Theories of Ageing and Memory

2.3.1 Speed of Processing

Older adults frequently require increased time compared to young adults to complete cognitive tasks (e.g. Salthouse, 1996; Verghaeghen and Salthouse, 1997). Speed of processing theory proposes that this increased time reflects a reduction in the speed at which many cognitive processes operate in the elderly, therefore reduced processing speed is a major contributing factor to age-related memory decline. Salthouse (1996) proposed that two general mechanisms underlie the relationship between speed of processing and memory performance. The first mechanism – the limited time

mechanism – states that the time to perform a later operation may be limited if most of the available time is occupied by the execution of earlier operations. The second mechanism – the simultaneity mechanism – proposes that information from different sources may not be available as fast as in the young, resulting in earlier information being lost by the time later processing is completed. Either of these mechanisms can result in slower performance on memory tasks, which can either be accurate or inaccurate. Earlier stages of processing may be slowed but accurate, but can result in poor performance if this slowing results in a failure to reach later stages.

Evidence in support of the speed of processing theory comes from a variety of sources that have found evidence that performance on perceptual speed tasks is an excellent indicator of age-related performance on episodic memory tasks. Perceptual speed tasks require rapid perceptual same-difference judgements about pairs of digits or letter strings. Speed of processing is measured by the number of correct judgements made within a certain time limit. Bunce and Macready (2005), for example, found that measures of perceptual processing speed accounted for older adults producing less

remember responses and more know responses. In addition, correlations between processing speed and accuracy on episodic memory tasks revealed that up to 70% of the age-related variance in accuracy may be related to variance in processing speed

(Salthouse, 1996; Verhaeghen and Salthouse, 1997; Park and Hedden, 2001).

2.3.2 Inhibition Deficit Hypothesis

The inhibition deficit hypothesis (Hasher and Zacks, 1988; see also Zacks and Hasher, 1994; Zacks et al., 1996; Hasher et al., 1999) proposes that we perform cognitive tasks through two mechanisms of selective attention: activation and inhibition. Inhibition suppresses the activation of goal irrelevant information so it is less likely to enter working memory, freeing space for goal relevant information; irrelevant information that does enter working memory is quickly removed. Attentional inhibition may also prevent the return of attention to a previously rejected item, such as an external stimulus or an internal thought. The inhibition deficit hypothesis states that the inhibitory

mechanism becomes deficient with ageing, resulting in older adults being slower, less able to focus on goal relevant information, and worse at remembering details than young adults.

Support for the inhibition deficit hypothesis comes from the negative priming effect.

When the brain is required to inhibit a response to a stimuli, having it uninhibit that response takes some time. The delay is called the Negative Priming Effect. The brain has been prepared to take extra time in producing a response, because that response was previously suppressed. Negative priming is frequently assessed using a variation of the Stroop Test, where the participant is presented with the name of a colour printed in a different colour font. The participant is told to ignore what the word says, and to instead say the colour of the font. On an immediately subsequent trial, the participant is

presented with the name of a colour printed in the previously suppressed font, and is again told to say the colour of the font: for example on the first trial the word GREEN is printed in blue font and the participant is required to say ‘blue’. In the subsequent trial, the word RED is printed in green font and the participant is required to say ‘green’.

Several studies have demonstrated that, in comparison to young people, older adults have a reduced Negative Priming Effect, indicating that their attentional inhibition to previous stimuli is poorer (Hasher et al., 1991; Kane et al., 1994; Stoltzfus et al., 1993).

In addition, the elderly show a range of effects consistent with the notion of reduced inhibition. These effects include increased susceptibility to concurrent environmental distracters (Connelly et al., 1991), and from concurrently activated goal irrelevant thoughts (Gerrard et al., 1991).

2.3.3 Reduced Processing Resources

While the inhibition deficit hypothesis proposed that attentional inhibition is deficient in the elderly, the reduced processing resources theory states that the attentional resources available to older adults for conscious processing are less than that available to the young (Craik and Simon, 1980; Craik and Byrd, 1982; Craik, 1983). Therefore,

retrieval tasks that require a great deal of self-initiated processing, or strategic retrieval, should be most susceptible to ageing because strategies are effortful and demand attentional resources (Light, 1991).

Examining the consequences of divided attention on memory performance has provided evidence in support of the reduced processing resources theory. Dividing attention between a primary task and a simultaneously performed secondary task is more demanding than performing the primary task alone, therefore requires increased

attentional resources. For example, dividing attention at encoding produces age-related

recognition memory deficits during retrieval, indicating that the attentional resources of the elderly were reduced compared to the young (Craik and McDowd, 1987; Anderson et al., 1998). In addition, Whiting and Smith (1997) found that dividing attention at retrieval reduced recognition memory performance more so for older than for younger adults.

The three general resource theories outlined above are not in competition. Reduced speed of processing, reduced inhibitory control and reduced processing resources are likely to operate together to produce the age-related memory impairment seen in many studies (Luo and Craik, 2008). These three theories provided the predominant

theoretical explanations of adult age differences in cognition in the 1980s and 1990s (Phillips and Henry, 2005). However, one major criticism of these theories is their lack of specification, i.e. their precise anatomical substrates are yet to be identified

(Salthouse, 1996). In the 1990s, attention shifted away from general resource theories to the frontal lobe hypothesis of cognitive ageing, which combined neuroanatomical and neuropsychological evidence to relate cognitive changes to neural changes in the frontal lobes.

2.3.4 The Frontal Lobe Hypothesis

Neuroanatomical research has consistently shown that decreases in brain volume and increases in cerebrospinal fluid occur with ageing (Stafford et al., 1988; Raz, 2000).

However, there are greater age-related changes in both the neuroanatomy and

neurochemistry of the frontal lobes than other cortical regions. The PFC shows a 10-17% reduction in volume compared to 1-8% reduction elsewhere (West, 1996). In addition to decreases in frontal lobe volume with ageing, alterations in glucose metabolism and cerebral blood flow have been observed in the region (Madden and

Hoffman, 1997; Raz, 2000), along with declines in neuronal synaptic density and dendritic arborisation (Esiri, 1994), and increases in white matter (Kawamura et al., 1993; Pantoni and Garcia, 1997).

The frontal lobes sub-serve our executive functions. Neuropsychological studies have revealed that patients with frontal lobe damage show a range of executive impairments and, as executive functions play an important role in controlling memory retrieval, these studies have also indicated that memory disruption is a key characteristic of frontal damage (Stuss and Benson, 1987). Essentially, the range of executive related deficits seen with normal ageing is highly similar to that found in frontal lobe patients

(Moscovitch and Winocur, 1995; but see Phillips and Henry, 2005 for discrepant findings). Consistent with frontal lobe patients, healthy adults demonstrate an age-related decline on tasks such as the Wisconsin Card Sorting Test, which is considered to tap executive functioning, including inhibition (Cohn et al., 1984; Daigneault et al., 1992; Shilling et al., 2002) and cognitive switching (Hughes and Bryan, 2002; Isingrini and Vazou, 1997; Parkin and Java, 1999).

While the aforementioned results convincingly substantiate the frontal lobe hypothesis of ageing, several important limitations must be highlighted: Firstly, just as the lack of uniformity of frontal lesions introduces a high degree of variability into patient data, the age-related changes in the neuroanatomy and neurochemistry of the frontal lobes will also show a high degree of variability from person to person, making a precise

characterisation of the frontal lobe hypothesis of ageing difficult. Secondly,

neuropsychological tests of frontal functioning only provide an indirect indication of age-related changes in the frontal lobes and, as it is impossible to measure executive functions without tapping other cognitive functions such as attention or perception (Phillips and Henry, 2005), performance on these tests may also reflect a combination

of impairment on the three general resource theories outlined previously (reduced speed of processing, reduced inhibitory control and reduced processing resources).

Nonetheless, over the last decade, neuroimaging techniques have provided a tool to directly investigate the relationship between the brain and cognitive ageing; combining the findings from neuroimaging and neuropsychology can provide a more complete picture of the brain regions affected by ageing and the consequences on cognition. The following section presents evidence from functional haemodynamic neuroimaging studies, showing the age-related changes in the frontal cortex that are involved in episodic retrieval.

2.3.4.1 Neuroimaging: Changes in Frontal Lobe Activation with Age

Neuroimaging studies have shown that age-related anatomical changes in the frontal lobes are associated with lower frontal activation in older adults during episodic retrieval (Grady, 2002). The more common finding, however, from neuroimaging studies is that age-related anatomical changes in the frontal lobes are related to more widespread and often equivalent levels of activation in frontal lobe regions than young adults, especially during episodic retrieval (Nyberg et al., 1996; Cabeza, 2002).

The pattern of frontal activity observed in the elderly, whether epitomized by lower frontal activation or more widespread and equivalent levels of activation, could reflect the neuroanatomical or neurochemical changes in the ageing frontal lobes that are associated with older adults being less effective at using retrieval strategies (Madden et al., 2002). Alternatively the age-related differences could reflect compensatory

processes to help counteract the cognitive decline related to changes in the frontal lobes.

Support for the compensation account comes from a variety of sources. For example, an event-related fMRI study showed that when episodic memory performance was

equated between young and older adults, older adults exhibited increased bilaterality (Morcom et al., 2003). In addition, younger adults also show increased bilateral frontal activation when the working memory load of a task increases (e.g. Jonides et al., 1997).

Furthermore, a PET study showed that older adults who performed well on a source memory task exhibited bilateral prefrontal activation, whereas those who performed less well produced only right-sided activation (Cabeza et al., 2002).

It seems unlikely that the effects of neuroanatomical and neurochemical changes in the ageing brain on cognition are restricted to the frontal lobes (Greenwood, 2000; Band et al., 2002). The following section therefore reviews evidence implicating structural and neurochemical changes in the hippocampus in the age-related decline in episodic memory.