Abstract

Autobiographical memories enable us to mentally reconstruct and relive past events, which is essential for one's personal identity. Unfortunately, this complex memory system is susceptible to age-related deterioration, possibly changing the way episodic information is being processed in older adults. The aim of this study was to investigate whether age influences the neural activity associated with content (episodic versus semantic) and remoteness (recent versus remote) of memories. Using functional magnetic resonance imaging in healthy older and young adults, we found significant age-dependent differences in the neural networks underlying memory content but not remoteness. Our data suggest an age-associated functional reorganization in the neural networks underlying long-term declarative memory. Relative increase in activity of posterior brain regions could reflect changes in visuospatial processing during episodic memory retrieval in older adults.

Introduction

Autobiographical memory is a highly complex memory system. It allows us to mentally travel back in time, re-experiencing our past. This ability is most probably unique to humans, as it is associated with autonoetic awareness of having a self that exists in subjective time (Tulving, 2002). Autobiographical memory formation requires a semantic contribution, thus, general knowledge of one's past, such as generic events and facts (e.g., having worked for a certain company, names, or addresses). The episodic component contains specific personal details of the circumstances surrounding a known autobiographical fact (Tulving, 2002; Tulving, Schacter, McLachlan, & Moscovitch, 1988). In a meta-analysis of functional imaging studies investigating autobiographical memory, Svoboda, McKinnon, and Levine (2006) identified a predominantly left-lateralized ‘core’ network consisting of prefrontal, temporal, and retrosplenial/posterior cingulate cortices. These ‘core’ regions have been found in the majority of studies investigating autobiographical memories, and may reflect essential components of autobiographical remembering. Examples for such essential components are self-referential processes (Conway & Pleydell-Pearce, 2000), associated with medial prefrontal activity, or the mechanisms of encoding and retrieval supported by medial temporal networks. ‘Core’ regions are complemented by ‘secondary’ regions that have been described less consistently, and include orbitofrontal, anterior cingulate, and occipital brain regions (Svoboda et al., 2006). Different patterns of activation among ‘core’ and ‘secondary’ regions could reflect the influence of modulating variables on autobiographical memory processing, such as the degree of visual imagery, as well as differential remoteness, emotionality, or personal relevance of memories (Conway, Pleydell-Pearce, Whitecross, & Sharpe, 2003; Svoboda et al., 2006). Neuroimaging studies investigating these factors can contribute to a better understanding of the complex networks of brain activity that support autobiographical remembering.

Among these modulating variables, differential remoteness of memories has received much attention. Activations of retrosplenial, prefrontal, and hippocampal regions have been reported for the retrieval of recent versus remote autobiographical memories (Maguire & Frith, 2003b; Oddo et al., 2010; Piefke, Weiss, Zilles, Markowitsch, & Fink, 2003; Poettrich et al., 2009). Specifically, hippocampal involvement has been studied extensively and two different theories emerged, suggesting a time-limited (Squire, Stark, & Clark, 2004) or time-independent (Moscovitch et al., 2005) role of the hippocampus in memory retrieval. Extending the classical memory consolidation model, it has been hypothesized that if memory retrieval would be gradually less dependant on the hippocampus, medial prefrontal cortical regions may become important to make retrieval sustainable by neocortical representations alone (Takashima et al., 2006). However, both greater (Ryan et al., 2001; Takashima et al., 2006) and fewer (Poettrich et al., 2009; Woodard et al., 2007) neocortical activations have been reported associated with recent versus remote memories.

Piefke and Fink (2005) have suggested that the process of aging may change the pattern of brain regions involved in autobiographical memory processing, which could contribute to the variability of findings across studies. Behavioral research across the adult lifespan revealed age-related cognitive decline in cued and free recall, recollection of context, working memory, or processing speed. Other cognitive domains, such as implicit memory, memory span, and recognition, remain relatively stable (for review, see Hedden & Gabrielli, 2004). With respect to autobiographical memories, the richness of details of a remembered episode decreases with age (Levine, Svoboda, Hay, Winocur, & Moscovitch, 2002), whereas abstract/semantic memory components of the same event become more important (episodic-semantic dissociation; Piolino et al., 2006). Several models have been proposed to explain age-related changes in brain activation patterns associated with long-term memory retrieval. Older adults activated prefrontal areas bilaterally whereas young people showed more lateralized activity during cognitive performance (hemispheric asymmetry reduction in old adults, Cabeza, 2002). Additional neural recruitment in older adults has been shown consistently (Davis, Dennis, Daselaar, Fleck, & Cabeza, 2008; Duverne, Habibi, & Rugg, 2008; Maguire & Frith, 2003a; Morcom, Li, & Rugg, 2007), which could reflect dedifferentiation (Grady & Craik, 2000) or compensatory processes (Cabeza, 2002; Cabeza, Anderson, Locantore, & McIntosh, 2002). The pattern of age-associated additional neural recruitment is modulated by task type and subject performance (Duverne et al., 2008; Morcom et al., 2007). Older adults have shown a reduced ability to recall sensory-perceptive event details (Piolino et al., 2006). Recollection of episodic memories is conceived as being largely sensory-perceptual in nature, involving visuospatial processing and vivid visual imagery. These functions are associated with posterior brain regions, especially occipital lobe networks (Conway, 2001; Greenberg & Rubin, 2003). Occipital brain regions are part of the ‘secondary’ network associated with autobiographical memory processing described by Svoboda and colleagues (2006).

The effects of aging on the neural networks supporting autobiographical memory have been studied rarely. One study focused on healthy older adults to reveal the cerebral structures associated with retrieval of autobiographical memories from five time periods covering the entire lifespan (Viard et al., 2007). The authors showed a similar activation pattern of frontal, posterior cingulate, and hippocampal cortices for all time periods. In the current study, we investigated older and younger adults. Using functional magnetic resonance imaging (fMRI), our aim was to reveal the possible age-related differences in the neural networks underlying long-term declarative memory content (autobiographical memories prepared from a semi-structured interview versus semantic event memories) and memory remoteness. We investigated memories from two specific time periods: The last five years (recent memories) and age 5–15 years (remote memories). We hypothesized that older adults compared with young people would show an increased neural activity associated with long-term memory retrieval. Given an experimental design that mainly relied on scenic imagination, we expected an additional neural recruitment in posterior brain regions (i.e., ‘secondary’ regions), associated with autobiographical memory retrieval among older adults due to age-related decline in sensory-perceptive recall abilities. On the basis of previous findings (Oddo et al., 2010), we also expected greater prefrontal cortical activations for recent versus remote memory retrieval among young participants.

Materials and Methods

Subjects

All participants were recruited using public advertisements. Written informed consent was obtained, and the University's Ethic Committee for Medical Research approved the study. Fourteen older adults (mean age 60.5 years; age range 55–69; 7 women) and 15 young subjects (mean age 28.0 years; age range 25–32; 8 women) participated. All subjects were Caucasian, cognitively healthy, right-handed, and equally educated (older adults: 11.1 ± 1.0 years [mean ± SD]; young subjects: 11.5 ± 0.9 years). Participants had normal or corrected to normal vision. Diagnostic evaluation included medical history assessment, physical and neurological examination, standard laboratory testing, neuropsychological testing, and structural brain MRI. The degree of white and subcortical grey matter lesions were rated using the ARWMC scale (Wahlund et al., 2001). Only subjects free of white matter lesions or with focal lesions only (ARWMC < 2 points), and free of focal lesions in grey matter areas (ARWMC 0 points) entered the study. Other exclusion criteria were education <8 years, history of alcohol or substance abuse, head trauma, psychiatric or neurological disorder, and major systemic disease affecting brain function. These criteria were assessed based on information provided by the participants in a clinical interview with a physician during the diagnostic evaluation process; and using direct measurements to additionally examine possible pathological conditions, such as dyslipoproteinemia or elevated blood glucose levels. All subjects were free of psychotropic medication.

Neuropsychological Assessments and Post-Scanning Debriefing

All participants underwent extensive neuropsychological testing to reveal whether they would show comparable performance and score within the normal age-corrected range. We investigated verbal memory (California Verbal Learning Test, Delis, Kramer, Kaplan, & Ober, 1987), visual memory (Rey–Osterrieth–Figure copy and delayed recall, Spreen & Strauss, 1998), working memory (Digit span and Block span forward and backward, Wechsler Memory Scale—revised, Härting, 2000), speed of information processing (Trail Making Part A and B, Reitan & Wolfson, 1993), language and verbal fluency (Token Test, Aachen Aphasia Test, Huber, Poeck, & Willmes, 1984; Animal Test, CERAD, Spreen & Strauss, 1998).

Using a questionnaire, post-scanning debriefing was performed to rate characteristic features of the memories as experienced during fMRI scanning, such as scene-likeness, emotionality, or perceptions of touch and temperature. This allowed further examination of whether group differences in brain activity were influenced by differences in these behavioral measures. Subjects were also asked to re-read and classify all cue sentences displayed during the experiment.

Study Materials and Procedure

A semi-structured autobiographical interview was conducted with each participant 4–6 weeks prior to scanning. Subjects were asked to specifically recollect events that they could vividly recall from the time when they were 5–15 years old (remote episodic memory), and from the last 5 years (recent episodic memory). These memories should refer to a particular time and place, lasting no longer than 1 day. This information was then used to prepare subject-specific stimulus sentences to be presented during the fMRI experiment (e.g., My mother said: Son, the war started today.). During the interview, participants were asked to report a total of 16 events (8 remote and 8 recent events). Participants were asked to report four emotionally positively and four emotionally negatively toned memories per category, based on a basic happy/sad rating of each memory reported. Five cue sentences were generated for each event. As emotional valence is known to affect neural processes underlying autobiographical memory (Piefke et al., 2003), but was not a factor of interest in this study, we balanced for emotional valence by obtaining an equal number of happy and sad memories.

For each individual autobiographical memory event, a fact of general knowledge (semantic event memory) was selected, matched for the time period of the respective episodic event. The facts of general knowledge had been collected in healthy volunteers (n = 30, mean age 61.0 years, SD 6.2, 15 women; and n = 28, mean age 29.6 years, SD 5.4, 16 women) not participating in the fMRI study. We adapted a remember-know-paradigm (Eldridge, Knowlton, Furmanski, Bookheimer, & Engel, 2000). From a total of 354 public facts (194 for the time period 1934–1969, 104 for the period 1975–1995, and 56 for the period 2000–2005), 260 semantic facts, rated as very familiar (known) but without a particular autobiographical context (remember), were used as experimental stimuli. Semantic items were not shown to the study participants prior to fMRI scanning to keep them devoid of context-specific information. For each subject, semantic stimuli included 40 remote (age 5–15) and 40 recent (last 5 years) stimuli (e.g., The Olympic Summer Games 2004 took place in Athens). We again controlled for emotional tone by balancing the number of positive and negative semantic stimuli. A total of 40 sentences, similar in length/number of syllables, were generated for each of the four memory conditions (episodic-recent, episodic-remote, semantic-recent, semantic-remote) for every individual participant.

Task and Experimental Design

During the fMRI experiment, stimuli were presented on a screen located behind the head of the participant using an LCD projector. Participants viewed the stimuli via a mirror positioned at a 45° angle on top of the head coil. Task presentation and recording of the behavioral responses was performed using Presentation® software (Version 9.5, Neurobehavioral Systems, Inc., Albany, CA).

A factorial block design (Fig. 1) was applied to evoke the neural responses associated with retrieval of different content (episodic vs. semantic) and remoteness (recent vs. remote) of memories. Five individual stimuli for each memory condition were blocked together (stimulus onset-time, SOT = 5 s, interstimulus interval, ISI = 1.5 s). Each block contained an additional stimulus (same SOT and ISI), presented at variable positions within the blocks, in response to which the subjects were instructed to press a button on a keypad as fast as possible. This reaction time task was used to control for alertness (Piefke et al., 2003). An additional second at the end of each block was used as a safety margin to compensate for temporal delays that may occur due to computational processing (e.g., data being written on a hard drive). Experimental conditions were separated by baselines lasting 12 s, during which the participants were instructed to repeatedly read the instruction: Please recall the following event and situation as vividly as possible. It was pointed out to the subjects that it was important to read this same instruction every time it appeared on the screen to standardize the baseline condition. Per baseline three whole-brain volumes (repetition time [TR] = 4 s) were acquired. We acquired 10 whole-brain volumes per block of trials for each memory condition. A total of four experimental runs, each consisting of 9 baselines and 8 memory blocks were performed, leading to the acquisition of a total of 428 volumes per subject (107 volumes per run, duration of a run 428 s). The scanning paradigm resulted in two repeats per condition per experimental run. The order of the memory conditions was counterbalanced across runs and individuals. Individual stimulus sentences were not repeated within or across experimental conditions to avoid habituation effects. Accordingly, the four conditions were repeated, but all blocks were unique. Prior to scanning, subjects were familiarized with the experimental set-up and tasks.

Fig. 1.

fMRI paradigm. The figure visualizes one experimental run of the fMRI paradigm. Four of these experimental runs, each lasting 428 s, were performed for every subject. The order of the memory conditions (SN, SO, EN, EO) was counterbalanced across runs and individuals (lower row). Each block (40 s) of a given memory condition consisted of five memory stimuli (S, stimulus on time, SOT = 5 s, interstimulus interval, ISI = 1.5 s) and one stimulus, presented at variable positions within the blocks (C, same SOT and ISI) controlling for the subject's alertness (upper row). SN = semantic new (recent), SO = semantic old (remote), EN = episodic new (recent), EO = episodic old (remote), B = baseline.

Fig. 1.

fMRI paradigm. The figure visualizes one experimental run of the fMRI paradigm. Four of these experimental runs, each lasting 428 s, were performed for every subject. The order of the memory conditions (SN, SO, EN, EO) was counterbalanced across runs and individuals (lower row). Each block (40 s) of a given memory condition consisted of five memory stimuli (S, stimulus on time, SOT = 5 s, interstimulus interval, ISI = 1.5 s) and one stimulus, presented at variable positions within the blocks (C, same SOT and ISI) controlling for the subject's alertness (upper row). SN = semantic new (recent), SO = semantic old (remote), EN = episodic new (recent), EO = episodic old (remote), B = baseline.

Image Acquisition and Processing

Scanning was performed using a 1.5 T whole-body scanner (Siemens Sonata, Erlangen, Germany). We acquired magnetization prepared rapid acquisition gradient-echo (MPRAGE) [TE = 4.39 ms, TR = 2.2 s, matrix = 320 × 320, slice thickness 0.79 mm, orientation sagittal (104 slices) and echo planar imaging (EPI)-sequences (TE = 60 ms, TR = 4 s, FOV 230 mm, voxel size = 3.6 × 3.6 × 4 mm, matrix = 64 × 64, orientation axial, 30 slices)]. Image processing and statistical calculations were performed using MATLAB (The Mathworks, Inc., Natick, MA) and statistical parametric mapping software (SPM2, Wellcome Department of Imaging Neuroscience, London, UK). The first five EPI images were discarded to allow the MRI signal to reach steady state. Individual data were spatially realigned to correct for head movement. For normalization, we used a standard EPI template (MNI brain). Functional data were smoothed using an isotropic Gaussian kernel of 10 mm full width at half maximum (FWHM).

At the single subject level, we modeled all four conditions of the paradigm (episodic-recent, episodic-remote, semantic-recent, semantic-remote) in the context of a general linear model (GLM). We used a flexible factorial modeling procedure for second (group)-level analyses with memory type (episodic/semantic) and memory remoteness (recent/remote) as between-group factors. In our group-level analyses, we first compared the neural network supporting all memory types combined (SN + SO + EN + EO) versus baseline; with SN = semantic new (recent), SO = semantic old (remote), EN = episodic new (recent), and EO = episodic old (remote). Then, we compared the neural network supporting both memory types (episodic/semantic) irrespective of remoteness within and between younger and older participants (SN + SO > EN + EO and EN + EO > SN + SO). Thereafter, we examined the effect of remoteness irrespective of memory type on the neural activation patterns within and between groups (SN + EN > SO + EO and SO + EO > SN + EN), as well as additional interaction effects between memory type and remoteness [(EN > EO) > (SN > SO) and (SN > SO) > (EN > EO)]. Voxels in MNI-space were considered statistically significant at a threshold of p < .05 (corrected at cluster level) using a height threshold of p < .001 uncorrected, corresponding to T = 3.93. Sociodemographic data and neuropsychological scores were compared using two-tailed t-tests. To investigate the possible activation changes in the hippocampus and the retrosplenial cortex, a region of interest (ROI) approach was chosen based on the coordinates of Piefke and colleagues (2003); −4, −46, +24 for the retrospenial cortex and +30, −20, −20 and −24, −22, −18 for the right and left hippocampus, respectively, applying small volume correction using a sphere centered at these coordinates with a radius of 10 mm corresponding to the size of the Gaussian kernel (full-width at half-maximum) used for smoothing.

Voxel-Based Morphometry

We performed voxel-based morphometry (VBM, Good et al., 2001) with T1-weighted MPRAGE scans to assess the possible age-related brain volume differences. Voxel-based morphometry analysis was applied to whole-brain and ROI analyses. After defining the anterior commissure in each image as the origin of the individual stereotactic space, and reorienting all scans to the axial view, images were segmented automatically into gray matter, white matter, and cerebrospinal fluid probability maps. After removing all non-brain voxels, gray and white matter maps were separately normalized to MNI space using SPM2. The transformation parameters derived from normalizing individual gray matter maps to MNI templates were used to normalize the individual anatomical T1 images. Normalized images of all subjects were averaged and smoothed to create a study-specific template. Then, individual images were locally deformed to this study-specific template using nonlinear spatial transformations. We corrected for non-uniformities in signal intensity and segmented the normalized images into gray and white matter and cerebrospinal fluid maps. After correction for possible volume changes as a result of spatial normalization (images were modulated by multiplying voxel values in the segmented images by the Jacobian determinants derived from the spatial normalization step), the resulting gray and white matter maps were smoothed with a Gaussian kernel of 12-mm FWHM.

To further examine the effects of possible group differences in brain structure (as revealed by VBM) on the BOLD signal differences observed when comparing young subjects and older adults, biological parametric mapping was applied (Casanova et al., 2007). Biological parametric mapping enables multimodal image analyses, thus incorporating information from different imaging modalities into a common statistical model. In our study, we included the VBM data of each subject as a covariate in the fMRI analysis, to test whether the changes in BOLD signal would be at least in part explained by brain structure differences across subjects.

Results

Neuropsychological Profiles and Behavioral Data

Based on age-adjusted Z scores using normative data, there was no significant difference in test performance between the groups in any of the neuropsychological tests. All participants were cognitively healthy and performed better than 1 SD below the mean in all tests. During fMRI scanning, there was no significant difference in the alertness task reaction times across memory conditions and between groups. For details on post-scanning debriefing results, see online Supplementary Material, Table S1.

Brain Activity Associated with All Memory Conditions versus Baseline

For details, see Table 1 and online Supplementary Material, Figure S1. When contrasting all memory conditions (SN + SO + EN + EO) versus baseline, both subject groups activated a large, predominantly left-lateralized neural network. Young subjects but not older adults showed an additional regional decrease of brain activity. Inter-group comparison revealed significantly greater activations among older adults in parieto-occipital cortices.

Table 1.

Relative increases/decreases in brain activity for all memory conditions (SN + SO + EN + EO) versus baseline

Region Side x y z t-score (z-value) k (voxel) 
Older adults 
Relative activity increases 
 Dorso-lateral prefrontal cortex (DLPFC) +54 −4 +46 8.06 (4.64) 1,344 
−46 +6 +34 15.36 (5.93) 14,111 
 Frontal operculum +58 +24 +8 6.89 (4.3) 796 
 Medial temporal gyrus/STS +48 −50 17.31 (6.16) 4,121 
 Sup. parietal cortex +26 −68 +56 5.19 (3.69) 222 
 Medial frontal cortex/ACC −8 −2 +64 10.00 (5.29) 2,757 
 Medial prefrontal cortex −12 +48 +28 5.91 (3.8) 348 
Young subjects 
Relative activity increases 
 Dorso-lateral prefrontal cortex (DLPFC) +62 +18 +28 4.72 (3.59) 474 
−52 +12 +30 7.77 (4.76) 10,442 
 Lateral extra-striate cortex (V5/MT) +46 −72 6.33 (4.92) 1,430 
−52 −72 7.77 (4.76) 10,442 
 Middle temporal gyrus/STS +42 −32 −4 6.33 (4.92) 1,430 
−48 −34 −6 9.07 (5.12) 10,442 
 Dorsal premotor cortex (DD FEF) +48 −4 +56 6.48 (4.34) 474 
−40 −2 +58 8.71 (5.02) 10,442 
 Striatum +22 −6 +18 6.63 (4.39) 307 
−24 +16 6.63 (4.39) 307 
 Cerebellar hemisphere +36 −58 −24 10.49 (5.45) 6,025 
 Supplementary motor area (SMA) +2 +2 +62 8.33 (4.92) 1,430 
 Cingulate cortex +8 +10 +42 4.57 (3.52) 1,430 
 Cerebellar vermis −2 −54 −36 10.49 (5.45) 6,025 
 Posterior inferior temporal cortex −40 −54 −14 9.07 (5.12) 10,442 
Relative activity decreases 
 Striatum −10 −90 −6 8.5 (4.98) 2,270 
 Occipital cortex +16 −86 +38 9.88 (5.20) 6,357 
 Supramarginal gyrus +66 −30 +40 8.5 (4.98) 1,152 
Old > young subjects 
 Lateral parieto-occipital cortex +34 −74 +24 6.66 (5.04) 933 
−22 −82 +38 6.66 (5.04) 933 
 Superior parietal cortex +22 −68 +60 4.44 (3.80) 237 
 Medial occipital cortex +4 −88 +16 6.66 (5.04) 933 
Region Side x y z t-score (z-value) k (voxel) 
Older adults 
Relative activity increases 
 Dorso-lateral prefrontal cortex (DLPFC) +54 −4 +46 8.06 (4.64) 1,344 
−46 +6 +34 15.36 (5.93) 14,111 
 Frontal operculum +58 +24 +8 6.89 (4.3) 796 
 Medial temporal gyrus/STS +48 −50 17.31 (6.16) 4,121 
 Sup. parietal cortex +26 −68 +56 5.19 (3.69) 222 
 Medial frontal cortex/ACC −8 −2 +64 10.00 (5.29) 2,757 
 Medial prefrontal cortex −12 +48 +28 5.91 (3.8) 348 
Young subjects 
Relative activity increases 
 Dorso-lateral prefrontal cortex (DLPFC) +62 +18 +28 4.72 (3.59) 474 
−52 +12 +30 7.77 (4.76) 10,442 
 Lateral extra-striate cortex (V5/MT) +46 −72 6.33 (4.92) 1,430 
−52 −72 7.77 (4.76) 10,442 
 Middle temporal gyrus/STS +42 −32 −4 6.33 (4.92) 1,430 
−48 −34 −6 9.07 (5.12) 10,442 
 Dorsal premotor cortex (DD FEF) +48 −4 +56 6.48 (4.34) 474 
−40 −2 +58 8.71 (5.02) 10,442 
 Striatum +22 −6 +18 6.63 (4.39) 307 
−24 +16 6.63 (4.39) 307 
 Cerebellar hemisphere +36 −58 −24 10.49 (5.45) 6,025 
 Supplementary motor area (SMA) +2 +2 +62 8.33 (4.92) 1,430 
 Cingulate cortex +8 +10 +42 4.57 (3.52) 1,430 
 Cerebellar vermis −2 −54 −36 10.49 (5.45) 6,025 
 Posterior inferior temporal cortex −40 −54 −14 9.07 (5.12) 10,442 
Relative activity decreases 
 Striatum −10 −90 −6 8.5 (4.98) 2,270 
 Occipital cortex +16 −86 +38 9.88 (5.20) 6,357 
 Supramarginal gyrus +66 −30 +40 8.5 (4.98) 1,152 
Old > young subjects 
 Lateral parieto-occipital cortex +34 −74 +24 6.66 (5.04) 933 
−22 −82 +38 6.66 (5.04) 933 
 Superior parietal cortex +22 −68 +60 4.44 (3.80) 237 
 Medial occipital cortex +4 −88 +16 6.66 (5.04) 933 

Notes: All clusters with increased or decreased brain activity are significant at p < .05, corrected for multiple comparisons at the cluster level (with a height threshold of p < .001, uncorrected at the voxel level). Brain regions showing relatively significant BOLD signal increase associated with all memory conditions versus baseline in the intra-group-comparison for the old and young group. For each region, the coordinates of the maximally activated voxel within the activation cluster are given in standard MNI space. x = distance (mm) to right (R; +) or left (L; −) of the midsagittal plane; y = distance anterior (+) or posterior (−) to vertical plane through the anterior commissure; z = distance above (+) or below (−) the inter-commissural (AC-PC) plane.

SN = Semantic-new (recent); SO = Semantic-old (remote); EN = Episodic-new (recent); EO = Episodic-old (remote); ACC = anterior cingulate cortex; DLPFC = dorsolateral prefrontal cortex; SMA = supplementary motor area.

Brain Activity Associated with Differential Content of Memories: Episodic versus Semantic

For details, see Table 2 and Fig. 2. When contrasting episodic versus semantic memories (EN + EO > SN + SO), between-group comparison revealed significant greater activations in fusiform and occipital areas when older adults processed episodic memories. The inverse contrast (SN + SO > EN + EO) did not reveal significant findings within and between groups.

Table 2.

Relative increases in brain activity during autobiographical versus semantic memory retrieval irrespective of remoteness (EN + EO > SN + SO)

Region Side x y z t-score (z-value) k (voxel) 
Older adults 
 Temporo-parietal junction +58 −52 +26 7.03 (4.35) 500 
−56 −56 +34 11.93 (5.45) 1758 
 Cerebellar hemisphere +18 −68 −24 7.04 (4.35) 853 
−14 −70 −34 5.93 (3.98) 217 
 Cerebellar vermis −62 −32 6.69 (4.24) 853 
 Insular cortex −32 +6 −6 8.01 (4.63) 490 
 Superior temporal gyrus, temporal pole −50 +12 −26 6.41 (4.15) 490 
 Posterior cingulate cortex/retrosplenial cortex −12 −52 +34 6.91 (4.31) 602 
 ACC/medial prefrontal cortex −6 +40 +16 7.12 (4.37) 618 
Young subjects 
 Temporo-parietal junction (TPJ) +48 −48 +20 7.42 (4.65) 1032 
−44 −46 +10 8.78 (5.05) 2188 
 Middle temporal gyrus/STS +48 −26 −4 7.42 (4.65) 1032 
−50 −24 −10 8.78 (5.05) 2188 
 Anterior insular cortex +50 +18 −10 7.42 (4.65) 1032 
−38 +18 −12 7.93 (4.81) 2441 
 Anterior temporal cortex +48 +12 −26 6.89 (4.48) 996 
−50 +4 −28 8.78 (5.05) 2188 
 Anterior thalamus +4 −8 +10 7.43 (4.66) 1409 
−8 −10 7.43 (4.66) 1409 
 Posterior cingulate cortex/precuneus −4 −56 +44 7.42 (4.65) 1255 
 Posterior cingulate cortex/retrosplenial cortex −12 −46 +36 7.42 (4.65) 1255 
 (Pre-) supplementary motor area (SMA) +8 +6 +62 7.44 (4.66) 5702 
 (Posterior) rostral medial prefrontal cortex −4 +12 +62 7.44 (4.66) 5702 
 Dorso-lateral prefrontal cortex −42 +8 +36 4.50 (3.48) 336 
 Anterior lateral prefrontal cortexL −26 +54 +24 7.44 (4.66) 5702 
 Hippocampus +26 −26 −6 4.79 (3.63) 45* 
Old > young subjects 
 Fusiform gyrus +26 −78 −4 4.61 (3.90) 143 
 Superior occipital gyrus +22 −94 +4 4.44 (3.80) 237 
 Middle occipital gyrus −26 −82 +2 6.66 (5.04) 933 
Region Side x y z t-score (z-value) k (voxel) 
Older adults 
 Temporo-parietal junction +58 −52 +26 7.03 (4.35) 500 
−56 −56 +34 11.93 (5.45) 1758 
 Cerebellar hemisphere +18 −68 −24 7.04 (4.35) 853 
−14 −70 −34 5.93 (3.98) 217 
 Cerebellar vermis −62 −32 6.69 (4.24) 853 
 Insular cortex −32 +6 −6 8.01 (4.63) 490 
 Superior temporal gyrus, temporal pole −50 +12 −26 6.41 (4.15) 490 
 Posterior cingulate cortex/retrosplenial cortex −12 −52 +34 6.91 (4.31) 602 
 ACC/medial prefrontal cortex −6 +40 +16 7.12 (4.37) 618 
Young subjects 
 Temporo-parietal junction (TPJ) +48 −48 +20 7.42 (4.65) 1032 
−44 −46 +10 8.78 (5.05) 2188 
 Middle temporal gyrus/STS +48 −26 −4 7.42 (4.65) 1032 
−50 −24 −10 8.78 (5.05) 2188 
 Anterior insular cortex +50 +18 −10 7.42 (4.65) 1032 
−38 +18 −12 7.93 (4.81) 2441 
 Anterior temporal cortex +48 +12 −26 6.89 (4.48) 996 
−50 +4 −28 8.78 (5.05) 2188 
 Anterior thalamus +4 −8 +10 7.43 (4.66) 1409 
−8 −10 7.43 (4.66) 1409 
 Posterior cingulate cortex/precuneus −4 −56 +44 7.42 (4.65) 1255 
 Posterior cingulate cortex/retrosplenial cortex −12 −46 +36 7.42 (4.65) 1255 
 (Pre-) supplementary motor area (SMA) +8 +6 +62 7.44 (4.66) 5702 
 (Posterior) rostral medial prefrontal cortex −4 +12 +62 7.44 (4.66) 5702 
 Dorso-lateral prefrontal cortex −42 +8 +36 4.50 (3.48) 336 
 Anterior lateral prefrontal cortexL −26 +54 +24 7.44 (4.66) 5702 
 Hippocampus +26 −26 −6 4.79 (3.63) 45* 
Old > young subjects 
 Fusiform gyrus +26 −78 −4 4.61 (3.90) 143 
 Superior occipital gyrus +22 −94 +4 4.44 (3.80) 237 
 Middle occipital gyrus −26 −82 +2 6.66 (5.04) 933 

Notes: All clusters with increased brain activity are significant at p < .05, cluster-corrected for multiple comparisons at the cluster level (with a height threshold of p < .001, uncorrected at the voxel level). *pSVC < .05 in a hypothesis-driven region-of-interest (ROI) analysis. For further explanations, see Table 1.

Fig. 2.

Episodic relative to semantic memory retrieval irrespective of remoteness. First row: Older adults (red, n = 14), young subjects (green, n = 15). Second row: Intergroup comparison (blue): Intergroup comparison revealed significant activation differences for the right superior occipital and left middle occipital cortex, and fusiform gyrus (not shown on this view). Activations are superimposed on a rendered standard single subject brain provided by spm2. First column: Lateral view on the left hemisphere, second column: View from above, and third column: Lateral view on the right hemisphere (R = right; L = left; A = anterior; p = posterior). The activation clusters shown are significant at a level of p < .05, corrected for multiple comparisons at the cluster level. The exact coordinates of the local maxima within the areas of activation and their Z-value and t-statistics are given in Table 2.

Fig. 2.

Episodic relative to semantic memory retrieval irrespective of remoteness. First row: Older adults (red, n = 14), young subjects (green, n = 15). Second row: Intergroup comparison (blue): Intergroup comparison revealed significant activation differences for the right superior occipital and left middle occipital cortex, and fusiform gyrus (not shown on this view). Activations are superimposed on a rendered standard single subject brain provided by spm2. First column: Lateral view on the left hemisphere, second column: View from above, and third column: Lateral view on the right hemisphere (R = right; L = left; A = anterior; p = posterior). The activation clusters shown are significant at a level of p < .05, corrected for multiple comparisons at the cluster level. The exact coordinates of the local maxima within the areas of activation and their Z-value and t-statistics are given in Table 2.

Brain Activity Associated with Differential Remoteness of Memories: Recent versus Remote

For details, see Table 3 and Fig. 3. Contrasting recent versus remote memories (SN + EN > SO + EO and SO + EO > SN + EN) did not lead to significant differences between groups at the predefined statistical threshold, for both whole-brain and ROI analyses.

Table 3.

Relative increases in brain activity during retrieval of recent memories versus remote memories irrespective of memory type (EN + SN > EO + SO)

Region Side x y z t-Score (z-value) k (voxel) 
Older adults 
 Retrosplenial cortex −6 −48 +24 4.83 (3.53) 33* 
Young subjects 
 Hippocampus −28 −26 −10 4.82 (3.64) 39* 
+26 −26 −6 4.67 (3.57) 47* 
 Retrosplenial cortex −8 −58 +14 6.74 (4.43) 1,676** 
 Orbitofrontal cortex −4 +48 −4 4.78 (3.62) 227** 
Region Side x y z t-Score (z-value) k (voxel) 
Older adults 
 Retrosplenial cortex −6 −48 +24 4.83 (3.53) 33* 
Young subjects 
 Hippocampus −28 −26 −10 4.82 (3.64) 39* 
+26 −26 −6 4.67 (3.57) 47* 
 Retrosplenial cortex −8 −58 +14 6.74 (4.43) 1,676** 
 Orbitofrontal cortex −4 +48 −4 4.78 (3.62) 227** 

Notes: **p < .05, cluster-corrected for multiple comparisons at the cluster level (with a height threshold of p < .001, uncorrected at the voxel level); *pSVC < .05 in a hypothesis-driven region-of-interest (ROI) analysis. For further explanations, see Table 1.

Fig. 3.

Retrieval of recent versus memories irrespective of memory type. First column: Relative increase in neural activity in the retrosplenial cortex of older adults (n = 14) associated with recent relative to remote memory retrieval, irrespective of memory type. The BOLD signal change at the local maximum within the retrosplenial cortex activation (−6, −48, +24) is statistically significant at the voxel level in an ROI analysis (pSVC < .05) based on previously reported co-ordinates by Piefke and colleagues (2003). The local maximum is superimposed on a sagittal section of the structural mean group image, spatially normalized into MNI space (x = −6 mm). Second column: Relative increase in neural activity in the retrosplenial cortex and frontal cortex of young subjects (n = 15) associated with recent relative to remote memory retrieval, irrespective of memory type. The upper histogram displays the percentage BOLD signal change for the local maximum (−8, −58, +14) in the area of significantly increased neural activity within the retrosplenial cortex as a function of the experimental memory conditions (mean, SD). The lower histogram displays the percentage BOLD signal change for the local maximum (−4, +48, −4) in the area of significantly increased neural activity within the right orbitofrontal cortex as a function of the experimental memory conditions (mean, SD). Third column: Relative increase in neural activity in bilateral hippocampus of young subjects (n = 15) associated with recent memories relative to remote memories, irrespective of the memory type (episodic or semantic). The upper histogram displays the percentage BOLD signal change for the local maximum (+26, −26, −6) in the area of significantly increased neural activity within the right hippocampus as a function of the experimental memory conditions (mean, SD). The lower histogram displays the percentage BOLD signal change for the local maximum (−28, −26, −10) in the area of increased neural activity within the left hippocampus as a function of the experimental memory conditions (mean, SD). SN = semantic-new, SO = semantic-old, EN = episodic-new, and EO = episodic-old.

Fig. 3.

Retrieval of recent versus memories irrespective of memory type. First column: Relative increase in neural activity in the retrosplenial cortex of older adults (n = 14) associated with recent relative to remote memory retrieval, irrespective of memory type. The BOLD signal change at the local maximum within the retrosplenial cortex activation (−6, −48, +24) is statistically significant at the voxel level in an ROI analysis (pSVC < .05) based on previously reported co-ordinates by Piefke and colleagues (2003). The local maximum is superimposed on a sagittal section of the structural mean group image, spatially normalized into MNI space (x = −6 mm). Second column: Relative increase in neural activity in the retrosplenial cortex and frontal cortex of young subjects (n = 15) associated with recent relative to remote memory retrieval, irrespective of memory type. The upper histogram displays the percentage BOLD signal change for the local maximum (−8, −58, +14) in the area of significantly increased neural activity within the retrosplenial cortex as a function of the experimental memory conditions (mean, SD). The lower histogram displays the percentage BOLD signal change for the local maximum (−4, +48, −4) in the area of significantly increased neural activity within the right orbitofrontal cortex as a function of the experimental memory conditions (mean, SD). Third column: Relative increase in neural activity in bilateral hippocampus of young subjects (n = 15) associated with recent memories relative to remote memories, irrespective of the memory type (episodic or semantic). The upper histogram displays the percentage BOLD signal change for the local maximum (+26, −26, −6) in the area of significantly increased neural activity within the right hippocampus as a function of the experimental memory conditions (mean, SD). The lower histogram displays the percentage BOLD signal change for the local maximum (−28, −26, −10) in the area of increased neural activity within the left hippocampus as a function of the experimental memory conditions (mean, SD). SN = semantic-new, SO = semantic-old, EN = episodic-new, and EO = episodic-old.

Interactions

There were no significant interactions ([EN > EO] > [SN > SO] and [SN > SO] > [EN > EO]) between the factors memory content and remoteness neither within nor between groups.

Voxel-Based Morphometry

Voxel-based morphometry analysis did not reveal significant differences (p < .05 at the cluster level) between the brain volumes of older adults (1,098 ± 98 ml) and young participants (1,222 ± 126 ml). We also performed an ROI analysis for the hippocampus bilaterally using the co-ordinates reported by Piefke and colleagues (2003), which also did not reveal a significant between-group difference. Even though there were no significant differences in brain structure between groups, we performed biological parametric mapping (Casanova et al., 2007) on those fMRI contrasts, which had revealed significant between-group differences. Analyses of covariance that included VBM data of each subject as a covariate in the voxel-wise analysis revealed the same pattern of results as the initial analyses. This further confirmed that the observed BOLD differences did not occur due to changes in brain structure per se.

Discussion

In this fMRI study, we investigated whether there are age-related differences in brain activity associated with content (episodic vs. semantic) and remoteness (recent vs. remote) of long-term memory retrieval. Among older adults and young participants, there were many commonalities in the neural networks subserving declarative long-term memory, which is in line with previous studies (Gilboa, Winocur, Grady, Hevenor, & Moscovitch, 2004; Levine et al., 2004; Maguire & Frith, 2003a; and for review: Svoboda et al., 2006). But there were also key differences as discussed subsequently.

Network Subserving Long-Term Declarative Memory Processing (All Memory Conditions versus Baseline)

First we investigated neural activity associated with all memory conditions combined compared with baseline (reading an instruction sentence), in order to characterize the common network supporting long-term declarative memory. Older adults and young participants activated a bilateral predominantly left-lateralized prefrontal and temporal network. Older adults activated larger areas of both hemispheres with significantly greater activations in the parieto-occiptal cortex bilaterally, and right superior parietal and medial occipital cortices when compared with the young. The pattern is consistent with a ‘hemispheric asymmetry reduction in older adults’ (Cabeza, 2002) that could be a task-independent general aging phenomenon (Logan, Sanders, Snyder, Morris, & Buckner, 2002). During post-scanning debriefing older adults when compared with young subjects reported higher emotionality associated with the stimuli presented. Using post-scanning data as additional regressors (of no interest) in the image analyses did not change our results, suggesting that the bilateral increase of activity was not due to higher emotionality among older adults. Young subjects but not older adults additionally showed specific brain regions with relative activity decrease during long-term memory processing. Miller and colleagues (2008) demonstrated that successful memory formation requires a coordinated pattern of activation and deactivation in a distributed memory network that may be altered by the process of aging.

Network Subserving Autobiographical Memory Processing Irrespective of Remoteness

In this study we were specifically interested in the changes of brain activity associated with episodic versus semantic memory retrieval. Irrespective of memory remoteness, we found activations in brain areas described as the ‘core’-network associated with autobiographical memory (Svoboda et al., 2006) in our young and older participants. This large predominantly left-lateralized network includes the medial and ventrolateral prefrontal cortex, as well as medial and lateral temporal and retrosplenial/posterior cingulate regions, the temporoparietal junction and the cerebellum—a pattern consistently found across the majority of imaging studies (for reviews, see Maguire, 2001; Svoboda et al., 2006). However, we additionally revealed age-related activation changes, suggesting that age itself is a factor modulating the networks underlying autobiographical memory processing: Between-group analysis revealed significant greater neural activity in right fusiform and occipital areas among older adults when compared with the young people. The right fusiform gyrus has been linked to face recognition and processing of object details (Simons, Koutstaal, Prince, Wagner, & Schacter, 2003), as well as to successful encoding of specific visual features that support episodic recognition (Kukolja, Thiel, Wilms, Mirzazade, & Fink, 2009). In a study applying electroencephalography during autobiographical memory retrieval, activity in posterior temporal and occipital regions corresponded to the formation of detailed memory (Conway, Pleydell-Pearce, & Whitecross, 2001). Similar posterior activations were found, when participants held the fully formed memory in mind (Conway et al., 2003). In line with our results, two recent studies investigating episodic retrieval revealed greater neural recruitment, including medial temporal and occipital regions, in older adults when compared with younger subjects (Duverne et al., 2008; Morcom et al., 2007). In both studies, additional neural recruitment was interpreted as decline in neural efficiency, but it remains largely unclear if age-associated ‘over-recruitment’ reflects dedifferentiation and loss of functional specificity (Grady & Craik, 2000) or compensatory and adaptive processes (Cabeza, 2002; Cabeza et al., 2002). Morcom and colleagues (2007) showed that the pattern of additional neural recruitment in older adults varied due to task type and subject performance, but was not explained by behavioral performance differences between older and younger subjects alone. Duverne and colleagues (2008) and Morcom and colleagues (2007) did not find hemispheric asymmetry reduction in older participants. The authors argued that this might not be an invariable finding, but one example of a general tendency toward an increased neural recruitment in older adults. In contrast to both studies, we investigated personal autobiographical memories, which may have contributed to different patterns of neural recruitment.

Older adults experience a decrease in their ability to recall sensory–perceptive or spatio-temporal event details (Piolino et al., 2006). Our data suggest that the relative activity increase in occipital cortical regions among older adults may indicate additional, possibly compensatory neural recruitment. Alternatively, the process of holding long-term memories in mind could necessitate greater visuospatial processing in the aging brain, reflected by an additional neural recruitment. This notion is supported by the fact that older compared with young participants attributed significantly higher scores of picture-likeness and vividness in the post-scanning debriefing. However, using these scores as additional regressors in our fMRI analysis did not change our results. Our post-scanning data differ from previous studies reporting a decrease in memory recollection quality with age (Piolino, Desgranges, Benali, & Eustache, 2002; Piolino et al., 2006). Different aspects of our study design could have contributed to this disparity. It has been shown that when retrieval support is given (e.g., cues), there are fewer differences in autobiographical memory retrieval quality between older adults and young subjects (Anderson, Craik, & Neveh-Benjamin, 1998; Cohen, 1998). Older adults also reported greater emotionality, which could additionally attenuate possible age-related differences in recollection quality (Reisberg, Heuer, McLean, & O'Shaughnessy, 1988). We did not find a posterior–anterior shift in cortical activity in our older participants as observed by Davis and colleagues (2008). This could be explained by differences in task performance or experimental design (Daselaar, Veltman, Rombouts, Raaijmakers, & Jonker, 2003; Morcom et al., 2007). Daselaar and colleagues (2003) showed an additional prefrontal cortical activity associated with episodic memory retrieval only in ‘low performing’ older adults, whereas ‘high performing’ older participants and young subjects did not show this frontal recruitment. Our data support the notion that age-related changes in sensory-perceptive detail recall abilities (Piolino et al., 2006) may be different from those due to cognitive decline. The activation of the right hippocampus in young subjects (within group analysis) was unexpected, as activation of the left hippocampus has been described more often (Maguire & Frith, 2003a; Piefke et al., 2003). The right hippocampus is involved in memory tasks that require spatial location processing and wayfinding through complex environments (Burgess, Maguire, & O'Keefe, 2003). Young in contrast to older subjects more often take the perspective of an ‘actor’ while reliving autobiographical memories (Piolino et al., 2006), which may have contributed to this finding.

Effect of Remoteness on Memory Networks

Finally, we were interested in whether the participants' age modulates the pattern of brain activity associated with differential remoteness of memory. The direct comparison between older adults and young subjects did not reveal significant differences associated with memory remoteness irrespective of memory type. We therefore only briefly highlight our within-group results: Those are in line with previous studies: Piefke and colleagues (2003) and Oddo and colleagues (2010) revealed a neural network of retrosplenial and orbitofrontal cortices associated with recent versus remote memory retrieval in young subjects. Both regions are involved in autobiographical memory processing (Markowitsch, Vandekerckhove, Lanfermann, & Russ, 2003; Piefke & Fink, 2005; Piefke et al., 2003; Svoboda et al., 2006). The orbitofrontal cortex may be important for placing memories on a time-line (Tranel & Jones, 2006), and is associated with self-reference (Northoff et al., 2006). Among young subjects, the additional hippocampal activation associated with retrieval of recent versus remote memories is in line with the classical memory consolidation model (Squire et al., 2004). In our study, older adults showed activation in the retrosplenial cortex for recent versus remote memory retrieval, as revealed by ROI analysis, which is in line with our previous data (Poettrich et al., 2009). Our results suggest that the factor remoteness (of memories) is associated with differential brain activity in ‘core’ and ‘secondary’ brain regions of autobiographical memory networks as described by Svoboda and colleagues (2006). Whereas our results suggest age-related differences in brain activity associated with different memory content, differential remoteness of memories may modulate the networks underlying autobiographical memory retrieval in a similar manner in young and older adults.

Study Limitations and Summary

This study has several limitations: In contrast to semantic facts, autobiographical memory traces were reactivated prior to the fMRI experiment in the pre-experimental interviews. Sometimes, third-party interviews are used to obtain information about the subjects' life events (Gilboa et al., 2005; Viard et al., 2007). These have the inherent disadvantage that the events reported are not as personal as self-reported events, and vivid re-experience may be difficult. Another limitation is that some of the semantic facts could have potentially triggered autobiographical information. This potential confounder was addressed in post-scan assessments. All subjects correctly assigned all stimuli to either semantic fact or autobiographical event category. Furthermore, childhood memories are necessarily more remote for the older adults. There are two general possibilities: One can choose memories from the same absolute age range (e.g., memories from the age of 5–15 years for both groups), resulting in different temporal distances between remote and recent memories for young and older subjects, or one keeps the temporal distance between recent and remote memories fixed, resulting in memories from different life periods in each group, which might be differently encoded/retrieved. We decided to use memories from the same absolute age. Our rationale was to avoid collecting memories from the 'reminiscence bump' period (Rubin, Wetzler, & Nebes, 1986). This period, which has been described in subjects over 40 years old, refers to particularly well-remembered memories, encoded in adolescence and young adulthood. Thus, in this study, remote memories from the same absolute age might best represent the same type of memories in both groups.

In summary, our data could reflect a functional reorganization of long-term declarative memory processing in older adults. Greater activity in posterior brain regions among older people during episodic memory retrieval could reflect compensatory neural recruitment due to impairment in vivid visual imagery. It could also indicate greater utilization of visuospatial processing during episodic memory retrieval in the aging brain. Differential remoteness of long-term memories did not elicit significant group differences surviving stringent statistical thresholds.

Supplementary Material

Supplementary material is available at Archives of Clinical Neuropsychology online.

Conflict of Interest

None declared.

Acknowledgements

MD and KP have equally contributed to the study and therefore share first authorship.

References

Anderson
N. D.
Craik
F. I. M.
Neveh-Benjamin
M.
The attentional demands of encoding and retrieval in younger and older adults: 1. Evidence from divided attention costs
Psychology and Aging
 , 
1998
, vol. 
13
 (pg. 
405
-
423
)
[PubMed]
Burgess
N.
Maguire
E. A.
O'Keefe
J.
The human hippocampus and spatial and episodic memory
Neuron
 , 
2003
, vol. 
35
 
4
(pg. 
625
-
641
)
Cabeza
R.
Hemispheric asymmetry reduction in older adults: the HAROLD model
Psychology and Aging
 , 
2002
, vol. 
17
 
1
(pg. 
85
-
100
)
[PubMed]
Cabeza
R.
Anderson
N.
Locantore
J.
McIntosh
A.
Aging gracefully: compensatory brain activity in high-performing older adults
NeuroImage
 , 
2002
, vol. 
17
 (pg. 
1394
-
1402
)
[PubMed]
Cabeza
R.
St Jacques
P.
Functional neuroimaging of autobiographical memory
Trends in Cognitive Sciences
 , 
2007
, vol. 
11
 
5
(pg. 
219
-
227
)
[PubMed]
Casanova
R.
Srikanth
R.
Baer
A.
Laurienti
P. J.
Burdette
J. H.
Hayasaka
S.
, et al.  . 
Biological parametric mapping: A statistical toolbox for multimodality brain image analysis
NeuroImage
 , 
2007
, vol. 
34
 (pg. 
137
-
143
)
[PubMed]
Cohen
G.
Thompson
C.
Herrmann
D.
Bruce
D.
Read
J. D.
Payne
J.
Toglia
M.
The effects of aging on autobiographical memory
Autobiographical memory: theoretical and applied perspectives
 , 
1998
London
Lawrence Earlbaum
(pg. 
105
-
123
)
Conway
M. A.
Sensory-perceptual episodic memory and its context: autobiographical memory
Philosophical Transactions of the Royal Society London B: Biological Sciences
 , 
2001
, vol. 
356
 
1413
(pg. 
1375
-
1384
)
Conway
M. A.
Pleydell-Pearce
C. W.
The construction of autobiographical memories in the self-memory system
Psychological Review
 , 
2000
, vol. 
107
 
2
(pg. 
261
-
288
)
[PubMed]
Conway
M. A.
Pleydell-Pearce
C. W.
Whitecross
S. E.
The neuroanatomy of autobiographical memory: a slow cortical potential study of autobiographical memory retrieval
Journal of Memory and Language
 , 
2001
, vol. 
45
 
3
(pg. 
493
-
524
)
Conway
M. A.
Pleydell-Pearce
C. W.
Whitecross
S. E.
Sharpe
H.
Neurophysiological correlates of memory for experienced and imagined events
Neuropsychologia
 , 
2003
, vol. 
41
 
3
(pg. 
334
-
340
)
[PubMed]
Daselaar
S. M.
Veltman
D. J.
Rombouts
S. A. R. B.
Raaijmakers
J. G. W.
Jonker
C.
Neuroanatomical correlates of episodic encoding and retrieval in young and elderly subjects
Brain
 , 
2003
, vol. 
126
 (pg. 
43
-
56
)
[PubMed]
Davis
S.
Dennis
N.
Daselaar
S.
Fleck
M.
Cabeza
R.
Qué PASA? The posterior-anterior shift in aging
Cerebral Cortex
 , 
2008
, vol. 
18
 
5
(pg. 
1201
-
1209
)
[PubMed]
Delis
D. C.
Kramer
J. H.
Kaplan
E.
Ober
B. A.
California verbal learning test, adult version.
 , 
1987
San Antonio
The Psychological Corporation
Duverne
S.
Habibi
A.
Rugg
M.
Regional specificity of age effects on the neural correlates of episodic retrieval
Neurobiology of Aging
 , 
2008
, vol. 
29
 
12
(pg. 
1902
-
1916
)
[PubMed]
Eldridge
L. L.
Knowlton
B. J.
Furmanski
C. S.
Bookheimer
S. Y.
Engel
S. A.
Remembering episodes: a selective role for the hippocampus during retrieval
Nature Neuroscience
 , 
2000
, vol. 
3
 
11
(pg. 
1149
-
1152
)
[PubMed]
Gilboa
A.
Ramirez
J.
Kohler
S.
Westmacott
R.
Black
S. E.
Moscovitch
M.
Retrieval of autobiographical memory in Alzheimer's disease: relation to volumes of medial temporal lobe and other structures
Hippocampus
 , 
2005
, vol. 
15
 
4
(pg. 
535
-
550
)
[PubMed]
Gilboa
A.
Winocur
G.
Grady
C. L.
Hevenor
S. J.
Moscovitch
M.
Remembering our past: Functional neuroanatomy of recollection of recent and very remote personal events
Cerebral Cortex
 , 
2004
, vol. 
14
 (pg. 
1214
-
1225
)
[PubMed]
Good
C. D.
Johnsrude
I. S.
Ashburner
J.
Henson
R. N.
Friston
K. J.
Frackowiak
R. S.
A voxel-based morphometric study of ageing in 465 normal adult human brains
NeuroImage
 , 
2001
, vol. 
14
 
1 Pt 1
(pg. 
21
-
36
)
[PubMed]
Grady
C. L.
Craik
F. I.
Changes in memory processing with age
Current Opinion in Neurobiology
 , 
2000
, vol. 
10
 
2
(pg. 
224
-
231
)
[PubMed]
Greenberg
D. L.
Rubin
D. C.
The neuropsychology of autobiographical memory
Cortex
 , 
2003
, vol. 
39
 (pg. 
687
-
728
)
[PubMed]
Härting
C.
WMS-R. Wechsler Gedächtnistest- revidierte Fassung. Deutsche Adaptation der revidierten Fassung der Wechsler Memory Scale.
 , 
2000
Bern
Verlag Hans Huber
Hedden
T.
Gabrielli
J. D. E.
Insights into the aging mind: a view from cognitive neuroscience
Nature Reviews Neuroscience
 , 
2004
, vol. 
5
 (pg. 
87
-
96
)
[PubMed]
Huber
W.
Poeck
K.
Willmes
K.
The Aachen Aphasia Test
Advances in Neurology
 , 
1984
, vol. 
42
 (pg. 
291
-
303
)
[PubMed]
Kukolja
J.
Thiel
C. M.
Wilms
M.
Mirzazade
S.
Fink
G. R.
Ageing-related changes of neural activity associated with spatial contextual memory
Neurobiology of Aging
 , 
2009
, vol. 
30
 
4
(pg. 
630
-
645
)
[PubMed]
Levine
B.
Svoboda
E.
Hay
J. F.
Winocur
G.
Moscovitch
M.
Aging and autobiographical memory: Dissociating episodic from semantic retrieval
Psychology and Aging
 , 
2002
, vol. 
17
 
4
(pg. 
677
-
689
)
[PubMed]
Levine
B.
Turner
G. R.
Tisserand
D.
Hevenor
S. J.
Graham
S. J.
McIntosh
A. R.
The functional neuroanatomy of episodic and semantic autobiographical remembering: A prospective functional MRI study
Journal of Cognitive Neuroscience
 , 
2004
, vol. 
16
 
9
(pg. 
1633
-
1646
)
[PubMed]
Logan
J. M.
Sanders
A. L.
Snyder
A. Z.
Morris
J. C.
Buckner
R. L.
Under-recruitment and nonselective recruitment: Dissociable neural mechanisms associated with aging
Neuron
 , 
2002
, vol. 
33
 
5
(pg. 
827
-
840
)
[PubMed]
Maguire
E. A.
Neuroimaging studies of autobiographical event memory
Philosophical Transactions of the Royal Society London B: Biological Sciences
 , 
2001
, vol. 
356
 
1413
(pg. 
1441
-
1451
)
Maguire
E. A.
Frith
C. D.
Aging affects the engagement of the hippocampus during autobiographical memory retrieval
Brain
 , 
2003
, vol. 
126
 
Pt 7
(pg. 
1511
-
1523
)
[PubMed]
Maguire
E. A.
Frith
C. D.
Lateral asymmetry in the hippocampal response to the remoteness of autobiographical memories
Journal of Neuroscience
 , 
2003
, vol. 
23
 
12
(pg. 
5302
-
5307
)
[PubMed]
Markowitsch
H. J.
Vandekerckhove
M. M. P.
Lanfermann
H.
Russ
M. O.
Engagement of lateral and medial prefrontal areas in the ecphory of sad and happy autobiographical memories
Cortex
 , 
2003
, vol. 
39
 
4–5
(pg. 
643
-
665
)
[PubMed]
Miller
S. L.
Celone
K.
DePeau
K.
Diamond
E.
Dickerson
B. C.
Rentz
D.
, et al.  . 
Age-related memory impairment associated with loss of parietal deactivation but preserved hippocampal activation
Proceedings of the National Academy of Sciences USA
 , 
2008
, vol. 
105
 
6
(pg. 
2181
-
2186
)
Morcom
A.
Li
J.
Rugg
M. O.
Age effects on the neural correlates of episodic retrieval: Increased cortical recruitment with matched performance
Cerebral Cortex
 , 
2007
, vol. 
17
 
11
(pg. 
2491
-
2506
)
[PubMed]
Moscovitch
M.
Rosenbaum
R. S.
Gilboa
A.
Addis
D. R.
Westmacott
R.
Grady
C.
Functional neuroanatomy of remote episodic, semantic and spatial memory: A unified account based on multiple trace theory
Journal of Anatomy
 , 
2005
, vol. 
207
 
1
(pg. 
35
-
66
)
[PubMed]
Northoff
G.
Heinzel
A.
de Greck
M.
Bermpohl
F.
Dobrowolny
H.
Panksepp
J.
Self-referential processing in our brain–a meta-analysis of imaging studies on the self
NeuroImage
 , 
2006
, vol. 
31
 
1
(pg. 
440
-
457
)
[PubMed]
Oddo
S.
Lux
S.
Weiss
P.
Schwab
A.
Welzer
H.
Markowitsch
H.
, et al.  . 
Specific role of medial prefrontal cortex in retrieving recent autobiographical memories: an fMRI study of young female subjects
Cortex
 , 
2010
, vol. 
46
 
1
(pg. 
29
-
39
)
[PubMed]
Piefke
M.
Fink
G. R.
Recollections of one's own past: the effects of aging and gender on the neural mechanisms of episodic autobiographical memory
Anatomy and Embryology (Berlin)
 , 
2005
, vol. 
10
 
5–6
(pg. 
497
-
512
)
Piefke
M.
Weiss
P. H.
Zilles
K.
Markowitsch
H. J.
Fink
G. R.
Differential remoteness and emotional tone modulate the neural correlates of autobiographical memory
Brain
 , 
2003
, vol. 
126
 
3
(pg. 
650
-
668
)
[PubMed]
Piolino
P.
Desgranges
B.
Benali
K.
Eustache
F.
Episodic and semantic remote autobiographical memory in ageing
Memory
 , 
2002
, vol. 
10
 
4
(pg. 
239
-
257
)
[PubMed]
Piolino
P.
Desgranges
B.
Clarys
D.
Guillery-Girard
B.
Taconnat
L.
Isingrini
M.
, et al.  . 
Autobiographical memory, autonoetic consciousness, and self-perspective in aging
Psychology and Aging
 , 
2006
, vol. 
21
 
3
(pg. 
510
-
525
)
[PubMed]
Poettrich
K.
Weiss
P. H.
Werner
A.
Lux
S.
Donix
M.
Gerber
J.
, et al.  . 
Altered neural network supporting declarative long-term memory in mild cognitive impairment
Neurobiology of Aging
 , 
2009
, vol. 
30
 
2
(pg. 
284
-
298
)
[PubMed]
Reisberg
D.
Heuer
F.
McLean
J.
O'Shaughnessy
M.
The quantity, not the quality, of affect predicts memory vividness
Bulletin of the Psychonomic Society
 , 
1988
, vol. 
26
 (pg. 
100
-
103
)
Reitan
R.
Wolfson
D.
The Halstead-Reitan Neuropsychological Test Battery: Theory and clinical interpretation
 , 
1993
Tucson, AZ
Neuropsychology Press
Rubin
D. C.
Wetzler
S. E.
Nebes
R. D.
Rubin
D. C.
Autobiographical memory across the lifespan
Autobiographical memory
 , 
1986
Cambridge, UK
Cambridge University Press
(pg. 
202
-
221
)
Ryan
L.
Nadel
L.
Keil
K.
Putnam
K.
Schnyer
D.
Trouard
T.
, et al.  . 
Hippocampal complex and retrieval of recent and very remote autobiographical memories: Evidence from functional magnetic resonance imaging in neurologically intact people
Hippocampus
 , 
2001
, vol. 
11
 (pg. 
707
-
714
)
[PubMed]
Simons
J.
Koutstaal
W.
Prince
S.
Wagner
A.
Schacter
D.
Neural mechanisms of visual object priming: Evidence for perceptual and semantic distinctions in fusiform cortex
NeuroImage
 , 
2003
, vol. 
19
 (pg. 
613
-
626
)
[PubMed]
Spreen
O.
Strauss
E. A.
Compendium of neuropsychological tests: Administration, norms, and commentary.
 , 
1998
New York
Oxford University Press
Squire
L. R.
Stark
C. E.
Clark
R. E.
Medial temporal lobe
Annual Reviews of Neuroscience
 , 
2004
, vol. 
27
 (pg. 
279
-
306
)
Svoboda
E.
McKinnon
M. C.
Levine
B.
The functional neuroanatomy of autobiographical memory: A meta-analysis
Neuropsychologia
 , 
2006
, vol. 
44
 (pg. 
2189
-
2208
)
[PubMed]
Takashima
A.
Petersson
K. M.
Rutters
F.
Tendolkar
I.
Jensen
O.
Zwarts
M. J.
, et al.  . 
Declarative memory consolidation in humans: A prospective functional magnetic resonance imaging study
Proceedings of the National Academy of Sciences USA
 , 
2006
, vol. 
103
 (pg. 
756
-
761
)
Tranel
D.
Jones
R. D.
Knowing “what” and knowing “when”
Journal of Clinical and Experimental Neuropsychology
 , 
2006
, vol. 
28
 
1
(pg. 
43
-
66
)
[PubMed]
Tulving
E.
Schacter
D. L.
McLachlan
D. R.
Moscovitch
M.
Priming of semantic autobiographical knowledge: A case study of retrograde amnesia
Brain and Cognition
 , 
1988
, vol. 
8
 
1
(pg. 
3
-
20
)
[PubMed]
Tulving
E.
Episodic memory: From mind to brain
Annual Reviews of Psychology
 , 
2002
, vol. 
53
 (pg. 
1
-
25
)
Viard
A.
Piolino
P.
Desgranges
B.
Chetelat
G.
Lebreton
K.
Landeau
B.
, et al.  . 
Hippocampal activation for autobiographical memories over the entire lifetime in healthy aged subjects: An fMRI study
Cerebral Cortex
 , 
2007
, vol. 
17
 
10
(pg. 
2453
-
2467
)
[PubMed]
Wahlund
L. O.
Barkhof
F.
Fazekas
F.
Bronge
L.
Augustin
M.
Sjogren
M.
, et al.  . 
A new rating scale for age-related white matter changes applicable to MRI and CT
Stroke
 , 
2001
, vol. 
32
 
6
(pg. 
1318
-
1322
)
[PubMed]
Woodard
J. L.
Seidenberg
M.
Nielson
K. A.
Miller
S. K.
Franczak
M.
Antuono
P.
, et al.  . 
Temporally graded activation of neocortical regions in response to memories of different ages
Journal of Cognitive Neuroscience
 , 
2007
, vol. 
19
 (pg. 
1113
-
1124
)
[PubMed]

Author notes

MD and KP have equally contributed to the study and therefore share first authorship.