Stability Of Default-Mode Network Activity In The Aging Brain

LL Beason-Held; MA Kraut; SM Resnick

doi:10.1007/s11682-008-9054-z

. Author manuscript; available in PMC: 2009 Jun 29.

Published in final edited form as: Brain Imaging Behav. 2009 Jan 14;3(2):123–131. doi: 10.1007/s11682-008-9054-z

Stability Of Default-Mode Network Activity In The Aging Brain

LL Beason-Held ^1,^2,^✉, MA Kraut ², SM Resnick ¹

PMCID: PMC2703608 NIHMSID: NIHMS85880 PMID: 19568331

Abstract

Activity attributed to the default-mode occurs during the resting state and is thought to represent self-referential and other intrinsic processes. Although activity in default-associated regions changes across the lifespan, little is known about the stability of default-mode activity in the healthy aging brain. We investigated changes in rest-specific activity across an 8 year period in older participants in the Baltimore Longitudinal Study of Aging (BLSA) neuroimaging study. Comparison of resting-state and recognition memory PET regional cerebral blood flow conditions from baseline and 8-year follow-up shows relative stability of rest-specific activity over time in medial frontal/anterior cingulate, hippocampal and posterior cingulate regions commonly associated with the default-mode. In contrast, prefrontal, parahippocampal and occipital cortical regions, which are not typically associated with default-mode activity, show changes over time Overall, activity in the major components of the default-mode network remains stable in healthy older individuals, a finding which may assist in identifying factors that discriminate between normal and pathological aging.

Keywords: functional imaging, brain function, activation, age, fMRI, human, default activity

INTRODUCTION

Assessment of the neural networks thought to be active when the brain is at rest is an emerging area of interest in functional neuroimaging. Based on distinct patterns of activity observed when examining resting brain function, several processes are thought to occur during this baseline state. These patterns have been theoretically linked to visuoperception, sensorimotor processes, memory, executive function, and functions referred to as `default-mode' (Damoiseaux et al., 2006; Mazoyer et al., 2001; Raichle et al., 2001).

Default-mode activity was first noted as task-related deactivations in experiments where task conditions were contrasted with a baseline, often resting-state, condition. These early studies found that medial frontal/anterior cingulate, posterior cingulate/precuneus and inferior parietal regions exhibited greater activity during the resting state than during task performance (Raichle et al., 2001; Shulman et al., 1997). The spatial pattern of these regions originally prompted the theory of a baseline or default network involved in self-referential processes such as conscious awareness of the internal and external environment. Subsequent identification of medial temporal regions associated with this network expanded the proposed function of the default-mode system to include memory processes (Esposito et al., 2006; Greicius & Menon, 2004), and more recent theories propose that more fundamental intrinsic processes are related to activity in default-mode regions (Buckner & Vincent, 2007; Raichle & Snyder, 2007).

Activity in this network appears generally stable in young adults when examined across repeated sessions (Damoiseaux et al., 2006). The older brain, however, shows subtle changes in regional rest-related default activity when compared to the younger brain. Older subjects decrease resting-state activity in medial frontal/anterior cingulate and posterior cingulate regions of the default-mode network relative to task performance to a lesser extent than younger subjects (Grady, Springer, Hongwanishkul, McIntosh, & Winocur, 2006; Lustig et al., 2003; Persson, Lustig, Nelson, & Reuter-Lorenz, 2007). Older subjects also exhibit default activity in orbital and inferior frontal, parahippocampal and lateral temporal regions that are not involved in younger subjects (Greicius, Srivastava, Reiss, & Menon, 2004). Both types of findings suggest that changes in default-mode regional activity occur from young to older adulthood.

The extent to which activity within the default-mode regions changes as older individuals continue to age, however, remains unclear. Here, we present data from the Baltimore Longitudinal Study of Aging (BLSA) neuroimaging substudy where positron emission tomography (PET) regional cerebral blood flow (rCBF) scans were collected during rest and recognition memory performance at baseline and 8-year follow-up. Because assessment of activity patterns specific to the resting state relative to the recognition memory conditions was used to define regions within the so-called default-mode network, our prior findings of widespread longitudinal changes in both resting-state and recognition memory conditions in cognitively stable healthy agers (Beason-Held, Kraut, & Resnick, 2008a, 2008b) suggest that changes in rest-specific activity may also occur with advancing age. However, because longitudinal changes occur in both resting and task conditions, the relative difference between the two conditions, reflecting rest-specific default-related activity, could actually remain stable over time. In the present study, regions typically associated with the default-mode network show relative stability of rest-specific rCBF patterns, while regions outside of the network show subtle longitudinal changes in activity.

METHODS

Subjects

PET data from 32 older participants in the neuroimaging substudy (S. M. Resnick et al., 2000) of the BLSA were used in this analysis (14 female, 18 male; mean baseline age = 68.4±6.8). All individuals remained in good cognitive and physical health through the follow-up period with no history of central nervous system disorders, major psychiatric disorders including depression, or severe cardiovascular disease. All participants also completed annual neuropsychological evaluations and were deemed cognitively normal by consensus diagnosis (Kawas, Gray, Brookmeyer, Fozard, & Zonderman, 2000) through the 8-year follow-up evaluation.

This study was approved by the local Institutional Review Board. All participants provided written informed consent prior to each assessment.

PET Scanning Conditions

Participants underwent PET scanning sessions at baseline (year 1) and at 8-year follow-up (year 9). During each imaging session, three PET scans were performed: rest, delayed verbal (word) recognition memory, and delayed figural (abstract design) recognition memory. During rest, participants were instructed to keep their eyes open and focused on a computer screen covered by a black cloth. During the delayed recognition tasks, the participants were asked to identify stimuli that had been presented at the beginning of the session. During the imaging session, twenty novel distractor items were intermixed with the previously shown 20 target items and participants were asked to indicate whether or not they had seen each item before (see (Beason-Held et al., 2008a) for task details). Accuracy and reaction times were recorded throughout both recognition tasks.

PET Scanning Parameters

PET measures of regional cerebral blood flow (rCBF) were obtained using [¹⁵O]water. For each scan, 75 mCi of [¹⁵O] water were injected as a bolus. Scans were performed on a GE 4096+ scanner, which provides 15 slices of 6.5 mm thickness. Images were acquired for 60 seconds from the time the total radioactivity counts in brain reached threshold level. Attenuation correction was performed using a transmission scan acquired prior to the emission scans.

MRI Scanning Parameters

A 3-D spoiled gradient refocused (SPGR) MRI scan (124 slices, 256×256 matrix, 0.93×0.93mm voxel size, 1.5 mm slice thickness) was obtained on a 1.5 T GE Signa scanner at each imaging visit.

PET Data Analysis

For each subject, the PET scans were realigned, resliced with a voxel size of 2×2×2mm, spatially normalized into standard stereotactic and smoothed to a full width at half maximum of 12, 12, and 12 mm in the x, y, and z planes. To control for variability in global flow, rCBF values at each voxel were ratio adjusted to the mean global flow of 50 ml/100g/min for each image. The image data were analyzed using Statistical Parametric Mapping (SPM2; Wellcome Department of Cognitive Neurology, London, England), where whole brain voxel by voxel comparisons determined significant differences in rCBF using a single group 2 factor (task and time) design. All contrasts were adjusted for baseline age at year 1.

The data were analyzed in 4 stages. In the first step, we defined the voxels involved in default-mode activity in our sample. Rest-specific activity patterns, defined as activity occurring in the resting-state alone relative to the task activation states, were determined at each year separately by contrasting the rest condition with the mean of the verbal and figural recognition memory conditions at year 1 and year 9. Significance for each contrast was based on a magnitude of p≤0.001. To decrease the chance of false positive results resulting from multiple comparisons, an additional spatial extent threshold of >100 voxels was also used. Second, conjunction analyses were performed on the year 1 and 9 contrast images (magnitude p≤0.001, spatial extent>100 voxels) to determine spatial patterns of activity common to both years 1 and 9, and regions with spatial patterns of activity exclusive to year 1 or year 9 (masking thresholds of p≤0.05). Third, rCBF values were extracted from a 4 mm sphere centered on the local maxima of each region of interest identified in the conjunction analyses to examine the pattern of rCBF change over time. The rest-specific pattern of blood flow change (stable, decreasing, increasing) from year 1 to 9 was then determined for each common and exclusive region (paired sample t-tests, significance level p≤0.05). Finally, to determine the influence of tissue volume on the rCBF findings, the MRI scans were segmented into gray matter, white matter and cerebrospinal fluid and spatially normalized into stereotactic space using a high-dimensional elastic warping method and a volume-preserving transformation (Shen & Davatzikos, 2002). For each participant, binary maps of the common and exclusive regions from the conjunction analyses were registered with the MRI image. Total volumes of gray + white matter were subsequently calculated for each region of activity. The PET analyses were then repeated using tissue volume of each region at each time point as an additional covariate. It should be noted that volumes were unavailable for 1 participant due to inability to tolerate the MRI scanner, and data from an additional participant was excluded from the analysis of the right insular and parahippocampal regions (areas showing rest-specific rCBF decreases over time) due to registration errors.

RESULTS

Neuropsychological Evaluation

Performance on the annual battery of neuropsychological tests (Beason-Held et al., 2008a) was compared across years (1 and 9) using repeated measures MANOVA with sex as a grouping factor. Only one change in performance levels was noted across time. There was a significant increase (p=0.02) in the number of errors on the Benton Visual Retention Test across the group as a whole. There was also a sex effect (p=0.02) on the Digits Forward test with males scoring higher than females at both years. There were no sex × time interactions in this group of participants. Performance levels are shown in Table 1.

Table 1.

Neuropsychological Performance (Mean(SD))

Domain	Test	Measure	Year 1	Year 9
Memory	CVLT BVRT	total correct 5 trials total errors	60.4 (7.4) 3.5 (2.6)	61.3(11.6) 4.4 (3.2)^*
Verbal Fluency	Letters Categories	mean correct mean correct	15.5 (3.6) 16.9 (2.3)	14.4 (3.8) 16.1 (3.2)
Attention & Working Memory	Digits Forward Digits Back Trails A Trails B	total correct total correct time in seconds time in seconds	8.9 (2.5) 9.0 (2.2) 32.7 (12.0) 70.9 (25.8)	9.6 (2.3) 8.3 (2.6) 33.3 (10.7) 79.6 (24.6)
Visuospatial	Card Rotation	correct-incorrect	95.3 (36.7)	100.0 (42.4)

Open in a new tab

Legend: CVLT = California Verbal Learning Test; BVRT = Benton Visual Retention Test.

significant change in performance levels over time (p=0.02).

PET Task Performance

Paired-sample t-tests were used to determine differences in performance levels between year 1 and 9. No significant differences were seen in task accuracy for either recognition memory condition (sensitivity scores: Verbal year1=0.45 (0.18 SD), year9=0.46 (0.19); Figural year1=0.50 (0.20), year9=0.43 (0.20)). Reaction times decreased over time for both tasks (Verbal year1=1384.4 (348.7), year9=1256.9 (409.2), p=0.04; Figural year1=1594.7 (316.3), year9=1432.8 (328.1), p=0.02), most likely reflecting effects of practice during repeating testing.

PET rCBF

We assessed rCBF activity patterns specific to the resting-state relative to the recognition memory conditions. The rest-specific patterns observed at years 1 and 9 were consistent with those reported to comprise the default-mode network of resting brain function. Over time, both similarities and differences in rest-specific activity patterns were observed between years 1 and 9.

In the voxel-based whole brain analysis, regions within frontal, temporal, parietal and occipital lobes show increased activity during rest relative to the recognition memory task conditions at both year 1 and year 9. Based on the conjunction analyses comparing years 1 and 9, the regional pattern of rest-specific activity was similar in medial frontal (Brodmann Area (BA) 11) and anterior cingulate (BA 24/32) areas over time, as were the patterns within lateral insular cortex, temporal pole (BA 38), inferior portions of the middle temporal gyrus (BA 21/39) and the hippocampus (Table 2, Figure 1). Posteriorly, mid-level middle occipital gyrus (BA 19/37) and inferior portions of the posterior cingulate/precuneus (BA 31) also show similar rest-specific spatial activity patterns across time. Using extracted rCBF values, rest-specific rCBF within these regions shows no significant change over time from Year 1 to Year 9 (paired-sample t-tests, p values≥0.14), suggesting stability of activity levels with increasing age in these areas (Figure 2).