Performance of cognitive tests, individually and combined, for the detection of cognitive disorders amongst community-dwelling elderly people with memory complaints: the EVATEM study C. Vannier-Nitenberg a , V. Dauphinot b , B. Bongue c,d , C. Sass c , A. Bathsavanis b , I. Rouch b,e , N. Deville c , O. Beauchet c,f , P. Krolak-Salmon b,g,h and B. Fantino f,i a Centre d’Examens de Sante de la CPAM du Rh^ one, Lyon; b Centre Memoire de Ressources et de Recherche (CMRR) Lyon, H^ opital des Charpennes, Hospices civils de Lyon, Villeurbanne; c Centre Technique d’Appui et de Formation des Centres d’Examens de Sante (CETAF), Saint-Etienne; d Laboratory EA4607 SNA-EPIS, PRES Lyon, Jean Monnet University of Saint Etienne, Saint Etienne; e CMRR de Saint Etienne, Service de Neurologie, CHU de Saint Etienne, Saint Etienne Cedex; f Department of Neuroscience, Division of Geriatric Medicine, Angers University Hospital, Angers; g Inserm U1028, CNRS UMR5292, Lyon Neuroscience Research Center, Brain Dynamics and Cognition Team, Lyon; h University Lyon 1, Lyon; and i Filieris, Caisse Autonome Nationale de la Securite Sociale dans les Mines, Paris, France Keywords: Alzheimer’s disease, cognitive disorders, dementia, early diagnosis, memory complaint, mild cognitive impairment Received 7 January 2015 Accepted 2 September 2015 European Journal of Neurology 2015, 0: 1–8 doi:10.1111/ene.12888 Background and purpose: Dementia is a leading cause of dependence amongst the aged population. Early identification of cognitive impairment could help to delay advanced stages of dependence. This study aimed at assessing the per- formance of three neuropsychological tests to detect cognitive disorders in elderly subjects with memory complaints. Methods: The EVATEM study is a prospective multicentre cohort with a 1- year follow-up. Subjects with memory complaints were selected during pre- ventive health examinations, and three neuropsychological tests (five-word, cognitive disorders examination, verbal fluency) were administered. Two groups were identified in memory clinics: (i) cognitively healthy individuals (CHI) and (ii) mild cognitive impairment or demented individuals (MCI-DI). Cross-sectional analyses were performed on data at inclusion. The relation- ship between the diagnosis of MCI-DI/CHI and the neuropsychological tests was assessed using logistic regressions. The performance of the neuropsycho- logical tests, individually and combined, to detect cognitive disorders was calculated. Results: Of 585 subjects, 31.11% had cognitive disorders (MCI, 176 subjects; DI, six subjects). Amongst the three tests studied, the odds ratio for MCI-DI was higher for the five-word test <10 [odds ratio 3.2 (1.81; 5.63)]. The best per- formance was observed when the three tests were combined: specificity 90.5% and sensitivity 42.4% compared to respectively 89.2% and 28.3% for the five- word test. Conclusions: Despite the poor sensitivity of the five-word test, it seems to be the most adapted for the diagnosis of MCI-DI in older adults with a memory complaint, in prevention centres, taking into account its high specificity and its rapid administration compared to the other tests. Background Dementia is a leading cause of disability and depen- dence amongst the aged population. In 2010, the total cost of dementia in the world was estimated at $604 billion, equivalent to 1% of gross domestic product. Correspondence: C. Vannier-Nitenberg, 15 rue Baraban, 69456 Lyon Cedex 06, France (tel.: 04 72 65 16 48; fax: 04 72 65 16 00; e-mail: christiane.nitenberg@cpam-rhone.cnamts.fr). Trial registration: ClinicalTrials.gov: NCT01316562. © 2015 EAN 1 ORIGINALARTICLE EUROPEANJOURNALOFNEUROLOGY