1257

Journal of Alzheimer’s Disease 45 (2015) 1257–1268 DOI 10.3233/JAD-142880 IOS Press

Neuropsychological Profiles and Verbal Abilities in Lifelong Bilinguals with Mild Cognitive Impairment and Alzheimer’s Disease Magdalena Eva Kowolla,∗ , Christina Degena , Saskia Gladisa and Johannes Schr¨odera,b a University

Hospital Heidelberg, Section of Geriatric Psychiatry, Heidelberg, Germany University, Institute of Gerontology, Heidelberg, Germany

b Heidelberg

Accepted 31 January 2015

Abstract. Bilingualism is associated with enhanced executive functioning and delayed onset of mild cognitive impairment (MCI) and Alzheimer’s disease (AD). Here, we investigated neuropsychological differences between mono- and bilingual patients with MCI and AD as well as the respective effects of dementia on the dominant and non-dominant language of bilinguals. 69 patients with MCI (n = 22) or AD (n = 47) and 17 healthy controls were included. 41 subjects were classified as lifelong bilinguals (mean age: 73.6; SD = 11.5) and 45 as monolinguals (mean age: 78.1; SD = 10.9). Neuropsychological performance was assessed on the CERAD-NP, the clock-drawing test, and the logical memory subscale of the Wechsler Memory Scale. Neuropsychological profiles showed only minor nonsignificant differences between mono- and bilingual subjects when compared between diagnostic groups. Bilingual MCI patients scored significantly lower on the verbal fluency and picture naming task in their dominant language than bilingual controls. Bilingual AD patients showed a reduced performance in their nondominant language when compared to bilingual MCI patients and bilingual controls (main effect language dominance: verbal fluency task p < 0.001; BNT p < 0.001). Bilingual MCI and AD patients show a similar pattern of neuropsychological deficits as monolingual patients do. The dominant language appears to be compromised first in bilingual MCI patients, while severe deficits of the nondominant language develop later in the course with manifestation of AD. These findings are important for the diagnostic work up of bilingual patients and the development of improved care concepts for bilingual patients such as migrant populations. Keywords: Alzheimer’s disease, bilingualism, mild cognitive impairment, picture naming, verbal fluency

INTRODUCTION Lifelong bilingualism is associated with both advantages and disadvantages for cognitive functioning, where upon a protective effect against Alzheimer’s disease (AD) [1] versus shortcomings with respect to lexical access and verbal retrieval [2, 3] are of par∗ Correspondence

to: Magdalena Eva Kowoll, Dipl.-Psych., Section of Geriatric Psychiatry, University of Heidelberg, Voßstr. 4, 69115 Heidelberg, Germany. Tel.: +49 6221 56 38706; Fax: +49 6221 56 5327; E-mail: [email protected].

ticular clinical importance. Studies have demonstrated that bilingualism is linked to relatively delayed onset of AD-related cognitive deficits [1, 4–6] and the manifestation of mild cognitive impairment (MCI) and AD [4, 7, 8] [for more results, see 9]. Bilinguals demonstrate enhanced executive functioning across the lifespan [10–16, review in 17, for a discussion of bilingual executive function advantages 18], especially in tasks involving inhibitory control and conflict resolution [19] but also in non-verbal task switching [20], auditory attention [21], and spatial working memory [22]. Moreover, when compared to monolinguals, bilinguals

ISSN 1387-2877/15/$35.00 © 2015 – IOS Press and the authors. All rights reserved

1258

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong

display faster reaction times in executive control tasks [10, 13, 23] and in congruent and non-congruent trials [review in 24]. It has been put forward that the lifelong demand to attend to two languages leads to reorganization of specific brain networks thereby enhancing executive control [25]. Using functional magnetic resonance imaging (fMRI), Gold et al. [23] provided evidence of a neural basis for the bilingual cognitive control boost in aging by analyzing neural efficiency and performance in younger mono- versus bilinguals (mean age: 32.2, SD: 3.3 versus 31.6, SD: 4.3) and older mono- versus bilinguals (mean age: 64.4, SD: 5.1 versus 63.9, SD: 4.0), while completing a perceptual task-switching experiment. Bilingual older subjects outperformed their monolingual peers, while displaying decreased activation in the left lateral frontal and cingulate cortex similar to younger subjects. Moreover, a diffusion tensor imaging study revealed more preserved white matter integrity and stronger anterior to posterior functional connectivity in lifelong bilinguals compared to monolingual subjects [26]. Bilinguals seem to recruit brain regions involved in language control while performing non-linguistic executive tasks as demonstrated in an fMRI study comparing brain activity in 40 healthy mono- and bilinguals [20]. Bilingual participants exhibited relatively smaller reaction times and error rates thereby activating the left inferior frontal cortex and the left striatum, which have previously been associated with language control. That bilinguals activate their brains more efficiently is further corroborated by Abutalebi et al. [27] who demonstrated a lower activation of the dorsal anterior cingulate cortex, i.e., a structure tightly bound to executive control functions, in the bilingual group although they outperformed monolinguals. Bilinguals outperform monolinguals in tests assessing nonverbal [28] and verbal episodic memory [29] reflecting differences observed in tests of executive functioning. Clinical and neuroimaging studies have demonstrated the involvement of frontal cortices in both executive functioning [30–33] and episodic memory in monolinguals [34–36]. In spite of clear advantages in executive functioning, naming ability and word-fluency seem to be affected negatively by bilingualism [12, 37]. Thus, bilinguals have a smaller vocabulary in each language [38], exhibit slower reaction times in tasks addressing word naming ability [3], even in their dominant language [2], produce fewer words in verbal fluency tasks, especially of semantic verbal fluency [37], and name fewer

figures correctly in the Boston Naming Test (BNT) [12, 39]. The weaker links hypothesis was proposed to explain this performance gap between bi- and monolinguals [3, 40]. Accordingly, weaker links between semantics and phonology exist in each lexical system, as bilinguals use each language less frequently than monolinguals [3]. Only very few studies have analysed neuropsychological differences between mono- and bilinguals with MCI and AD, in particular with respect to differential effects on the dominant and nondominant language. Gollan et al. [41] reported greater losses in the dominant than the nondominant language in a group of 29 AD patients. A similar dissociation was found by Ivanova and collegues [42] in picture naming. However, their studies were limited to naming performance. In the present study, we sought to investigate neuropsychological performance in monolinguals and bilinguals with MCI, AD, and healthy controls as well as differential effects on dominant and nondominant languages. We expected to find more preserved executive functioning at more compromised verbal functions in the bi- versus the monolingual patients. Moreover, the existence of richer connections from concepts to words in the dominant relative to the nondominant language may make the dominant language more vulnerable to brain damage [41]. Furthermore, impairments in verbal fluency and word naming ability occur in early and even preclinical stages of AD [43–45]. It is therefore plausible that the dominant language is disrupted earlier by brain damage than the nondominant language. Therefore, losses of verbal abilities should first affect dominant rather than nondominant languages. MATERIALS AND METHODS Participants A total of 86 subjects were recruited between June 2012 and March 2014 from the Memory Clinic of the University of Heidelberg or from local nursing homes. 45 of the subjects were classified as monolinguals, 41 as lifelong bilinguals following Bialystok et al.’s [p. 460, 1] criterion: patients were classified as bilingual if they “ . . . had spent the majority of their lives, at least from early adulthood, regularly using at least two languages”. 17 subjects were cognitively healthy, 22 were diagnosed with MCI according to the aging-associated cognitive decline criteria (AACD, International Psychogeriatric Association working Party) [46], and 47 individuals were

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong

diagnosed with AD using the NINCDS–ADRDA criteria [47]. Diagnoses were established by consensus between an experienced geriatric psychiatrist and an experienced psychologist. The bilingual participants consisted of speakers of 14 different first languages, of which the most common were Arabic (n = 2), English (n = 2), French (n = 2), German (n = 22), Hungarian (n = 2), Spanish (n = 2), and Turkish (n = 2). There were 11 different second languages spoken; the most common were English (n = 7), French (n = 4), German (n = 17), Hungarian (n = 3), Polish (n = 2), and Russian (n = 3). 58.5% (n = 24) of bilinguals were multilingual and able to use more than two languages. 33 bilinguals were immigrants to Germany. Their countries of orgin were Hungary (n = 6), Ukraine (n = 3), Romania (n = 3), France (n = 2), Czechoslovakia (n = 2), Poland (n = 2), Turkey (n = 2), England, Eritrea, Finland, Italy, Jericho, Palestine, Peru, Portugal, Serbia, Slovakia, Spain, Taiwann and Trinidad (each n = 1). Immigration occurred predominantly between the age of 21 and 30 (n = 12), 11 and 20 (n = 6), and 41 and 50 (n = 5). 11 monolinguals were not born in Germany. The monolingual migrants immigrated to Germany between the age of 2 and 20 (n = 3), 21 and 30 (n = 1), 31 and 50 (n = 4), and 51 and 60 (n = 1). From two monolingual migrants no information about their age of immigration could be compiled due to their cognitive deficits. Two monolingual migrants spoke only Italian (n = 1) and Turkish (n = 1). 18 monolinguals reported to speak a further language, while 27 monolinguals could speak only their mother tongue. 18 monolinguals (of whom n = 12 were born in Germany) who knew another language were not classified as bilinguals because they did not match Bialystok et al.’s [1] criterion of lifelong and regular bilingualism. Their proficiency level based on selfratings using a scale of 1–7 with 1 being “little to no knowledge” and 7 being “like a native speaker” according to Gollan et al. [8] was: for German: M = 6.44, SD = 1.36; for the other language: M = 3.87, SD = 1.89.

Procedure The study was approved by the Ethical Committee of the University of Heidelberg. After complete description of the study to the participants, informed consent was obtained. Participants were carefully screened for language history, occupational history, fluency in German and other languages, place of birth, and date of immigration.

1259

For bilinguals, a language history questionnaire was applied and the 15-items version of the BNT and the 1-min semantic verbal fluency tasks (animals) of the CERAD were conducted to define language dominance following an objective method by Gollan et al. [8] determining that language as dominant, in which higher scores in the BNT and the verbal fluency task were achieved. For comparison, Gollan et al.’s [8] subjective measure of language dominance was slightly modified asking participants, in which language they feel most fluent. The objective and subjective measures of language dominance differed for 12 participants. 53.7% conducted the BNT and the verbal fluency task first in their dominant language and subsequently in their nondominant language. This proportion was calculated after the neuropsychological testing session. For language dominance analyses only bilinguals (n = 35) were selected who obtained the BNT and the verbal fluency task in both languages. Bilinguals with at least one missing value in one of these subtests (n = 3) and bilinguals who scored diametrical in two languages in the BNT and the verbal fluency task (n = 3) were excluded from language dominance analyses. The dominant language of two of those bilinguals with at least one missing value was German, the dominant language of one of those bilinguals was Spanish. The dominant language of 60.0% (n = 21) of the bilinguals was German. The other dominant languages were: English (5.7%, n = 2), Russian (5.7%, n = 2), Turkish (5.7%, n = 2), Finnish, French, Portuguese, Slovakian, Spanish, Taiwanese, Tigrinya, and Transylvanian Saxon dialect (each 2.9%, n = 1). German was the dominant language of 16 (of overall n = 33) bilingual migrants. All bilinguals whose dominant language was not German (n = 14) were immigrants to Germany. Three bilingual migrants scored diametrically in the the BNT and the verbal fluency task. 14 of the German-dominant bilinguals were immigrants, and 7 German-dominant bilinguals were born in Germany. In Table 1, more information about the age of acquisition and proficiency level for each of the two languages of bilinguals is provided. Because all bilinguals were proficient in German, the instruction to conduct the BNT and the verbal fluency task in their second language could be given in German. This procedure was also chosen to avoid negative stereotypes potentially associated with the usage of another language. If the respective language was not spoken by the investigator, a translater (in most cases a colleague or a relative) was called upon when necessary.

1260

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong Table 1 Mean and standard deviation of language proficiency and age of language acquisition in the 41 bilinguals

Characteristic M ± S.D./n n German self-rated speakinga Other language self-rated speakinga Frequency of German useb Frequency of other language useb Age of acquisition of German Age of acquisition of other language

German dominant (A)

Other language dominant (B)

Scored diametrical (C)

24 6.6 (0.8) 3.7 (1.5) 1.0 (0) 3.1 (1.2) 3.2 (6.8) 7.2 (6.3)

14 5.4 (1.7) 6.1 (2.0) 1.0 (0) 1.4 (0.8) 17.6 (12.8) 1.5 (4.6)

3 5.3 (1.5) 6.0 (1.7) 1.0 (0) 3.0 (1.7) 16.7 (14.6) 0 (0)

ANOVA/ Games-Howell

Post-Hocc

F(2,38) = 4.79∗ F(2,37) = 9.96∗∗∗

AB A B, C

a Proficiency

level based on self-ratings using a scale of 1–7 with 1 being “little to no knowledge” and 7 being “like a native speaker” according to Gollan et al. [8]. b Frequency based on self-ratings using a scale of 1–4 with 1 being “daily” and 4 being “seldom” according to Craik et al. [5]. c only significant associations are reported.

Naming and verbal fluency trials were administered according to the standardized BNT and verbal fluency instructions. According to Gollan et al. [8], alternative responses were allowed in the non German language in the BNT to adjust for the problem that the BNT was not designed for use with bilinguals. Neuropsychological test battery Neuropsychological assessment consisted of the German version of the CERAD- NP neuropsychological assessment battery [48–51] including the Mini-Mental State Examination (MMSE), the Trail Making Test (TMT) [52], and the Clock-Drawing Test [53] as well as the subtests logical memory and digit span of the German version of Wechsler Memory Scale (WMS-R and WMS IV) [54, 55]. Furthermore, the short version of the Geriatric Depression Scale [56] was obtained to exclude depressive symptoms. Analysis For statistical analyses, raw data from the individual CERAD, WMS, and TMT subscores were transformed into z-scores that were adjusted for age, gender, and years of education. SPSS for Windows version 22 was used for statistical analyses; p-values equal to or less than 0.05 were considered significant. Analyses of variance with post hoc Games-Howell tests (with repeated measures for language dominance) were calculated in order to compare demographic and clinical data between diagnostic groups. Likelihood-Ratio tests were used where appropriate. RESULTS In a first step, subjects were classified into six diagnostic groups: healthy monolinguals (n = 6) and bilinguals (n = 11), monolinguals (n = 14) and bilin-

guals (n = 8) with MCI, and monolinguals (n = 25) and bilinguals (n = 22) with AD. Demographic and clinical characteristics of the six groups are provided in Table 2. There were significant main effects of diagnoses regarding age, scores on the MMSE and the Clock drawing test (refer to Table 2). There were significant main effects of language group regarding years of education. Bilinguals differed significantly from monolinguals with AD in terms of educational background, since they had enjoyed more years of education (bilinguals: M = 13.10, SD = 4.32; monolinuals: M = 10.71, SD = 3.78; t(84) = 2.73, p = 0.008). With respect to gender only minor, non-significant differences by trend arose (refer to Table 2). Moreover, bilinguals were more likely than monolinguals to be immigrants. There were no significant main effects or differences on the Geriatric Depression Scale. No patient was excluded from analyses due to clinical depression. In a second step, neuropsychological performance was compared between diagnostic groups (Fig. 1). Performances of patients with MCI and AD were reduced in comparison to controls, but did not differ between patient groups. Significant main effects of “diagnoses” arose in all tests applied except constructional praxis, TMT-B, and digit span forward (refer to Table 3). We performed the analyses once again with the subsample, in which the bilinguals whose dominant language was not German (n = 14) were excluded. There were no changes in the main effects and the interactions on all neuropsychological tests in this subsample. For further analysis we contrasted performance of bilinguals in their dominant versus nondominant language on the BNT and the verbal fluency task of the CERAD (Figs. 2 and 3). When comparing diagnostic groups, the lowest scores were obtained by patients with manifest AD followed by those with

42/44 11.8 (4.2)

2.1 (1.4)

3.1 (3.2)

Born in Germany/ immigrant to Germany Years of education

Clock- drawing Test

Geriatric Depression Scale (GDS)

5.0 (5.2)

1.0 (0.0)

6/0 12.2 (1.2)

2/4 29.6 (0.9)

6 70.2 (8.2)

2.6 (3.3)

1.4 (0.8)

1/7 13.4 (3.2)

3/5 26.8 (1.8)

8 71.3 (8.4)

2.4 (2.7)

1.5 (0.8)

10/4 12.2 (3.3)

5/9 26.4 (3.0)

14 77.5 (12.4)

4.1 (3.4)

2.4 (1.2)

4/18 12.5 (4.9)

11/11 22.0 (3.3)

22 77.2 (10.6)

2.6 (3.1)

3.7 (1.2)

18/7 9.5 (4.1)

9/16 18.9 (6.3)

25 80.3 (10.0)

Post-Hoca / t-Test

LR(5) = 32.99∗∗∗ main effect diagnosis: F(2,80) = 2.53, p = 0.086 main effect language: F(1,80) = 4.36, p = 0.040 interaction: F(2,80) = 0.366, p = 0.695 main effect diagnosis: F(2,57) = 25.26, p < 0.001 main effect language: F(1,57) = 2.39, p = 0.128 interaction: F(2,57) = 3.45, p = 0.039 main effect diagnosis: F(2,62) = 0.589, p = 0.558 main effect language: F(1,62) = 2.33, p = 0.630 interaction: F(2,62) = 0.233, p = 0.106

Controls, MCI < AD

Bil > Mono

main effect diagnosis: F(2,80) = 4.89, Controls < MCI, AD p = 0.010 main effect language: F(1,80) = 2.03, p = 0.158 interaction: F(2,80) = 0.208, p = 0.813 LR(5) = 3.48 main effect diagnosis: Controls > MCI > AD F(2,72) = 30.31, p < 0.001 main effect language: F(1,72) = 1.19, p = 0.278 interaction: F(2,72) = 0.991, p = 0.376

Monolingual Bilinguals Monolinguals Bilinguals Monolinguals ANOVA/Likelihood-Ratio Controls (B) with MCI with MCI (D) with AD with AD (C) (E)

S.D., standard deviation; LR, Likelihood-Quotient; Bil, Bilingual; Mon, Monolingual. a Duncan Test; ∗ p ≤ 0.05; ∗∗ p ≤ 0.01; ∗∗∗ p ≤ 0.001.

2.1 (1.4)

1.1 (0.4)

3/8 14.2 (3.9)

7/4 29.0 (1.0)

37/49 23.9 (5.5)

♂/♀ MMSE

Bilingual Controls (A)

86 11 76.0 (11.3) 68.2 (13.2)

Total sample

n Age

M ± S.D./n

Table 2 Demographic and clinical characteristics of bilingual and monolingual controls, bilingual and monolingual patients with MCI, and bilingual and monolingual patients with AD

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong 1261

1262

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong

Fig. 1. z-scores (means) for the subscales ‘logical memory’ and ‘digit span’ of the Wechsler-Memory-Scale and TMT and the subscales of CERAD-NP in bilingual and monolingual controls, bilinguals and monolinguals with MCI and bilinguals and monolinguals with AD. ∗ Verbal Fluency and BNT-scores in bilinguals are illustrated for the language in which higher scores were obtained.

MCI and controls (main effect “diagnosis” on the verbal fluency task: F(2,32) = 25.10, p ≤ 0.001 and BNT: F(2,32) = 4.65, p = 0.017). Moreover, subjects achieved significantly higher scores in their dominant than their nondominant language (main effect “language dominance” in the verbal fluency task F(1,32) = 42.26, ≤0.001 and BNT F(1,32) = 42.39, ≤0.001). With respect to the semantic fluency task this effect was more pronounced in the AD patients (interaction effect “diagnosis*language dominance” in the verbal fluency task F(2,32) = 3.23, p = 0.05, but not in the BNT F(2,32) = 1.38, p = 0.27). Interestingly, for verbal fluency, the difference between bilingual controls and bilingual MCI patients was significantly larger for dominant (controls: 95% CI [21.18, 28.32], patients with MCI: 95% CI [11.55, 16.20]) than nondominant language scores (controls: 95% CI [12.13, 21.37], patients with MCI: 95% CI : [8.65, 14.35]). These results were also statistically significant for the BNT: dominant language scores: controls: 95% CI [14.63, 15.12], patients with MCI: 95% CI [12.72, 14.53], nondominant language scores: controls: 95% CI [9.52, 14.23], patients with MCI: 95% CI : [10.06, 13.19]. The reverse applied for the difference between bilinguals with MCI and AD which was significantly larger for the nondominant (patients with MCI: 95% CI [8.65, 14.35], patients with AD: 95% CI [5.41, 8.49]) than dominant idiom (patients with MCI: 95% CI [11.55, 16.20], patients with AD: [10.68, 14.27]) in the verbal fluency task and by trend in the BNT:

nondominant language scores: patients with MCI: 95% CI [10.06, 13.19], patients with AD: 95% CI [8.24, 10.60], dominant language scores: patients with MCI: 95% CI [12.72, 14.53], patients with AD: 95% CI [12.47, 13.74] Language dominance analyses were repeated using a subjective assessment of the dominant and nondominant language, but did not change the main results (verbal fluency task: main effect „diagnosis”: F(2,35) = 28.37, p ≤ 0.001; main effect “language dominance”: F(1,35) = 8.65, p = 0.006; interaction: F(2,35) = 0.67, p = 0.519 n.s.; BNT: main effect “diagnosis”: F(2,35) = 6.29, p = 0.005; main effect “language dominance”: F(1,35) = 11.15, p = 0.002; interaction: F(2,35) = 0.37, p = 0.693, n.s.).

DISCUSSION The present study yielded two major findings: (i) neuropsychological profiles of bilingual patients with MCI and AD parallel those of monolingual patients; and (ii) an indication that the dominant language is already compromised early in the disease process in patients with MCI while the nondominant language is affected later in the course with manifestation of AD. Neuropsychological profiles demonstrated only minor, non-significant differences between bi-and monolingual subjects when compared across diagnostic groups. In particular, bilingual MCI and AD

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong

1263

Table 3 z-scores for neuropsychological tests of bilingual and monolingual controls, bilingual and monolingual patients with MCI, and bilingual and monolingual patients with AD. Cognitive Domain/Test M ± S.D./n

Bilingual Monolingual Bilinguals Monolinguals Bilinguals Monolinguals ANOVA Controls Controls with MCI with MCI with AD with AD

n Word list immediate recall

11 −0.4 (1.3)

Word list delayed recall

Word list Recognition

Constructional Praxis

−0.3 (0.4)

0.5 (0.1)

0.5 (0.6)

Constructional Praxis Recall 0.8 (0.9)

Verbal fluency

BNT

TMT-A

TMT-B

−0.1 (1.0)

0.8 (0.5)

0.0 (0.9)

6 0.1 (1.0)

0.0 (0.9)

0.4 (0.4)

0.1 (0.8)

8 14 22 25 −1.8 (1.2) −1.5 (1.6) −2.1 (1.5) −2.8 (0.8)

−1.3 (1.3) −1.8 (1.5) −1.8 (1.4) −2.6 (0.7)

−0.7 (1.9) −2.6 (2.8) −2.0 (2.9) −1.4 (1.3)

0.9 (0.2)

−0.5 (1.7) −0.7 (1.6) −0.8 (2.2)

−0.9(1.7) −0.5 (0.8) −1.6 (2.0) −2.9 (1.7) −2.5 (0.7)

0.6 (0.7)

0.4 (0.6)

−1.4 (0.7) −0.7 (1.0) −1.7 (0.7) −2.0 (1.1)

−0.3 (1.3) −0.7 (1.8) −0.8 (1.3) −2.0 (1.9)

−2.1(3.6) −1.6 (3.0) −0.5 (1.0) −5.7 (4.7) −4.6 (3.4)

−0.5 (2.4) −1.7(2.8) −3.0 (4.4) −2.4 (4.5) −3.2 (1.8) −5.7 (4.9)

main effect diagnosis: F(2,42) = 10.61, p ≤ 0.001 main effect language: F(1,42) = 0.006, p = 0.940 interaction: F(2,42) = 0.924, p = 0.405 main effect diagnosis: F(2,42) = 10.40, p ≤ 0.001 main effect language: F(1,42) = 0.82, p = 0.369 interaction: F(2,42) = 0.69, p = 0.508 main effect diagnosis: F(2,42) = 3.93, p = 0.027 main effect language: F(1,42) = 0.52, p = 0.473 interaction: F(2,42) = 1.46, p = 0.244 main effect diagnosis: F(2,46) = 2.76, p = 0.074 main effect language: F(1,46) = 1.87, p = 0.178 interaction: F(2,46) = 0.96, p = 0.390 main effect diagnosis: F(2,46) = 12.66, p ≤ 0.001 main effect language: F(1,46) = 3.27, p = 0.077 interaction: F(2,46) = 2.32, p = 0.110 main effect diagnosis: F(2,71) = 26.78, p ≤ 0.001 main effect language: F(1,71) = 2.74, p = 0.102 interaction: F(2,71) = 2.91, p = 0.061 main effect diagnosis: F(2,70) = 9.84, p ≤ 0.001 main effect language: F(1,70) = 3.27, p = 0.075 interaction: F(2,70) = 0.63, p = 0.538 main effect diagnosis: F(2,63) = 12.55, p ≤ 0.001 main effect language: F(1,63) = 0.002, p = 0.961 interaction: F(2,63) = 1,38, p = 0.259 main effect diagnosis: F(2,41) = 3.05, p = 0.058 main effect language: F(1,41) = 0.92, p = 0.343 interaction: F(2,41) = 0.68, p = 0.513

η2

0.157 0.000

0.151 0.006

0.113 0.007

0.098 0.033

0.176 0.023

0.209 0.011

0.168 0.028

0.167 0.000

0.084 0.013

1264

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong Table 3 (Continued)

Cognitive Domain/Test Bilingual Monolingual Bilinguals Monolinguals Bilinguals Monolinguals ANOVA M ± S.D./n Controls Controls with MCI with MCI with AD with AD Logical Memory I

Logical Memory II

Digit Span Forward

Digit Span Backward

−0.8 (1.6)

−0.8 (1.3)

−0.3 (1.2)

−0.4 (0.9)

0.3 (1.3)

0.0 (1.9)

−0.1(1.7)

0.4 (1.2)

−1.3 (1.1)

−1.4 (1.3)

−0.8 (1.1)

−0.5 (0.8)

−1.4 (1.2)

−2.3 (0.9)

−1.0 (0.7)

−0.3 (0.6)

−2.4 (1.1)

−2.7 (0.6)

−1.1 (1.2)

−1.2 (1.1)

−2.6 (0.9)

−2.7 (1.1)

−0.7 (0.9)

−1.4 (0.7)

main effect diagnosis: F(2,42) = 13.30, p ≤ 0.001 main effect language: F(1,42) = 0.53, p = 0.469 interaction: F(2,42) = 1.11, p = 0.340 main effect diagnosis: F(2,42) = 19.73, p ≤ 0.001 main effect language: F(1,42) = 0.001, p = 0.972 interaction: F(2,42) = 2.03, p = 0.144 main effect diagnosis: F(2,46) = 1.87 p = 0.166 main effect language: F(1,46) = 0.10, p = 0.748 interaction: F(2,46) = 0.22, p = 0.807 main effect diagnosis: F(2,46) = 8.46 p ≤ 0.001 main effect language: F(1,46) = 1.16, p = 0.287 interaction: F(2,46) = 1.44, p = 0.248

η2 0.172 0.003

0.149 0.000

0.050 0.001

0.180 0.012

S.D., standard deviation; η2 , effect size.

patients showed similar neuropsychological profiles as monolingual patients. This finding also included verbal fluency and the TMT B which both address aspects of frontal executive functioning. In contrast, the bilingual healthy controls scored better on the TMT-A and TMT-B than their monolingual counterparts. While this differences did not reach statistical significance, it corresponds to studies which have demonstrated that bilingualism is linked to enhanced executive functions in bilinguals [10, 12, 14–16, review in 17, 57]. Bialystok and colleagues [4] found a significantly smaller Stroop effect in bilingual than monolingual AD- patients but a significantly lower performance in the TMT B in bilingual MCI patients than monolingual ones. That this difference was not significant in the present study may refer to the lower number of MCI patients recruited here. Bialystok et al.’s [4] neuropsychological test battery comprised the TMT, the color-word interference test, and verbal fluency tests (letter fluency, category fluency, and category switching). As in the present study, performance on both parts of the TMT and verbal fluency was more reduced in patients with AD than in patients with MCI. Similarly, performance in the color-word interference test was significantly reduced in patients with AD. Besides a larger sample size, their study differed from our

examination with respect to a number of important methodological aspects; testing sessions on three occasions over a period of approximately one year were of particular importance. From a clinical standpoint, these findings indicate that bilingualism does not necessarily lead to cross-sectional neuropsychological differences in patients with MCI or AD, indicating that the same neuropsychological test may be applied to them. Apart from this, longitudinal [4, 58] and cross sectional studies [1, 4–8, 59] found bilingualism to be associated with a more favorable course in agerelated decline. This beneficial effect is generally explained by higher cognitive reserve associated with bilingualism. Since the neuropsychological tests were administered in German, 34.1% of the participants performed these tests in their nondominant language, which might have led to overall lower performance for the bilingual group (under the assumption that these participants might have performed better if tested in their dominant language). Therefore, we performed the analyses excluding these bilinguals whose dominant language was not German (n = 14). The results remained the same. There were no changes in the main effects or interactions.

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong 30

16

Dominant language Non-dominant language

Dominant language 14

Number of correct named figures

Non-dominant language

Number of correct animals

25

20

15

1265

12

10

8

6

4

2

0

10

Controls

MCI

AD

Fig. 3. Performance (means ± se) in the BNT in healthy bilinguals (n = 8) and bilingual patients with MCI (n = 8) or AD (n = 19). Main effect “language dominance”: F(1,32) = 42.39, p < 0.001, η2 = 0.295. Main effect “diagnosis”: F(2,32) = 4.65, p = 0.017, η2 = 0.135; Interaction: F(2,32) = 1.38, p = 0.27, n.s., η2 = 0.019.

5

0 Controls

MCI

AD

Fig. 2. Performance (means ± s.e.) in the verbal fluency task (number of animals) in healthy bilinguals (n = 8) and bilingual patients with MCI (n = 8) or AD (n = 19). Main effect “language dominance”: F(1,32) = 42.26, p < 0.00, η2 = 0.250; Main effect “diagnosis”: F(2,32) = 25.10, p < 0.001, η2 = 0.821; Interaction: F(2,32) = 3.23, p = 0.05, η2 = 0.038.

In the verbal fluency and the word finding tasks, performance of bilingual subjects was significantly higher in the dominant than the nondominant language. However, this discrepancy was greatest in controls in the verbal fluency task and ameliorated with development of AD. Hence, patients with MCI showed greater losses in their dominant than nondominant languages, while the reverse held true in the MCI-AD comparison for both, the BNT and the verbal fluency task. A similar finding was communicated by Gollan and colleagues [41] who reported a greater sensitivity of dominant than nondominant languages to the effects of AD. For Gollan et al. [41], this finding was consistent with the concept of enhanced connectivity of lexical and conceptual representations in dominant relative to nondominant languages. The integrity of semantic representations is primarily reduced by AD pathology [overview in 41, 60, 61]. At the same time, there is a greater number of conceptual associations in dominant, compared to nondominant language names. It is

therefore plausible that the dominant language is more easily disrupted by brain damage than the nondominant language, making the dominant language more sensitive to AD-related changes than the nondominant language [41]. There also exists a contrasting hypothesis suggesting a more pronounced vulnerability of the nondominant language to the effects of AD [overview in 41]. It is plausible, that retrieval deficits accompanying AD [62–65] may be more pronounced in nondominant than dominant languages due to heightened vulnerability. Accordingly, studies have illustrated an increased interference between competing languages as well as a general tendency to retreat to one’s dominant language in aging [overview in 41, overview in 66]. Likewise, cognitive deficits in AD seem to be more pronounced when tested in nondominant languages [overview in 41, overview in 66]. Further support for this hypothesis comes from caregivers reporting a more frequent use of phrases and words from patients’ primarily acquired languages as dementia progresses [67]. In addition, impairments in verbal fluency and word naming ability occur in early and even preclinical stages of AD already [43–45]. Languages acquired in later life are represented in episodic memory mainly [68–71] which is particularly vulnerable to the disease process. Nevertheless, for the purpose of this study only those bilinguals who have acquired their second languages in early adulthood at

1266

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong

the latest and used it regulary were included. It might be possible that a bias occured, because 18 monolinguals had some knowledge of one other language. However, a complete exclusion of knowing elements of other languages is hardly possible, because foreign languages are taught in school and speaking of dialects within one language is partly considered as bilingualism. Beside phenomena like code switching [72, 73] and borrowing [72] languages contain elements from other languages. These difficulties in delineating mono- and bilingual participants may have increased the risk of false negative (type II) but not false positive (type I error) results. In our study, bilinguals were more likely than monolinguals to be immigrants, representing a potentially confounding factor. Immigration status and not bilingualism may have been responsible for the minor non-significant differences between language groups on neuropsychological performance. However, that possibility is ruled out by the large heterogeneity of the bilingual migrant participants including 20 different home countries. Most notably, in the study by Gollan et al. [41], bilinguals’ language dominance was determined rather subjectively. Participants who felt they would obtain higher neuropsychological test scores if tested primarily in English were classified as Englishdominant and participants who preferred to be tested in Spanish were classified as Spanish-dominant. The current study applied an objective assessment of language dominance [8]. An analysis taking a subjective scoring method of language dominance into account by asking participants, in which language they feel most fluent, revealed similar results. Our findings confirm the presence of language impairments in patients with AD and MCI, as well as an increase of deficits as the disease aggravates. Bilingual and monolingual patients with MCI and AD showed similar neuropsychological profiles at a given severity of cognitive deficits. Primarily, verbal fluency in the nondominant language was affected. This could be an important indication for nonmedical therapy. Our finding that the nondominant language is particularly affected with manifestation of AD underlines the rational for preventive measures, in particular language courses for bilinguals who immigrated to Germany and whose dominant language is not German. The respective courses should be tailored to this group of “young-old” citizens who are typically just about to be pensioned. Along with this, the usage of non verbal material such as sketches, etc., to facilitate the communication with the patients should be encouraged. The potential input of translators in the communication appears to be

limited given the decline in both, the dominant and the nondominant language. This differentiation is important for the development of improved care concepts for bilingual patients such as migrant populations. ACKNOWLEDGMENTS The study was supported by the “Ministry for Work, Social Order, Family, Women and Senior Citizen” of the State Baden-W¨urttemberg, Germany and the “Robert Bosch Foundation/Stuttgart”, Germany. Authors’ disclosures available online (http://jalz.com/manuscript-disclosures/14-2880r1). REFERENCES [1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

Bialystok E, Craik FIM, Freedman M (2007) Bilingualism as a protection against the onset of symptoms of dementia. Neuropsychologia 45, 459-464. Gollan TH, Montoya RI, Fennema-Notestine C, Morris SK (2005) Bilingualism affects picture naming but not picture classification. Mem Cogn 33, 1220-1234. Gollan TH, Montoya RI, Cera C, Sandoval TC (2008) More use almost always means a smaller frequency effect: Aging, bilingualism, and the weaker links hypothesis. J Mem Lang 58, 787-814. Bialystok E, Craik FI, Binns MA, Ossher L, Freedman M (2014) Effects of bilingualism on the age of onset and progression of MCI and AD: Evidence from executive function tests. Neuropsychology 28, 290-304. Craik FIM, Bialystok E, Freedman M (2010) Delaying the onset of Alzheimer disease: Bilingualism as a form of cognitive reserve. Neurology 75, 1726-1729. Alladi S, Bak TH, Duggirala V, Surampudi B, Shailaja M, Shukla AK, Chaudhuri JR, Kaul S (2013) Bilingualism delays age at onset of dementia, independent of education and immigration status. Neurology 81, 1938-1944. Ossher L, Bialystok E, Craik FIM, Murphy KJ, Troyer AK (2013) The Effect of Bilingualism on Amnestic Mild Cognitive Impairment. J Gerontol B Psychol Sci Soc Sci 68, 8-12. Gollan TH, Salmon DP, Montoya RI, Galasko DR (2011) Degree of bilingualism predicts age of diagnosis of Alzheimer’s disease in low-education but not in highly educated Hispanics. Neuropsychologia 49, 3826-3830. Zahodne LB, Schofield PW, Farrell MT, Stern Y, Manly JJ (2014) Bilingualism does not alter cognitive decline or dementia risk among Spanish-speaking immigrants. Neuropsychology 28, 238-246. Bialystok E, Craik FIM, Klein R, Viswanathan M (2004) Bilingualism, aging, and cognitive control: Evidence from the Simon task. Psychol Aging 19, 290-303. Bialystok E, Martin MM, Viswanathan M (2005) Bilingualism across the lifespan: The rise and fall of inhibitory control. Int J Bilingualism 9, 103-119. Bialystok E, Craik F, Luk G (2008) Cognitive control and lexical access in younger and older bilinguals. J Exp Psychol Learn Mem Cogn 34, 859-873. Costa A, Hernandez M, Costa-Faidella J, Sebastian-Galles N (2009) On the bilingual advantage in conflict processing: Now you see it, now you don’t. Cognition 113, 135-149.

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong [14]

[15]

[16] [17] [18]

[19]

[20]

[21]

[22]

[23]

[24]

[25] [26]

[27]

[28]

[29]

[30] [31]

[32] [33] [34]

[35]

Christoffels IK, Haan AM, Steenbergen L, Wildenberg WPM, Colzato LS (2014) Two is better than one: Bilingual education promotes the flexible mind. Psychol Res. doi: 10.1007/s00426-014-0575-3 Costa A, Hern´andez M, Sebasti´an-Gall´es N (2008) Bilingualism aids conflict resolution: Evidence from the ANT task. Cognition 106, 59-86. ´ Kov´acs AM, Mehler J (2009) Cognitive gains in 7-month-old bilingual infants. Proc Natl Acad Sci U S A 106, 6556-6560. Costa A, Sebastian-Galles N (2014) How does the bilingual experience sculpt the brain? Nat Rev Neurosci 15, 336-345. de Bruin A, Treccani B, Della Sala S (2014) Cognitive advantage in bilingualism: An example of publication bias? Psychol Sci 26, 99-107. Blumenfeld HK, Marian V (2011) Bilingualism influences inhibitory control in auditory comprehension. Cognition 118, 245-257. Garbin G, Sanjuan A, Forn C, Bustamante JC, RodriguezPujadas A, Belloch V, Hernandez M, Costa A, Avila C (2010) Bridging language and attention: Brain basis of the impact of bilingualism on cognitive control. Neuroimage 53, 12721278. Bak TH, Vega-Mendoza M, Sorace A (2014) Never too late? An advantage on tests of auditory attention extends to late bilinguals. Front Psychol 5, 485. Luo L, Craik FIM, Moreno S, Bialystok E (2013) Bilingualism interacts with domain in a working memory task: Evidence from aging. Psychol Aging 28, 28-34. Gold BT, Kim C, Johnson NF, Kryscio RJ, Smith CD (2013) Lifelong bilingualism maintains neural efficiency for cognitive control in aging. J Neurosci 33, 387-396. Hilchey MD, Klein RM (2011) Are there bilingual advantages on nonlinguistic interference tasks? Implications for the plasticity of executive control processes. Psychon Bull Rev 18, 625-658. Bialystok E, Craik FIM, Luk G (2012) Bilingualism: Consequences for mind and brain. Trends Cogn Sci 16, 240-250. Luk G, Bialystok E, Craik FIM, Grady CL (2011) Lifelong bilingualism maintains white matter integrity in older adults. J Neurosci 31, 16808-16813. Abutalebi J, Della Rosa PA, Green DW, Hernandez M, Scifo P, Keim R, Cappa SF, Costa A (2012) Bilinguatism tunes the anterior cingulate cortex for conflict monitoring. Cereb Cortex 22, 2076-2086. Schroeder SR, Marian V (2012) A bilingual advantage for episodic memory in older adults. J Cogn Psychol (Hove) 24, 591-601. Ljungberg JK, Hansson P, Andres P, Josefsson M, Nilsson LG (2013) A longitudinal study of memory advantages in bilinguals. PLoS One 8, e73029. Elliott R (2003) Executive functions and their disorders. Br Med Bull 65, 49-59. Daniels K, Toth J, Jacoby L (2006) The aging of executive functions. In Lifespan cognition: Mechanisms of Change, Bialystok E, Craik FIM, eds. Oxford University Press, New York, NY, US, pp. 96-111. Miller EK, Cohen JD (2001) An integrative theory of prefrontal cortex function. Annu Rev Neurosci 24, 167-202. Stuss DT, Alexander MP (2000) Executive functions and the frontal lobes: A conceptual view. Psychol Res 63, 289-298. Nolde SF, Johnson MK, Raye CL (1998) The role of prefrontal cortex during tests of episodic memory. Trends Cogn Sci 2, 399-406. Schroder J, Buchsbaum MS, Shihabuddin L, Tang C, Wei TC, Spiegel-Cohen J, Hazlett EA, Abel L, Luu-Hsia C, Ciaravolo

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

1267

TM, Marin D, Davis KL (2001) Patterns of cortical activity and memory performance in Alzheimer’s disease. Biol Psychiatry 49, 426-436. Wheeler MA, Stuss DT, Tulving E (1995) Frontal lobe damage produces episodic memory impairment. J Int Neuropsychol Soc 1, 525-536. Gollan TH, Montoya RI, Werner GA (2002) Semantic and letter fluency in Spanish-English bilinguals. Neuropsychology 16, 562-576. Portocarrero JS, Burright RG, Donovick PJ (2007) Vocabulary and verbal fluency of bilingual and monolingual college students. Arch Clin Neuropsychol 22, 415-422. Roberts PM, Garcia LJ, Desrochers A, Hernandez D (2002) English performance of proficient bilingual adults on the Boston Naming Test. Aphasiology 16, 635-645. Gollan TH, Acenas LA (2004) What is a TOT? Cognate and translation effects on tip-of-the-tongue states in SpanishEnglish and Tagalog-English bilinguals. J Exp Psychol Learn Mem Cogn 30, 246-269. Gollan TH, Salmon DP, Montoya RI, da Pena E (2010) Accessibility of the nondominant language in picture naming: A counterintuitive effect of dementia on bilingual language production. Neuropsychologia 48, 1356-1366. Ivanova I, Salmon DP, Gollan TH (2014) Which language declines more? Longitudinal versus cross-sectional decline of picture naming in bilinguals with Alzheimer’s disease. J Int Neuropsychol Soc 20, 534-546. Adlam A-LR, Bozeat S, Arnold R, Watson P, Hodges JR (2006) Semantic knowledge in mild cognitive impairment and mild Alzheimer’s disease. Cortex 42, 675-684. Willers IF, Feldman ML, Allegri RF (2008) Subclinical naming errors in mild cognitive impairment: A semantic deficit? Dement Neuropsychol 2, 217-222. Hodges JR, Salmon DP, Butters N (1992) Semantic memory impairment in Alzheimer’s disease: Failure of access or degraded knowledge? Neuropsychologia 30, 301-314. Levy R (1994) Aging-associated cognitive decline. Working Party of the International Psychogeriatric Association in collaboration with the World Health Organization. Int Psychogeriatr 6, 63-68. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM (1984) Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 34, 939-944. Morris JC, Heyman A, Mohs RC, Hughes JP, van Belle G, Fillenbaum G, Mellits ED, Clark C (1989) The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology 39, 1159-1165. Welsh KA, Butters N, Mohs RC, Beekly D, Edland S, Fillenbaum G, Heyman A (1994) The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part V. A normative study of the neuropsychological battery. Neurology 44, 609-614. Thalmann B, Monsch AU, Bernasconi F, Schneitter M, Aebi C, Camachova-Davet Z, St¨ahelin HB (2000) The CERAD Neuropsychological Assessment Battery (CERAD-NAB) – A Minimal Data Set as a Common Tool for German-speaking Europe. International Conference on Alzheimer Disease and Related Disorders, Washington DC, USA. Aebi C (2002) PhD Thesis: Validierung der neuropsychologischen Testbatterie CERAD-NP: Eine Multi-Center Studies. Advisor: Klaus Opwis, Faculty of Humanities, University of Basel, Basel, pp. 214, http://edoc.unibas.ch/diss/DissB 6279

1268 [52]

[53]

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

M.E. Kowoll et al. / Neuropsychological Profiles and Verbal Abilities in Lifelong Tombaugh TN (2004) Trail Making Test A and B: Normative data stratified by age and education. Arch Clin Neuropsychol 19, 203-214. Shulman KI, Gold DP, Cohen CA, Zucchero CA (1993) Clock-drawing and dementia in the community: A longitudinal study. Int J Geriatr Psychiatry 8, 487-496. Petermann F, Lepach AC (2012) Wechsler Memory Scale ¨ - Fourth Edition Deutsche Ubersetzung und Adaption des WMS-IV von David Wechsler, Pearson, Frankfurt. H¨arting C, Markowitsch H-J, Neufeld H, Calabrese P, Deisinger K, Kessler J (2000) Wechsler Memory Scale Revised Edition, German Edition, Huber, Bern. Sheikh JI, Yesavage JA (1986) Geriatric Depression Scale (GDS): Recent evidence and development of a shorter version. Clin Gerontol 5, 165-173. Costa A, Hern´andez M, Costa-Faidella J, Sebasti´an-Gall´es N (2009) On the bilingual advantage in conflict processing: Now you see it, now you don’t. Cognition 113, 135-149. Kav´e G, Eyal N, Shorek A, Cohen-Mansfield J (2008) Multilingualism and cognitive state in the oldest old. Psychol Aging 23, 70-78. Chertkow H, Whitehead V, Phillips N, Wolfson C, Atherton J, Bergman H (2010) Multilingualism (but not always bilingualism) delays the onset of Alzheimer disease: Evidence from a bilingual community. Alzheimer Dis Assoc Disord 24, 118-125. Chertkow H, Bub D (1990) Semantic memory loss in Alzheimer-type dementia. In Modular Deficits in AlzheimerType Dementia. Schwartz MF, ed. The MIT Press, Cambridge, MA, US, pp. 207-244. Butters N, Granholm E, Salmon DP, Grant I, Wolfe J (1987) Episodic and semantic memory: A comparison of amnesic and demented patients. J Clin Exp Neuropsychol 9, 479-497. Pekkala S, Wiener D, Himali JJ, Beiser AS, Obler LK, Liu Y, McKee A, Auerbach S, Seshadri S, Wolf PA, Au R

[63]

[64]

[65]

[66] [67]

[68]

[69]

[70]

[71]

[72] [73]

(2013) Lexical retrieval in discourse: An early indicator of Alzheimer’s dementia. Clin Linguist Phon 27, 905-921. Levinoff EJ, Phillips NA, Verret L, Babins L, Kelner N, Akerib V, Chertkow H (2006) Cognitive estimation impairment in Alzheimer disease and mild cognitive impairment. Neuropsychology 20, 123-132. Jacobs DM, Sano M, Dooneief G, Marder K, Bell KL, Stern Y (1995) Neuropsychological detection and characterization of preclinical Alzheimer’s disease. Neurology 45, 957-962. Rubin EH, Storandt M, Miller P, Kinscherf DA, Grant EA, Morris JC, Berg L (1998) A prospective study of cognitive function and onset of dementia in cognitively healthy elders. Arch Neurol 55, 395-401. Ardila A, Ramos E (2008) Normal and abnormal aging in bilinguals. Dement Neuropsychol 2, 242-247. Mendez MF, Perryman KM, Pont´on MO, Cummings JL (1999) Bilingualism and dementia. J Neuropsychiatry Clin Neurosci 11, 411-412. Jiang N, Forster KI (2001) Cross-language priming asymmetries in lexical decision and episodic recognition. J Mem Lang 44, 32-51. Ullman MT (2001) The neural basis of lexicon and grammar in first and second language: The declarative/procedural model. Biling (Camb Engl) 4, 105-122. Ullman MT (2004) Contributions of memory circuits to language: The declarative/procedural model. Cognition 92, 231-270. Witzel NO, Forster KI (2012) How L2 words are stored: The episodic L2 hypothesis. J Exp Psychol Learn Mem Cogn 38, 1608-1621. Grosjean F (2010) Bilingual: Life and Reality, Harvard University Press, Cambridge, Mass. [u.a.]. Poplack S (1980) Sometimes I’ll start a sentence in Spanish y termino en Espa˜nol: Toward a typology of code-switching. Linguistics 18, 581-618.