John Libbey Eurotext

Gériatrie et Psychologie Neuropsychiatrie du Vieillissement


Narrative discourse in dementia with Lewy bodies and Alzheimer's disease Volume 17, issue 2, Juin 2019


  • Figure 1


1 Centre mémoire de ressources et de recherche (CMRR), Hôpital de jour de gériatrie, Pôle de gériatrie et unité de neuropsychologie, Pôle tête et cou, Hôpitaux Universitaires de Strasbourg, France
2 Équipe IMIS/Neurocrypto, Fédération de médecine translationnelle de Strasbourg (FMTS), Laboratoire ICube, Université de Strasbourg et CNRS, France
* Correspondance : C. Bouvet, V. Borel
a Auteurs principaux

Dementia with Lewy bodies (DLB) is the second most common form of dementia, after Alzheimer's disease (AD). Subjects with DLB remain under-diagnosed, especially in the early stage of the disease, when they can show subtle neurocognitive disorders similar to subjects with AD. In order to refine the differential diagnosis between these two neurodegenerative diseases and to improve patients’ care, our speech therapy study aimed to analyze their ability to tell a story by producing a narrative discourse (ND). Method: 25 participants with DLB and 12 participants with AD underwent a ND test based on an illustrated story. The test was selected from the French language assessment corpus GREMOTS. The grading of the ND was done according to the following six parameters: lexicon, syntax, pragmatics, presence of the main actions, quality of discourse and informativeness. Results: In the early stage, a quarter of the participants with DLB are under cut-off score for ND, and this proportion strongly increases in the advanced stage. In contrast, all the participants with AD show a pathological ND in both stages. In the early stage, the ND abilities appear significantly better preserved in participants with DLB than in participants with AD. No difference is found in the advanced stage. This result highlights two distinct language profiles, with the participants with AD being significantly less informative than the participants with DLB in the early stage. Furthermore, the participants with DLB show pathological scores spreading over the six parameters, whereas the participants with AD have more selective impairments. Indeed, the informativeness of participants with AD is 100% pathological while their syntax is 100% preserved. Discussion: These results are encouraging as they could enable speech therapists to better adjust their follow-up according to the linguistic profile of the patients and to educate caregivers more appropriately. Further research on language issues in DLB and AD is essential.