Objective

The Brief Assessment of Cognition in Schizophrenia (BACS) is a cognitive assessment tool used to measure the broad aspects of cognition that are most frequently impaired in patients with schizophrenia. This study aims to develop the normative data of the Chinese version of the BACS among the Mandarin-speaking population.

Method

This cross-sectional study included 382 healthy participants (age range: 19–79 years; mean age: 48.0 ± 16.7 years, 47.6% men) in Taiwan, who were evaluated with the BACS. Means and standard deviations of subtests and composite scores were arranged by age group and gender. The Z-scores calculated based on the U.S. norms were compared to our scores based on the norms established in this study.

Results

The raw scores of all the BACS tests (verbal memory, digit sequencing, token motor test, verbal fluency, symbol coding, and Tower of London) were negatively correlated with participants’ age. Women were superior to men in verbal memory, but inferior to them in executive function. Furthermore, applying the U.S. norms of the BACS to determine the performance of the Chinese BACS results in bias with regard to verbal memory, token motor test, verbal fluency, symbol coding, Tower of London and composite score.

Conclusions

These findings demonstrate that directly applying western cognitive norms to a Mandarin-speaking population can cause biased interpretations. The results of this study can be an important reference for clinical settings and research related to cognitive assessments in Mandarin-speaking Chinese populations.

Introduction

Neurocognitive deficit is a key manifestation of schizophrenia (Harvey, Bowie, & Friedman, 2001; Tosato, & Dazzan, 2005). Patients who suffer from chronic schizophrenia are often impaired in a variety of cognitive functions, including memory, motor speed, attention, reasoning, verbal fluency, and executive functions (Bortolato, Miskowiak, Kohler, Vieta, & Carvalho, 2015; Kitchen, Rofail, Heron, & Sacco, 2012). Such impairments significantly influence patients’ daily functions and longitudinal outcomes (Barch, & Ceaser, 2012; Hoe, Nakagami, Green, & Brekke, 2012). As a result, a standard and easily administered battery of tests could be useful for specifically and efficiently evaluating the most important aspects of cognitive deficits in patients with schizophrenia (Bakkour et al., 2014). The ideal battery of tests should be reliable, demonstrate validity and be able to be repeated to assess evolution over periods of long-term follow-up (Keefe, & Harvey, 2012).

The Brief Assessment of Cognition in Schizophrenia (BACS) was developed before initiating the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) (Kern, Green, Nuechterlein, & Deng, 2004) process to provide clinical trial investigators with a brief battery of tests for measuring cognition. The included tests refer to the domains of cognition most blatantly impaired in schizophrenia (i.e., verbal memory, working memory, motor speed, verbal fluency, attention & processing speed, and executive function) (Keefe et al., 2004) and strongly correlated with schizophrenic patients’ real-world functioning (Keefe, Poe, Walker, & Harvey, 2006; Keefe, Poe, Walker, Kang, & Harvey, 2006). The BACS can be simply administered by various specialties as it requires only paper, pencils, and a stopwatch. The test session lasts approximately 30–35 minutes. Normative data of the English version of the BACS were established by collecting tests of 404 healthy English-speakers in Durham, North Carolina, U.S. (Keefe et al., 2008). The BACS has been translated into various languages, and the Japanese (Kaneda et al., 2007), Spanish (Segarra et al., 2011), French (Bralet et al., 2007), German (Sachs, Winklbaur, Jagsch, & Keefe, 2011), Italian (Anselmetti et al., 2008), Persian (Mazhari et al., 2014), Brazilian Portuguese (Araujo et al., 2015), and Chinese versions (Wang et al., 2016) have been validated.

An English-speaking ethnic Chinese sample in Singapore has also developed normative data for the BACS (Eng et al., 2013), but notable differences in subtest performances were indicated when compared to a Western sample. Our research team has previously used the Chinese version of the BACS as the primary cognitive assessment tool to study the differential cognitive functions between methamphetamine users with psychosis and patients with schizophrenia in Taiwan (Chen et al., 2015), as well as to compare the cognitive differences between methamphetamine users with and without concomitant ketamine use (Chen, Wang, Lin, & Chen, 2015). Because the normative data of the Chinese BACS has yet to be established, we consulted the U.S. BACS norms to determine the cognitive function of the Mandarin-speaking participants in these studies. We found that the standardized T-scores of the subset of verbal memory and verbal fluency in healthy control subjects were 1.5–2 standard deviations (SD) below the normal range (T-score of 50) (Chen et al., 2015). Therefore, we believe that indiscriminately applying U.S. norms of the BACS to evaluate the performance of the Chinese BACS could have biased results. As a result, appropriate norms for the Chinese BACS are vital for reliably evaluating the cognitive profiles of Mandarin-speaking populations.

Based on the data mentioned earlier, we hypothesize that the U.S. normative data, especially in the verbal domain subtests, are not appropriate for Mandarin-speaking populations. The norms of the Chinese version of the BACS must be established in order to deliver an accurate neurocognitive assessment to Chinese populations. This study aims to develop the normative data of the BACS in the Mandarin-speaking population in Taiwan and compare the difference between these norms and those of the USA.

Materials and Methods

Study Participants

This cross-sectional study was conducted at Chang Gung Memorial Hospital upon receiving approval from the respective Institutional Review Board (IRB No: 103–6849C). We posted information about study recruitment through advertising posters that were approved by the IRB. These posters were hung in two general hospitals (Kaohsiung Chang Gung Memorial Hospital and Keelung Chang Gung Memorial Hospital), three community centers, and a local health bureau in Kaohsiung City, Taiwan (TW). We recruited participants based on the following criteria: (a) age ≥ 18 years; (b) without history of major psychiatric disorders (e.g., psychosis, bipolar disorder, major depressive disorder, organic mental disorders, and substance use disorder) confirmed using the Diagnostic and Statistical Manual of Mental Disorders, fourth edition criteria and the Chinese Version of the Mini International Neuropsychiatric Interview (MINI) (Sheehan et al., 1998). The inter-rater reliability of the Chinese version of the MINI, which was translated by the Taiwan Society of Psychiatry, has already been established (Kuo et al., 2003); (c) without any known systemic or neurological diseases that could influence cognitive performance; (d) ethnic Han Chinese; and (e) capacity to speak Mandarin, read Chinese, and give their informed consent. Ultimately, we recruited a total of 382 healthy subjects between the ages of 19 and 79 years old (mean age: 48.0 ± 16.7 years; 47.6% men; education level: ≥64.7% graduated college).

Cognitive Assessment

All the participants’ cognitive functions were assessed by a research team member properly trained to administer the BACS. Both the neuropsychological testing and clinical assessments were conducted in Mandarin. The BACS consisted of the following subtests:

List Learning Test (verbal memory): Participants are presented with 15 phrases (containing two Chinese characters) in Mandarin and then are asked to remember as many as possible. This procedure is repeated five times. Of the eight alternative forms designed to minimize the practice effect, the first form was chosen to test all the participants in the current study. The outcome measure is the total number of phrases remembered.

Digit Sequencing Task (working memory): Participants are given clusters of numbers of increasing length and asked to repeat the numbers in order from the lowest to the highest. The trials are increasingly difficult. The outcome measure is the total number of correct items.

Token Motor Task (motor speed): Participants are given 100 plastic tokens and requested to place them into a container as quickly as they can over a period of 60 s. The outcome measure is the total number of tokens placed in the container.

Verbal Fluency. Category Instances Test (semantic fluency): Participants have 60s to name as many words as possible within the category of animals. The outcome measure is the total number of unique animals given. Controlled Oral Word Association Test (letter fluency): In two separate trials, patients have 60s to come up with as many “phrases” as possible that begin with the Mandarin words “Kou” and “Jing.” The outcome measure is the total number of unique phrases given. The performance of Verbal Fluency is the sum of the total numbers of words/phrases adequately provided in both the aforementioned tests.

Symbol Coding (attention & processing speed): Participants write the numerals 1–9 to match to symbols on a response sheet as quickly as possible for 90s. The outcome measure is the total number of correct responses.

Tower of London Test (executive function): Participants are asked to look at two pictures simultaneously, each of which shows three differently colored balls arranged on three pegs with the balls in a unique arrangement in each picture. The participants are explained the rules of the task and are asked to provide the least number of times the balls in one picture would have to be moved to make the arrangement of balls identical to that of the opposing picture. The outcome measure is the number of trials on which the correct response is provided.

Finally, a Composite Score is calculated by comparing each participant's performance on each measure to the performance of a healthy comparison group, which shall be the Z-score of that sum (Keefe et al., 2008). In this study, the Z-scores for each scale are used for analysis. A-Z-score of 0 for each scale indicates an average functioning with reference to the normal population of the same age range and gender, while every 1 point represents 1 SD.

Statistical Analyses

We used the statistical software package SPSS (Version 21.0; SPSS Inc., Chicago, IL) to analyze the data. The variables are shown as either mean (±SD) or frequency (in percentage). Categorical variables among age groups were compared using the chi-square test. In a two-tailed test, p < .05 was considered statistically significant.

We calculated the scaled subset scores and composite scores as follows: The Z-score has a mean of 0 and a standard deviation of 1; Zij = (Rij– Mj)/SDj, where Zij is the scaled score of the ith subject for subtest j; Rij is the raw score of the ith subject on subtest j; Mj and SDj are the mean and the SD for test j of each sample, respectively. The BACS composite score for each subject has been calculated as follows: composite i is the BACS composite score for the ith subject, 6j = 1 Zij—the sum of the six scaled test scores of the ith subject, and SD—the standard deviation of the sum of the scaled scores of the sample.

We applied linear regression to study the effects of age, gender, and educational level on the performance of each BACS subtest. The raw scores from the aforementioned six domains were set as dependent variables, and B-value and 95% confidence interval were reported.

We categorized our normative sample into six age groups (with a 10-year range) and male and female gender groups. We developed a program file to calculate age–gender stratum-specific norms, and users can make use of the normative data by inputting the observed scores into the program files to determine the Z- or T-scores that have been adjusted for the effects of age and gender. The raw scores of each BACS subtest in our study population were input into this program file to calculate the Z-scores (Z-scores-TW). The raw scores were also transformed into Z-scores based on the U.S. norms (Z-scores-US) (Keefe et al., 2008). We performed paired t-tests and Cohen's d to explore the effect sizes of the difference between Z-scores-TW and Z-scores-US.

Results

A total of 382 healthy subjects between the ages of 19 and 79 years old participated in this study. The distributions of their age groups, gender, and educational levels are summarized in Table 1. No significant differences were found in gender distribution between the six age groups. Older participants tended to have lower educational levels (p < .001). The raw scores, Z-scores, and composite scores of each subtest of the BACS were presented by the age and gender groups (Table 2), and the raw scores of all the subtests were significantly correlated to one another (p < .001).

Table 1.

Demographic characteristics of normative sample

 Age group Statistics 
19–29 (N = 71) 30–39 (N = 65) 40–49 (N = 63) 50–59 (N = 67) 60–69 (N = 61) 70–79 (N = 55) 
Sex 
 Male 35 (49.3%) 31 (47.7%) 29 (46.0%) 29 (43.3%) 31 (50.8%) 27 (49.1%) χ2 = 0.947 
 Female 36 (50.7%) 34 (52.3%) 34 (54.0%) 38 (56.7%) 30 (49.2%) 28 (50.9%) p = .967 
Education 
 High school or lower 9 (12.7%) 2 (3.1%) 11 (17.5%) 29 (43.3%) 43 (70.5%) 41 (74.5%) χ2 = 126.213 
 College or above 62 (87.3%) 63 (96.9%) 52 (82.5%) 38 (56.7%) 18 (29.5%) 14 (25.5%) p < .001* 
 Age group Statistics 
19–29 (N = 71) 30–39 (N = 65) 40–49 (N = 63) 50–59 (N = 67) 60–69 (N = 61) 70–79 (N = 55) 
Sex 
 Male 35 (49.3%) 31 (47.7%) 29 (46.0%) 29 (43.3%) 31 (50.8%) 27 (49.1%) χ2 = 0.947 
 Female 36 (50.7%) 34 (52.3%) 34 (54.0%) 38 (56.7%) 30 (49.2%) 28 (50.9%) p = .967 
Education 
 High school or lower 9 (12.7%) 2 (3.1%) 11 (17.5%) 29 (43.3%) 43 (70.5%) 41 (74.5%) χ2 = 126.213 
 College or above 62 (87.3%) 63 (96.9%) 52 (82.5%) 38 (56.7%) 18 (29.5%) 14 (25.5%) p < .001* 

*p < .05

Table 2.

Mean performance of each subset of the BACS across age and sex groups

 19–29 30–39 40–49 50–59 60–69 70–79 
Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD 
Verbal memory 
 Male 43.71 7.65 0.49 44.52 7.42 0.57 40.24 8.09 0.19 37.55 8.63 −0.44 30.81 8.90 −0.63 24.59 6.37 −1.16 
 Female 50.39 7.13 1.07 40.24 8.09 0.19 44.59 3.04 0.56 37.53 8.77 −0.05 25.80 7.06 −1.06 24.39 6.96 −1.18 
Working memory 
 Male 23.80 3.23 0.31 23.77 3.40 0.37 21.45 3.93 0.27 20.62 3.84 −0.72 20.84 11.23 −0.05 13.81 4.20 −0.90 
 Female 24.08 2.82 0.35 21.45 3.93 0.27 22.21 3.04 0.12 20.61 12.98 −0.07 17.83 9.41 −0.41 18.75 14.90 −0.30 
Motor speed 
 Male 81.74 10.75 0.26 85.52 8.41 0.12 84.93 8.39 0.18 83.00 11.77 0.35 71.29 14.40 −0.44 65.81 13.58 −0.81 
 Female 81.94 10.69 0.28 84.93 8.39 0.48 82.42 11.40 0.31 78.11 14.50 0.02 71.93 17.97 −0.40 62.68 18.56 −1.02 
Verbal fluency 
 Male 33.31 7.00 0.29 36.19 8.17 0.38 33.48 6.64 0.31 32.21 8.11 0.15 28.94 6.01 −0.26 25.81 5.48 −0.65 
 Female 35.17 6.22 0.52 33.48 6.64 0.31 34.65 7.70 0.45 28.89 7.76 −0.26 23.57 5.97 −0.93 22.93 5.82 −1.01 
Attention & processing speed 
 Male 68.63 8.77 0.79 67.00 11.28 0.85 59.00 8.00 0.21 52.59 12.65 −0.17 44.26 10.13 −0.66 34.15 10.54 −1.26 
 Female 72.22 7.77 1.00 59.00 8.00 0.21 61.12 9.87 0.34 54.16 10.60 −0.73 38.33 11.90 −1.01 29.86 11.68 −1.51 
Executive function                  
 Male 17.66 3.42 0.43 18.06 2.39 0.46 18.23 1.70 0.52 16.76 3.77 0.22 15.39 4.43 −0.10 14.11 4.08 −0.39 
 Female 17.94 2.50 0.50 18.03 1.70 0.52 16.50 3.60 0.16 13.05 5.00 −0.64 12.30 4.50 −0.81 11.11 4.72 −1.09 
Composite score 
 Male — — 0.55 — — 0.69 — — 0.37 — — 0.09 — — −0.45 — — −1.10 
 Female — — 0.79 — — 0.58 — — 0.41 — — −0.23 — — −0.99 — — −1.30 
 19–29 30–39 40–49 50–59 60–69 70–79 
Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD 
Verbal memory 
 Male 43.71 7.65 0.49 44.52 7.42 0.57 40.24 8.09 0.19 37.55 8.63 −0.44 30.81 8.90 −0.63 24.59 6.37 −1.16 
 Female 50.39 7.13 1.07 40.24 8.09 0.19 44.59 3.04 0.56 37.53 8.77 −0.05 25.80 7.06 −1.06 24.39 6.96 −1.18 
Working memory 
 Male 23.80 3.23 0.31 23.77 3.40 0.37 21.45 3.93 0.27 20.62 3.84 −0.72 20.84 11.23 −0.05 13.81 4.20 −0.90 
 Female 24.08 2.82 0.35 21.45 3.93 0.27 22.21 3.04 0.12 20.61 12.98 −0.07 17.83 9.41 −0.41 18.75 14.90 −0.30 
Motor speed 
 Male 81.74 10.75 0.26 85.52 8.41 0.12 84.93 8.39 0.18 83.00 11.77 0.35 71.29 14.40 −0.44 65.81 13.58 −0.81 
 Female 81.94 10.69 0.28 84.93 8.39 0.48 82.42 11.40 0.31 78.11 14.50 0.02 71.93 17.97 −0.40 62.68 18.56 −1.02 
Verbal fluency 
 Male 33.31 7.00 0.29 36.19 8.17 0.38 33.48 6.64 0.31 32.21 8.11 0.15 28.94 6.01 −0.26 25.81 5.48 −0.65 
 Female 35.17 6.22 0.52 33.48 6.64 0.31 34.65 7.70 0.45 28.89 7.76 −0.26 23.57 5.97 −0.93 22.93 5.82 −1.01 
Attention & processing speed 
 Male 68.63 8.77 0.79 67.00 11.28 0.85 59.00 8.00 0.21 52.59 12.65 −0.17 44.26 10.13 −0.66 34.15 10.54 −1.26 
 Female 72.22 7.77 1.00 59.00 8.00 0.21 61.12 9.87 0.34 54.16 10.60 −0.73 38.33 11.90 −1.01 29.86 11.68 −1.51 
Executive function                  
 Male 17.66 3.42 0.43 18.06 2.39 0.46 18.23 1.70 0.52 16.76 3.77 0.22 15.39 4.43 −0.10 14.11 4.08 −0.39 
 Female 17.94 2.50 0.50 18.03 1.70 0.52 16.50 3.60 0.16 13.05 5.00 −0.64 12.30 4.50 −0.81 11.11 4.72 −1.09 
Composite score 
 Male — — 0.55 — — 0.69 — — 0.37 — — 0.09 — — −0.45 — — −1.10 
 Female — — 0.79 — — 0.58 — — 0.41 — — −0.23 — — −0.99 — — −1.30 

The correlations between the performance of each subscale and the demographic variables (age, gender, and educational level) are listed in Table 3. We found that more advanced age was significantly correlated to reduced performance of all the BACS subtests (p < .001). Controlling for the effects of age and education, male participants performed worse in verbal memory (B = −1.76, p = .033) but better in executive function (B = 1.69, p < .001). Compared to participants whose educational levels were high school or lower, their counterparts who attended college or above performed better in verbal memory (B = 5.72, p < .001), verbal fluency (B = 4.45, p < .001), attention & processing speed (B = 9.02, p < .001), and executive function (B = 1.75, p < .001).

Table 3.

Effects of age, sex and education level on neurocognitive performance

 Age (A) Sex (S) Education (E) 
 95% CI p 95% CI p 95% CI p 
Verbal memory 
 Model 1 (A) −0.49 −0.54 to −0.44 <.001*       
 Model 2 (S)    −1.34 −3.68 to 1.01 .262    
 Model 3 (E)       13.17 11.11 to 15.24 <.001* 
 Model 4 (A+S+E) −0.40 −0.46 to −0.34 <.001* −1.76 −3.37 to −0.14 .033* 5.72 3.71 to 7.74 <.001* 
Working memory 
 Model 1 (A) −0.16 −0.21 to −0.11 <.001*       
 Model 2 (S)    −0.49 −2.16 to 1.17 .560    
 Model 3 (E)       4.25 2.56 to 5.93 <.001* 
 Model 4 (A+S+E) −0.29 −0.19 to −0.07 <.001* −0.63 −2.21 to 0.95 .434 1.83 −0.14 to 3.79 .069 
Motor speed 
 Model 1 (A) −0.33 −0.41 to −0.25 <.001*       
 Model 2 (S)    2.28 −0.72 to 5.28 .064    
 Model 3 (E)       8.39 5.36 to 11.42 <.001* 
 Model 4 (A+S+E) −0.29 −0.39 to −0.19 <.001* 2.06 −0.73 to 4.85 .148 2.81 −0.67 to 6.28 .114 
Verbal fluency 
 Model 1 (A) −0.23 −0.27 to −0.19 <.001*       
 Model 2 (S)    1.53 −0.09 to 3.15 .064    
 Model 3 (E)       7.57 6.05 to 9.09 <.001* 
 Model 4 (A+S+E) −0.16 −0.21 to −0.11 <.001* 1.23 −0.16 to 2.61 .083 4.45 2.72 to 6.17 <.001* 
Attention & processing speed 
 Model 1 (A) −0.81 −0.87 to −0.75 <.001*       
 Model 2 (S)    −0.53 −3.98 to 2.92 .764    
 Model 3 (E)       21.67 18.79 to 24.54 <.001* 
 Model 4 (A+S+E) −0.67 −0.60 to −0.74 <.001* −1.20 −3.19 to 0.79 .237 9.02 6.54 to 11.50 <.001* 
Executive function 
 Model 1 (A) −0.12 −0.12 to −0.10 <.001*       
 Model 2 (S)    1.81 0.96 to 2.66 <.001*    
 Model 3 (E)       3.65 2.82 to 4.48 <.001* 
 Model 4 (A+S+E) −0.09 −0.12 to −0.07 <.001* 1.69 0.95 to 2.42 <.001* 1.75 0.83 to 2.67 <.001* 
 Age (A) Sex (S) Education (E) 
 95% CI p 95% CI p 95% CI p 
Verbal memory 
 Model 1 (A) −0.49 −0.54 to −0.44 <.001*       
 Model 2 (S)    −1.34 −3.68 to 1.01 .262    
 Model 3 (E)       13.17 11.11 to 15.24 <.001* 
 Model 4 (A+S+E) −0.40 −0.46 to −0.34 <.001* −1.76 −3.37 to −0.14 .033* 5.72 3.71 to 7.74 <.001* 
Working memory 
 Model 1 (A) −0.16 −0.21 to −0.11 <.001*       
 Model 2 (S)    −0.49 −2.16 to 1.17 .560    
 Model 3 (E)       4.25 2.56 to 5.93 <.001* 
 Model 4 (A+S+E) −0.29 −0.19 to −0.07 <.001* −0.63 −2.21 to 0.95 .434 1.83 −0.14 to 3.79 .069 
Motor speed 
 Model 1 (A) −0.33 −0.41 to −0.25 <.001*       
 Model 2 (S)    2.28 −0.72 to 5.28 .064    
 Model 3 (E)       8.39 5.36 to 11.42 <.001* 
 Model 4 (A+S+E) −0.29 −0.39 to −0.19 <.001* 2.06 −0.73 to 4.85 .148 2.81 −0.67 to 6.28 .114 
Verbal fluency 
 Model 1 (A) −0.23 −0.27 to −0.19 <.001*       
 Model 2 (S)    1.53 −0.09 to 3.15 .064    
 Model 3 (E)       7.57 6.05 to 9.09 <.001* 
 Model 4 (A+S+E) −0.16 −0.21 to −0.11 <.001* 1.23 −0.16 to 2.61 .083 4.45 2.72 to 6.17 <.001* 
Attention & processing speed 
 Model 1 (A) −0.81 −0.87 to −0.75 <.001*       
 Model 2 (S)    −0.53 −3.98 to 2.92 .764    
 Model 3 (E)       21.67 18.79 to 24.54 <.001* 
 Model 4 (A+S+E) −0.67 −0.60 to −0.74 <.001* −1.20 −3.19 to 0.79 .237 9.02 6.54 to 11.50 <.001* 
Executive function 
 Model 1 (A) −0.12 −0.12 to −0.10 <.001*       
 Model 2 (S)    1.81 0.96 to 2.66 <.001*    
 Model 3 (E)       3.65 2.82 to 4.48 <.001* 
 Model 4 (A+S+E) −0.09 −0.12 to −0.07 <.001* 1.69 0.95 to 2.42 <.001* 1.75 0.83 to 2.67 <.001* 

Sex: reference group is female; Education level: college and above versus high school or lower; CI, confidence interval. *p < .05.

The Z-scores of each BACS subset standardized using both the U.S. norms (Z-scores-US) and the current norm (Z-scores-TW) are compared and summarized in Table 4. Compared to participants’ Z-scores-US, their Z-scores-TW for verbal memory (Cohen's d = −2.94, p < .001), verbal fluency (Cohen's d = −6.94, p < .001), attention & processing speed (Cohen's d = −0.50, p < .001), executive function (Cohen's d = −0.94, p < .001), and composite score (Cohen's d = −1.58, p < .001) were all significantly lower. With regard to motor speed, Z-scores-TW were significantly higher than Z-scores-US (Cohen's d = 3.30, p < .001). No difference was found between Z-scores-US and Z-scores-TW with regard to working memory.

Table 4.

Comparison of neurocognitive performance (Z-score) standardized using the U.S. norms and which standardized using TW norms

 Raw scores Z-scores standardized using US norm Z-scores standardized using TW norm Statistical value 
 Mean SD Mean SD Mean SD Mean difference t Effect size (Cohen's dp 
Verbal memory 38.00 11.62 −0.92 1.13 0.00 0.99 −0.92 −28.722 −2.943 <.001* 
Working memory 21.20 8.23 0.02 2.11 0.00 0.99 0.02 0.222 0.023 .824 
Motor speed 77.76 14.89 0.71 1.11 −0.03 0.99 0.74 32.184 3.298 <.001* 
Verbal fluency 30.98 8.05 −1.75 0.66 −0.00 0.99 −1.74 −67.731 −6.940 <.001* 
Attention & processing speed 55.24 17.08 −0.21 1.31 −0.01 1.00 −0.20 −4.827 −0.495 <.001* 
Executive function 15.79 4.31 −0.32 1.31 0.03 0.96 −0.35 −9.134 −0.936 <.001* 
Composite score — — −0.67 1.20 −0.00 0.97 −0.67 −15.453 −1.583 <.001* 
 Raw scores Z-scores standardized using US norm Z-scores standardized using TW norm Statistical value 
 Mean SD Mean SD Mean SD Mean difference t Effect size (Cohen's dp 
Verbal memory 38.00 11.62 −0.92 1.13 0.00 0.99 −0.92 −28.722 −2.943 <.001* 
Working memory 21.20 8.23 0.02 2.11 0.00 0.99 0.02 0.222 0.023 .824 
Motor speed 77.76 14.89 0.71 1.11 −0.03 0.99 0.74 32.184 3.298 <.001* 
Verbal fluency 30.98 8.05 −1.75 0.66 −0.00 0.99 −1.74 −67.731 −6.940 <.001* 
Attention & processing speed 55.24 17.08 −0.21 1.31 −0.01 1.00 −0.20 −4.827 −0.495 <.001* 
Executive function 15.79 4.31 −0.32 1.31 0.03 0.96 −0.35 −9.134 −0.936 <.001* 
Composite score — — −0.67 1.20 −0.00 0.97 −0.67 −15.453 −1.583 <.001* 

*p < .05.

Discussion

This study provides normative data for the BACS in a sample of Mandarin-speaking Chinese people between the ages of 19 and 79 years old in Taiwan. We found advanced age to be strongly associated with decreased cognitive function in all dimensions among the general population, which is in line with the previous studies (Hayat et al., 2014; Kaneda, Katagai, & Yasui-Furukori, 2013; Reynolds, & Finkel, 2015). Notably, we did not collect normative data for adolescents younger than 18 years old in this study. Future research should collect normative data of adolescents younger than 18 years old to provide a reference for cognitive assessment in patients with early onset schizophrenia. Regarding gender differences, we found that women performed better in verbal memory, whereas men performed better in executive function. This finding generally agrees with the normative data for the BACS in the English-speaking Chinese population in Singapore (Eng et al., 2013). However, men performed better in working memory in the Italian sample (Anselmetti et al., 2008), whereas women performed better in the token motor test in the U.S. population (Keefe et al., 2008). One study has suggested that gender differences in cognitive functions may change over time and be related to societal improvements in living conditions and educational opportunities (Weber, Skirbekk, Freund, & Herlitz, 2014). Cognitive gender differences may vary among different social and cultural environments (Miller, & Halpern, 2014).

In addition to age and gender, we found that participants’ educational levels positively correlated to four of the six BACS domains, which were verbal memory, verbal fluency, attention & processing speed, and executive function. However, we had decided not to include education-corrected norms for the following reasons: (a) educational opportunities may change over time and may not necessarily be related to an individual's cognitive function. For example, people attending university or graduate school in Taiwan has become easier over the past decades (Ministry of Education, 2016); (b) if normative tables contained stratification based upon education levels, the case numbers per cell would be too small to yield reliable estimates; (c) patients with schizophrenia seldom attain their potential education levels, so adjusting for education in schizophrenia will deflate the magnitude of patients’ cognitive deficits. Therefore, we did not use education-corrected norms that were also not recommended in previous studies that have investigated similar topics (Anselmetti et al., 2008; Eng et al., 2013; Keefe et al., 2008).

We compared the U.S. norms and the current TW norms of each BACS subtest within each sample. The Z-scores-US of verbal memory and verbal fluency were significantly lower than those based on our normative data (Z-scores-TW). One possible explanation for such a phenomenon is the grammar discrepancy of different versions of the BACS. In the List Learning Test (verbal memory) of the English-language BACS, participants are presented with 15 “words” and then asked to remember as many as possible. In contrast, in the same subtest of the Chinese BACS, participants are presented with 15 “phrases” with two Chinese characters and then asked to remember them. Meanwhile, in the Controlled Oral Word Association Test (letter fluency) in the English BACS, participants have 60 s to generate as many words as possible that begin with the letter F or S (Keefe, Poe, Walker, & Harvey, 2006; Keefe, Poe, Walker, Kang, & Harvey, 2006). In contrast to English (a phonographic language), Mandarin is an ideogram, and a Chinese phrase with identifiable meaning generally contains at least two Chinese characters. Therefore, the Controlled Oral Word Association Test in the Chinese BACS asks patients to generate as many “phrases” as possible that begin with the words “Kou” or “Jing.” Our results indicate that verbal domains in the Chinese BACS are more difficult than those in the English BACS. Therefore, directly applying U.S. norms to determine the performance of the Chinese BACS results in the underestimation of verbal abilities.

Compared to the U.S. population, performance of the Symbol Coding (attention & processing speed) and Tower of London Test (executive function) in our sample was inferior, whereas performance of the Token Motor Task (motor speed) was superior. In general, these three tests are less influenced by language characteristics and cultural differences. However, neurocognitive functions are most likely affected by living environment, educational system, and cultural values (Norbury, & Sparks, 2013). Compared to Western populations, this study's population in Taiwan may have better fine motor skills that correspond to a better motor speed, but the reasoning and processing abilities in our study sample are poorer than those of the U.S. individuals. In view of these findings, using the U.S. norms on a Mandarin-speaking Chinese sample increases the likelihood of false impairments with regard to language, attention & processing speed, and executive function. This bias could result in incorrectly interpreting cognitive performance in clinical research or intervention trials. Altogether, appropriate norms are essential for evaluating the cognitive profiles of Mandarin-speaking Chinese patients with schizophrenia, as well as for comparing the observed performances of healthy controls.

Recruitment of the participants in this study was not through random sampling. Instead, we conveniently recruited the volunteers from hospitals and nearby communities in Taiwan. Therefore, some of the study's participants were hospital employees, so the study sample may not perfectly represent the demographics of Taiwan's Han Chinese population. For example, hospital personnel may be generally younger and better educated, so their performance on the BACS tests will probably be better than that of an average person. Still, we found that directly applying Western BACS norms to a Mandarin-speaking population of ethnic Chinese people creates biased interpretations, particularly with regard to the verbal domains of cognitive function. Furthermore, this is the first study to provide Chinese BACS data of a large-scale Mandarin-speaking population. Because Chinese Mandarin is one of the most widely spoken languages in the world, this study's findings provide an important reference for clinical use and research purposes related to cognitive assessment in schizophrenic patients.

Acknowledgements

The authors express their deepest gratitude to Professor Richard S.E. Keefe for granting us the use of the Chinese version of the BACS. We also thank Joanne Lo for helping participant recruitment and thank all of the individuals who participated in this study.

Conflict of Interest

None declared.

Funding

Supported by Chang Gung Memorial Hospital, Taiwan (CMRPG8C1051, CMRPG8C1291, and CMRPG8E1351). The funding sources had no involvement in the study design, collection, analysis and interpretation of data, writing of the report or the decision to submit the article for publication.

References

Anselmetti
,
S.
,
Poletti
,
S.
,
Ermoli
,
E.
,
Bechi
,
M.
,
Cappa
,
S.
,
Venneri
,
A.
, et al
. (
2008
).
The Brief Assessment of Cognition in Schizophrenia
.
Normative data for the Italian population. Neurological Sciences
 ,
29
(
2
),
85
92
.
Araujo
,
G. E.
,
Resende
,
C. B.
,
Cardoso
,
A. C.
,
Teixeira
,
A. L.
,
Keefe
,
R. S.
, &
Salgado
,
J. V.
(
2015
).
Validity and reliability of the Brazilian Portuguese version of the BACS (Brief Assessment of Cognition in Schizophrenia)
.
Clinics (Sao Paulo)
 ,
70
(
4
),
278
282
.
Bakkour
,
N.
,
Samp
,
J.
,
Akhras
,
K.
,
El Hammi
,
E.
,
Soussi
,
I.
,
Zahra
,
F.
, et al
. (
2014
).
Systematic review of appropriate cognitive assessment instruments used in clinical trials of schizophrenia, major depressive disorder and bipolar disorder
.
Psychiatry Research
 ,
216
(
3
),
291
302
.
Barch
,
D. M.
, &
Ceaser
,
A.
(
2012
).
Cognition in schizophrenia: core psychological and neural mechanisms
.
Trends in Cognitive Sciences
 ,
16
(
1
),
27
34
.
Bortolato
,
B.
,
Miskowiak
,
K. W.
,
Kohler
,
C. A.
,
Vieta
,
E.
, &
Carvalho
,
A. F.
(
2015
).
Cognitive dysfunction in bipolar disorder and schizophrenia: A systematic review of meta-analyses
.
Neuropsychiatric Disease and Treatment
 ,
11
,
3111
3125
.
Bralet
,
M. C.
,
Falissard
,
B.
,
Neveu
,
X.
,
Lucas-Ross
,
M.
,
Eskenazi
,
A. M.
, &
Keefe
,
R. S.
(
2007
).
Validation of the French version of the BACS (the brief assessment of cognition in schizophrenia) among 50 French schizophrenic patients
.
European Psychiatry
 ,
22
(
6
),
365
370
.
Chen
,
C. K.
,
Lin
,
S. K.
,
Chen
,
Y. C.
,
Huang
,
M. C.
,
Chen
,
T. T.
,
Ree
,
S. C.
, et al
. (
2015
).
Persistence of psychotic symptoms as an indicator of cognitive impairment in methamphetamine users
.
Drug and Alcohol Dependence
 ,
148C
,
158
164
.
Chen
,
Y. C.
,
Wang
,
L. J.
,
Lin
,
S. K.
, &
Chen
,
C. K.
(
2015
).
Neurocognitive profiles of methamphetamine users: comparison of those with or without concomitant ketamine use
.
Substance Use & Misuse
 ,
50
(
14
),
1778
1785
.
Ministry of Education, Republic of China (Taiwan)
. (
2016
). http://english.moe.gov.tw/lp.asp?CtNode=11429&CtUnit=1345&BaseDSD=16&mp=1.
Eng
,
G. K.
,
Lam
,
M.
,
Bong
,
Y. L.
,
Subramaniam
,
M.
,
Bautista
,
D.
,
Rapisarda
,
A.
, et al
. (
2013
).
Brief Assessment of Cognition in Schizophrenia: Normative data in an English-Speaking Ethnic Chinese sample
.
Archives of Clinical Neuropsychology
 ,
28
(
8
),
845
858
.
Harvey
,
P. D.
,
Bowie
,
C. R.
, &
Friedman
,
J. I.
(
2001
).
Cognition in schizophrenia
.
Current Psychiatry Reports
 ,
3
(
5
),
423
428
.
Hayat
,
S. A.
,
Luben
,
R.
,
Moore
,
S.
,
Dalzell
,
N.
,
Bhaniani
,
A.
,
Anuj
,
S.
, et al
. (
2014
).
Cognitive function in a general population of men and women: A cross sectional study in the European Investigation of Cancer-Norfolk cohort (EPIC-Norfolk)
.
BMC Geriatrics
 ,
14
,
142
.
Hoe
,
M.
,
Nakagami
,
E.
,
Green
,
M. F.
, &
Brekke
,
J. S.
(
2012
).
The causal relationships between neurocognition, social cognition and functional outcome over time in schizophrenia: a latent difference score approach
.
Psychological Medicine
  (pp.
1
13
).
Kaneda
,
A.
,
Katagai
,
T.
, &
Yasui-Furukori
,
N.
(
2013
).
Comparing the influences of age and disease on the Brief Assessment of Cognition in Schizophrenia in Japanese patients with schizophrenia
.
Neuropsychiatric Disease and Treatment
 ,
9
,
1203
1208
.
Kaneda
,
Y.
,
Sumiyoshi
,
T.
,
Keefe
,
R.
,
Ishimoto
,
Y.
,
Numata
,
S.
, &
Ohmori
,
T.
(
2007
).
Brief assessment of cognition in schizophrenia: Validation of the Japanese version
.
Psychiatry and Clinical Neurosciences
 ,
61
(
6
),
602
609
.
Keefe
,
R. S.
,
Goldberg
,
T. E.
,
Harvey
,
P. D.
,
Gold
,
J. M.
,
Poe
,
M. P.
, &
Coughenour
,
L.
(
2004
).
The Brief Assessment of Cognition in Schizophrenia: Reliability, sensitivity, and comparison with a standard neurocognitive battery
.
Schizophrenia Research
 ,
68
(
2–3
),
283
297
.
Keefe
,
R. S.
, &
Harvey
,
P. D.
(
2012
).
Cognitive impairment in schizophrenia
.
Handbook of Experimental Pharmacology
 , (
213
),
11
37
.
Keefe
,
R. S.
,
Harvey
,
P. D.
,
Goldberg
,
T. E.
,
Gold
,
J. M.
,
Walker
,
T. M.
,
Kennel
,
C.
, et al
. (
2008
).
Norms and standardization of the Brief Assessment of Cognition in Schizophrenia (BACS)
.
Schizophrenia Research
 ,
102
(
1–3
),
108
115
.
Keefe
,
R. S.
,
Poe
,
M.
,
Walker
,
T. M.
, &
Harvey
,
P. D.
(
2006
).
The relationship of the Brief Assessment of Cognition in Schizophrenia (BACS) to functional capacity and real-world functional outcome
.
Journal of Clinical and Experimental Neuropsychology
 ,
28
(
2
),
260
269
.
Keefe
,
R. S.
,
Poe
,
M.
,
Walker
,
T. M.
,
Kang
,
J. W.
, &
Harvey
,
P. D.
(
2006
).
The Schizophrenia Cognition Rating Scale: An interview-based assessment and its relationship to cognition, real-world functioning, and functional capacity
.
The American Journal of Psychiatry
 ,
163
(
3
),
426
432
.
Kern
,
R. S.
,
Green
,
M. F.
,
Nuechterlein
,
K. H.
, &
Deng
,
B. H.
(
2004
).
NIMH-MATRICS survey on assessment of neurocognition in schizophrenia
.
Schizophrenia Research
 ,
72
(
1
),
11
19
.
Kitchen
,
H.
,
Rofail
,
D.
,
Heron
,
L.
, &
Sacco
,
P.
(
2012
).
Cognitive impairment associated with schizophrenia: A review of the humanistic burden
.
Advances in Therapy
 ,
29
(
2
),
148
162
.
Kuo
,
C. J.
,
Tang
,
H. S.
,
Tsay
,
C. J.
,
Lin
,
S. K.
,
Hu
,
W. H.
, &
Chen
,
C. C.
(
2003
).
Prevalence of psychiatric disorders among bereaved survivors of a disastrous earthquake in Taiwan
.
Psychiatric Services
 ,
54
(
2
),
249
251
.
Mazhari
,
S.
,
Parvaresh
,
N.
,
Eslami Shahrbabaki
,
M.
,
Sadeghi
,
M. M.
,
Nakhaee
,
N.
, &
Keefe
,
R. S.
(
2014
).
Validation of the Persian version of the brief assessment of cognition in schizophrenia in patients with schizophrenia and healthy controls
.
Psychiatry and Clinical Neurosciences
 ,
68
(
2
),
160
166
.
Miller
,
D. I.
, &
Halpern
,
D. F.
(
2014
).
The new science of cognitive sex differences
.
Trends in Cognitive Sciences
 ,
18
(
1
),
37
45
.
Norbury
,
C. F.
, &
Sparks
,
A.
(
2013
).
Difference or disorder? Cultural issues in understanding neurodevelopmental disorders
.
Developmental Psychology
 ,
49
(
1
),
45
58
.
Reynolds
,
C. A.
, &
Finkel
,
D.
(
2015
).
A meta-analysis of heritability of cognitive aging: minding the “missing heritability” gap
.
Neuropsychology Review
 ,
25
(
1
),
97
112
.
Sachs
,
G.
,
Winklbaur
,
B.
,
Jagsch
,
R.
, &
Keefe
,
R. S.
(
2011
).
Validation of the German version of the Brief Assessment of Cognition in Schizophrenia (BACS)—Preliminary results
.
European Psychiatry
 ,
26
(
2
),
74
77
.
Segarra
,
N.
,
Bernardo
,
M.
,
Gutierrez
,
F.
,
Justicia
,
A.
,
Fernadez-Egea
,
E.
,
Allas
,
M.
, et al
. (
2011
).
Spanish validation of the Brief Assessment in Cognition in Schizophrenia (BACS) in patients with schizophrenia and healthy controls
.
European Psychiatry
 ,
26
(
2
),
69
73
.
Sheehan
,
D. V.
,
Lecrubier
,
Y.
,
Sheehan
,
K. H.
,
Amorim
,
P.
,
Janavs
,
J.
,
Weiller
,
E.
, et al
. (
1998
).
The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10
.
The Journal of Clinical Psychiatry
 ,
59
(
Suppl 20
),
22
33
; quiz 34–57.
Tosato
,
S.
, &
Dazzan
,
P.
(
2005
).
The psychopathology of schizophrenia and the presence of neurological soft signs: a review
.
Current Opinion in Psychiatry
 ,
18
(
3
),
285
288
.
Wang
,
L. J.
,
Lin
,
P. Y.
,
Lee
,
Y.
,
Huang
,
Y. C.
,
Hsu
,
S. T.
,
Hung
,
C. F.
, et al
. (
2016
).
Validation of the Chinese Version of Brief Assessment of Cognition in Schizophrenia (BACS)
.
Neuropsychiatric Disease and Treatment
 ,
12
,
2819
2826
.
Weber
,
D.
,
Skirbekk
,
V.
,
Freund
,
I.
, &
Herlitz
,
A.
(
2014
).
The changing face of cognitive gender differences in Europe
.
Proceedings of the National Academy of Sciences of the United States
 ,
111
(
32
),
11673
11678
.