Double dissociation between familiarity and recollection in Parkinson's disease as a function of encoding tasks

Abstract

Parkinson's disease (PD) is associated with episodic memory deficits, but their exact nature is unclear. Some dual-process studies have suggested that recollection is impaired and familiarity is spared in PD, yet others have found the opposite. Our goal was to investigate these memory processes in PD and determine whether the inconsistency among existing findings is related to differences in encoding conditions. We used a process-dissociation procedure with word pairs to estimate familiarity and recollection. In Experiment 1, we used a directed, deep, relational encoding condition (i.e., sentence generation), and in Experiment 2, we contrasted this encoding condition with a shallower, non-directed encoding condition (i.e., read condition). We found a double dissociation as a function of the encoding task: In the sentence generation encoding condition, recollection was impaired in the PD patients, but familiarity was spared. In contrast, in the read encoding condition, there was no group difference in recollection, but familiarity was impaired in the PD group. Within-subject comparisons revealed that both control and PD participants benefitted from the provision of a directed, deep relational encoding strategy. However, this benefit was manifested as an increase in recollection in the controls, but an increase in familiarity in the PD patients. These findings help to reconcile the extant literature and suggest that episodic memory deficits in PD are two-fold, involving: (1) difficulties instantiating encoding strategies independently, leading to deficits in familiarity, and (2) impaired recollection when encoding strategies are equated across groups. Our results highlight the importance of controlling encoding conditions between groups and of taking account of other variables that may influence the participants’ performance, such as deficits associated with normal aging, which may mask deficits in neurodegenerative diseases in particular situations. More generally, our study raises the possibility that deficits in recollection or familiarity in patient populations are not immutably linked to the structure that is affected, as is typically assumed, but that such deficits may interact with type of encoding, and possibly with the nature of the retrieval process.