Brazilian version of Northwestern university famous faces test (NUFFACE‐Br): Evidence of validity between control and clinical groups

Abstract

AbstractBackgroundNUFFACE was created to identify the locus of name‐face failure common in healthy aging and dementia. Psychometric studies are necessary for the development of the Brazilian version. Therefore, we intend to verify the evidence of predictive (or diagnostic) validity of this instrument.Methodsa total 24 participants were included in this study and divided into 12 cognitively healthy elderly (NC) and 12 with dementia due to Alzheimer’s Disease (AD; CDR‐1). All responded to the NUFFACE‐Br, which has four tasks: 1) facial naming; 2) verbal description of the facts related to the face; 3) verbal description in response to the printed name; and 4) face‐name matching (multiple‐choice format). Different types of errors, such as Semantically Related (SR), Semantically Unrelated (SU), Phonological Similarity (Ph), “I don’t know” (DK), and Others (O), were developed as part of the correction criteria. The comparison of sociodemographic variables and NUFFACE‐Br total tasks were performed by Mann‐Whitney U test, adopting a significance level of 5%.ResultsThe groups did not differ by age, education, and sex (p<0.05). The NC performed better than the AD in almost all tasks of this instrument (p<0.01), except for task 4 (p<0.266). The AD presented more difficulty in the naming faces tasks (average = 14±6.6) than the others and similar performance to the NC in the face‐name correspondence (average score = 30.7±2.1). A higher frequency of all types of errors was verified in the AD group for the first three tasks of this instrument, except for the “phonological similarity,” in which the frequency of errors was the same between the groups. Task 4 presented a higher frequency in SR, SU, and O errors.ConclusionSignificant differences were verified between the NC and AD groups, indicating the predictive (or diagnostic) evidence of NUFFACE‐Br. The face‐name matching task was the only one that did not show differences between the groups. The AD presented a higher frequency of errors when compared to the NC, especially in the task of naming faces. Further psychometric studies are being conducted to elucidate the clinical evidence of this instrument to old Brazilian adults.