Unfamiliar facial identity discrimination and recognition impairment in prodromal Alzheimer's disease: A behavioral pattern separation and completion study

Background

Impairment of face identity recognition (FIR) in prodromal Alzheimer's disease (AD) is not clearly established, and the cognitive processes underpinning a potential FIR impairment remain elusive. The influential pattern separation (PS) and completion (PC) framework may offer a fascinating insight in this respect. Indeed, efficient FIR implies PS to encode facial identities in distinct memory representations, while PC is involved in matching the perceived facial patterns to these memory representations at retrieval. Based on this functional dissociation, the present study investigated FIR using a PS and PC paradigm in prodromal AD patients.

Method

Thirty-one cognitively unimpaired (CU) older individuals and 16 amyloid-positive patients with amnestic Mild Cognitive Impairment (Aβ+ aMCI) were familiarized with 40 facial identities, using a viewpoint-matching task. They then performed a forced-choice recognition task assessing the extent to which these unfamiliar faces were implicitly learned. Subsequently, they underwent separate Yes/No recognition tasks involving parametrically controlled blurred or morphed faces, designed to solicit PC and PS processes respectively. Finally, in two separate discrimination tasks, participants had to determine whether two simultaneously displayed faces, a proportion of them being blurred or morphed, corresponded to the same identity or not.

Results

Aβ+ aMCI patients obtained lower performance than CU older individuals in each task, including the tasks that did not involve FIR demands in episodic memory. There was no disproportionate performance decrease with increasing levels of PC and PS requirements.

Conclusions

No isolated PS/PC deficit was evidenced in FIR in prodromal AD patients. Importantly, besides a general FIR deficit, discrimination of simultaneously presented unfamiliar facial identities was impaired.