Prevalence and determinants of language impairment in non-demented amyotrophic lateral sclerosis patients

Background and purpose: This study aimed at estimating the prevalence of language impairment (LI) in a large, clinic-based cohort of non-demented amyotrophic lateral sclerosis (ALS) patients and assessing its underpinnings at motor and non-motor levels. Methods: Non-demented ALS patients (N = 348) underwent the Edinburgh Cognitive and Behavioural ALS Screen (ECAS), as well as an assessment of behavioural/psychiatric and motor-functional features. The prevalence of LI was estimated based on the proportion of patients showing a performance below the age- and education-adjusted cut-off on the ECAS-Language. Multiple regression models were run to assess the determinants of language functioning and impairment. Results: The prevalence of LI was 22.7%. 46.6% of the variance of ECAS-Language scores remained unexplained, with only the ECAS-Executive positively predicting them (p < 0.001; eta(2) = 0.07). Similarly, only a lower score on the ECAS-Executive predicted a higher probability of a below cut-off ECAS-Language performance (p < 0.001). Spelling and Naming tasks were the major drivers of ECAS-Language performance. Conclusions: This study suggests that, in non-demented ALS patients, LI occurs in asymptotic to 23% of cases, is significantly driven by executive dysfunction but, at the same time, partially independent of it and is not associated with other motor or non-motor features.