Age estimation from faces in Alzheimer disease

Although studies on age estimation showed that the performance of estimation is fairly accurate, this performance can be influenced by group biases such as the own-age bias (George & Hole, 1995). Moreover this bias occurs both in young and older adults (Moyse & Brédart, 2012). Because difficulties in face processing have been reported in Alzheimer disease (Della Sala et al., 1995), the aim of this study was to examine the performance of age estimation from faces in patients with Alzheimer disease (mild to moderate) compared with normal aging persons. Moreover to test the preservation of the occurrence of an own-age bias, stimuli belonging to different age groups (young, middle age and older adults) were used. We observed a main effect of Group indicating that patients were less accurate than control whatever the age of faces. In addition a main effect of Age of faces was obtained; the percentage of accuracy was better for older faces than for the two other age groups of faces. Consequently although patients’ performance in age estimation of faces is impaired, an own-age bias was still present. These results have two main interests: a clinical interest (expanding the diagnostic criteria of the Alzheimer disease) and a forensic interest (assessing the credibility of eyewitness testimony in older adults with a possible Alzheimer disease).