Regional correlates of linguistic error types using the Mini Linguistic State Examination (MLSE)

Authors: Nikil Patel, Thomas R. Barrick, Katie A. Peterson, Ruth Ingram, Peter Garrard

Published: 2020-12-07

DOI: 10.1002/alz.042829

Source: Full article


Abstract

AbstractBackgroundPrimary progressive aphasia is a clinically and pathologically heterogeneous condition. The currently recognised subsyndromes are non‐fluent, fluent and logopenic variants. The MLSE adopts a novel error‐scoring system to classify the PPA syndromes, and has previously been shown to distinguish the syndromic subtypes based on performance in five linguistic domains: motor speech, phonology, semantics, working memory and syntax. Further validity of this method can be achieved with brain MR correlates specific to the MLSE language domains. For instance, shortfalls in semantics would be expected for correlate with atrophy in bilateral anterior temporal gyrus, and errors in motor speech and syntax with regions in the left inferior frontal gyrus.MethodsThe cohort compromised of 105 participants, 61 patients with a root diagnosis of primary progressive aphasia (PNFA, 20; LPA, 24; SD, 17) and 44 matched controls. The MLSE was administrated on all participants who also had 3T MR imaging. Group level analysis was carried out using voxel‐based morphometry and Statistical Parametric Mapping with a region of interest approach.ResultsWe observed a pattern of high sensitivity and specificity for the proposed atrophy patterns (bilateral anterior temporal cortex) associated with semantic errors. Similarly, correlations with atrophy in left inferior frontal and insular regions were observed for deficits in motor speech, phonology and syntax. Regional correlates with working memory deficits were more diffuse.ConclusionsThe results of this study provide evidence that assessment using error types aids classification and differentiation of the profile of PPA patients and conforms with patterns of MR atrophy that would be expected based on previous research.