Overgeneral autobiographical memory in Parkinson's disease

Abstract

Autobiographical memory (AM) concerns the ability to remember past events from one's own life and consists of autobiographical knowledge (personal facts) and autobiographical incidents (personal events). The novelty of this research was to assess both personal factual and personal event AM in Parkinson's disease (PD) for specified lifetime periods. An autobiographical fluency task was used in which participants were asked to recall personal events and personal facts from five separate lifetime periods. Previous findings as well the brain regions affected in PD lead to the hypothesis that Parkinson's patients would recall less autobiographical memories especially for the most recent lifetime periods. Sixteen non-demented and non-depressed Parkinson's patients and sixteen age-education-matched controls participated. The results showed a temporal gradient for the recall of personal events in Parkinson's patients as they recalled fewer events for recent time periods. The PD group also had more difficulties in recalling autobiographical events rather than an autobiographical knowledge. The difficulty in recalling autobiographical events was characterized by overgenerality, with PD patients failing to generate specific episodic memories.

Introduction

Autobiographical memory (AM), despite its clear everyday importance for the individual (for example, Conway, 2005) has not been extensively investigated in Parkinson's disease (PD) (Sagar et al., 1988). According to the Self Memory System (SMS) proposed by Conway and Pleydell-Pearce (2000) and recently reviewed by Conway (2005), AM contains different types of representations organized in a hierarchical structure, which range from conceptual knowledge (autobiographical facts) to event-specific knowledge (ESK). Autobiographical facts refer to lifetime periods, e.g., being a student at university, that can be used to access more specific autobiographical knowledge called general events, including repeated or categoric events e.g., writing lab reports. These general events can cue access to ESK, defined as sensory-perceptual episodic memories e.g., submitting your first lab report, just before the deadline, and panicking because you did not know where to go to hand it in. This study will explore for the first time both personal facts and personal events (ESK) for different lifetime periods in PD.

Previous studies exploring memory for public events have yielded conflicting results (Fama et al., 2000, Leplow et al., 1997, Sagar et al., 1988, Venneri et al., 1997). When impairments were found, results showed that PD patients had difficulties recalling public events and their dates for more recent decades (Sagar et al., 1988). Furthermore, PD patients were significantly impaired in recalling remote public events, but only when a recall task was used in comparison to a cued-recall or a recognition task (Sagar et al., 1988, Venneri et al., 1997). The different patterns of performance between recall and recognition tasks lead Sagar et al. (1988) to suggest that PD patients' difficulties in recalling memories from recent periods could be due to a retrieval failure.

Remote memory for personal events has been less explored in PD. Sagar et al. (1988) asked PD patients, with and without dementia, to give personally experienced events from any time period in relation to ten cues such as bird, car, boy (a procedure modified from Crovitz and Schiffman, 1974). Results showed that PD patients recalled fewer memories than control participants but did not show a clear temporal gradient. Analysis of the age of episodes recalled revealed conflicting results with some PD patients recalling more memories from the most recent five year period and others recalling more memories from the most remote periods. However, the results clearly showed that impaired performance was characterized by a failure to recall time-specific events i.e., patients could describe a specific instantiation of the cue word but were likely to generalise across similar episodes. These results suggest that AM in PD might be characterized by overgenerality. Overgenerality of AM (Williams and Broadbent, 1986) is a memory deficit found in depression and patients with a dysexecutive syndrome (Baddeley and Wilson, 1986; and see Williams et al., 2007 for a review). It is defined as the tendency to recall repeated events (occurring more than one time) or extended events (lasting more than one day) rather than specific events (Williams et al., 1996).

Neuropsychological data supports the hypothesis that AM might be impaired in PD. Critically, the brain regions affected in PD, i.e., the prefrontal cortex (PFC) (for reviews, see Owen, 2004 and Zgaljardic et al., 2003) and the mediotemporal lobe (MTL) (Braak et al., 2003, Bruck et al., 2004, Double et al., 1996, Laakso et al., 1996, Nagano-Saito et al., 2005, Tam et al., 2005) are also the brain regions that have been identified as part of the brain network underlying AM (default network, Buckner et al., 2008). This network includes several regions: the PFC, involved in the retrieval of AM information (for review see Cabeza and St Jacques, 2007), and the MTL, in particular the hippocampus involved in the retrieval of emotional AM (Addis et al., 2004) or memories associated with visuospatial imagery (Cabeza et al., 2004). Many of these brain regions also subserve a number of cognitive processes such as retrieval processes or visual memory (Cabeza et al., 2004, Dubois et al., 1991), which could explain AM deficits in PD. Indeed, neuropsychological studies clearly support the idea that PD is characterized by difficulties in retrieving information (Buytenhuijs et al., 1994, Ivory et al., 1999, Taylor et al., 1986) and by visuospatial deficits (Lavernhe et al., 1989, Levin et al., 1991, Lee et al., 1998). Finally, the severe depletion of dopamine in the striatum observed in PD (Lewis and Barker, 2009) that directly affects the dopaminergic projections in fronto-striatal circuits (Lewis et al., 2005) has an impact on PD patients' performance on tests measuring executive function (Dubois and Pillon, 1997; Hart et al., 1998, Owen et al., 1992). Executive function is implicated in AM (Conway and Pleydell-Pearce, 2000, Dalgeish et al., 2007, Williams et al., 2007), thus, the depletion of dopaminergic could lead to AM failures in PD.

To summarize, the novelty of this research was to assess AM in PD by referring to recent AM models (Conway and Pleydell-Pearce, 2000, Tulving et al., 1988) distinguishing between two forms of AMs: personal facts and personal events. Previous findings exploring memory for public events showed impairments for more recent events (Sagar et al., 1988) suggesting the existence of a temporal gradient. However, whether or not PD patients show a temporal gradient while recalling personal events was unclear (Sagar et al., 1988) and never explored for personal factual knowledge in this population. To explore AM recall in PD and the existence of a temporal gradient, an autobiographical fluency task was used (Dritschel et al., 1992) in which participants were asked to recall personal events and personal facts from five separate lifetime periods. Finally, previous research converges on the hypothesis that AM retrieval is impaired in Parkinson's patients, and specifically, that it is overgeneral (Sagar et al., 1988). Thus, for personal events, we assessed whether or not Parkinson's patients would show impairments in recalling specific autobiographical memories (such as ‘I remember my first day at high school the weather was fantastic but I was sad because I forgot my lunchbox.’).