Facial Emotion Recognition in Patients with Amnesic Mild Cognitive Impairment and Mild-Moderate Alzheimer's Disease

Background: Although several studies have found evidence of impairment in facial emotion recognition in Alzheimer's disease, current understanding regarding which specific emotions are preserved and disrupted is inconsistent. Moreover, facial emotion recognition has been little explored in subjects with amnesic mild cognitive impairment. Objective: To investigate processing of human faces identity and emotional expressions in patients with probable mild-moderate Alzheimer disease (AD), amnesic mild cognitive impairment (a-MCI) and healthy control subjects (CS). Methods: Thirty subjects were included in the study: 10 AD (mean MMSE corrected score=20.94; DS=1.96), 10 a-MCI (mean MMSE corrected score=25.98; DS=0.69), and 10 CS (mean MMSE corrected score=29.85; DS=0.4). The three groups did not differ for age and education. All patients underwent an extensive neuropsychological test battery. Geriatric Depression Scale was employed to exclude depressed patients. A new battery for assessing face emotion processing was developed. It included 48 faces pictures of 6 models balanced for sex and age (young, adult and old). For each model there were poses corresponding to seven emotions (anger, disgust, fear, happiness, sadness, surprise, boredom) as well as neutral expressions. Subjects had to perform four different tasks: 1) deciding the emotion label that best described the facial expression shown; 2) choosing the picture that matched the target emotion verbal label; 3) sorting the faces displaying the same facial expression; 4) sorting the faces displaying the same identity. Results: Recognition accuracy in all three groups was better for positive emotions and neutral expressions than negative emotions, consistent with previous studies. AD patients were more impaired in the recognition of overall emotions and neutral faces than a-MCI and CS subjects. Compared with CS, a-MCI did not differ significantly in their emotion recognition abilities. When segregated by emotions, we found significant differences in emotion recognition between the diagnostic groups for fearful and sad faces. In particular, AD patients and a-MCI subjects differed significantly from CS in fearful face recognition; AD patients also had impairment in recognizing facial expressions of sadness. Only patients with AD were impaired on the facial identity task. The predominant pattern across all groups and emotions was of a better recognition of emotions when displayed by young faces instead of either adult or old faces. Conclusions: A selective impairment in recognition of facial expressions of fear is already present in patients with a-MCI. An additional deficit in processing of sad faces emerge with AD progression and may be related to the degeneration progression towards structures implicated in emotional processing systems. An early detection of emotional impairment in MCI phases of dementia may have clinical impact and prognostic value.