Which language is more affected in bilinguals with Alzheimer’s disease? Diagnostic sensitivity of the Multilingual Naming Test.

Abstract Background The present study aimed to determine if Alzheimer’s disease (AD) affects bilinguals’ ability to name pictures in both languages equally, and to determine what testing procedures best distinguish bilingual patients from controls. Method Sixty Spanish‐English bilinguals, including 19 with dementia and 41 controls with equivalent age, education, and bilingual proficiency level, were tested on the full 68‐items of the Multilingual Naming Test (MINT; Gollan et al., 2012) first in their dominant language and then in the nondominant language. Four naming scores were derived for each participant: one for each language, one for either‐language that counted items as correct if produced only in one of the two languages, and one for both‐languages that only counted as correct items that bilinguals named in both languages. We also considered whether item difficulty modulates which naming score is most affected by AD by dividing the MINT test into difficulty tertiles (easy = items 1–20; medium = 21‐44; difficult = 45‐68). Result Bilinguals with AD named fewer pictures than controls, but this group effect varied with naming score and item difficulty. With difficult and medium items the difference between patients and controls was significant only in the dominant‐language, whereas with easy items the opposite was true (i.e., the group effect was significant only in the nondominant‐language), a significant interaction between participant group, language‐dominance, and item difficulty ( F (1,58) = 5.449, partial‐eta‐squared = .086, p = .023). The either‐language scoring procedure increased naming scores especially in balanced bilinguals, and to an equal extent in patients and controls. Critically, ROC analyses revealed that dominant‐language and either‐language naming scores are superior for classifying bilinguals as patients versus controls, and significantly better than nondominant‐language and both‐languages naming scores. Conclusion For assessment purposes, testing primarily or exclusively in the dominant language is best for detecting bilingual naming impairments. However, AD likely affects the ability to access names in both languages, and simple descriptions of language decline as “similar in both languages” or as affecting one language more than the other may be misleading in terms of theoretical implications for how two languages are represented and processed in the bilingual brain.