Bilingual Spanish-English speakers’ performance on the Montreal Cognitive Assessment (MoCA) in English: implications for test bias and score adjustments

BackgroundThe Montreal Cognitive Assessment (MoCA) is a commonly used cognitive screener for detecting cognitive impairment. However, the originally suggested cutoff score of 26/30 has not held up in other diagnostic accuracy studies, and it is unclear if bilingualism impacts this score. The test items have not been sufficiently examined for test bias, particularly in bilingual speakers. Aims The aims of this study were to investigate the performance of Spanish-English bilingual speakers on MoCA total score and item scores, factors influencing total and item scores, and the need for score adjustments for bilingual speakers. Methods & Procedures One hundred and one bilingual neurotypical Spanish-English speaking individuals (age range = 25-76 years, 27M / 74F) were virtually administered the Montreal Cognitive Assessment in English. Language proficiency was measured using objective vocabulary measures as well as self-ratings using a validated tool. Subgroups of participants with high proficiency in both languages, and mid-proficiency in English were identified and separately analyzed. A four-predictor model (age, English education, socioeconomic status, and English proficiency) was used to examine factors that influence the MoCA total score and individual item accuracies. Outcomes & ResultsMoCA total scores were significantly lower than the normative scores of the Anglophone/Francophone population originally published by the developers. Seven and fifteen items had low passing rates (below 87%) for the entire group and the mid-proficiency English groups respectively. English language proficiency emerged as the single significant predictor of total scores and of over one-fourth (8 out of 28) of the item accuracies. ConclusionsThe normative data obtained from the present study shows that Spanish-English bilingual speakers’ total scores on the English MoCA need to be adjusted and raw scores might not be sensitive to identify cognitive impairment. Measuring language proficiency is critical for determining the need for and type of score adjustments for bilingual speakers.