Models of Memory in Ageing

One of the main aims of ageing research is to be able to account for the effects of ageing on cognitive abilities by offering an explanation for age-related changes in brain anatomy and performance (S. W. Davis et al., 2008). Different researchers in cognitive neuroscience have proposed models of ageing in an attempt to explain the age-related deficits in cognitive performance. The most prominent of these models are outlined below.

Jacoby’s Capture Model (Jacoby, Bishara, Hessels, & Toth, 2005) posits that an interference or inhibitory control defect causes improper engagement in the recollection of information. According to Jacoby (1999), capture is composed of two stages: an early-selection stage and a later evaluation stage. The early-selection stage consists of identifying the characteristics of an episode in order to allow for retrieval of that episode (Schacter, Norman, & Koutstaal, 1998). The later evaluation stage involves accepting a recollection as “remembered” once it passes a particular accessibility threshold (Dockree et al., 2006). The model offers an explanation for the decline in cognition by suggesting that an increase in interference can lead to a decrease in the ability to recall characteristics, or the interference can cause a decrease in the accessibility threshold and items are accepted as remembered. The model further suggests that both of these occur without a corresponding change in the other stage of capture (Jacoby, Bishara, et al., 2005). This leads to one identifying a recollection as “remembered” with a high accessibility level when the recollection is incorrect.

Naveh-Benjamin’s Associative Deficit Hypothesis (Naveh-Benjamin, 2000) works under the postulation that episodes are composed of individual units of information which have connections with one another. According to the model, the ageing process causes a breakdown in creating and retrieving these links. Furthermore,

this breakdown leads to poorer memory recall in older adults (Naveh-Benjamin et al., 2003). The Associative Deficit Hypothesis suggests that the discrepancy in ageing is caused by older adults having more difficulty than young with binding the elements of an episodic memory together, and this is particularly problematic when these elements appear to be unrelated (Naveh-Benjamin, Guez, & Shulman, 2004). However, if the elements of an episode are previously linked, the age-related decrement in memory is less severe (Naveh-Benjamin et al., 2003).

The most prominent and influential neuroanatomical models of ageing suggest a compensatory function in the brains of older adults. These include the HAROLD model, the CRUNCH model and the PASA model. Research has indicated that the left PFC is more involved during the process of encoding an item into memory while the right PFC has more responsibility in the recall of our memories (Tulving et al., 1994). This led to the development of the Hemispheric Encoding/Retrieval Asymmetry (HERA) model (Nyberg, Cabeza, & Tulving, 1996; Tulving et al., 1994) postulating that the right and left hemispheres of the brain play different roles in human memory, i.e. left for encoding, right for recall. However, with the emergence of further studies, it appears as though the HERA model is specific to healthy young adults, and studies have indicated that our brain anatomy and physiology changes with ageing (Cabeza, 2002). The Hemispheric Asymmetry Reduction in Older Adults (HAROLD) model offers a rationalization for why the HERA model may only be seen in younger adults by

brain regions become active, rather than by the ineffective recruitment of the required areas (dedifferentiation) (Berlingeri, Danelli, Bottini, Sberna, & Paulesu, 2013). The HAROLD model also postulates that the deviation to bilateral activity may stem from a change in cognitive strategies, or that it may reflect changes in neural mechanisms globally or locally, and that this pattern may be task dependent (Cabeza, 2002).

However, more recent research has indicated that the HAROLD model may not reflect a compensatory function within the ageing brain (Duverne, Motamedinia, &

Rugg, 2009; Manenti, Cotelli, & Miniussi, 2011). These findings led to the development of the Compensatory-Related Utilisation of Neural Circuits Hypothesis (CRUNCH) model. The CRUNCH model posits that more activation occurs in the brain in order to allow older adults to have similar recollections to those of young adults (Reuter-Lorenz & Cappell, 2008). The main idea behind this model is that older adults recruit more neural resources than younger adults for simpler tasks (Grady, 2012;

Rypma, Eldreth, & Rebbechi, 2007). This additional recruitment causes an increase in activity in the brain of older adults compared to young, and occurs irrespective of hemisphere (Berlingeri et al., 2013). This pattern of overactivation has been observed by many researchers during different aspects of cognition, such as working memory (Cappell, Gmeindl, & Reuter-Lorenz, 2010; V S Mattay et al., 2006; Schneider-Garces et al., 2010) and source memory (Spaniol & Grady, 2012). The model further postulates that as the demands of a task increase, these overactivations plateau, leading to activity that is equivalent to or less than that of young adults (Grady, 2012; Reuter-Lorenz & Cappell, 2008).

Another widely cited model is Cabeza’s PASA model, which describes an age-related reduction in occipitotemporal activity which is accompanied by a corresponding increase in frontal activity (S. W. Davis et al., 2008). Thus, there is a shift in activity

from the posterior to the anterior regions of the brain (Posterior-Anterior Shift in Ageing, PASA). The model proposes that this occurs as a compensatory function in the brain due to the ageing process (Grady, 2012). Grady et al. (1994) first reported a shift from posterior to anterior activity in older adults. Since then, this pattern has been observed in many cognitive tasks involving older adults (Anderson et al., 2000; Cabeza et al., 2004; Gutchess et al., 2005; Rypma & D’Esposito, 2000). Since these initial studies, Davis and colleagues (2008) have indicated that this shift is independent of task difficulty and, when performance levels are matched between young and older adults, this pattern remains stable irrespective of the confidence levels of groups. This research has indicated that the PASA model may be indicative of a compensatory effect whereby the anterior region of the brain is recruited in order to balance the decline in the posterior regions which may have been caused by ageing (Davis et al., 2008).

Each of these models offers an explanation for the age-related decline seen in memory. However, as the HAROLD, CRUNCH and PASA models discuss changes in neural activation due to the ageing process, these offer the most beneficial explanation for cognitive decline in ageing, and each of the models posits a compensatory effect in ageing, suggesting that the ageing brain can benefit from compensation. As the primary focal point of this thesis is source memory, a more comprehensive account of source memory and ageing is necessary to further understand the processes of change in source memory in the ageing brain.

1.6 Source Memory and Ageing