7 (0.6) 19.7 (0.7) 19.6 (0.9) 19.8 (0.6) 19.9 (0.4) 
Story Recall 8.9 (1.7) 9.2 (1.7) 9.6 (1.9) 9.9 (2.0) 10.6 (1.5) 8.5 (1.7) 8.7 (1.9) 9.3 (1.8) 9.8 (1.5) 10.5 (1.5) 
Figure Recall 14.4 (4.1) 14.1 (3.1) 15.0 (2.7) 15.7 (2.8) 16.8 (2.4) 12.7 (3.7) 13.5 (3.4) 14.4 (3.4) 15.6 (2.6) 15.9 (2.7) 
 n = 12 n = 24 n = 79 n = 96 n = 33 n = 30 n = 39 n = 112 n = 106 n = 31 
 65–69 years
 
70–74 years
 
 0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
 26.3 (4.7) 26.9 (4.1) 27.6 (4.4) 27.1 (4.5) 28.6 (4.1) 25.9 (4.4) 26.6 (4.2) 25.9 (4.4) 25.6 (4.6) 26.4 (4.2) 
 14.3 (3.1) 15.6 (3.7) 16.2 (3.3) 18.0 (2.8) 19.3 (2.8) 15.2 (3.7) 15.5 (2.9) 16.3 (3.3) 15.7 (4.1) 18.2 (2.9) 
 15.6 (3.3) 17.1 (2.2) 17.0 (2.9) 17.6 (2.6) 18.3 (1.7) 15.7 (2.7) 16.6 (2.5) 16.9 (2.5) 17.6 (2.5) 18.8 (1.8) 
 12.9 (3.7) 14.3 (2.9) 14.6 (3.3) 16.1 (2.6) 16.7 (3.3) 12.2 (3.3) 13.4 (3.5) 14.2 (3.3) 15.6 (2.3) 15.7 (3.3) 
 9.8 (0.5) 9.9 (0.2) 10.0 (0.2) 9.9 (0.3) 10.0 (0.2) 9.7 (0.6) 9.9 (0.3) 9.9 (0.3) 9.8 (0.4) 9.9 (0.3) 
 16.9 (3.6) 16.0 (3.3) 16.1 (3.8) 16.5 (3.3) 15.2 (4.3) 15.1 (2.6) 15.7 (2.5) 16.1 (3.1) 15.1 (3.4) 16.4 (3.4) 
 13.7 (1.9) 14.0 (2.0) 13.9 (2.1) 13.2 (2.6) 13.8 (2.3) 13.0 (2.2) 13.4 (1.7) 14.0 (1.9) 12.7 (2.4) 13.4 (2.3) 
 16.1 (7.5) 23.9 (10.5) 27.5 (7.6) 34.7 (10.6) 41.1 (8.4) 13.7 (6.6) 22.1 (8.6) 26.0 (9.2) 32.7 (9.6) 37.0 (7.4) 
 6.1 (2.5) 6.6 (2.1) 6.7 (2.1) 6.6 (2.3) 6.9 (2.2) 5.9 (2.7) 6.1 (2.5) 5.7 (2.5) 6.0 (2.7) 6.9 (1.6) 
 19.4 (0.9) 19.4 (1.1) 19.6 (0.8) 19.5 (0.9) 19.5 (0.9) 19.3 (0.8) 19.3 (1.0) 19.2 (1.0) 19.2 (1.0) 19.4 (1.1) 
 7.4 (2.1) 8.4 (2.1) 9.3 (1.8) 9.8 (1.8) 10.1 (1.4) 8.2 (1.9) 8.6 (1.9) 9.0 (1.9) 9.0 (2.2) 10.2 (1.8) 
 12.5 (3.6) 13.5 (3.1) 13.9 (3.0) 14.8 (2.9) 15.6 (3.0) 11.9 (3.5) 13.1 (3.4) 13.1 (3.3) 14.8 (3.3) 15.9 (2.4) 
 n = 38 n = 53 n = 82 n = 69 n = 24 n = 50 n = 35 n = 55 n = 37 n = 17 
 75 years and above
 
     
 0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
     
 24.6 (4.8) 24.7 (5.2) 24.4 (4.9) 25.9 (2.4) 25.8 (5.7)      
 13.9 (3.2) 15.7 (2.5) 15.4 (3.3) 15.1 (3.4) 16.5 (4.7)      
 14.4 (3.0) 17.0 (2.6) 16.8 (3.0) 17.0 (3.0) 19.3 (1.0)      
 11.9 (3.8) 13.5 (4.1) 14.4 (3.3) 16.6 (2.7) 17.3 (2.5)      
 9.8 (0.4) 9.9 (0.3) 9.9 (0.4) 10.0 (0.0) 10.0 (0.0)      
 15.2 (3.5) 15.0 (3.6) 15.3 (3.4) 16.6 (2.4) 15.3 (2.2)      
 13.7 (1.9) 14.1 (2.4) 13.9 (2.1) 14.3 (2.1) 13.0 (3.6)      
 12.2 (6.7) 22.6 (11.0) 23.7 (8.9) 24.3 (10.0) 30.0 (6.8)      
 5.6 (2.2) 5.2 (2.6) 5.9 (2.3) 5.6 (1.8) 6.5 (2.6)      
 19.3 (1.0) 19.2 (1.2) 19.0 (1.3) 19.4 (0.7) 19.8 (0.5)      
 7.3 (2.2) 8.2 (1.9) 8.1 (1.9) 7.9 (2.2) 8.0 (3.5)      
 11.6 (3.2) 13.0 (3.7) 12.5 (3.0) 14.5 (3.1) 15.3 (2.8)      
 n = 46 n = 20 n = 44 n = 11 n = 4      
Subtests 54–59 years
 
60–64 years
 
 0 Edu 1–3 Edu 4–6 Edu 7–10 Edu >10 Edu 0 Edu 1–3 Edu 4–6 Edu 7–10 Edu >10 Edu 
List Learning 29.2 (3.7) 28.3 (3.9) 28.7 (3.7) 29.2 (3.8) 29.2 (4.7) 26.7 (4.2) 27.2 (3.6) 27.2 (4.6) 28.6 (4.1) 28.9 (3.9) 
Story Memory 16.8 (1.8) 15.4 (3.2) 16.7 (3.3) 18.0 (3.4) 18.9 (3.2) 15.6 (3.9) 16.7 (3.8) 16.8 (3.6) 17.8 (3.0) 19.5 (3.1) 
Figure Copy 15.6 (3.7) 17.2 (2.5) 17.6 (1.9) 18.2 (1.9) 18.4 (2.2) 15.9 (3.3) 16.7 (2.6) 17.3 (2.8) 18.1 (1.9) 18.4 (1.6) 
Line Orientation 14.4 (4.3) 14.8 (3.2) 16.1 (2.7) 17.1 (2.4) 17.9 (2.3) 12.8 (2.9) 14.1 (4.1) 15.3 (3.2) 16.5 (3.2) 17.6 (2.6) 
Picture Naming 10.0 (0.0) 9.8 (0.5) 9.9 (0.2) 9.9 (0.3) 9.9 (0.2) 9.9 (0.3) 9.9 (0.4) 9.9 (0.4) 10.0 (0.2) 10.0 (0.2) 
Semantic Fluency 16.6 (2.8) 16.8 (3.2) 17.7 (3.0) 17.4 (3.1) 18.4 (3.7) 15.9 (3.2) 16.8 (2.4) 16.5 (3.4) 17.1 (3.3) 18.7 (3.4) 
Digit Span 12.9 (2.3) 13.7 (2.6) 14.1 (1.9) 13.8 (2.3) 13.5 (2.9) 13.4 (2.1) 13.5 (2.0) 43.0 (1.9) 13.4 (2.3) 13.7 (2.4) 
Coding 25.5 (9.9) 31.2 (13.4) 36.4 (8.4) 44.1 (11.0) 47.3 (11.8) 17.6 (8.6) 27.2 (13.0) 32.3 (8.1) 39.5 (8.8) 48.0 (10.7) 
List Recall 8.8 (1.5) 7.2 (2.1) 7.7 (1.8) 7.6 (1.7) 7.6 (1.7) 7.2 (1.6) 7.1 (2.0) 6.9 (2.2) 7.4 (2.1) 7.7 (1.8) 
List Recognition 19.8 (0.6) 19.7 (0.8) 19.7 (0.7) 19.7 (0.7) 19.7 (0.6) 19.7 (0.6) 19.7 (0.7) 19.6 (0.9) 19.8 (0.6) 19.9 (0.4) 
Story Recall 8.9 (1.7) 9.2 (1.7) 9.6 (1.9) 9.9 (2.0) 10.6 (1.5) 8.5 (1.7) 8.7 (1.9) 9.3 (1.8) 9.8 (1.5) 10.5 (1.5) 
Figure Recall 14.4 (4.1) 14.1 (3.1) 15.0 (2.7) 15.7 (2.8) 16.8 (2.4) 12.7 (3.7) 13.5 (3.4) 14.4 (3.4) 15.6 (2.6) 15.9 (2.7) 
 n = 12 n = 24 n = 79 n = 96 n = 33 n = 30 n = 39 n = 112 n = 106 n = 31 
 65–69 years
 
70–74 years
 
 0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
 26.3 (4.7) 26.9 (4.1) 27.6 (4.4) 27.1 (4.5) 28.6 (4.1) 25.9 (4.4) 26.6 (4.2) 25.9 (4.4) 25.6 (4.6) 26.4 (4.2) 
 14.3 (3.1) 15.6 (3.7) 16.2 (3.3) 18.0 (2.8) 19.3 (2.8) 15.2 (3.7) 15.5 (2.9) 16.3 (3.3) 15.7 (4.1) 18.2 (2.9) 
 15.6 (3.3) 17.1 (2.2) 17.0 (2.9) 17.6 (2.6) 18.3 (1.7) 15.7 (2.7) 16.6 (2.5) 16.9 (2.5) 17.6 (2.5) 18.8 (1.8) 
 12.9 (3.7) 14.3 (2.9) 14.6 (3.3) 16.1 (2.6) 16.7 (3.3) 12.2 (3.3) 13.4 (3.5) 14.2 (3.3) 15.6 (2.3) 15.7 (3.3) 
 9.8 (0.5) 9.9 (0.2) 10.0 (0.2) 9.9 (0.3) 10.0 (0.2) 9.7 (0.6) 9.9 (0.3) 9.9 (0.3) 9.8 (0.4) 9.9 (0.3) 
 16.9 (3.6) 16.0 (3.3) 16.1 (3.8) 16.5 (3.3) 15.2 (4.3) 15.1 (2.6) 15.7 (2.5) 16.1 (3.1) 15.1 (3.4) 16.4 (3.4) 
 13.7 (1.9) 14.0 (2.0) 13.9 (2.1) 13.2 (2.6) 13.8 (2.3) 13.0 (2.2) 13.4 (1.7) 14.0 (1.9) 12.7 (2.4) 13.4 (2.3) 
 16.1 (7.5) 23.9 (10.5) 27.5 (7.6) 34.7 (10.6) 41.1 (8.4) 13.7 (6.6) 22.1 (8.6) 26.0 (9.2) 32.7 (9.6) 37.0 (7.4) 
 6.1 (2.5) 6.6 (2.1) 6.7 (2.1) 6.6 (2.3) 6.9 (2.2) 5.9 (2.7) 6.1 (2.5) 5.7 (2.5) 6.0 (2.7) 6.9 (1.6) 
 19.4 (0.9) 19.4 (1.1) 19.6 (0.8) 19.5 (0.9) 19.5 (0.9) 19.3 (0.8) 19.3 (1.0) 19.2 (1.0) 19.2 (1.0) 19.4 (1.1) 
 7.4 (2.1) 8.4 (2.1) 9.3 (1.8) 9.8 (1.8) 10.1 (1.4) 8.2 (1.9) 8.6 (1.9) 9.0 (1.9) 9.0 (2.2) 10.2 (1.8) 
 12.5 (3.6) 13.5 (3.1) 13.9 (3.0) 14.8 (2.9) 15.6 (3.0) 11.9 (3.5) 13.1 (3.4) 13.1 (3.3) 14.8 (3.3) 15.9 (2.4) 
 n = 38 n = 53 n = 82 n = 69 n = 24 n = 50 n = 35 n = 55 n = 37 n = 17 
 75 years and above
 
     
 0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
     
 24.6 (4.8) 24.7 (5.2) 24.4 (4.9) 25.9 (2.4) 25.8 (5.7)      
 13.9 (3.2) 15.7 (2.5) 15.4 (3.3) 15.1 (3.4) 16.5 (4.7)      
 14.4 (3.0) 17.0 (2.6) 16.8 (3.0) 17.0 (3.0) 19.3 (1.0)      
 11.9 (3.8) 13.5 (4.1) 14.4 (3.3) 16.6 (2.7) 17.3 (2.5)      
 9.8 (0.4) 9.9 (0.3) 9.9 (0.4) 10.0 (0.0) 10.0 (0.0)      
 15.2 (3.5) 15.0 (3.6) 15.3 (3.4) 16.6 (2.4) 15.3 (2.2)      
 13.7 (1.9) 14.1 (2.4) 13.9 (2.1) 14.3 (2.1) 13.0 (3.6)      
 12.2 (6.7) 22.6 (11.0) 23.7 (8.9) 24.3 (10.0) 30.0 (6.8)      
 5.6 (2.2) 5.2 (2.6) 5.9 (2.3) 5.6 (1.8) 6.5 (2.6)      
 19.3 (1.0) 19.2 (1.2) 19.0 (1.3) 19.4 (0.7) 19.8 (0.5)      
 7.3 (2.2) 8.2 (1.9) 8.1 (1.9) 7.9 (2.2) 8.0 (3.5)      
 11.6 (3.2) 13.0 (3.7) 12.5 (3.0) 14.5 (3.1) 15.3 (2.8)      
 n = 46 n = 20 n = 44 n = 11 n = 4      

Notes: Edu = years of education.

Missing information on educational level (n = 18).

Table 7.

Means and standard deviations for RBANS index and Total Scale raw scores stratified by age and education (N = 1,147)

Subtests 54–59 years
 
60–64 years
 
0 Edu 1–3 Edu 4–6 Edu 7–10 Edu >10 Edu 0 Edu 1–3 Edu 4–6 Edu 7–10 Edu >10 Edu 
Immediate Memory 103.5 (7.7) 99.2 (11.0) 102.6 (10.7) 105.9 (11.6) 107.9 (12.9) 96.9 (12.2) 99.8 (11.8) 100.2 (12.6) 104.5 (10.3) 108.6 (11.3) 
Visuospatial/Constructional 93.9 (15.1) 99.2 (11.3) 103.3 (8.5) 107.2 (8.5) 109.2 (9.0) 91.2 (12.2) 96.3 (13.4) 100.4 (12.1) 105.4 (9.2) 108.4 (7.8) 
Language 102.6 (6.3) 99.0 (13.2) 103.8 (8.7) 102.0 (11.1) 105.3 (9.8) 99.5 (9.2) 99.6 (11.7) 100.0 (11.5) 102.6 (8.9) 105.1 (10.0) 
Attention 94.5 (10.3) 100.3 (10.9) 104.6 (8.3) 108.1 (9.9) 108.7 (14.2) 91.8 (9.2) 97.4 (10.8) 101.8 (8.6) 104.1 (9.8) 109.7 (10.8) 
Delayed Memory 104.6 (11.7) 101.6 (10.0) 104.1 (9.1) 105.6 (9.9) 107.5 (9.2) 98.9 (9.9) 100.1 (9.8) 101.2 (12.8) 105.3 (8.6) 108.1 (8.4) 
Total Scale 99.8 (7.3) 99.8 (6.7) 103.7 (6.0) 105.8 (6.2) 107.7 (8.2) 95.7 (6.4) 98.6 (7.5) 100.8 (7.5) 104.4 (5.7) 108.3 (6.3) 
 n = 12 n = 24 n = 79 n = 96 n = 33 n = 30 n = 39 n = 112 n = 106 n = 31 
 65–69 years
 
70–74 years
 
 0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
 93.4 (11.8) 97.0 (12.1) 100.9 (11.5) 102.6 (11.5) 107.8 (11.0) 94.4 (12.7) 96.5 (10.0) 96.7 (11.7) 95.5 (14.2) 101.6 (10.6) 
90.5 (12.4) 97.9 (9.9) 98.2 (11.2) 103.2 (10.2) 106.3 (9.7) 89.4 (11.7) 94.6 (11.3) 97.0 (11.0) 102.4 (9.6) 105.7 (10.7)  
 97.8 (15.2) 100.0 (10.1) 100.9 (9.7) 100.5 (8.9) 99.5 (9.6) 93.4 (14.6) 98.0 (9.9) 99.4 (9.6) 94.3 (12.6) 99.5 (11.0) 
 91.7 (8.0) 97.4 (10.3) 98.8 (8.6) 100.6 (9.9) 106.4 (9.6) 88.0 (8.6) 94.1 (8.9) 98.3 (8.1) 97.9 (9.0) 102.7 (8.7) 
 93.7 (12.2) 97.0 (13.0) 100.0 (10.9) 100.6 (12.2) 102.6 (12.6) 93.5 (11.8) 95.3 (13.3) 94.2 (12.0) 96.9 (15.5) 102.8 (13.5) 
 93.4 (7.1) 97.9 (7.3) 99.7 (6.6) 101.3 (6.5) 104.7 (7.5) 91.9 (8.1) 95.7 (7.3) 97.0 (6.6) 97.2 (8.2) 102.5 (6.0) 
 n = 38 n = 53 n = 82 n = 69 n = 24 n = 50 n = 35 n = 55 n = 37 n = 17 
 75 years and above
 
     
 0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
     
 89.7 (10.9) 93.4 (12.2) 92.2 (12.3) 94.4 (9.1) 97.1 (19.4)      
  85.1 (14.3) 95.8 (14.7) 97.3 (13.0) 102.5 (9.3) 110.4 (6.0)      
 95.0 (11.7) 96.5 (11.2) 96.7 (13.5) 102.5 (5.4) 99.6 (5.0)      
 89.7 (7.9) 96.9 (12.4) 97.0 (9.8) 99.9 (8.5) 97.3 (15.9)      
 90.3 (13.8) 91.9 (16.5) 90.7 (15.4) 95.6 (7.9) 101.4 (9.7)      
 90.1 (7.5) 95.8 (7.8) 94.8 (8.6) 98.4 (4.4) 101.1 (8.8)      
 n = 46 n = 20 n = 44 n = 11 n = 4      
Subtests 54–59 years
 
60–64 years
 
0 Edu 1–3 Edu 4–6 Edu 7–10 Edu >10 Edu 0 Edu 1–3 Edu 4–6 Edu 7–10 Edu >10 Edu 
Immediate Memory 103.5 (7.7) 99.2 (11.0) 102.6 (10.7) 105.9 (11.6) 107.9 (12.9) 96.9 (12.2) 99.8 (11.8) 100.2 (12.6) 104.5 (10.3) 108.6 (11.3) 
Visuospatial/Constructional 93.9 (15.1) 99.2 (11.3) 103.3 (8.5) 107.2 (8.5) 109.2 (9.0) 91.2 (12.2) 96.3 (13.4) 100.4 (12.1) 105.4 (9.2) 108.4 (7.8) 
Language 102.6 (6.3) 99.0 (13.2) 103.8 (8.7) 102.0 (11.1) 105.3 (9.8) 99.5 (9.2) 99.6 (11.7) 100.0 (11.5) 102.6 (8.9) 105.1 (10.0) 
Attention 94.5 (10.3) 100.3 (10.9) 104.6 (8.3) 108.1 (9.9) 108.7 (14.2) 91.8 (9.2) 97.4 (10.8) 101.8 (8.6) 104.1 (9.8) 109.7 (10.8) 
Delayed Memory 104.6 (11.7) 101.6 (10.0) 104.1 (9.1) 105.6 (9.9) 107.5 (9.2) 98.9 (9.9) 100.1 (9.8) 101.2 (12.8) 105.3 (8.6) 108.1 (8.4) 
Total Scale 99.8 (7.3) 99.8 (6.7) 103.7 (6.0) 105.8 (6.2) 107.7 (8.2) 95.7 (6.4) 98.6 (7.5) 100.8 (7.5) 104.4 (5.7) 108.3 (6.3) 
 n = 12 n = 24 n = 79 n = 96 n = 33 n = 30 n = 39 n = 112 n = 106 n = 31 
 65–69 years
 
70–74 years
 
 0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
 93.4 (11.8) 97.0 (12.1) 100.9 (11.5) 102.6 (11.5) 107.8 (11.0) 94.4 (12.7) 96.5 (10.0) 96.7 (11.7) 95.5 (14.2) 101.6 (10.6) 
90.5 (12.4) 97.9 (9.9) 98.2 (11.2) 103.2 (10.2) 106.3 (9.7) 89.4 (11.7) 94.6 (11.3) 97.0 (11.0) 102.4 (9.6) 105.7 (10.7)  
 97.8 (15.2) 100.0 (10.1) 100.9 (9.7) 100.5 (8.9) 99.5 (9.6) 93.4 (14.6) 98.0 (9.9) 99.4 (9.6) 94.3 (12.6) 99.5 (11.0) 
 91.7 (8.0) 97.4 (10.3) 98.8 (8.6) 100.6 (9.9) 106.4 (9.6) 88.0 (8.6) 94.1 (8.9) 98.3 (8.1) 97.9 (9.0) 102.7 (8.7) 
 93.7 (12.2) 97.0 (13.0) 100.0 (10.9) 100.6 (12.2) 102.6 (12.6) 93.5 (11.8) 95.3 (13.3) 94.2 (12.0) 96.9 (15.5) 102.8 (13.5) 
 93.4 (7.1) 97.9 (7.3) 99.7 (6.6) 101.3 (6.5) 104.7 (7.5) 91.9 (8.1) 95.7 (7.3) 97.0 (6.6) 97.2 (8.2) 102.5 (6.0) 
 n = 38 n = 53 n = 82 n = 69 n = 24 n = 50 n = 35 n = 55 n = 37 n = 17 
 75 years and above
 
     
 0 Edu
 
1–3 Edu
 
4–6 Edu
 
7–10 Edu
 
>10 Edu
 
     
 89.7 (10.9) 93.4 (12.2) 92.2 (12.3) 94.4 (9.1) 97.1 (19.4)      
  85.1 (14.3) 95.8 (14.7) 97.3 (13.0) 102.5 (9.3) 110.4 (6.0)      
 95.0 (11.7) 96.5 (11.2) 96.7 (13.5) 102.5 (5.4) 99.6 (5.0)      
 89.7 (7.9) 96.9 (12.4) 97.0 (9.8) 99.9 (8.5) 97.3 (15.9)      
 90.3 (13.8) 91.9 (16.5) 90.7 (15.4) 95.6 (7.9) 101.4 (9.7)      
 90.1 (7.5) 95.8 (7.8) 94.8 (8.6) 98.4 (4.4) 101.1 (8.8)      
 n = 46 n = 20 n = 44 n = 11 n = 4      

Discussion

To our knowledge this is the largest normative study of the RBANS in cognitively normal Chinese elderly. In addition to providing a set of age- and education-adjusted norms for the RBANS subtests that can be used in Asia and in other parts of the world where ethnic Chinese live, we also provide evidence of differential performances within very low educated or uneducated subjects (0 years, 1–3 years, and 4–6 years) and include age- and education-adjusted norms for this group. We also investigated the effects of language and gender on the performance of the RBANS.

The influence of age and education on neuropsychological performance has been well documented (Beatty et al., 2003; Gontkovsky et al., 2002; Vanderploeg et al., 1997; von Gunten et al., 2008; Welsh-Bohmer et al., 2009). The question of whether age or education is a primary predictor of performance was the subject of a previous large-scale community study. Gontkovsky and colleagues (2002) hierarchical regression analyses in a sample of 631 healthy elderly aged 64–94 showed education to be a significant primary predictor of performance across all RBANS indices. Contrary to their original hypotheses, age entered as a significant secondary predictor for the Visuospatial/Constructional, Attention, Delayed Memory, and Total Scale Scores. In our analysis, Education was a primary predictor for the Attention, Visuospatial/Constructional, Immediate Memory indexes, and Total Scale and accounted for 18.3%, 17.6%, 7.5%, and 18.6% of the total variance, respectively. Education was also a primary predictor of Coding, Line Orientation, Story Recall, Story Memory, Figure Recall, Figure Copy, and Picture Naming subtests and accounted for 15.4%, 10.8%, 9.7%, 9.4%, 8.8%, and 0.7% variance, respectively. In contrast to samples in the west where most people have at least a full primary education, it is possible that education is a more important predictor of performance than age. Evidence suggests that in low-educated samples even 2 or 3 years of primary education will result in performance differences later in life (Lee et al., 2012).

Compared with education, age was a primary predictor of List Learning, Semantic Fluency, List Recall and List Recognition subtests, and Language and Delayed Memory indexes. Previous studies have identified a general age-related decline in cognitive speed (Houx & Jolles, 1993; Jolles, Houx, Vreeling, & Verhey, 1993), working memory (Salthouse, 1994) and the efficient consolidation of newly learned information (Salthouse, 1998) as well as mental organization and search strategies (Bryan & Luszcz, 2000; Craik & Salthouse, 2000). Neuropsychological tests such as verbal learning, verbal fluency, list recall, and recognition are heavily dependent on these cognitive abilities. Test performance deteriorates in an age-dependent manner for the Rey's Auditory Verbal Learning Test and semantic fluency tasks (animals and fruits) in healthy elderly participants (Acevedo et al., 2000; Geffen, Moar, O'Hanlon, Clarck, & Geffen, 1990; Van der Elst, Van Boxtel, Van Breukelen, & Jolles, 2005). Similarly, findings in the present study found age-dependent deterioration in performances on the subtests that are known to be susceptible to the effects of age in the general literature. In general, performance decreased linearly, although significant differences between participants in the 70–74-age group and the 75 years and above age group were found only for the Coding subtest. On the other hand, RBANS performance generally increased with more education linearly, although significant differences between the 7–10 years and >10 years of education groups were found only for the Coding subtest. No significant interaction between age and education was found in the present study indicating that RBANS performance was uniform across different age and education groups.

Very low levels of education are known to be associated with the lowest levels of performance on neuropsychological tests (Ardila et al., 2000; Brucki & Nitrini, 2007; Lee et al., 2012). However, we found that the influence of education is not homogeneous among the lowest education stratum. Significant differences in performances were found between participants without any formal education and those with 1–3 or 4–6 years of education across 6 RBANS subtests and 4 RBANS indices and the Total Scale suggesting that, in terms of neuropsychological test taking ability, even a little education is beneficial compared with none at all. Ardila and colleagues (2000) previously proposed that this educational effect is non-linear, but rather a negatively accelerated curve, tending to a plateau. Our results do not support this proposition. Such effects suggest that future studies should avoid collapsing low-education groups (e.g., 6 years or lower) to avoid the loss of subtle but important differences in performance.

The apparent protection that education affords as people age yielded further findings. Two main patterns were identified: Parallelism (age-related decline changed in a parallel fashion across the low- and high-education groups) and protection (age-related decline is reduced due to a protective effect of education). Confluence (initial advantage of education on age-related decline is reduced later in life) was observed only for the Story Memory and Digit Span subtests. While the protective effects of education have been consistently demonstrated on tests of word recall (Ardila et al., 2000; Capitani et al., 1996), in the present study we found that education was also associated with better performance on Figure Copy, Line Orientation, Story Recall, and Figure Recall subtests–tests that assess visuospatial abilities and delayed auditory and visuospatial recall. The concept of ‘cognitive reserve capacity’ is partly based on evidence showing the protective effects of education in delaying cognitive deterioration with age (Albert et al., 1995; Butler, Ashford, & Snowdon, 1996; Chodosh, Reuben, Albert, & Seeman, 2002; Christensen et al., 1997; Farmer, Kittner, Rae, Bartko, & Regier, 1995; Lyketsos, Chen, & Anthony, 1999). The current pattern of findings generally support this concept but performance advantages on neuropsychological tests also occur in English speaking and educated subjects due to cultural familiarity. This effect may have been greater in our sample as there is a much wider variation in education experience here than in most Western countries which may have lead to the stronger association of education and performance across the board (see also Lam et al., 2013; Lee et al., 2012).

Testing subjects in their own homes in their dominant language or dialect helped to reduce culture effects but some language differences nonetheless remained. In general, the English- and Mandarin-speaking groups outperformed the dialect speakers (Hokkien, Teochew, and Cantonese). Among the dialects-speaking groups, the Cantonese-speakers had superior performance to the Hokkien- and Teochew-speaking groups. No significant differences were found between the Hokkien- and Teochew-speaking groups. The reverse of this general trend was true for the Digit Span subtest, where poorer performance was found for the English-speaking group when compared with the Mandarin- and dialects-speaking groups.

Cross-linguistic differences in digit memory spans have been well-established and Mandarin-speaking people have been previously found to have a larger digit span than English speakers (Hedden et al., 2002; Hoosain, 1979), Arabic (Naveh-Benjamin & Ayres, 1986), or Malay speakers (Chan & Elliot, 2011). This effect can be explained by reference to the Baddeley's Working Memory Model (Baddeley & Logie, 1999). Within this model, differences in digit memory span arise from differences in articulation times for digits between languages. A language with short articulation times for digits makes fewer demands upon the limited temporal resources of working memory; consequently, digit span is likely to be greater compared with a language with relatively longer articulatory duration for digits.

A significant interaction between language and education was found for the List Learning subtest. The majority of the English-speaking participants were relatively younger (between 55 and 59 years) and most had 7–10 years of education or more. In contrast, the majority of the Mandarin-, Teochew-, and Cantonese-speaking groups were aged between 60 and 64 years and most had 4–6 years of education. It is therefore not surprizing that the language × education interaction favored English-speaking participants on this subtest.

Significant gender differences were found after controlling for the effects of age and education, for all subtests, except for the Story Memory, Figure Copy and Story Recall subtests and for all RBANS indices, except for the Total Scale. Consistent with the previous RBANS study of Beatty and colleagues (2003), males performed significantly better females on the Visuospatial/Constructional and Attention indices. On the other hand, the females were found to perform significantly better than the males on the Immediate Memory, Language, and Delayed Memory indices.

Several limitations of this study should be noted. First, this is a cross-sectional study and there are clear limits to how far the age-related data can be interpreted. Cohort effects (i.e., cohort-specific educational experiences of a group of people born at the same time or same interval) and time effects (historical events that impact on developmental abilities and may include changes in educational policies or technique of education) may have influenced the data in unaccountable ways (Ardila et al., 2000). Education experiences have been found to vary with population and cohort differences (Manly and Jacobs, 2002). Lam et al. (2013) found that compared with using the number of years of education to reflect an individual's exposure to education, an index combining education with age (expressed as the percentage of education with age) showed the greatest sensitivity to neuropsychological performance compared with other age and education measurements. The authors also suggested that this index is a proxy of an individual's cognitive reserve. Despite its sensitivity, we did not use this index so that these norms have wide international applicability to the ethnic Chinese of other countries. Other proxy measures of education (such as reading ability) may be used in other international contexts to assess the protective effects of education. However, to address the possibility of cohort effects in our study, we are currently conducting longitudinal follow-up of the sample so as to draw true trajectories of cognitive protection and decline.

Another limitation of the study is that is the focus on Chinese only due to small sample sizes for the other ethnic groups of Singapore (Malay, Indian, and Eurasian), we therefore cannot be sure that the current norms are entirely applicable to SE Asian people other than ethnic Chinese. Third, the raw score distributions of some of the RBANS subtests were considerably skewed. For example, the raw scores for Figure Copy, Picture Naming, List Recall, List Recognition, and Story Recall were largely leptokurtic and negatively skewed mainly due to ceiling effects. Finally, when the RBANS subtest, Index, and Total Scale scores were further stratified by education, the cell sizes for a number of age and language (especially English speaking) groups were small. Therefore, cautious interpretation is advised when using this normative data for diagnostic decision making.

In conclusion, this large normative dataset of elderly Chinese people indicates that individual differences have a large effect on performance of the RBANS. Significant age, education, gender, and linguistic differences influence RBANS performance emphasizing the importance of using culturally appropriate normative data. In view of the surge in dementia prevalence across Asia and those who emigrated from Asia in the last 50 years—the so-called ‘Chinese diaspora’ comprising over 40 million people presently living outside of mainland China—it is hoped that the present data will serve as an useful reference source for clinicians and researchers using the RBANS.

References

Acevedo
A.
Loewenstein
D. A.
Barker
W. W.
Harwood
D. G.
Luis
C.
Bravo
M.
et al.  
Category fluency test: Normative data for English- and Spanish speaking elderly
Journal of the International Neuropsychological Society
 
2000
6
7
760
769
Albert
M. S.
Jones
K.
Savage
C. R.
Berkman
L.
Seeman
T.
Blazer
D.
et al.  
Predictors of cognitive change in older persons: MacArthur studies of successful aging
Psychology and Aging
 
1995
10
578
589
Allen
J. S.
Bruss
J.
Brown
C. K.
Damasio
H.
Normal neuroanatomical variation due to age: The major lobes and a parcellation of the temporal region
Neurobiology of Aging
 
2005
26
9
1245
1260
Ardila
A.
A note of caution: Normative neuropsychological test performance: Effects of age, education, gender and ethnicity: A comment on Saykin et al. (1995)
Applied Neuropsychology
 
1998
5
51
53
Ardila
A.
Ostrosky-Solis
F.
Rosselli
M.
Gómez
C.
Age-related cognitive decline during normal aging: The complex effect of education
Archives of Clinical Neuropsychology
 
2000
15
6
495
513
Ardila
A.
Rosselli
M.
Neuropsychological characteristics of normal aging
Developmental Neuropsychology
 
1989
5
307
320
Baddeley
A. D.
Logie
R. H.
Miyake,
A.
Shah
P.
Working memory: The multiple-component model
Models of working memory: Mechanisms of active maintenance and executive control
 
1999
Cambridge, UK
Cambridge University Press
28
61
Beatty
W. W.
Mold
J. W.
Gontkovsky
S. T.
RBANS performance: Influences of sex and education
Journal of Clinical and Experimental Neuropsychology
 
2003
25
1065
1069
Belle
S. H.
Seaberg
E. C.
Ganguli
M.
Ratcliff
G.
DeKosky
S.
Kuller
L. H.
Effect of education and gender adjustment on the sensitivity and specificity of a cognitive screening battery for dementia: Results from the MoVIES project
Neuroepidemiology
 
1996
15
321
329
Boone
K. B.
Victor
T. L.
Wen
J.
Razani
J.
Pontón
M.
The association between neuropsychological scores and ethnicity, language, and acculturation variables in a large patient population
Archives of Clinical Neuropsychology
 
2007
22
3
355
365
Brayne
C.
Ince
P. G.
Keage
H. A.
McKeith
I. G.
Matthews
F. E.
Polvikoski
T.
et al.  
Education, the brain and dementia: Neuroprotection or compensation?
Brain
 
2010
133
8
2210
2216
Brucki
S. M.
Nitrini
R.
Cancellation task in very low educated people
Archives of Clinical Neuropsychology
 
2007
23
139
147
Bryan
J.
Luszcz
M. A.
Measurement of executive function: Considerations for detecting adult age differences
Journal of Clinical and Experimental Neuropsychology
 
2000
22
40
55
Butler
S. M.
Ashford
J. W.
Snowdon
D. A.
Age, education, and changes in the Mini-Mental State Exam scores of older women: Findings from the Nun Study
Journal of the American Geriatrics Society
 
1996
44
675
681
Capitani
E.
Barbarotto
R.
Laicana
M.
Does education influence age-related cognitive decline? A further inquiry
Developmental Neuropsychology
 
1996
4
17
22
Chan
M. E.
Elliott
J. M.
Cross-linguistic differences in digit memory span
Australian Psychologist
 
2011
46
1
25
30
Chodosh
J.
Reuben
D. B.
Albert
M. S.
Seeman
T. E.
Predicting cognitive impairment in high-functioning community-dwelling older persons: MacArthur Studies of Successful Aging
Journal of the American Geriatrics Society
 
2002
50
1051
1060
Christensen
H.
Korten
A. E.
Jorm
A. F.
Henderson
A. S.
Jacomb
P. A.
Rodgers
B.
et al.  
Education and decline in cognitive performance: Compensatory but not protective
International Journal of Geriatric Psychiatry
 
1997
12
323
330
Collinson
S. L.
Yeo
D.
Fuji
D.
Neuropsychology in Singapore: History, development and future directions
the neuropsychology of Asian Americans
 
2010
London
Psychology Press
Craik
F. I. M.
Salthouse
T. A.
The handbook of aging and cognition
 
2000
Mahwah, NJ
Lawrence Erlbaum Associates, Inc
Del Arco
A.
Segovia
G.
de Blas
M.
Garrido
P.
Acuña-Castroviejo
D.
Pamplona
R.
et al.  
Prefrontal cortex, caloric restriction and stress during aging: Studies on dopamine and acetylcholine release, BDNF and working memory
Behavioural Brain Research
 
2011
216
1
136
145
2010
Department of Statistics, Singapore Population: Census From http://www.singstat.gov.sg/publications/publications_and_papers/cop2010/cop2010_sr1.html
Duff
K.
Pattern
D.
Schoenberg
M. R.
Mold
J.
Scott
J. G.
Adams
R. L.
Age- and education-corrected independent normative data for the RBANS in a community dwelling elderly sample
The Clinical Neuropsychologist
 
2003
17
351
366
Erixon-Lindroth
N.
Farde
L.
Robins Wahlin
T. B.
Sovago
J.
Halldin
C.
Bäckman
L.
The role of the striatal dopamine transporter in cognitive aging
Psychiatry Research: Neuroimaging
 
2005
138
1
1
12
Farmer
M. E.
Kittner
S. J.
Rae
D. S.
Bartko
J. J.
Regier
D. A.
Education and change in cognitive function: The epidemiologic catchment area study
Annals of Epidemiology
 
1995
5
1
7
Feng
L.
Chiu
H.
Chong
M.-Y.
Yu
X.
Kua
E. H.
Dementia in Chinese populations: Current data and future research
Asia-Pacific Psychiatry
 
2011
3
3
109
114
Fotenos
A. F.
Snyder
A. Z.
Girton
L. E.
Morris
J. C.
Buckner
R. L.
Normative estimates of cross-sectional and longitudinal brain volume decline in aging and AD
Neurology
 
2005
64
6
1032
1039
Fuji
D. E. M.
The neuropsychology of Asian-Americans
 
2010
London
Psychology Press
Geffen
G.
Moar
K. J.
O'Hanlon
A. P.
Clarck
C. R.
Geffen
L. N.
Performance measures of 16- to 86-year-old males and females on the Auditory Verbal Learning Test
The Clinical Neuropsychologist
 
1990
4
45
63
Gontkovsky
S. T.
Mold
J. W.
Beatty
W. W.
Age and educational influences on RBANS index scores in a non- demented geriatric sample
The Clinical Neuropsychologist
 
2002
16
258
263
Hall
C. B.
Derby
C.
LeValley
A.
Katz
M. J.
Verghese
J.
Lipton
R. B.
Education delays accelerated decline on a memory test in persons who develop dementia
Neurology
 
2007
69
17
1657
1664
Hedden
T.
Park
D. C.
Nisbett
R.
Ji
L. J.
Jing
Q.
Jiao
S.
Cultural variation in verbal versus spatial neuropsychological function across the life span
Neuropsychology
 
2002
16
1
65
73
Hoosain
R.
Forward and backward digit span in the languages of the bilingual
The Journal of Genetic Psychology
 
1979
135
2
263
268
Houx
P. J.
Jolles
J.
Age-related decline of psychomotor speed: Effects of age, brain health, sex, and education
Perceptual and Motor Skills
 
1993
76
195
211
Hsu
J. L.
Leemans
A.
Bai
C. H.
Lee
C. H.
Tsai
Y. F.
Chiu
H. C.
et al.  
Gender differences and age-related white matter changes of the human brain: A diffusion tensor imaging study
Neuroimage
 
2008
39
2
566
577
Jolles
J.
Houx
P. J.
Vreeling
F. W.
Verhey
F. R. J.
Cognitive aging, biological life events and primitive reflexes
Neuroscience Research Communications
 
1993
13
Suppl. 1
S47
S50
Retrieved from http://arno.unimaas.nl/show.cgi?fid=4924
Kadota
T.
Horinouchi
T.
Kuroda
C.
Development and aging of the cerebrum: Assessment with proton MR spectroscopy
American Journal of Neuroradiology
 
2001
22
128
135
Kruggel
F.
MRI-based volumetry of head compartments: Normative values of adults
NeuroImage
 
2006
30
1
11
Lam
M.
Eng
G. K.
Rapisarda
A.
Subramaniam
M.
Kraus
M.
Keefe
R. S.
et al.  
Formulation of the age-education index: Measuring age and education effects in neuropsychological performance
Psychological assessment
 
2013
25
1
61
71
Lee
C. K.
Collinson
S. L.
Feng
L.
Ng
T. P.
Preliminary normative neuropsychological data for an elderly chinese population
The Clinical Neuropsychologist
 
2012
26
2
321
334
Lim
M. L.
Collinson
S. L.
Feng
L.
Ng
T. P.
Cross-cultural application of the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS): Performances of elderly Chinese Singaporeans
The Clinical Neuropsychologist
 
2010
24
5
811
826
Long
C. L.
Klein
K.
Decision strategies on neuropsychology II: Determination of age effects on neuropsychological performance
Archives of Clinical Neuropsychology
 
1990
5
335
345
Lyketsos
C. G.
Chen
L.-S.
Anthony
J. C.
Cognitive decline in adulthood: An 11.5-year follow-up of the baltimore epidemiologic catchment area study
American Journal of Psychiatry
 
1999
156
58
65
Magnotta
V. A.
Andreasen
N. C.
Schultz
S. K.
Harris
G.
Cizadlo
T.
Heckel
D.
et al.  
Quantitative in vivo measurement of gyrification in the human brain: Changes associated with aging
Cerebral Cortex
 
1999
9
151
160
Mangione
C. M.
Seddon
J. M.
Cook
E. F.
Krug
J. H.
Sahagian
C. R.
Campion
E. W.
et al.  
Correlates of cognitive function scores in elderly outpatients
Journal of the American Geriatrics Society
 
1993
41
491
497
Manly
J. J.
Critical issues in cultural neuropsychology: Profit from diversity
Neuropsychology Review
 
2008
18
3
179
183
Manly
J. J.
Jacobs
D. M.
Ferraro
F. R.
Future directions in neuropsychological assessment with African Americans
Minority and cross-cultural aspects of neuropsychological assessment
 
2002
Lisse
Swets & Zeitlinger
79
96
Marcopulos
B. A.
McLain
C. A.
Guiliano
A. J.
Cognitive impairment or inadequate norms? A study of healthy, rural, older adults with limited education
The Clinical Neuropsychologist
 
1997
11
111
131
McDowell
I.
Xi
G.
Lindsay
J.
Tierney
M.
Mapping the connections between education and dementia. Journal of Clinical and Experimental Neuropsychology
2007
29
2
127
141
Naveh-Benjamin
M.
Ayres
T. J.
Digit span, reading rate, and linguistic relativity
Quarterly Journal of Experimental Psychology
 
1986
38A
739
751
Parry
J.
Cui
W. Y.
Looming dementia epidemic in Asia
Bulletin of the World Health Organization
 
2011
89
166
167
Pieperhoff
P.
Hömke
L.
Schneider
F.
Habel
U.
Shah
N. J.
Zilles
K.
et al.  
Deformation field morphometry reveals agerelated structural differences between the brains of adults up to 51 years
The Journal of Neuroscience
 
2008
28
828
842
Randolph
C.
Repeatable battery for the assessment of neuropsychological status
 
1998
San Antonio, TX
The Psychological Corporation
Randolph
C.
Tierney
M. C.
Mohr
E.
Chase
T. N.
The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS): preliminary clinical validity
Journal of Clinical and Experimental Neuropsychology
 
1998
20
3
310
319
Roe
C. M.
Xiong
C.
Miller
J. P.
Morris
J. C.
Education and Alzheimer disease without dementia support for the cognitive reserve hypothesis
Neurology
 
2007
68
3
223
228
Rosselli
M.
Ardila
A.
Effects of age, gender and socioeconomic level on the Wisconsin Card Sorting Test
The Clinical Neuropsychologist
 
1993
7
145
154
Salthouse
T. A.
The aging of working memory
Neuropsychology
 
1994
8
535
543
Salthouse
T. A.
Nordhus
I. H.
VandenBos
G. R.
Berg
S.
Fromholt
P.
Cognitive and information-processing perspectives on aging
Clinical geropsychology
 
1998
Washington, DC
American Psychological Association
49
59
Sewell
M. C.
Vigario
A.
Sano
M.
Fillit
H. M.
Rockwood
K.
Woodhouse
K.
Neuropsychology in the diagnosis and treatment of dementia
Brocklehursts textbook of geriatric medicine and gerontology
 
2010
7th ed.
Philadelphia, PA
Saunders
402
410
Shan
I. K.
Chen
Y. S.
Lee
Y. C.
Su
T. P.
Adult normative data of the Wisconsin card sorting test in Taiwan
Journal of the Chinese Medical Association
 
2008
71
10
517
522
Toothaker
L. E.
Multiple comparisons for researchers
 
1991
Newbury Park, CA: Sage Publications Inc.
Valenzuela
M. J.
Sachdev
P.
Brain reserve and dementia: A systematic review
Psychological Medicine
 
2005
35
1
14
Van Der Elst
W. I. M.
Van Boxtel
M. P.
Van Breukelen
G. J.
Jolles
J.
Rey's verbal learning test: Normative data for 1855 healthy participants aged 24–81 years and the influence of age, sex, education, and mode of presentation
Journal of the International Neuropsychological Society
 
2005
11
3
290
302
Vanderploeg
R. D.
Axelrod
B. N.
Sherer
M.
Scott
J. G.
Adams
R. L.
The importance of demographic adjustments on neuropsychological test performance: A response to Reitan and Wolfson (1995)
The Clinical Neuropsychologist
 
1997
11
210
217
von Gunten
A.
Ostos-Wiechetek
M.
Brull
J.
Vaudaux-Pisquem
I.
Cattin
S.
Duc
R.
Clock-drawing test performance in the normal elderly and its dependence on age and education
European Neurology
 
2008
60
73
78
Welsh-Bohmer
K. A.
Østbye
T.
Sanders
L.
Pieper
C. F.
Hayden
K. M.
Tschanz
J. T.
et al.  
Neuropsychological performance in advanced age- influences of demographic factors and apolipoprotein E: Findings from the Cache County Memory Study
The Clinical Neuropsychologist
 
2009
23
1
77
99
Wharton
S. B.
Brayne
C.
Savva
G. M.
Matthews
F. E.
Forster
G.
Simpson
J.
et al.  
Epidemiological neuropathology: The MRC Cognitive Function and Aging Study experience
Journal of Alzheimer's Disease
 
2009
25
2
359
372