Does multilingualism act as a protective factor (cognitive reserve) against cognitive impairment in multiple sclerosis?: A retrospective cohort study.

BACKGROUND: Cognitive impairment (CI) is common in multiple sclerosis (MS), affecting half of persons with MS (PwMS). Cognitive reserve has been associated with delaying the onset and slowing the progression of CI in PwMS. Multilingualism has been demonstrated to be a protective factor against CI in Alzheimer's disease (AD) but has never been studied in PwMS.

OBJECTIVE: To explore if multilingualism is a protective factor against CI in PwMS.

METHODS: This is a retrospective cohort study of PwMS aged 18-59, with a confirmed diagnosis of relapsing MS, fluent in English, who completed the Minimal Assessment of Cognitive Function in MS (MACFIMS) at the London (ON) MS Clinic. Any PwMS with a history of dementia or developmental delay, daily marijuana use, a major psychiatric disorder, or less than grade 9 education was excluded. We focused on the Brief Visuospatial Memory Test (BVMTR), immediate recall (-IR) and delayed recall (-DR) as it would be the least affected by language, as well as the Symbol Digit Modalities Test (SDMT), as information processing speed is the most commonly affected domain in PwMS. One-way ANOVA was used to compare raw scores on the BVMTR and SDMT between groups (uni- vs. multillingual), while chi-square was used to compare impairment on BVMTR and SDMT between groups.

RESULTS: The cohort consisted of 678 subjects. The mean age was 39.7 (± 9.6) years with 501 (73.9 %) females (sex at birth), the mean duration of disease of 5.9 (± 6.9) years, and mean years of education was 13.9 (±2.2). The majority of subjects (563, 83 %) were unilingual and (115, 17 %) were multilingual; 102 subjects were bilingual and 13 subjects fluent in ≥ three languages. English was the first language was in most of subjects (614, 90.6 %). There was no significant difference on the BVMTR-IR scores (p = 0.189) or BVMTR-DR (p = 0.096) between groups. Similarly, there was no difference in the number of subjects impaired on the BVMTR-IR (X2 (1, N = 678) = 3.167, p = 0.057) or BVMT-DR between groups (X2 (1, N = 678) = 2.996, p = 0.083). Further, there was no significant difference on the SDMT (p = 0.506) between groups, or in number of subjects impaired on the SDMT between groups (X2 (1, N = 678) = 1.023, p = 0.312).

CONCLUSION: This study shows that multilingualism does not have a protective effect against CI in PwMS and does not play a role in enriching the cognitive reserve, in contrast to studies in AD. This difference may be explained by a different underlying pathological mechanism in these diseases and warrants further study.