Episodic Memory Decline and Healthy Aging

Episodic Memory Decline and Healthy Aging

3.28 Episodic Memory Decline and Healthy Aging S. Daselaar and R. Cabeza, Duke University, Durham, NC, USA ª 2008 Elsevier Ltd. All rights reserved. ...

710KB Sizes 0 Downloads 3 Views

3.28 Episodic Memory Decline and Healthy Aging S. Daselaar and R. Cabeza, Duke University, Durham, NC, USA ª 2008 Elsevier Ltd. All rights reserved.

3.28.1 3.28.1.1 3.28.1.1.1 3.28.1.1.2 3.28.1.2 3.28.1.2.1 3.28.1.2.2 3.28.1.3 3.28.1.4 3.28.2 3.28.2.1 3.28.2.2 3.28.2.3 3.28.2.4 3.28.3 3.28.3.1 3.28.3.1.1 3.28.3.1.2 3.28.3.2 3.28.3.2.1 3.28.3.2.2 3.28.3.2.3 3.28.3.3 3.28.3.4 3.28.3.5 3.28.3.6 References

Effects of Aging on Memory Performance: Executive Functions Versus Binding Deficits Resource Deficit Hypothesis Encoding Retrieval Binding Deficit Hypothesis Encoding Retrieval Summary Assumptions Regarding Brain Regions Underlying Resource and Binding Deficits Effects of Aging on PFC and MTL Anatomy and Physiology Brain Atrophy Declining White Matter Integrity Dopamine Deficits Summary Effects of Aging on PFC and MTL Activity PFC Encoding Retrieval MTL Encoding Retrieval Linking cognitive theories to age-related changes in PFC and MTL Resource Deficit Hypothesis and PFC Function Binding Deficit Hypothesis and MTL Function Healthy versus Pathological Aging Summary

Aging is accompanied by continuing degradation of the anatomy and function of our brain. While our brain shrinks and its functions decline, cognitive processes also slow down and begin to falter. Among the most prevalent cognitive problems in older adults are deficits in memory function. Difficulties in learning and memory can be found to a certain degree in all older adults. Understanding age-associated memory decline is important for two reasons. First, in view of the mounting number of older adults in today’s society, cognitive aging is increasingly becoming a problem in our health care system, and therapeutic intervention methods can only be developed on the basis of knowledge obtained through basic research. Second, there is a subgroup of elderly whose memory impairments are more severe, preventing normal

578 578 578 579 579 579 580 581 581 582 582 583 584 585 585 586 586 588 590 590 591 592 592 594 595 596 596

functioning in their environment. In these persons, such impairments can be the earliest manifestation of pathological age-related conditions, such as Alzheimer’s disease (AD). Particularly in the early stages of this disease, the differentiation from normal age-related memory impairments is very difficult. Thus, it is important to delineate which memory impairments can be regarded as correlates of normal aging and which impairments are associated with age-related pathology. One type of memory that is particularly affected by the aging process is our memory for personally experienced past events, or episodic memory (EM) (Tulving, 1983; Gabrieli, 1998). Clinical studies have shown that EM is primarily dependent on the integrity of the medial temporal lobe (MTL) memory system 577

578 Episodic Memory Decline and Healthy Aging

(Milner, 1972; Squire et al., 2001). However, the prefrontal cortex (PFC) also plays an important role in EM (Stuss and Alexander, 2000; Fletcher and Henson, 2001; Maril et al., 2003). As a result of increased accessibility of brain imaging techniques, such as positron emission tomography (PET) and functional magnetic resonance imaging (fMRI), aging research has now started to focus on the relationship between age-related changes in memory performance and changes in brain function. Functional neuroimaging (PET and fMRI) provides an ideal method to study patterns of neurocognitive decline, because changes in brain activity can be directly related to the effects of aging on behavioral measures, providing a link between cerebral aging and cognitive aging. One popular neurocognitive view is that the decline in EM associated with healthy aging results from a selective deterioration of PFC regions, whereas degradation of other brain regions, such as the MTL, is a hallmark of pathological age-related conditions (West, 1996; Buckner, 2004; Hedden and Gabrieli, 2004). However, there is now substantial evidence that MTL decline contributes to EM deficits in healthy older adults, and hence, it is no longer possible to attribute these deficits exclusively to PFC decline. In this chapter, we review recent PET and fMRI studies of healthy aging and EM with a focus on age-related changes in PFC and MTL regions. In relation to these findings, we address two major factors thought to underlie age-related memory decline and strongly linked to PFC and MTL function, namely, deficits in executive function and deficits in binding processes. The term ‘executive function’ describes a set of cognitive abilities that control and regulate other abilities and behaviors. With respect to EM, executive functions are necessary to keep information available online so that it can be encoded in, or retrieved from, EM. The PFC is generally thought to be the key brain region underlying executive functions (Miller and Cohen, 2001). Binding refers to our capacity to bind into one coherent representation the individual elements that together make up an episode in memory, such as sensory inputs, thoughts, and emotions. According to relational memory theory, different subregions of MTL are differentially involved in memory for relations between items and memory for individual items (Cohen and Eichenbaum, 1993; Eichenbaum et al., 1994). In particular, the hippocampal formation is more involved in binding or relational memory operations, whereas the surrounding parahippocampal cortex is more involved in individual item memory.

This chapter has four main sections. In the first section, we begin with a brief overview of behavioral evidence indicating how EM performance changes as we age. In the second section, we discuss the anatomical and physiological changes in MTL and PFC that accompany the aging process. In the third section, we focus on functional neuroimaging studies of EM that reveal age-related changes in MTL and/or PFC regions. In the final section, we discuss different interpretations of age-related memory decline that have emerged from these findings, and how they relate to deficits in executive function and binding capacity.

3.28.1 Effects of Aging on Memory Performance: Executive Functions Versus Binding Deficits Age-related deficits in EM may reflect difficulties during the formation of new episodic memory traces (encoding) and/or during the recovery of stored memory traces (retrieval). Two main theories of cognitive aging have been put forward to account for age-related deficits in EM encoding and retrieval. According to the resource deficit hypothesis (Craik, 1986), age-related cognitive impairments, including EM deficits, are the result of a general reduction in attentional resources. As a result, older adults have greater difficulties with cognitive tasks that provide less environmental support and, hence, require greater self-initiated processing. According to the binding deficit hypothesis ( Johnson et al., 1993; Naveh-Benjamin, 2000), older adults are impaired in forming and remembering associations between individual items and between items and their context. As a result, age-related EM impairments are more pronounced on tasks that require binding between study items. In separate sections below, we discuss some examples of behavioral studies of encoding, retrieval, and aging in relation to these two theories of cognitive aging. Within each of these sections, we group studies according to whether they manipulated factors affecting mainly encoding or retrieval. However, it is important to keep in mind that behavioral studies cannot make a clear distinction between these two phases of episodic memory. 3.28.1.1

Resource Deficit Hypothesis

3.28.1.1.1

Encoding In terms of encoding, the resource deficit hypothesis is supported by evidence that age-related deficits are

Episodic Memory Decline and Healthy Aging

3.28.1.1.2

Retrieval Similar to encoding, retrieval studies of aging have shown that age-related deficits in retrieval become larger on tasks that put a greater demand on executive function. Older adults experience more difficulties on recall and context memory tasks, which are more dependent on self-initiated search processes, than on recognition tasks, in which candidate targets are provided by the experimenter and the participant only needs to decide whether or not they were part of the study list (Rabinowitz, 1984; Rabinowitz and Craik, 1986; Craik and McDowd, 1987). For example, in a study by Craik and McDowd, young and older adults performed cued recall and recognition tasks, both of which also included a secondary reaction time task. As expected, older adults performed less well than the young adults on the recall task, but not on the recognition task (see Figure 1). Moreover, relative to recognition, older adults were much more impaired during recall on the secondary task than were the young adults. These findings are in line with the idea that recall requires more resources than

0.9 Young

0.85

Old

0.8 Proportion correct

most pronounced when there is little environmental support to help encoding. For instance, age-related differences tend to be larger when young and older adults intentionally try to encode a list of study items than when a deep semantic processing task is used to guide encoding (Craik and Simon, 1980; Burke and Light, 1981). As an example, in a recent study (Troyer et al., 2006), older adults showed significantly impaired recognition memory performance following intentional encoding of a list of proper names, but they performed equally well as young adults when encoding was incidental and involved making semantic associations to the names. Likewise, age differences in encoding can be reduced by providing additional semantic context with the study items. For example, in one study (Craik et al., 1987), young and older adults encoded a list of words either with or without a short descriptive phrase. Age-related deficits were much smaller when the encoded words were accompanied by the descriptive phrases. Also, age-related deficits in encoding become greater when there is no apparent link between the elements. Smith et al. (1998) found that older adults’ difficulties in recalling associations between two pictures were attenuated when the two pictures were linked by preexistent relationships. The authors interpreted these findings in terms of age-related differences in self-initiated processing.

579

0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4

Recall

Recognition

Figure 1 Reduced memory performance in older adults compared with young adults during cued recall, but not during recognition. Adapted from Craik FIM and McDowd JM (1987) Age differences in recall and recognition. J. Exp. Psychol. Learn. Mem. Cogn. 13: 474–479.

recognition and that these resources are reduced in older adults. One explanation for the greater age-related decline in recall than in recognition is that older adults do not spontaneously produce retrieval cues to guide the search process. In line with this idea, several studies have found that age-related deficits are reduced when cues are provided (Park and Shaw, 1992; Naveh-Benjamin and Craik, 1995). For example, age differences in EM become smaller when the encoding context is reinstated during retrieval. In a study by Naveh-Benjamin and Craik, young and older adults encoded words that were either shown in one of two different fonts or spoken by a female or a male voice. Age differences in recognition performance were much smaller when the font or voice was reinstated during recognition. In other words, older adults benefit when provided with environmental support at retrieval. 3.28.1.2

Binding Deficit Hypothesis

3.28.1.2.1

Encoding In addition to deficits in self-initiated processing, several studies have shown that older adults consistently exhibit age-related deficits in encoding tasks that require binding different pieces of information together, even when memory for individual features is intact (Schacter et al., 1994; Chalfonte and Johnson, 1996; Mitchell et al., 2000b; Naveh-Benjamin, 2000). Chalfonte and Johnson (1996) tested age-related differences in encoding of

580 Episodic Memory Decline and Healthy Aging

3.28.1.2.2

Retrieval Age-related associative deficits are present during retrieval as well. According to dual-process models, recognition memory can be based on the recovery of specific contextual details (recollection) or on the mere feeling that an event is old or new in the absence of confirmatory contextual information (familiarity). Older adults are more impaired in recollection than in familiarity, which has been demonstrated using the Remember/Know (R/K) paradigm (Parkin and Walter, 1992; Mantyla, 1993; Java, 1996; Davidson and Glisky, 2002; Bastin and Van der Linden, 2003), ROC curves (Howard et al., 2006), and the processdissociation procedure ( Jennings and Jacoby, 1993). For example, Parkin and Walter (1992) used the R/K

1 Young: full attention Older adults Young: divided attention

0.95

Memory performance

both color and item information across three experiments. While no effect of age was found for individual features, older adults exhibited significant deficits in recognition for combined features compared with younger adults. Furthermore, age-related differences in memory for bound features persisted across various encoding instructions. Similar results were reported by Mitchell et al. (2000b), who also interpreted results as an age-related deficit in binding features in memory. Recent studies have indicated that this associative deficit is not limited to arbitrary associations (e.g., word–color, item–location) but is also found in more ecologically valid pairings (Naveh-Benjamin, 2000; Naveh-Benjamin et al., 2004). For instance, in a study by Naveh-Benjamin and colleagues (2004), young and older adults studied a list of face–name pairs. Memory for individual faces and names was tested using forced-choice recognition tests in which the faces and names were paired with other names and faces not seen at study. Associative memory for the face–name pairs was tested in a separate test that included only studied faces and names. Participants paired a face or name with the corresponding name or face shown at study by choosing from two alternatives. Although older adults performed similarly to the young adults on individual face and name recognition, they were considerably impaired on the associative recognition test. These age-related differences persisted even when the young adults encoded information under divided attention (see Figure 2). These results indicate that binding deficits in older adults cannot be explained solely by a reduction in attentional resources. Thus, available evidence suggests that, even though older adults show deficits in general memory performance, this deficit is greater for associative tasks.

0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.55 0.5

Face recognition

Name recognition

Name–face associative recognition

Figure 2 Reduced memory performance in older adults compared with young adults during associative recognition, but not during recognition of single faces and names; binding deficits in older adults cannot be explained solely by a reduction in attentional resources. Adapted from NavehBenjamin M, Guez J, and Shulman S (2004) Older adults’ associative deficit in episodic memory: Assessing the role of decline in attentional resources. Psychon. Bull. Rev. 11: 1067–1073.

paradigm in which participants indicate whether their recognition judgment was based on recollection (R) or familiarity (K). They found that older adults made fewer R responses and more K responses than younger adults. Given that recollection involves the retrieval of associations between core and contextual elements of an episode, this evidence is consistent with the view that binding deficits are an important factor in agerelated EM decline (Naveh-Benjamin, 2000). In line with the aforementioned deficits in recollection, older adults are also less accurate in recalling information associated with the context in which an item was encoded, including the color, case, or font of words (Park and Puglisi, 1985); modality of presentation (e.g., auditory or visual) (Light et al., 1992); and speaker gender (Kausler and Puckett, 1981). For example, when given a list of made-up facts (e.g., Bob Hope’s father was a fireman) and tested on them 1 week later, both younger and older adults could successfully recall the items, but older adults were impaired at knowing where they had first learned the information (McIntyre and Craik, 1987). Similarly, older adults are impaired at knowing where on a computer screen an item was presented, though they have little problem in recognition of items themselves (Parkin et al., 1995). Thus, impaired memory for context appears to exceed memory impairments for individual items. A prevailing

Episodic Memory Decline and Healthy Aging

theory accounting for these recollection and context memory deficits involves the binding deficit hypothesis. 3.28.1.3

Summary

To summarize, behavioral evidence of EM encoding and retrieval has provided support for both resource and binding deficit hypotheses. In support of the resource deficit hypothesis, age-related deficits in encoding and retrieval are greater on EM tasks with little environmental support, which require more self-initiated processing. The resource deficit hypothesis is in line with the idea that a general reduction in executive function plays an important role in age-related memory decline. In support of the binding deficit hypothesis, age deficits are more pronounced when encoding and retrieval tasks require the formation or recollection of item–item or item– context associations. The binding deficit hypothesis supports the idea that a decline in relational memory functions plays an important role in age-related EM deficits. 3.28.1.4 Assumptions Regarding Brain Regions Underlying Resource and Binding Deficits Adding brain assumptions to the resource and binding deficit hypotheses is straightforward, because different neural substrates have been proposed for executive and binding operations. As noted in the introduction, the PFC is considered the key brain region underlying executive functions and the managing of attentional resources (Miller and Cohen, 2001). Based on this generally accepted idea, Craik (1983) proposed that older adults’ deficits in executive processing are related to a reduction in the efficiency of PFC functioning. The strong relation between PFC function and executive processes is exemplified by patient, animal, and functional neuroimaging studies using the wellknown Wisconsin Card Sorting Test (WCST) – a task that draws heavily on top-down attentional and executive processes (Miller and Cohen, 2001). In this task, participants sort a set of cards according to number, color, and symbol without being told the correct sorting rule, and this rule changes periodically. The experimenter only indicates whether a response was correct or not. Thus, any given card can be associated with several possible actions, and the correct one is determined by whichever rule is

581

currently in effect. Several researchers have found that humans with PFC damage show clear deficits in the WCST. They are able to acquire the initial rule but are unable to adapt their behavior when the rule changes (Milner, 1963). Similar findings have been reported in animal studies. For instance, Dias et al. found that monkeys with experimental lesions to the PFC show very similar deficits in analog task of the WCST (Dias et al., 1996b, 1997). Finally, functional neuroimaging studies have also indicated the importance of PFC in the WCST. For example, a recent fMRI study found that different PFC subregions mediate different types of executive operations during WCST performance. Whereas ventrolateral PFC activity in the WCST was related to simple working memory operations, dorsolateral PFC activity was associated with more complex/manipulative working memory operations (Lie et al., 2006). Similar to the strong relation between PFC and executive function, there is substantial evidence that the MTL is critical for binding. Recent evidence indicates that within MTL, binding is specifically associated with the function of the hippocampus, whereas the surrounding cortical regions (e.g., perirhinal cortex) are more involved in individual item memory (Brown and Aggleton, 2001; Eichenbaum, 2006). The selective role of the hippocampus in binding operations can be exemplified with patient, animal, and functional neuroimaging studies that capitalized on the distinction between recollection-based (item– context binding) and familiarity-based (individual item memory) EM retrieval. Regarding patient data, using several different measures including ROC curves, Yonelinas and colleagues (2002) found that hypoxic patients, who typically have greater hippocampal than parahippocampal damage, showed a decline in recollection measures but normal familiarity. This was also reflected in their ROC curves, which were less asymmetrical (recollection) and more curvilinear (familiarity). These findings suggest that an intact hippocampus is necessary for recollection but not for familiarity. Regarding animal data, Fortin et al. (2004) also examined the contribution of recollection and familiarity processes using ROC curves in an odor recognition memory test designed for rodents. Normal rats showed ROC curves that had asymmetrical and curvilinear components, indicating the existence of both recollection and familiarity. However, following selective damage to the hippocampus, the ROC curve was not asymmetrical

582 Episodic Memory Decline and Healthy Aging

(recollection) but completely curvilinear (familiarity), supporting the view that the hippocampus specifically mediates the capacity for recollection. Finally, regarding functional neuroimaging data, a recent study by Daselaar et al. (2006a) used confidence ratings to distinguish between recollection and familiarity. Confirming patient and animal findings, they found that the hippocampus showed selective activity for high confidence retrieval (recollection), whereas parahippocampal and rhinal activity showed a gradual function (familiarity) with confidence. Thus, based on patient, animal, and functional neuroimaging evidence, we can infer that resource and binding deficits are tied to age-related changes in PFC and MTL (hippocampus), respectively. In the remainder of this chapter, we focus on how the structure and function of PFC and MTL change as a result of the aging process, and in the final section, we discuss a possible link between these behavioral and neurobiological changes.

3.28.2 Effects of Aging on PFC and MTL Anatomy and Physiology The effects of aging on the brain occur at many levels from genes to gross anatomy. Reviewing this large research domain is beyond the scope of this chapter. Here, we focus on PFC and MTL and mention only three examples of cerebral aging measures that have been directly related to cognitive decline in aging humans: brain atrophy, declining white matter integrity, and dopamine deficits. 3.28.2.1

Brain Atrophy

In postmortem and in vivo studies, the brains of older adults tend to have lower volumes of gray matter than young adult brains (Resnick et al., 2003). These volume declines are not always related to a loss of cells but can also be ascribed to lower synaptic densities in older adults (Terry, 2000). Cross-sectional studies suggest that the volume of gray matter declines linearly with age, whereas white matter volume increases during childhood, plateaus during young adulthood and middle age, and declines during old age, an inverted U function (Raz et al., 2005). Apart from differential age effects on gray and white matter volume, the relation between age and brain volume is also not uniform across different brain regions. The region most affected is the PFC, whereas other

regions, such as the occipital cortex, are relatively unaffected by the aging process (Raz et al., 1997). With an average decline rate of between 0.9% and 1.5% per year, the frontal lobes show the steepest rate of atrophy (Pfefferbaum et al., 1998; Resnick et al., 2003; Raz et al., 2005). The disproportionate effect of aging on PFC volume, together with the finding that age-related differences tend to be larger on tasks assumed to depend on PFC function, has led to the proposal that age-related deficits are primarily the result of frontal decline (for a review, see West, 1996). Indeed, frontal atrophy has been shown to correlate with cognitive deficits mediated by frontal regions. For example, Gunning-Dixon and Raz (2003) found that in a large group of older adults, perseveration errors on the WCST negatively correlated with prefrontal volume. However, other brain regions, including the basal ganglia (Bugiani et al., 1978; Schwartz et al., 1985) and the thalamus (Petersen et al., 2000; Gallo et al., 2004), also show a pronounced decline in brain volume with increasing age. In fact, in the last decades of normal life, volumetric changes in PFC do not differ from those in other neocortical areas (Resnick et al., 2000; Salat et al., 1999). MTL volume also declines with age, though not all subregions (e.g., entorhinal cortex, hippocampus, parahippocampal gyrus) are equally affected. For example, as shown in Figure 3, a recent longitudinal study found that in healthy older adults, the hippocampus showed substantial atrophy, whereas the entorhinal cortex did not (Raz et al., 2005). Furthermore, studies have shown that the rate of hippocampal atrophy increases with age (Scahill et al., 2003; Raz, 2004). In one study, for example, this rate was an average of 0.86% per year in the whole sample (26–82 years) but 1.18% when considering only individuals over 50 years of age (Raz, 2004). A review of 12 studies estimated that after the age of 70 this rate may be as high as 1.85% per year (see Raz et al., 2005). The differential effects of aging on hippocampus and entorhinal cortex is very interesting, because the entorhinal cortex is one of the regions first affected by AD (Braak et al., 1993). As discussed later, together with recent fMRI evidence of dissociations between hippocampal and rhinal functions in aging (Daselaar et al., 2006b), these findings have implications for the early diagnosis of AD. Several studies have suggested that the decline in hippocampal volume contributes to age-related deficits in EM (Golomb et al., 1993, 1994; Lupien et al., 1998; Sperling et al., 2003). For instance, Golomb et al. (1993)

Episodic Memory Decline and Healthy Aging

Rhinal cortex 4.0

9.0 8.5 8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 0.0

Adjusted volume (cm3)

Adjusted volume (cm3)

Hippocampus

583

3.5 3.0 2.5 2.0 1.5 0.0

0 20

30

40 50 60 70 Age at baseline (years)

80

90

0 20

30

40 50 60 70 Age at baseline (years)

80

90

Figure 3 Longitudinal changes in volumes of hippocampus and rhinal cortex as a function of baseline age. From Raz N, Lindenberger U, Rodrigue KM, et al. (2005) Regional brain changes in aging healthy adults: general trends, individual differences and modifiers. Cereb. Cortex 15: 1676–1689.

investigated the link between hippocampal atrophy and memory performance in a group of healthy older adults. They found that, after controlling for such factors as age, education, and vocabulary skills, individuals with hippocampal atrophy performed less well on memory tests compared with those with no decline. Furthermore, as part of another longitudinal study on memory function, Golomb et al. (1996) found a significant correlation between hippocampal atrophy across a 3.8-year span and decline in memory performance in a group of older adults (mean age, 68.4 years). Additionally, Persson and colleagues (2005) found reduced hippocampal volume in a group of older adults whose episodic memory performance declined across a decade compared with that of a group whose memory performance remained stable.

3.28.2.2

Declining White Matter Integrity

In addition to declining gray matter, aging is also accompanied by reduced white matter integrity. White matter is composed of myelinated axons, which are essential for efficient neural transmission. When we age, myelin undergoes a number of significant changes including splitting of the myelin layers, increase in myelin volume resulting in redundant myelin sheets around axons, and finally, a breakdown of myelin (Wozniak and Lim, 2006). A method to study the integrity of white matter in vivo is an MRI technique termed diffusion tensor imaging (DTI). This technique measures the magnitude and directionality of water diffusion within the brain. Without barriers, the movement of water is uniform in all directions. In the presence of barriers

such as cell membranes, fibers, and myelin, the diffusion is greater in a certain direction and is termed anisotropic. Using DTI, the fractional anisotropy (FA) can be computed, indicating the degree of anisotropy in a particular area. Since degradation of white matter structure will result in a lower FA value, this measure can be used as an index of white matter integrity (Sullivan and Pfefferbaum, 2006). Several DTI studies have shown that the integrity of white matter deteriorates with advancing age (Salat et al., 2005; Sullivan et al., 2006). For instance, Salat and colleagues calculated whole-brain maps of FA to determine whether particular fiber systems of the brain are preferentially vulnerable to white matter degeneration. The results showed significant age-related decline in FA in frontal white matter, the internal capsule, and the genu of the corpus callosum. However, FA in temporal and posterior regions was relatively preserved (Figure 4). These findings indicate that fiber tracts in PFC are more vulnerable to agerelated degeneration than those within MTL regions. Similarly, Sullivan and colleagues (2006) found that frontocallosal fibers showed a much steeper rate of age-related decline than fibers in more posterior regions of the brain. Reductions in FA have also been linked to a decline in cognitive abilities in older adults (Sullivan et al., 2001; Madden et al., 2004; Charlton et al., 2006). For example, a recent study by Charlton et al. (2006) investigated the relationship between white matter structure and cognition in 106 healthy middle-aged and elderly adults. They calculated correlations between DTI measures and performance on a test battery assessing executive function, working

584 Episodic Memory Decline and Healthy Aging

Young adults

Older adults

FA = 0

FA = 0.3

FA = 0.5

Deep frontal

Deep temporal 1 Fractional anisotropy

Fractional anisotropy

1 0.8 0.6 0.4 0.2

Left: r = –0.67, p < 0.01 Right: r = –0.56, p < 0.01

0 10

20

30

40 50 Age

0.8 0.6 0.4 0.2

Left: r = –0.15, NS Right: r = –0.04, NS

0 60

70

80

10

20

30

40 50 Age

60

70

80

Figure 4 Steeper age-related decline in fractional anisotropy (FA) in frontal than in temporal white matter tracts. Adapted from Salat DH, Tuch DS, Greve DN, et al. (2005) Age-related alterations in white matter microstructure measured by diffusion tensor imaging. Neurobiol. Aging 26: 1215–1227.

memory, and information-processing speed. The results indicated a reduction in FA with age, which correlated with reduced performance in all three cognitive domains. However, after controlling for age, FA parameters correlated with working memory but not with the other two cognitive domains. These results indicate that white matter damage is an important factor in age-related cognitive decline and that working memory operations are particularly vulnerable to this type of cerebral aging. 3.28.2.3

Dopamine Deficits

Aging affects not only brain anatomy but also brain physiology, including the function of neurotransmitter systems, such as serotonin, acetylcholine, and dopamine (Strong, 1998). Associated with volume decrements in PFC are decreases in dopamine (DA) concentration and transporter availability (Volkow et al., 2000). Additionally, dopamine D2 receptor

density declines at a rate of 8% per decade, beginning in the 40s. There is abundant evidence that DA systems play an important role not only in motor operations but also in higher-order cognitive processes. DA function can be measured in vivo using PET (Ba¨ckman and Farde, 2005). There is strong evidence of age-related losses in pre- and postsynaptic DA markers, which may reflect decreases in the number of neurons, the number of synapses per neuron, and/or the expression of receptor proteins in each neuron. D1 and D2 receptor binding declines from early adulthood at a rate of 4–10% per decade, and this decline is correlated with the decline of dopamine transporter, possibly reflecting a common causal mechanism. DA loss with aging has been observed in frontal, temporal, and occipital cortices as well as in hippocampus and thalamus (Kaasinen et al., 2000; Inoue et al., 2001). The magnitude of extrastriatal DA decline mirrors that observed within the striatum itself. Given the

Episodic Memory Decline and Healthy Aging

cognitive role of frontostriatal loops, age-related striatal DA deficits could also account for age-related cognitive deficits associated with PFC dysfunction. Moreover, age-related deficits in DA binding have been observed in PFC, as well as in posterior cortical and hippocampal regions. Evidence is mixed as to whether these declines are linear (see Reeves et al., 2002) or exponential (Bannon and Whitty, 1997; Rinne et al., 1998; Ghilardi et al., 2000) across adulthood. The relationship between age-related changes in DA and age-related cognitive differences has been examined in only a small number of studies. Despite the paucity of data, findings are remarkably consistent. Age deficits in striatal DA have been associated with reduction in episodic memory (Ba¨ckman et al., 2000; Erixon-Lindroth et al., 2005), executive function (Volkow et al., 1998; Mozley et al., 2001; Erixon-Lindroth et al., 2005), and motor performance (Wang et al., 1998; Mozley et al., 2001). Furthermore, several studies have also found that striatal DA markers serve as a significant predictor of cognitive performance, after controlling for the effects of age (Ba¨ckman et al., 2000; Volkow et al., 1998), as well as that age-related cognitive deficits are mediated by reductions in striatal DA functioning (Erixon-Lindroth et al., 2005). 3.28.2.4

Summary

In sum, results indicate that the brain undergoes significant structural changes with age, but agerelated atrophy differs across and within regions. Regarding volumetric measures of gray matter, studies have shown that the frontal lobes exhibit the highest rate of decline, and posterior regions the most moderate decline. The MTL also shows substantial atrophy in healthy aging, but the rate of decline differs for different subregions. For instance, whereas the hippocampus shows a marked decline, the rhinal cortex is relatively preserved in healthy aging. Measures of white matter volume and integrity (i.e., FA) also show differential aging effects throughout the brain. Similar to gray matter decline, age-related white matter degradation and DA dysfunction are greater in anterior compared with posterior regions. Correlations between these measures and cognitive function emphasize the importance that these changes have on cognitive functions. Finally, age-related decline in both preand postsynaptic DA markers also tends to follow this anterior–posterior gradient of decline. These

585

patterns coincide with behavioral findings that show greater age-related performance decrements in cognitive functions mediated by frontal functioning (see West, 1996). In general, age-related decline in brain anatomy and physiology is most pronounced within PFC but is also present in MTL regions, particularly in the hippocampus. Based on existing evidence that PFC and MTL play key roles in executive and binding functions, these changes can readily account for the deficits reported in behavioral EM studies. The next section discusses functional neuroimaging studies, which provide a bridge between behavioral and anatomical findings by directly measuring agerelated differences in brain activity during the performance of cognitive tasks.

3.28.3 Effects of Aging on PFC and MTL Activity Before reviewing the findings of functional neuroimaging studies of EM, it is useful to first describe two patterns of age-related differences in brain activity that are consistently found in PFC and MTL. In the case of PFC, the most consistent finding has been an age-related reduction in lateralization. This evidence has been conceptualized in a model called Hemispheric Asymmetry Reduction in Older Adults (HAROLD), which states that, under similar conditions, PFC activity tends to be less lateralized in older than in younger adults (Cabeza et al., 2002). This model is supported by functional neuroimaging, electrophysiological, and behavioral evidence in the domains of episodic memory, semantic memory, working memory, perception, and inhibitory control (Cabeza, 2002). In the case of MTL, the most consistent finding has been an age-related reduction in activity. However, recent studies suggest that not all MTL regions show reduced activity in older adults. Indeed, some MTL regions show preserved or increased activity in older adults, possibly reflecting differential age effects on various EM processes. In this section, we first review the effects of aging on PFC activation in encoding and retrieval tasks. Then we turn to the effects of aging on MTL activation as it relates to these cognitive processes. Although the number of studies is still too limited to identify clear patterns in the data, and a considerable amount of variability still remains unexplained, we try to emphasize the most consistent findings across studies.

586 Episodic Memory Decline and Healthy Aging

3.28.3.1

PFC

3.28.3.1.1

Encoding Functional imaging studies of encoding and aging have investigated a wide variety of stimuli, including words, word pairs, faces, and scenes. Despite this variety of stimuli, PFC findings have been quite consistent: older adults typically show reduced left PFC activity during encoding compared with younger adults. However, there is some evidence suggesting that this effect may be modulated by whether encoding is intentional or incidental. In intentional encoding conditions, participants are asked to learn information for a subsequent memory test, while in incidental encoding conditions, a memory test is not mentioned and participants are only asked to perform a certain task (semantic or nonsemantic). In terms of the resource deficit hypothesis, intentional encoding tasks provide less environmental support than incidental encoding tasks and, thus, are more dependent on PFC-mediated executive functions. Following, we consider first intentional encoding studies, then incidental encoding studies, and finally, studies comparing intentional vs. incidental encoding. Intentional encoding studies In a PET study by Grady et al. (1995), young and older adults intentionally encoded pictures of faces followed by a recognition test. Regarding encoding-related PFC activity, they found that older adults showed less activity in the left PFC than younger adults. In another PET study by Cabeza et al. (1997), young and older adults intentionally encoded a list of word pairs for subsequent cued recall and recognition tests. During encoding, older adults showed increased activity in several regions, including the insula, but reduced activity in occipitotemporal and left prefrontal regions. Importantly, Cabeza and colleagues noted that younger adults selectively recruited the left PFC, whereas older adults showed equivalent activity in left and right PFC (i.e., HAROLD). Since younger adults and older adults had similar memory scores on both cued recall and recognition tasks, they interpreted the additional recruitment of right PFC by older adults as compensatory. In another PET study of encoding and aging, Anderson et al. (2000) assessed intentional encoding of moderately associated word-pairs under conditions of full and divided attention. During full attention, they found a pattern similar to the one

3.28.3.1.1.(i)

observed in the PET study by Cabeza and colleagues (1997). Whereas the older adults showed reduced activity in left PFC, they showed increased activity in right PFC, leading to bilateral frontal activity in older adults. Incidental encoding studies Agerelated differences in PFC activation have also been shown when investigating levels of processing at encoding. When comparing deep and shallow encoding of words, Stebbins et al. (2002) reported greater activity in both younger adults and older adults for the deep relative to the shallow encoding condition. However, the older adults showed decreased activation in left PFC. Furthermore, decreased performance on neuropsychological tests correlated with reduced PFC activity. As a result of the reduced left PFC activity, PFC activity in older adults was more symmetric than in younger adults, again in accord with the HAROLD model. Rosen et al. (2002) also studied deep and shallow encoding of words in younger adults and older adults. However, they distinguished between older adults with high and low memory scores based on a neuropsychological test battery. They reported equivalent left PFC activity but greater right PFC activity in the old-high memory group relative to younger adults. In contrast, the old-low memory group showed reduced activity in both left and right PFC. As a result, the old-high group showed a more bilateral pattern of PFC activity than younger adults (HAROLD). Similarly, using a verbal encoding/recognition task, Daselaar et al. (2003b) compared groups of high- and low-performing older adults, divided post hoc based on their memory scores. During the semantic encoding task (pleasant/unpleasant decisions), all groups showed left lateralized activations patterns, but PFC activity was slightly less lateralized in the low-performing elderly, and even less so in the high-performing elderly. Consistent with the results of Cabeza et al. (2002), these findings support the compensatory interpretation of HAROLD. Morcom et al. (2003) used event-related fMRI to study subsequent memory for semantically encoded words. Recognition memory for these words was tested after a short and a longer delay. At the short delay, performance in older adults was equal to that of younger adults at the long delay. Under these conditions, activity in left inferior PFC was greater for subsequently recognized than forgotten words and was equivalent in both age groups. However,

3.28.3.1.1.(ii)

Episodic Memory Decline and Healthy Aging

older adults showed greater right PFC activity than younger adults, again resulting in a more bilateral pattern of frontal activity (HAROLD). Gutchess et al. (2005) studied subsequent picture memory using a deep processing task. While young and older adults showed equivalent activity in right PFC, the older adults showed increased activity in the left PFC. Since picture encoding in younger adults was associated with bilateral PFC activity, these findings suggest a selective recruitment of left PFC, which may be compensatory. A recent study by Dennis et al. (2006) used hybrid blocked/event-related analyses to distinguish between transient and sustained subsequent memory effects during deep incidental encoding of words. Subsequent memory was defined as parametric increases in encoding activity as a function of a combined subsequent memory/confidence scale. This parametric response was measured in each trial (transient activity) and in blocks of eight trials (sustained activity). Similar to the study conducted by Gutchess et al., subsequent memory analyses of transient activity showed agerelated increases in left PFC. At the same time, subsequent memory analyses of sustained activity showed age-related reductions in right PFC. Contrary to Stebbins et al.’s study (2002), these findings suggest that during semantic classification, older adults can show even greater PFC-mediated semantic processing than young adults. Additionally, the decline in sustained subsequent memory activity in PFC may involve age-related deficits in sustained attention that impact encoding processes. The results underline the importance of investigating aging effects on both transient and sustained neural activity. Finally, Persson and colleagues (2006) used longitudinal behavioral data to identify two groups of older adults that differed with regard to whether their performance on tests of episodic memory remained stable or declined over a decade. During incidental encoding of words, both groups showed equivalent activation in left PFC, but the elderly subjects with the greatest decline in memory performance showed additional right PFC activity. Moreover, mean DTI measures (FA) in the anterior corpus callosum correlated negatively with activation in right PFC. These results demonstrate that cognitive decline is associated with differences in the structure as well as function of the aging brain and suggest that contralateral PFC recruitment is either caused by structural disruption or is a compensatory response to such disruption.

587

3.28.3.1.1.(iii) Intentional versus incidental encoding Grady and colleagues (1999) compared

intentional vs. incidental encoding conditions across various stimuli (e.g., words, pictures). They scanned participants during shallow (uppercase/lowercase, picture size), deep (living/nonliving), and intentional encoding conditions. Overall, picture encoding resulted in greater activity in visual and MTL regions, while word encoding yielded greater activity in left PFC and left lateral temporal cortex. However, deep encoding produced greater left PFC activity, while intentional encoding yielded greater right PFC activity. Though older adults showed the same patterns, the overall level of activity was reduced. Interestingly, researchers did not find a difference in deep vs. intentional encoding of pictures, indicating that age-related differences were greater for words than for pictures. The same research group conducted another study comparing intentional and incidental encoding conditions, this time using faces as study items (Grady et al., 2002). Convergent with their earlier study investigating intentional face encoding (Grady et al., 1995), older adults showed decreased activity in left PFC compared with younger adults and diminished connectivity between frontal and MTL areas during both encoding conditions. Despite the lack of activation differences across encoding instructions in the studies reported by Grady et al., a more recent study did find age-related frontal differences for intentional encoding instructions. Logan et al. (2002) reported that during selfinitiated, intentional encoding instructions, older adults compared with younger adults showed less activity in left PFC but greater activity in right PFC, resulting in a more bilateral activity pattern (HAROLD). Results were similar for intentional encoding of both verbal and nonverbal material. Interestingly, further exploratory analyses revealed that this pattern was present in a group of old-older adults (mean age, 80), but not in a group of young-older adults (mean age, 67), suggesting that contralateral recruitment is associated with more pronounced age-related cognitive decline. At the same time, the decrease in left PFC was not present in older adults during incidental encoding instructions, suggesting that frontal reductions can be remediated by providing environmental support during encoding (Figure 5). Summary To summarize encoding studies, the most consistent finding was an age-related reduction in left PFC activity. This

3.28.3.1.1.(iv)

588 Episodic Memory Decline and Healthy Aging

Young adults

Older adults

high-performing group (Rosen et al., 2002; Daselaar et al., 2003b). These findings provide direct support for the compensation account of HAROLD. 3.28.3.1.2

Intentional encoding

Retrieval In line with the resource deficit hypothesis, agerelated deficits in episodic retrieval tend to be more pronounced for recall and context memory tasks than for recognition tasks (Spencer and Raz, 1995). However, considerable differences in activity have also been observed during simple recognition tasks. We first review studies looking at recognition processes and then examine those that focus on recall and different forms of context memory. Recognition memory The face encoding study by Grady et al. (1995) also included a face recognition test. During this task, older adults showed reduced activity in parietal and occipital regions but equivalent activity to young adults in right PFC. This last finding contrasts with the agerelated reduction in left PFC activity found in the same study during face encoding. Based on these results, the authors suggested that age effects are more pronounced on encoding than on retrieval. As noted below, however, many subsequent PET and fMRI studies have found reliable age-related changes in PFC activity during episodic retrieval. The word-pair encoding study by Cabeza et al. (1997) also included a word-pair recognition task. During this task, older adults showed reduced activity in right PFC but increased activity in other brain regions, such as the precuneus. The age-related reduction in right PFC contrasts with the lack of age effects in this region’s activity in Grady et al.’s (1995) study. This inconsistency could reflect differences in stimuli (faces vs. words) or retrieval processes (recognition of items vs. recognition of pairs). The age-related reduction in right PFC activity during recognition was replicated by Madden et al. (1999), using a single-word recognition task. Additionally, this study found an age-related increase in left PFC. This age-related increase extended to recognition a finding previously reported for recall by Cabeza et al. (1997), which is reviewed here. As in the previous recall study, the age-related increase in left PFC led to a more bilateral pattern in older adults (i.e., HAROLD). In a subsequent study, Madden et al. (1999) reanalyzed the recognition data using a stepwise regression method that distinguished between exponential (tau) and

3.28.3.1.2.(i)

Incidental encoding

Figure 5 Reduced left PFC activity in older compared with younger adults during intentional encoding, but similar activity during incidental encoding. Adapted from Logan JM, Sanders AL, Snyder AZ, Morris JC, and Buckner RL (2002) Under-recruitment and nonselective recruitment: Dissociable neural mechanisms associated with aging. Neuron 33: 827–840.

finding was more frequent for intentional than for incidental encoding studies, suggesting that, in line with the resource deficit hypothesis, the environmental support provided by a deep semantic encoding task may attenuate the age-related decrease in left PFC activity. This effect was found within subjects in the study by Logan et al. (2002). The difference between intentional vs. incidental encoding conditions suggests an important strategic component in age-related memory decline. The reduction in left PFC activity was often coupled with an increase in right PFC activity, leading to a bilateral pattern of PFC activity in older adults (HAROLD). Importantly, extending to encoding a finding originally reported for retrieval (Cabeza et al., 2002), two studies that divided older adults into high and low performers found the HAROLD pattern only in the

Episodic Memory Decline and Healthy Aging

Gaussian (mu) components of RT distributions. Young adults showed a correlation between mu and right PFC activity, whereas the older adults showed correlations in left and right PFC regions related to both mu and tau. Since tau is associated with task-specific decision processes, and mu with residual sensory coding and response processes, the authors concluded that attentional demands were greater for older adults, leading to the recruitment of additional regions. These findings suggest that the retrieval network is more widely distributed in older adults. Daselaar et al. (2003b) used event-related fMRI to study recognition of words in younger adults and older adults. Based on recognition performance in the scanner, older adults were divided into old-high and old-low groups. During recognition, compared with baseline, the old-low group showed much increased activity throughout the brain relative to the other groups. In addition, the old-low group and younger adults showed bilateral PFC activity, whereas the old-high group showed a more left lateralized pattern of frontal activity. In other words, the old-low group showed a nonselective increase in global brain activity, whereas the old-high group showed a selective recruitment of left PFC. The authors interpreted these findings in terms of strategic retrieval differences. Interestingly, when correctly recognized old words were compared with correctly rejected new words, these group differences disappeared. The difference in activity between these two trial types is generally considered to be a correlate of retrieval success. Hence, these findings suggest that age-related differences in episodic recognition primarily reflect strategic search deficits. Velanova et al. (2006) also investigated differences between young and older adults during word retrieval in an easy (15 study repetitions) and a difficult (one study repetition) recognition task. Although many correlates of retrieval were similar between the groups, including medial and lateral parietal responses to successful recognition, older adults showed increased recruitment of PFC regions relative to young adults during the difficult condition. This effect was not significant during the easy condition. Moreover, the timing of increased recruitment in older adults occurred at relatively late stages of the retrieval event. These findings suggest that older adults fail to engage appropriate PFCmediated executive processes at early stages of retrieval, and as a result, PFC involvement is extended at late stages to compensate.

589

Recall and context memory The aforementioned word-pair study by Cabeza et al. (1997) included not only a recognition test but also a cued-recall test. During recall, older adults showed weaker activity in the anterior cingulate and left temporal cortex. In addition, older adults showed weaker activations in right PFC than the younger adults. Conversely, older adults showed greater activity than younger adults in left PFC. The net result was that PFC activity during recall was right lateralized in younger adults but bilateral in older adults. The authors noted this change in hemispheric asymmetry and interpreted it as compensatory. This was the first study identifying the HAROLD pattern and the first one suggesting the compensatory interpretation of this finding. As noted, this study also compared age-related changes in activity during recall and recognition. These changes were more pronounced during recall than during recognition, consistent with behavioral evidence that recall is more sensitive to aging. Ba¨ckman et al. (1997) found a result similar to Cabeza et al. (1997), using word-stem cued recall instead of word-pair cued recall: younger adults activated right PFC, whereas older adults activated both left and right PFC (HAROLD). Also using word pairs, Anderson et al. (2000) investigated the effects of divided attention on cued recall. They reported negligible effects of divided attention in both groups. However, under full attention conditions, older adults showed weaker activations primarily in right PFC but stronger activations primarily in left PFC, suggesting an attenuation of the right-lateralized pattern shown by younger adults (HAROLD). Cabeza et al. (2000) investigated item and temporalorder memory tasks. In the item task, a word pair was presented consisting of one studied word and one new word, and participants indicated which word was studied. In the temporal-order task, both words were studied, and participants indicated which of the two words appeared later in the study list. They reported that younger adults showed increased activation in right PFC for temporal-order compared with item memory, whereas older adults did not. In contrast, the activations during item memory were relatively unaffected by age. These findings are in line with the resource deficit hypothesis, indicating that memory deficits in older adults are a result of PFC dysfunction and that context memory is more heavily dependent on the frontal lobes than item memory is. In another study of context memory by Cabeza and colleagues (2002), younger adults, high-performing 3.28.3.1.2.(ii)

590 Episodic Memory Decline and Healthy Aging

(a) Young

(b) Old-low

(c) Old-high

Figure 6 PFC activity during EM retrieval was right-lateralized in young (a) and low-old (b) participants, but bilateral in oldhigh (c) subjects. From Cabeza R, Anderson ND, Locantore JK, and McIntosh AR (2002) Aging gracefully: Compensatory brain activity in high-performing older adults. Neuroimage 17: 1394–1402.

older adults (old-high), and low-performing older adults (old-low) studied words presented auditorily or visually. During scanning, they were presented with words visually and made either old/new decisions (item memory) or heard/seen decisions (context memory). Consistent with their previous results, younger adults showed right PFC activity for context trials, whereas older adults showed bilateral PFC activity (HAROLD). Importantly, however, this pattern was only seen for the old-high adults, supporting a compensation account of the HAROLD pattern (Figure 6). Summary Summarizing the studies on PFC and retrieval, the HAROLD pattern has been found more frequently in studies using tasks with little environmental support, including recall and context memory tasks than during simple item recognition. This was exemplified in the study by Cabeza et al. (1997), which included both recall and recognition tasks. These findings suggest a three-way interaction between age, executive demand, and frontal laterality. Importantly, distinguishing between oldhigh and old-low adults, the study by Cabeza et al. (2002) provided direct evidence for the compensation account of HAROLD.

3.28.3.1.2.(iii)

3.28.3.2

MTL

Frontal activations in aging showed both reductions and increases across aging, as well as shifts in lateralization of activation. On the other hand, activation within the MTL generally shows age-related decreases compared with that seen in younger adults. However, some studies show a shift in the foci of activation from the hippocampus proper to more parahippocampal regions in aging. Evidence for such a shift is presented later, where appropriate, and is discussed further in the conclusions.

3.28.3.2.1

Encoding Though the binding deficit hypothesis would predict age-related reductions in MTL activity, particularly during associative encoding, reductions have been found during encoding of individual items as well. We first discuss studies focusing on individual item encoding and then discuss the results of the only study reporting MTL differences using an associative encoding task. Individual item encoding In their study examining face encoding, Grady et al. (1995) found that older adults showed less activity not only in the left PFC but also in MTL compared with younger adults. Furthermore, they found a highly significant correlation between hippocampus and left PFC activity in younger adults, but not in older adults. Based on these results, they concluded that encoding in older adults is accompanied by reduced neural activity and diminished connectivity between PFC and MTL areas. Daselaar et al. (2003a) investigated levels of processing in aging using a deep (living/nonliving) vs. shallow (uppercase/lowercase) encoding task. Despite seeing common activation of regions involved in a semantic network across both age groups, activation differences were seen when comparing levels of processing. Older adults revealed significantly less activation in left anterior hippocampus during deep relative to shallow classification. The researchers concluded that under-recruitment of MTL regions contributes, at least in part, to age-related impairments in encoding. Similarly, the same group showed decreased activity in the MTL for poor-performing older adults compared with young and high-performing elderly. Despite similar PFC activation, Daselaar et al. (2003b) found that the older adults showed decreased activity in the left hippocampus during successful 3.28.3.2.1.(i)

Episodic Memory Decline and Healthy Aging

encoding of words. Based on these findings, they concluded that MTL dysfunction during encoding is an important factor in age-related memory decline. Similar to Daselaar et al. (2003b), Gutchess et al.’s (2005) study of subsequent picture memory observed reduced activity in the MTL for subsequently remembered items, even when older adults were not divided into high- and low-memory groups. Additionally, older adults exhibited a significant negative correlation between inferior frontal and parahippocampal activity, whereas younger adults did not. These results suggest that those older adults exhibiting the least involvement of the parahippocampal region conversely activated PFC areas the most. Similar results were reported in the study by Dennis et al. (2006), which distinguished between transient and sustained subsequent memory for words. In line with Gutchess et al., they found that older adults showed a reduced transient subsequent memory effect in the hippocampus coupled with an increased effect in left PFC. These data suggest that PFC regions could be activated in a compensatory manner to offset declines in MTL activations in older adults. Associative encoding Mitchell et al. (2000a) conducted the only fMRI study reporting MTL differences in an associative encoding task. In each trial, participants were presented with an object in a particular screen location and had to hold in working memory the object, its location, or both (combination trials). Combination trials can be assumed to involve the binding of different information into an integrated memory trace. Older adults showed a deficit in accuracy in the combination condition but not in the object or location conditions. Two regions were differentially involved in the combination condition in younger adults but not in older adults: a left anterior hippocampal region and an anteromedial PFC region (right BA 10). According to the authors, disruption of a hippocampal–PFC circuit may underlie binding deficits in older adults.

3.28.3.2.1.(ii)

Summary In line with the binding deficit hypothesis, Mitchell and collegues found age-related reductions in MTL activity specifically during associative encoding. Yet several other studies also found age-related MTL reductions during encoding of individual items (Daselaar et al., 2003a,b; Gutchess et al., 2005; Dennis et al., 2006). Moreover, the study by Daselaar et al. (2003b) directly linked reduced MTL activity during single-word encoding to impaired performance on a subsequent recognition

3.28.3.2.1.(iii)

591

test. These findings suggest that age-related binding deficits play a role not only in complex associative memory tasks but also in simpler item memory tasks. One explanation for these results is that, in general, item memory tasks also have an associative component. In fact, the deep processing tasks used in these tasks are specifically designed to invoke semantic associations in relation to the study items. As discussed in the next section, recollection of these associations can be used as confirmatory evidence during EM retrieval. The results from two studies suggest that PFC regions might be activated in a compensatory manner to offset declines in MTL activations in older adults (Gutchess et al., 2005; Dennis et al., 2006). In particular, using correlational analyses, Gutchess et al. (2005) found that the older adults who showed the least MTL activity also showed the greatest PFC activity. These findings suggest that older adults may try to compensate for deficits in binding mediated by MTL by recruiting additional executive processes mediated by PFC. 3.28.3.2.2

Retrieval In contrast to encoding, MTL differences during retrieval involve not only decreases but also agerelated increases. This pattern has been observed during recognition memory tasks as well as cued recall and autobiographical retrieval tasks. Recognition memory Cabeza et al. (2004) investigated the effects of aging on several cognitive tasks including a verbal recognition task. Within the medial temporal lobes, they found a dissociation between a hippocampal region, which showed weaker activity in older adults than in younger adults, and a parahippocampal region, which showed the converse pattern. Given evidence that hippocampal and parahippocampal regions are, respectively, more involved in recollection vs. familiarity (Aggleton and Brown, 1999; Yonelinas, 2002), this finding is consistent with the notion that older adults are more impaired in recollection than in familiarity (e.g., Parkin and Walter, 1992; Jennings and Jacoby, 1993). Indeed, the agerelated increase in parahippocampal cortex suggests that older adults may be compensating for recollection deficits by relying more on familiarity. Supporting this idea, older adults had a larger number of K responses than younger adults, and these responses were positively correlated with the parahippocampal activation. A recent follow-up study by the same group showed a similar pattern of results (Prince et al., 2005). Young and older adults made old/new

3.28.3.2.2.(i)

592 Episodic Memory Decline and Healthy Aging

judgments about previously studied words, followed by a confidence judgment from low to high. On the basis of previous research (Yonelinas, 2001), recollection was measured as an exponential change in brain activity as a function of confidence and familiarity, and as a linear change. The results revealed a clear double dissociation within MTL: Whereas recollectionrelated activity in the hippocampus was reduced by aging, familiarity-related activity in rhinal cortex was increased by aging (see Figure 7(a)). These results suggested that older adults compensated for deficits in recollection processes mediated by the hippocampus by relying more on familiarity processes mediated by rhinal cortex. Supporting this interpretation, withinparticipants regression analyses based on single-trial activity showed that recognition accuracy was determined by only hippocampal activity in young adults but by both hippocampal and rhinal activity in older adults. Also consistent with the notion of compensation, functional connectivity analyses showed that correlations between the hippocampus and posterior regions associated with recollection were greater in younger adults, whereas correlations between rhinal cortex and bilateral PFC regions were greater in older adults (see Figure 7(b)). The latter effect suggests a top-down modulation of PFC on rhinal activity in older adults. The finding of preserved rhinal function in healthy older adults has important clinical implications, because this region is impaired early in AD (Killiany et al., 2000; Pennanen et al., 2004). 3.28.3.2.2.(ii) Cued recall, autobiographical retrieval, and context memory In addition to

increases in left PFC during word-stem cued recall, Ba¨ckman and colleagues (1997) found increased MTL activation in older adults. Just as the increased left PFC activation resulted in more bilateral frontal activation for older adults, increased activation in left MTL had the same effect: Compared with younger adults, older adults show more bilateral MTL activity. It should be noted that this bilateral activation was not accompanied by increased performance, as older adults recalled only about half as many words as did younger adults. This HAROLD pattern within MTL was also observed in a recent study using event-related fMRI. Maguire and Frith (2003) investigated the recall of autobiographical events gathered in a prescan interview. Although the groups activated largely the same regions, they observed a striking difference in the MTL. Younger adults showed left lateralized hippocampal activity, whereas older adults showed

bilateral activity in the MTL. These findings suggest that HAROLD extends beyond the PFC not only to other cortical regions (as shown by several studies) but also to subcortical areas. Summary In sum, retrieval studies have found both increases and decreases in MTL activity. The findings by Daselaar and colleagues suggest that at least some of these increases reflect a shift from recollection-based (hippocampus) to familiarity-based (rhinal cortex) retrieval. Furthermore, their functional connectivity findings suggest that the greater reliance on familiarity processes in older adults may be mediated by a top-down frontal modulation.

3.28.3.2.2.(iii)

3.28.3.2.3 Linking cognitive theories to age-related changes in PFC and MTL

In the first part of this chapter, we discussed behavioral evidence indicating how EM encoding and retrieval performance changes as a function of aging. We also discussed two important cognitive hypotheses that have been put forward to account for age-related deficits in EM – the resource deficit hypothesis and the binding deficit hypothesis. In the following sections, we discussed how aging affects the anatomy and function of two brain regions thought to be critical to EM function, PFC and MTL. Here, we connect these behavioral and neurobiological findings by linking the resource and binding deficit hypotheses to PFC and MTL function in older adults. Finally, returning to the beginning of the chapter, we discuss the relevance of these findings in terms of the clinical distinction between healthy and pathological deficits in EM. 3.28.3.3 Resource Deficit Hypothesis and PFC Function As described in the behavioral section of this chapter, the resource deficit hypothesis postulates that aging reduces attentional resources, and as a result, older adults have greater difficulties with cognitive tasks, including EM tasks, that require greater self-initiated processing. This hypothesis predicts that age-related differences should be smaller when the task provides a supportive environment that reduces attentional demands. Among other findings, the resource deficit hypothesis is supported by evidence that when attentional resources are reduced in younger adults, they tend to show EM deficits that resemble those of older adults ( Jennings and Jacoby, 1993; Anderson et al., 1998).

Episodic Memory Decline and Healthy Aging

Familiarity: O > Y

Recollection: Y > O 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

0.7 HC

0.6 Rhinal slope

Hipp. exp. rate (λ)

(a)

593

Rhinal Ctx

0.5 0.4 0.3 0.2 0.1

Young

0

Older

Young

Older

(b)

L PFC

R PFC 0.3 0.25 0.2 0.15 0.1 0.05 0 –0.05 –0.1 –0.15 –0.2

0.25 0.2 0.15 0.1 0.05 0 –0.05 –0.1 –0.15 –0.2

L Rhinal Ctx 0.3 0.25 0.2 0.15 0.1 0.05 0 –0.05 –0.1 –0.15

L Hipp

L Pariet

0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 0.02 0.04 0.06

Retrosplenial

Figure 7 (a) The effects of aging yielded a double dissociations between two MTL subregions: Whereas recollection-related activity (exponential increase) in the hippocampus was attenuated by aging, familiarity-related activity (linear decrease) in the rhinal cortex was enhanced by aging. The hippocampal exponential rate parameter () provides a measure of the sharpness of the exponential increase of the perceived oldness function in the hippocampus. The rhinal slope parameter provides a measure of the steepness of the perceived oldness function in the rhinal cortex. (b) Younger adults showed greater connectivity between the hippocampus and posterior regions associated with recollection, whereas older adults showed greater correlations between rhinal cortex and bilateral PFC regions. The latter effect suggests a top-down modulation of PFC on rhinal activity in older adults. Adapted from Daselaar SM, Fleck MS, Dobbins IG, Madden DJ, and Cabeza R (2006b) Effects of healthy aging on hippocampal and rhinal memory functions: An event-related fMRI study. Cereb. Cortex 16: 1771–1782.

594 Episodic Memory Decline and Healthy Aging

Regarding neural correlates, Craik (1983) proposed that older adults’ deficits in processing are related to a reduction in the efficiency of PFC functioning. As discussed in this chapter, there is much evidence in support of this idea. First, anatomical and physiological studies indicate that PFC is one of the brain regions most affected by the aging process. This region shows the most prominent gray matter atrophy and the most widespread degeneration of white matter integrity, as well as a substantial loss of DA modulation. Moreover, functional neuroimaging studies have found agerelated changes in PFC activity that are generally in line with the resource deficit hypothesis. Given the critical role of PFC in managing attentional resources, the resource deficit hypothesis predicts that age-related changes in PFC activity will be larger for tasks involving greater self-initiated processing and/or less environmental support. The results of functional neuroimaging studies are generally consistent with this prediction. During EM encoding, agerelated decreases in left PFC activation were found frequently during intentional encoding conditions (which provide less environmental support) but rarely during incidental encoding conditions (which provide greater environmental support). Similarly, during EM retrieval, age-related differences in PFC activity were usually larger for recall and context memory tasks (which require greater cognitive resources) than for recognition memory tasks (which require fewer cognitive resources). Thus, in general, age effects on PFC activity tend to increase as a function of the demands placed on cognitive resources. However, not all age-related changes in PFC activity suggested decline; on the contrary, many studies found age-related increases in PFC that suggested compensatory mechanisms in the aging brain. In particular, several encoding and retrieval studies found activations in contralateral PFC regions in older adults that were not seen in young adults. Importantly, experimental comparisons between high- and lowperforming older adults (Cabeza et al., 2002; Rosen et al., 2002) demonstrated the beneficial contribution of contralateral PFC recruitment to memory performance in older adults. Moreover, a recent study using transcranial magnetic stimulation (TMS) found that in younger adults, episodic retrieval performance was impaired by TMS of right PFC but not of left PFC, whereas in older adults it was impaired by either right or left PFC stimulation (Rossi et al., 2004). This result indicates that the left PFC was less critical for younger adults and was used more by older adults, consistent with the compensation hypothesis.

It is important to note that resource-deficit and compensatory interpretations are not incompatible. In fact, it is reasonable to assume that the recruitment of additional brain regions (e.g., in the contralateral PFC hemisphere) reflects an attempt to compensate for reduced cognitive resources. One way in which older adults could counteract deficits in the particular pool of cognitive resources required by a cognitive task is to tap into other pools of cognitive resources. If one task is particularly dependent on cognitive processes mediated by one hemisphere, the other hemisphere represents an alternative pool of cognitive resources. Thus, in the case of PFC-mediated cognitive resources, if older adults have deficits in PFC activity in one hemisphere, they may compensate for these deficits by recruiting contralateral PFC regions. Moreover, age-related decreases suggestive of resource-deficits and age-related increases suggestive of compensation have often been found in the same conditions. For example, intentional encoding studies have shown age-related decreases in left PFC activity coupled with age-related increases in right PFC, leading to a dramatic reduction in hemispheric asymmetry in older adults (ie., HAROLD).

3.28.3.4 Binding Deficit Hypothesis and MTL Function As noted in the first part of the chapter, the binding deficit hypothesis postulates that age-related memory deficits are primarily the result of difficulties in encoding and retrieving novel associations between items. This hypothesis predicts that older adults are particularly impaired in EM tasks that involve relations between individual items or between items and their context. Given that relational memory has been strongly associated with the hippocampus (Eichenbaum et al., 1994), this hypothesis also predicts that older adults will show decreased hippocampal activity during memory tasks, particularly when they involve relational information. As noted in the preceding parts of the chapter, anatomical, physiological, and functional neuroimaging studies have identified considerable age-related changes not only in PFC but also in MTL regions. For instance, the MTL also shows substantial atrophy in aging. Yet the rate of decline differs for different subregions. Whereas the hippocampus shows a marked decline, the rhinal cortex is relatively preserved in healthy aging (Figure 3). This finding is in line with the idea that age-related memory deficits are

Episodic Memory Decline and Healthy Aging

particularly pronounced during relational memory tasks, which depend on the hippocampus. In line with anatomical findings, functional neuroimaging studies have found substantial age-related changes in MTL activity during both encoding and retrieval. Several studies have found age-related decreases in both hippocampal and parahippocampal regions. During encoding, however, declines in hippocampal activation are also seen for encoding of individual features in healthy older adults (Grady et al., 1995; Schiavetto et al., 2002; Daselaar et al., 2003a,b; Gutchess et al., 2005). Finally, during retrieval, some studies found decreases in hippocampal activity (Cabeza et al., 2004, 2005), but also greater activity in older than younger adults in parahippocampal (Cabeza et al., 2004), rhinal (Cabeza et al., 2005), or contralateral MTL (Ba¨ckman et al., 1997; Maguire and Frith, 2003) regions, which may be compensatory. In general, age-related changes in MTL activity are consistent with the binding deficit hypothesis. Consistent with this hypothesis, age-related reductions in hippocampal activity were found during the encoding of complex scenes, which involve associations among picture elements (Gutchess et al., 2005), and during deep encoding of words, which involves identification of semantic associations (Daselaar et al., 2003a,b; Dennis et al., 2006). Finally, a recent study specifically associated age-related reductions in hippocampal activity with recollection, which involves recovery of item-context associations (Daselaar et al., 2006b). Yet it should be noted that age-related changes in MTL activity were often accompanied by concomitant changes in PFC activity. Hence, in these cases, it is unclear whether such changes signal MTL dysfunction or whether they are the result of a decline in executive processes mediated by PFC regions. However, studies using incidental encoding tasks with minimal self-initiated processing requirements have also identified age-related differences in MTL activity without significant changes in PFC activity (Daselaar et al., 2003a,b). As in the case of PFC, not all age-related changes in MTL activity suggest decline; several findings suggest compensation. First, similar to the bilateral pattern frequently observed in PFC, older adults have also demonstrated bilateral hippocampal recruitment while performing memory retrieval tasks (Maguire and Frith, 2003). Second, during retrieval, older adults have been found to show reduced activity in the hippocampus but increased activity in other brain regions such as the parahippocampal gyrus (Cabeza et al., 2004) and the rhinal cortex (Daselaar et al., 2006b). These results were

595

interpreted as a recruitment of familiarity processes mediated by parahippocampal regions in order to compensate for the decline of recollection processes that are dependent on the hippocampus proper. These results fit well with the relational memory view (Cohen and Eichenbaum, 1993; Eichenbaum et al., 1994), which states that the hippocampus is involved in binding an item with its context (recollection), whereas the surrounding parahippocampal cortex mediates itemspecific memory processes (familiarity). As noted in the behavioral section of the chapter, this distinction is supported not only by functional neuroimaging data in healthy young adults (Daselaar et al., 2006a) but also by lesion data from both humans (Yonelinas et al., 2002) and animals (Fortin et al., 2004). 3.28.3.5

Healthy versus Pathological Aging

As mentioned at the beginning of this chapter, one of the biggest challenges in cognitive aging research is to isolate the effects of healthy aging from those of pathological aging. A general review of the structural neuroimaging literature suggests that healthy aging is accompanied by greater declines in frontal regions compared with MTL (Raz et al., 2005). In contrast, pathological aging is characterized by greater decline in MTL than in frontal regions (Braak et al., 1993; Kemper, 1994). In fact, functional neuroimaging evidence suggests that prefrontal activity tends to be maintained or even increased in early AD (Grady, 2005). Thus, these findings suggest that memory decline in healthy aging is more dependent on frontal than MTL deficits, whereas the opposite pattern is more characteristic of pathological aging (see West, 1996; Buckner, 2004, for reviews). In view of these findings, clinical studies aimed at an early diagnosis of age-related pathology have mainly targeted changes in MTL (Nestor et al., 2004). Yet the studies reviewed in this chapter clearly indicate that healthy older adults are also prone to MTL decline. Hence, rather than focusing on MTL deficits alone, diagnosis of age-related pathology may be improved by employing some type of composite score reflecting the ratio between MTL and frontal decline. In terms of MTL dysfunction in healthy and pathological aging, it is also critical to assess the specific type or loci of MTL dysfunction. Critically, a decline in hippocampal function can be seen in both healthy aging and AD. Thus, even though hippocampal volume decline is an excellent marker of concurrent AD (Scheltens et al., 2002), it is not a reliable measure for distinguishing normal aging from early stages of

596 Episodic Memory Decline and Healthy Aging

the disease (Raz et al., 2005). In contrast, changes in the entorhinal cortex are not apparent in healthy aging (Figure 3), but they are present in early AD patients with only mild impairments (Dickerson et al., 2004). In a discriminant analysis, Pennanen and colleagues (2004) showed that, although hippocampal volume is indeed the best marker to discriminate AD patients from normal controls, the volume of the entorhinal cortex is much better in distinguishing between incipient AD (mild cognitive impairment, MCI) and healthy aging. Finally, it should be noted that, despite the rigorous screening procedures typical of functional neuroimaging studies of healthy aging, it remains possible that early symptoms of age-related pathology went undetected in some of the studies reviewed in this chapter. 3.28.3.6

Summary

In this chapter, we reviewed behavioral, anatomical, and functional neuroimaging evidence highlighting the role of PFC and MTL regions in age-related decline in EM function. The chapter focused on two major factors thought to underlie age-related memory decline and strongly linked to PFC and MTL function, namely, deficits in executive function and deficits in binding processes. In line with a reduction in PFC and MTL function, we discussed behavioral studies indicating that age deficits are most pronounced on tasks that put a great demand on executive and binding operations, respectively. We also discussed anatomical studies indicating a general decline with age in the anatomy and physiology of PFC and MTL. Linking these behavioral and anatomical findings, we discussed functional neuroimaging studies that generally showed age-related decreases in PFC and MTL activity during both EM encoding and retrieval. Yet some of these studies also found preserved or increased levels of PFC or MTL activity in older adults, which may be compensatory. Regarding PFC, several EM studies have found an age-related increase in contralateral PFC activity leading to an overall reduction in frontal asymmetry in older adults (HAROLD). As discussed, studies that divided older adults into high and low performers provided strong support for the idea that HAROLD reflects a successful compensatory mechanism. Regarding MTL, several EM studies reported age-related decreases in MTL activity, particularly during EM encoding. Yet studies of EM retrieval have also found age-related increases in MTL activity. Recent findings suggest that at least some of these increases reflect a compensatory shift from

hippocampal-based recollection processes to parahippocampal-based familiarity processes. In sum, in view of the substantial changes in PFC and MTL that take place when we grow older, a reduction in EM function seems inevitable. Yet our review also suggests that some older adults may cope well with this reduction by shifting to alternative brain resources within PFC and MTL that can compensate for the general deficits in executive and binding operations underlying agerelated EM decline.

References Aggleton JP and Brown MW (1999) Episodic memory, amnesia, and the hippocampal-anterior thalamic axis. Behav. Brain Sci. 22: 425–444. Anderson ND, Craik FIM, and Naveh-Benjamin M (1998) The attentional demands of encoding and retrieval in younger and older adults: I. Evidence from divided attention costs. Psychol. Aging 13: 405–423. Anderson ND, Iidaka T, McIntosh AR, Kapur S, Cabeza R, and Craik FIM (2000) The effects of divided attention on encoding- and retrieval-related brain activity: A PET study of younger and older adults. J. Cogn. Neurosci. 12: 775–792. Ba¨ckman L and Farde L (2005) The role of dopamine systems in cognitive aging. In: Cabeza R, Nyberg L, and Park DC (eds.) Cognitive Neuroscience of Aging, pp. 58–84. New York: Oxford University Press. Ba¨ckman L, Almkvist O, Andersson J, et al. (1997) Brain activation in young and older adults during implicit and explicit retrieval. J. Cogn. Neurosci. 9: 378–391. Ba¨ckman L, Ginovart N, Dixon RA, et al. (2000) Age-related cognitive deficits mediated by changes in the striatal dopamine system. Am. J. Psychiatry 157: 635–637. Bannon MJ and Whitty CJ (1997) Age-related and regional differences in dopamine transporter mRNA expression in human midbrain. Neurology 48: 969–977. Bastin C and Van der Linden M (2003) The contribution of recollection and familiarity to recognition memory: A study of the effects of test format and aging. Neuropsychology 17: 14–24. Braak H, Braak E, and Bohl J (1993) Staging of Alzheimerrelated cortical destruction. Eur. Neurol. 33: 403–408. Brown MW and Aggleton JP (2001) Recognition memory: What are the roles of the perirhinal cortex and hippocampus? Nat. Rev. Neurosci. 2: 51–61. Buckner RL (2004) Memory and executive function in aging and AD: Multiple factors that cause decline and reserve factors that compensate. Neuron 44: 195–208. Bugiani O, Salvarani S, Perdelli F, Mancardi GL, and Leonardi A (1978) Nerve cell loss with aging in the putamen. Eur. Neurol. 17: 286–291. Burke DM and Light LL (1981) Memory and aging: The role of retrieval processes. Psychol. Bull. 90: 513–514. Cabeza R (2002) Hemispheric asymmetry reduction in older adults: the HAROLD model. Psychol. Aging 17: 85–100. Cabeza R, Grady CL, Nyberg L, et al. (1997) Age-related differences in neural activity during memory encoding and retrieval: A positron emission tomography study. J. Neurosci. 17: 391–400. Cabeza R, Anderson ND, Houle S, Mangels JA, and Nyberg L (2000) Age-related differences in neural activity during item

Episodic Memory Decline and Healthy Aging and temporal-order memory retrieval: A positron emission tomography study. J. Cogn. Neurosci. 12: 1–10. Cabeza R, Anderson ND, Locantore JK, and McIntosh AR (2002) Aging gracefully: Compensatory brain activity in high-performing older adults. Neuroimage 17: 1394–1402. Cabeza R, Daselaar SM, Dolcos F, Prince SE, Budde M, and Nyberg L (2004) Task-independent and task-specific age effects on brain activity during working memory, visual attention and episodic retrieval. Cereb. Cortex 14: 364–375. Cabeza R, Fleck MS, Madden DJ, Dobbins IG, and Daselaar SM (2005) Effects of aging on the neural correlates of recollection vs. familiarity: An event-related fMRI study. J. Cogn. Neurosci. Suppl.: S233. Chalfonte BL and Johnson MK (1996) Feature memory and binding in young and older adults. Mem. Cogn. 24: 403–416. Charlton RA, Barrick TR, McIntyre DJ, et al. (2006) White matter damage on diffusion tensor imaging correlates with agerelated cognitive decline. Neurology 66: 217–222. Cohen NJ and Eichenbaum H (1993) Memory, Amnesia, and the Hippocampal System. Cambridge, MA: 1993. Craik FIM (1983) On the transfer of information from temporary to permanent memory. Philos. Trans. R. Soc. Lond. B 302: 341–359. Craik FIM (1986) A functional account of age differences in memory. In: Klix F and Hagendorf H (eds.) Human Memory and Cognitive Capabilities, pp. 409–422. New York: Elsevier. Craik FIM and McDowd JM (1987) Age differences in recall and recognition. J. Exp. Psychol. Learn. Mem. Cogn. 13: 474–479. Craik FIM and Simon D (1980) Age differences in memory: The roles of attention and depth of processing. In: Poon LW, Fozard JL, Cermak LS, and Arenberg D (eds.) New Directions in Memory and Aging: Proceedings of the George Talland Memorial Conference, pp. 95–112. Hillsdale, NJ: Lawrence Erlbaum. Craik FI, Byrd M, and Swanson JM (1987) Patterns of memory loss in three elderly samples. Psychol Aging 2: 79–86. Daselaar SM, Veltman DJ, Rombouts SA, Raaijmakers JG, and Jonker C (2003a) Deep processing activates the medial temporal lobe in young but not in old adults. Neurobiol. Aging 24: 1005–1011. Daselaar SM, Veltman DJ, Rombouts SA, Raaijmakers JG, and Jonker C (2003b) Neuroanatomical correlates of episodic encoding and retrieval in young and elderly subjects. Brain 126: 43–56. Daselaar SM, Fleck MS, and Cabeza R (2006a) Triple dissociation in the medial temporal lobes: Recollection, familiarity, and novelty. J. Neurophysiol. 96: 1902–1911. Daselaar SM, Fleck MS, Dobbins IG, Madden DJ, and Cabeza R (2006b) Effects of healthy aging on hippocampal and rhinal memory functions: An event-related fMRI study. Cereb. Cortex 16: 1771–1782. Davidson PS and Glisky EL (2002) Neuropsychological correlates of recollection and familiarity in normal aging. Cogn. Affect. Behav. Neurosci. 2: 174–86. Dennis NA, Daselaar S, and Cabeza R (2006) Effects of aging on transient and sustained successful memory encoding activity. Neurobiol. Aging 2006 Aug 16. Dias R, Robbins TW, and Roberts AC (1996) Primate analogue of the Wisconsin Card Sorting test: Effects of excitotoxic lesions of the prefrontal cortex in the marmoset. Behav. Neurosci. 110: 872–876. Dias R, Robbins TW, and Roberts AC (1997) Dissociable forms of inhibitory control within prefrontal cortex with an analogue of the Wisconsin Card Sorting test: Restriction to novel situations and independence from online Processing. J. Neurosci. 17: 9285–9287.

597

Dickerson BC, Salat DH, Bates JF, et al. (2004) Medial temporal lobe function and structure in mild cognitive impairment. Ann. Neurol. 56: 27–35. Eichenbaum H (2006) Remembering: Functional organization of the declarative memory system. Curr. Biol. 16: R643–5. Eichenbaum H, Otto T, and Cohen NJ (1994) Two component functions of the hippocampal memory system. Behav. Brain Sci. 17: 449–517. Erixon-Lindroth N, Farde L, Wahlin TB, Sovago J, Halldin C, and Ba¨ckman L (2005) The role of the striatal dopamine transporter in cognitive aging. Psychiatry Res. 138: 1–12. Fletcher PC and Henson RNA (2001) Frontal lobes and human memory – Insights from functional neuroimaging. Brain 124: 849–881. Fortin NJ, Wright SP, and Eichenbaum H (2004) Recollectionlike memory retrieval in rats is dependent on the hippocampus. Nature 431: 188–191. Gabrieli JD (1998) Cognitive neuroscience of human memory. Annu. Rev. Psychol. 49: 87–115. Gallo DA, Sullivan AL, Daffner KR, Schacter DL, and Budson AE (2004) Associative recognition in Alzheimer’s disease: Evidence for impaired recall-to-reject. Neuropsychology 18: 556–563. Ghilardi MF, Ghez C, Dhawan V, et al. (2000) Patterns of regional brain activation associated with different forms of motor learning. Brain Res. 871: 127–145. Golomb J, de Leon MJ, Kluger A, George AE, Tarshish C, and Ferris SH (1993) Hippocampal atrophy in normal aging. An association with recent memory impairment. Arch. Neurol. 50: 967–973. Golomb J, Kluger A, de Leon MJ, et al. (1994) Hippocampal formation size in normal human aging: A correlate of delayed secondary memory performance. Learn. Mem. 1: 45–54. Golomb J, Kluger A, de Leon MJ, et al. (1996) Hippocampal formation size predicts declining memory performance in normal aging. Neurology 47: 810–813. Grady CL (2005) Functional connectivity during memory tasks in healthy aging and dementia. In: Cabeza R, Nyberg L, and Park D (eds.) Cognitive Neuroscience of Aging, pp. 286–308. New York: Oxford University Press. Grady CL, McIntosh AR, Horwitz B, et al. (1995) Age-related reductions in human recognition memory due to impaired encoding. Science 269: 218–221. Grady CL, McIntosh AR, Raja MN, Beig S, and Craik FIM (1999) The effects of age on the neural correlates of episodic encoding. Cereb. Cortex 9: 805–814. Grady CL, Bernstein LJ, Beig S, and Siegenthaler AL (2002) The effects of encoding strategy on age-related changes in the functional neuroanatomy of face memory. Psychol. Aging 17: 7–23. Gunning-Dixon FM and Raz N (2003) Neuroanatomical correlates of selected executive functions in middle-aged and older adults: A prospective MRI study. Neuropsychologia 41: 1929–1941. Gutchess AH, Welsh RC, Hedden T, et al. (2005) Aging and the neural correlates of successful picture encoding: Frontal activations compensate for decreased medial-temporal activity. J. Cogn. Neurosci. 17: 84–96. Hedden T and Gabrieli JD (2004) Insights into the ageing mind: A view from cognitive neuroscience. Nat. Rev. Neurosci. 5: 87–96. Howard MW, Bessette-Symons B, Zhang Y, and Hoyer WJ (2006) Aging selectively impairs recollection in recognition memory for pictures: Evidence from modeling and receiver operating characteristic curves. Psychol. Aging 21: 96–106. Inoue M, Suhara T, Sudo Y, et al. (2001) Age-related reduction of extrastriatal dopamine D2 receptor measured by PET. Life Sci. 69: 1079–1084.

598 Episodic Memory Decline and Healthy Aging Java RI (1996) Effects of age on state of awareness following implicit and explicit word-association tasks. Psychol. Aging 11: 108–111. Jennings JM and Jacoby LL (1993) Automatic versus intentional uses of memory: Aging, attention, and control. Psychol. Aging 8: 283–293. Johnson MK, Hashtroudi S, and Lindsay DS (1993) Source monitoring. Psychol. Bull. 114: 3–28. Kaasinen V, Vilkman H, Hietala J, et al. (2000) Age-related dopamine D2/D3 receptor loss in extrastriatal regions of the human brain. Neurobiol. Aging 221: 683–688. Kausler DH and Puckett JM (1981) Adult age differences in memory for sex of voice. J. Gerontol. 36: 44–50. Kemper T (1994) Neuroanatomical and neuropathological changes during aging and in dementia. In: Albert ML and Knoepfel EJE (eds.) Clinical Neurology of Aging, 2nd edn., pp. 3–67. New York: Oxford University Press. Killiany RJ, Gomez-Isla T, Moss M, et al. (2000) Use of structural magnetic resonance imaging to predict who will get Alzheimer’s disease. Ann. Neurol. 47: 430–439. Lie CH, Specht K, Marshall JC, and Fink GR (2006) Using fMRI to decompose the neural processes underlying the Wisconsin Card Sorting Test. Neuroimage 30: 1038–1049. Light LL, LaVoie D, Valencia-Laver D, Owens SA, and Mead G (1992) Direct and indirect measures of memory for modality in young and older adults. J. Exp. Psychol. Learn. Mem. Cogn. 18: 1284–1297. Logan JM, Sanders AL, Snyder AZ, Morris JC, and Buckner RL (2002) Under-recruitment and nonselective recruitment: Dissociable neural mechanisms associated with aging. Neuron 33: 827–840. Lupien SJ, de Leon M, de Santi S, et al. (1998) Cortisol levels during human aging predict hippocampal atrophy and memory deficits. Nat. Neurosci. 1: 69–73. Madden DJ, Turkington TG, Provenzale JM, et al. (1999) Adult age differences in functional neuroanatomy of verbal recognition memory. Hum. Brain Mapp. 7: 115–135. Madden DJ, Whiting WL, Huettel SA, White LE, MacFall JR, and Provenzale JM (2004) Diffusion tensor imaging of adult age differences in cerebral white matter: Relation to response time. Neuroimage 21: 1174–1181. Maguire EA and Frith CD (2003) Aging affects the engagement of the hippocampus during autobiographical memory retrieval. Brain 126: 1511–1523. Mantyla T (1993) Knowing but not remembering: Adult age differences in recollective experience. Mem. Cogn. 21: 379–388. Maril A, Simons JS, Mitchell JP, Schwartz BL, and Schacter DL (2003) Feeling-of-knowing in episodic memory: An eventrelated fMRI study. Neuroimage 18: 827–836. McIntyre JS and Craik FIM (1987) Age differences in memory for item and source information. Can. J. Psychol. 41: 175–192. Miller EK and Cohen JD (2001) An integrative theory of prefrontal cortex function. Annu. Rev. Neurosci. 24: 167–202. Milner B (1963) Effects of different brain lesions on card sorting. Arch. Neurol. 9: 90. Milner B (1972) Disorders of learning and memory after temporal lobe lesions in man. Clin. Neurosurg. 19: 421–446. Mitchell KJ, Johnson MK, Raye CL, and D’Esposito M (2000a) fMRI evidence of age-related hippocampal dysfunction in feature binding in working memory. Brain Res. Cogn. Brain Res. 10: 197–206. Mitchell KJ, Johnson MK, Raye CL, Mather M, and D’Esposito M (2000b) Aging and reflective processes of working memory: Binding and test load deficits. Psychol. Aging 15: 527–541. Morcom AM, Good CD, Frackowiak RS, and Rugg MD (2003) Age effects on the neural correlates of successful memory encoding. Brain 126: 213–229.

Mozley LH, Gur RC, Mozley PD, and Gur RE (2001) Striatal dopamine transporters and cognitive functioning in healthy men and women. Am. J. Psychiatry 158: 1492–1499. Naveh-Benjamin M (2000) Adult age differences in memory performance: tests of an associative deficit hypothesis. J. Exp. Psychol. Learn. Mem. Cogn. 26: 1170–1187. Naveh-Benjamin M and Craik FI (1995) Memory for context and its use in item memory: Comparisons of younger and older persons. Psychol. Aging 10: 284–293. Naveh-Benjamin M, Guez J, and Shulman S (2004) Older adults’ associative deficit in episodic memory: Assessing the role of decline in attentional resources. Psychon. Bull. Rev. 11: 1067–1073. Nestor PJ, Scheltens P, and Hodges JR (2004) Advances in the early detection of Alzheimer’s disease. Nat. Med. 10(supplement): S34–S41. Park DC and Puglisi JT (1985) Older adults’ memory for the color of pictures and words. J. Gerontol. 40: 198–204. Park DC and Shaw RJ (1992) Effect of environmental support on implicit and explicit memory in younger and older adults. Psychol. Aging 7: 632–642. Parkin AJ and Walter BM (1992) Recollective experience, normal aging, and frontal dysfunction. Psychol. Aging 7: 290–298. Parkin AJ, Walter BM, and Hunkin M (1995) Relationships between normal aging, frontal-lobe function, and memory for temporal and spatial information. Neuropsychology 9: 304–312. Pennanen C, Kivipelto M, Tuomainen S, et al. (2004) Hippocampus and entorhinal cortex in mild cognitive impairment and early AD. Neurobiol. Aging 25: 303–310. Persson J, Nyberg L, Lind J, et al. (2005) Structure-function correlates of cognitive decline in aging. Cereb. Cortex 16: 907–915. Petersen RC, Jack CR Jr, Xu YC, et al. (2000) Memory and MRIbased hippocampal volumes in aging and AD. Neurology 54: 581–587. Pfefferbaum A, Sullivan EV, Rosenbloom MJ, Mathalon DH, and Lim KO (1998) A controlled study of cortical gray matter and ventricular changes in alcoholic men over a 5-year interval. Arch. Gen. Psychiatry 55: 905–912. Prince SE, Daselaar SM, and Cabeza R (2005) Neural correlates of relational memory: Successful encoding and retrieval of semantic and perceptual associations. J. Neurosci. 25: 1203–1210. Rabinowitz JC (1984) Aging and recognition failure. J. Gerontol. 39: 65–71. Rabinowitz JC and Craik FI (1986) Prior retrieval effects in young and old adults. J. Gerontol. 41: 368–375. Raz N (2004) The aging brain observed in vivo: Differential changes and their modifiers. In: Cabeza R, Nyberg L, and Park DC (eds.) Cognitive Neuroscience of Aging: Linking Cognitive and Cerebral Aging, pp. 19–57. New York: Oxford University Press. Raz N, Gunning FM, Head D, et al. (1997) Selective aging of the human cerebral cortex observed in vivo: Differential vulnerability of the prefrontal gray matter. Cereb. Cortex 7: 268–282. Raz N, Lindenberger U, Rodrigue KM, et al. (2005) Regional brain changes in aging healthy adults: general trends, individual differences and modifiers. Cereb. Cortex 15: 1676–1689. Reeves S, Bench C, and Howard R (2002) Ageing and the nigrostriatal dopaminergic system. Int. J. Geriatr. Psychiatry 17: 359–370. Resnick SM, Goldszal AF, Davatzikos C, et al. (2000) One-year age changes in MRI brain volumes in older adults. Cereb. Cortex 10: 464–472.

Episodic Memory Decline and Healthy Aging Resnick SM, Pham DL, Kraut MA, Zonderman AB, and Davatzikos C (2003) Longitudinal magnetic resonance imaging studies of older adults: A shrinking brain. J. Neurosci. 23: 3295–2301. Rinne JO, Sahlberg N, Ruottinen H, Nagren K, and Lehikoinen P (1998) Striatal uptake of the dopamine reuptake ligand [11C]beta-CFT is reduced in Alzheimer’s disease assessed by positron emission tomography. Neurology 50: 152–156. Rosen AC, Prull MW, O’Hara R, et al. (2002) Variable effects of aging on frontal lobe contributions to memory. Neuroreport 13: 2425–2428. Rossi S, Miniussi C, Pasqualetti P, Babiloni C, Rossini PM, and Cappa SF (2004) Age-related functional changes of prefrontal cortex in long-term memory: A repetitive transcranial magnetic stimulation study. J. Neurosci. 24: 7939–7944. Salat DH, Kaye JA, and Janowsky JS (1999) Prefrontal gray and white matter volumes in healthy aging and Alzheimer disease. Arch. Neurol. 56: 338–344. Salat DH, Tuch DS, Greve DN, et al. (2005) Age-related alterations in white matter microstructure measured by diffusion tensor imaging. Neurobiol. Aging 26: 1215–1227. Scahill RI, Frost C, Jenkins R, Whitwell JL, Rossor MN, and Fox NC (2003) A longitudinal study of brain volume changes in normal aging using serial registered magnetic resonance imaging. Arch. Neurol. 60: 989–994. Schacter DL, Osowiecki D, Kaszniak AW, Kihlstrom JF, and Valdiserri M (1994) Source memory: Extending the boundaries of age-related deficits. Psychol. Aging 9: 81–89. Scheltens P, Fox N, Barkhof F, and De Carli C (2002) Structural magnetic resonance imaging in the practical assessment of dementia: Beyond exclusion. Lancet Neurol. 1: 13–21. Schiavetto A, Kohler S, Grady CL, Winocur G, and Moscovitch M (2002) Neural correlates of memory for object identity and object location: Effects of aging. Neuropsychologia 40: 1428–1442. Schwartz M, Creasey H, Grady CL, et al. (1985) Computed tomographic analysis of brain morphometrics in 30 healthy men, aged 21 to 81 years. Ann. Neurol. 17: 146–157. Smith AD, Park DC, Earles JL, Shaw RJ, and Whiting WLT (1998) Age differences in context integration in memory. Psychol. Aging 13: 21–28. Spencer WD and Raz N (1995) Differential effects of aging on memory for content and context: A meta-analysis. Psychol. Aging 10: 527–539. Sperling RA, Bates JF, Chua EF, et al. (2003) fMRI studies of associative encoding in young and elderly controls and mild Alzheimer’s disease. J. Neurol. Neurosurg. Psychiatry 74: 44–50. Squire LR, Schmolck H, and Stark SM (2001) Impaired auditory recognition memory in amnesic patients with medial temporal lobe lesions. Learn. Mem. 8: 252–256.

599

Stebbins GT, Carrillo MC, Dorman J, et al. (2002) Aging effects on memory encoding in the frontal lobes. Psychol. Aging 17: 44–55. Strong R (1998) Neurochemical changes in the aging human brain: Implications for behavioral impairment and neurodegenerative disease. Geriatrics 53(supplement 3): S9–S12. Stuss DT and Alexander MP (2000) Executive functions and the frontal lobes: A conceptual view. Psychol. Res. 63: 289–298. Sullivan EV and Pfefferbaum A (2006) Diffusion tensor imaging and aging. Neurosci. Biobehav. Rev. 30: 749–761. Sullivan EV, Adalsteinsson E, Hedehus M, et al. (2001) Equivalent disruption of regional white matter microstructure in ageing healthy men and women. Neuroreport 12: 99–104. Sullivan EV, Adalsteinsson E, and Pfefferbaum A (2006) Selective age-related degradation of anterior callosal fiber bundles quantified in vivo with fiber tracking. Cereb. Cortex 16: 1030–1039. Terry RD (2000) Cell death or synaptic loss in Alzheimer disease. J. Neuropathol. Exp. Neurol. 59: 1118–1119. Troyer AK, Hafliger A, Cadieux MJ, and Craik FI (2006) Name and face learning in older adults: Effects of level of processing, self-generation, and intention to learn. J. Gerontol. B Psychol. Sci. Soc. Sci. 61: P67–P74. Tulving E (1983) Elements of Episodic Memory. Oxford: Oxford University Press. Velanova K, Lustig C, Jacoby LL, and Buckner RL (2006) Evidence for frontally mediated controlled processing differences in older adults. Cereb. Cortex 17: 1033–1046. Volkow ND, Wang GJ, Fowler JS, et al. (1998) Parallel loss of presynaptic and postsynaptic dopamine markers in normal aging. Ann. Neurol. 44: 143–147. Volkow ND, Logan J, Fowler JS, et al. (2000) Association between age-related decline in brain dopamine activity and impairment in frontal and cingulate metabolism. Am. J. Psychiatry 157: 75–80. Wang Y, Chan GL, Holden JE, et al. (1998) Age-dependent decline of dopamine D1 receptors in human brain: A PET study. Synapse 30: 56–61. West RL (1996) An application of prefrontal cortex function theory to cognitive aging. Psychol. Bull. 120: 272–292. Wozniak JR and Lim KO (2006) Advances in white matter imaging: A review of in vivo magnetic resonance methodologies and their applicability to the study of development and aging. Neurosci. Biobehav. Rev. 30: 762–774. Yonelinas AP (2001) Components of episodic memory: The contribution of recollection and familiarity. Philos. Trans. R. Soc. Lond. B Biol. Sci. 356: 1363–1374. Yonelinas AP (2002) The nature of recollection and familiarity: A review of 30 years of research. J. Mem. Lang. 46: 441–517. Yonelinas AP, Kroll NE, Quamme JR, et al. (2002) Effects of extensive temporal lobe damage or mild hypoxia on recollection and familiarity. Nat. Neurosci. 5: 1236–1241.