Neurobiology of Aging 30 (2009) 717–730

Neural correlates of incidental memory in mild cognitive impairment: An fMRI study Jennifer L. Mandzia a,∗ , Mary Pat McAndrews c,e , Cheryl L. Grady d,e,f , Simon J. Graham d,g , Sandra E. Black a,b,d a

L.C. Campbell Cognitive Neurology Research Unit, Neurosciences Research, Sunnybrook Health Sciences Centre, Canada b Department of Medicine, Division of Neurology, University of Toronto, Canada c University Health Network, Toronto Western Hospital and Research Institute, Canada d Rotman Research Institute and Baycrest Centre for Geriatric Care, Canada e Department of Psychology, University of Toronto, Canada f Department of Psychiatry, University of Toronto, Canada g Department of Medical Biophysics, University of Toronto, Faculty of Medicine, Canada Received 12 November 2006; received in revised form 28 August 2007; accepted 31 August 2007 Available online 25 October 2007

Abstract Behaviour and fMRI brain activation patterns were compared during encoding and recognition tasks in mild cognitive impairment (MCI) (n = 14) and normal controls (NC) (n = 14). Deep (natural vs. man-made) and shallow (color vs. black and white) decisions were made at encoding and pictures from each condition were presented for yes/no recognition 20 min later. MCI showed less inferior frontal activation during deep (left only) and superficial encoding (bilaterally) and in both medial temporal lobes (MTL). When performance was equivalent (recognition of words encoded superficially), MTL activation was similar for the two groups, but during recognition testing of deeply encoded items NC showed more activation in both prefrontal and left MTL region. In a region of interest analysis, the extent of activation during deep encoding in the parahippocampi bilaterally and in left hippocampus correlated with subsequent recognition accuracy for those items in controls but not in MCI, which may reflect the heterogeneity of activation responses in conjunction with different degrees of pathology burden and progression status in the MCI group. © 2007 Elsevier Inc. All rights reserved. Keywords: Alzheimer’s disease; Mild cognitive impairment; fMRI; Memory; Medial temporal lobe

1. Introduction Amnestic mild cognitive impairment (MCI) is a term used to describe individuals who have objective and subjective memory loss greater than normal for their age, do not meet criteria for dementia and yet are at high risk for progressing to AD (Petersen et al., 2001). Longitudinal studies suggest

∗ Corresponding author at: c/o Sunnybrook Health Sciences Centre, 2075 Bayview Avenue, A421, Toronto, Ontario, Canada M4N 3M5. Tel.: +1 416 480 4551; fax: +1 416 480 4552. E-mail address: [email protected] (J.L. Mandzia).

0197-4580/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.neurobiolaging.2007.08.024

that there is a bimodal distribution in terms of progression, where 10–16% of MCI progress to Alzheimer’s disease (AD) annually, but the larger proportion remains stable over time (Petersen et al., 2001, 2005). With the development of new therapies that aim to modify disease progression, there will be an increasing need to identify those individuals at high risk for progression to AD. Functional magnetic resonance imaging (fMRI) can examine early functional brain abnormalities “in vivo” that may not be apparent using more conventional neuroimaging techniques and neuropsychological testing, as significant pathological changes consistent with AD have been found in patients with MCI prior to clinical dementia, suggesting

718

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

some cognitive reserve in the brain’s capacity for adaptation to neural insult (DeKosky et al., 2002; Morris et al., 2001). Several functional neuroimaging studies have examined taskinduced activation patterns in individuals with presumed MCI (Celone et al., 2006; Dickerson et al., 2004, 2005; Johnson et al., 2004, 2006; Machulda et al., 2003; Petrella et al., 2006; Ries et al., 2006; Rombouts et al., 2005; Saykin et al., 2004; Small et al., 1999); The studies in MCI have revealed both increases (Celone et al., 2006; Dickerson et al., 2004, 2005) and decreases in regional activation (Johnson et al., 2004, 2006; Machulda et al., 2003; Petrella et al., 2006; Ries et al., 2006) as well as spatial differences in activation (Celone et al., 2006), when comparing MCI patients to controls. A recent study demonstrated that individuals with MCI and greater clinical impairment recruited a larger extent of right parahippocampal gyrus during an fMRI encoding task suggesting that in the pre-dementia stage intact neurons may actually be “revving up” activity to compensate for pathological changes that are underway (Dickerson et al., 2004). A subsequent study by the same group revealed that patients with very mild MCI showed increased hippocampal activity during associative encoding, but with no differences in hippocampal or entorhinal volumes compared to normal controls demonstrating the uniqueness of fMRI in revealing earlier changes (Dickerson et al., 2005). Therefore, there may exist a continuum of task-related brain activity related to pathological load and host factors. In the early pre-dementia stage compensation may occur in order to maintain task performance, reflected as increased regional activation in similar areas to normal controls or in the use of alternate pathways. As the disease progresses compensatory mechanisms begin to fail and neural reserve would become insufficient, leading to performance deficits and a reduction in the extent and/or magnitude of task-induced activation. Here, fMRI was used to examine brain activation patterns associated with incidental encoding of photographs and subsequent retrieval in a group of amnestic MCI patients and normal controls (NC). By manipulating the depth of memory

encoding (Craik and Lockhart, 1972), we hoped to provide a “cognitive stress test” to probe the ability of MCI patients to activate regions involved in episodic memory—particularly the MTL. To probe for a direct relationship between MTL activity and memory performance, we also performed a region of interest (ROI) analysis to examine differences in the extent of activation – operationalized as the percent of activated voxels – in the hippocampus and parahippocampal gyri between NC and MCI, as well as the relationship between fMRI activity during deep encoding in these regions and subsequent recognition performance, as MTL activation during encoding has been shown to be predictive of successful retrieval (Morcom et al., 2003; Otten et al., 2001).

2. Methods 2.1. Participant recruitment and selection criteria All participants were recruited through the Sunnybrook & Women’s (S&W) Cognitive Neurology Unit. All studies were performed after approval from the S&W Research Ethics Board and with informed consent from each individual. Seventeen MCI patients and 16 normal controls participated in the experiment. However due to problems with motion in the scanner, three MCI and two NC were dropped from the study. The remaining sample consisted of 14 normal controls (NC) (7F/7M) and 14 (7F/7M) MCI patients. Demographic and neuropsychological assessment data are shown in Table 1. Controls were healthy community volunteers from the larger Sunnybrook Dementia Study. All underwent comprehensive neuropsychological testing as well as structural MRI to ensure they were normal for their age and had no concomitant neurological or psychiatric disorders. Twelve of the normal controls were from the original study and an additional two were recruited for the present study (Mandzia et al., 2004). Amnestic MCI patients were recruited from the Cognitive

Table 1 Group demographics of participants and neuropsychological performance

Mean age, range Years of education, range MMSE CDR CDR, sum of boxes Mattis Dementia Rating Scale CVLT, acquisition* Visual reproduction, delayed recall* Boston naming* Semantic fluency* FAS Backward digit span Wisconsin card sort categories

Normal controls, N = 14 (7F/7M)

MCI, N = 14 (7F/7M)

72.2 ± 15.4 ± 28.6 ± – – 140.9 ± 50 ± 26.7 ± 28.7 ± 21.6 ± 51.3 ± 7.8 ± 3.4 ±

68.6 ± 7.4 (59–81) 13.4 ± 2.8 (10–18) 27.7 ± 1.1 0.5 1.5 ± 0.99 135.1 ± 4.3 31.6 ± 11.2 11.9 ± 8.5 25.7 ± 2.2 16.2 ± 3.5 35.3 ± 11.2 7.9 ± 3.3 2.8 ± 1.3

6.4 (63–81) 2.8 (11–19) 1.1

2.1 6 7.3 1.4 4.8 15.4 1.9 1.3

MMSE, mini mental status exam; CDR, clinical dementia rating scale; CVLT, California verbal learning test. A CDR for normal controls was not done and a presumed score of 0 was given. * Significant group differences p < 0.05 from GLM neuropsychological testing. Select test results are shown here.

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

Neurology Clinic where a standardized assessment and investigation protocol is in place to rule out secondary causes of cognitive impairment. MCI patients were categorized as such based on a structured interview, neuropsychological tests and clinical exam. They exhibited subjective memory complaints verified by an informant, with performance greater than or equal to 1 standard deviation below age and educationadjusted norms on memory tests (acquisition or delayed recall of the California Verbal Learning Test (Delis, 1987) or on long delay of logical memory from the Wechsler Memory Scale) (Wechsler, 1987), MMSE score ≥ 24 and an overall Clinical Dementia Rating (CDR) = 0.5 (Morris, 1993). It was not feasible to perform the CDR on the NC group, but by pre-assessment telephone screen they had no cognitive complaints or functional limitations. The presumed CDR score was 0. In seven MCI cases the CDR interview was administered by telephone, with the patient and informant (except for two participants who had no informant). Secondary causes of MCI such as vascular, metabolic, nutritional or mood disorders were all excluded, through the standardized work-up. 2.2. fMRI task To examine differences in memory-related activations we used a blocked design incidental encoding and recognition task (Mandzia et al., 2004). Stimuli consisted of color and black and white photographs of objects and animals, which were taken from a CD-Rom catalogue of photos and a set of digital photos obtained by J.M. All tasks contained a mixture of both color and black and white photographs. For most participants, stimuli were presented using MRI-compatible goggles (Silent Vision, Avotec, Inc., Jensen Beach, FL) that included adjustments for visual acuity, except the first five NC participants who viewed stimuli projected onto a screen at the foot of the MRI patient table. Individual runs consisted of six 31 s blocks of the “memory task” interleaved with five 26 s blocks of the fixation condition. There were five stimuli (photographs or repeated scrambled patterns) per block with a 6 s set of text instructions at the beginning of each block to remind subjects of the task. Each photograph was presented for 4 s with an interstimulus interval (ISI) of 1 s. The incidental encoding task consisted of two imaging runs differing in the level of processing. During the deep encoding run, participants decided whether each photograph they saw was “natural” or “man-made”. During the shallow encoding run, participants decided whether the photograph was “in color” or “black and white”. The baseline control condition consisted of a colored scrambled pattern, presented for 3 s with an ISI of 1 s (five stimuli per block). Following the two encoding runs, three-dimensional T1-weighted anatomical scanning (see below) was performed during the 12 min delay between encoding and recognition tasks. The deep and shallow recognition test runs followed. During these tasks participants decided whether the photograph they saw was “old” (previously seen in the encoding runs) vs. “new” (novel — not seen in previous encoding runs). The deep recognition

719

run consisted of 15 photographs from the deep encoding run and 15 new photographs. The shallow recognition run consisted of 15 photographs from the shallow encoding run and 15 new photographs. New and old stimuli were interspersed within blocks for the deep and shallow recognition runs. Both recognition conditions were also contrasted to the same baseline condition used in the encoding runs. For consistency with our previous study (Mandzia et al., 2004), the retrieval of pictures that were deeply encoded will be referred to as deep retrieval and shallow retrieval will denote the retrieval condition for shallowly encoded items. Accuracy and reaction time were recorded using an MRIcompatible keypad (Lumitouch, Lightwave Technologies, Surrey, BC). Participants indicated their choice by pressing the right button (for natural, color and old photographs) or the left button (for man-made, black and white and new photographs) with the middle or index finger of their right hand, respectively. During the fixation condition they pressed both buttons when the colored scrambled pattern appeared. Stimuli were counterbalanced for color and type of photograph (natural vs. man-made) in all four memory conditions. Deep and shallow imaging runs were counterbalanced across subjects to control for task-order related effects. 2.3. fMRI imaging protocol Experiments were conducted on a research dedicated whole-body MRI system operating at 1.5 T magnetic field strength (Signa, GE Medical Systems, Waukesha, WI; CV/I hardware, LX8.4 software). Single shot spiral imaging was performed with imaging parameters to optimize BOLD contrast (Ogawa et al., 1990) and included offline gridding, reconstruction, and correction for magnetic field inhomogeneity and Maxwell gradient terms (TR = 3 s, TE = 40 ms, 64 × 64 matrix, 30–34 slices, 3.1 × 3.1 × 5 mm) (Glover and Lai, 1998). Coronal scans were performed to maximize signal intensity in the MTL. Three-dimensional anatomical MRI scanning was performed with higher spatial resolution (fast spoiled gradient echo imaging, field of view = 22 cm × 18 cm, flip angle/TE/TR/ = 35◦ /6 ms/15 ms, 256 × 192, 128 axial slices 1.5 mm thick). These images were used as an anatomical underlay of color maps of brain activity. 2.4. fMRI analysis The fMRI data were analyzed using the Analysis of Functional NeuroImages (AFNI version 2.50) statistical package (Cox, 1996). Following motion correction with 3D image realignment (Cox and Jesmanowicz, 1999) and time series shifting by 6 s to account for hemodynamic delay, a voxel by voxel box-car correlation analysis was performed to estimate brain activation as a percentage BOLD signal change from baseline. As part of the analysis, the time series data were detrended to correct for baseline drift and activation was estimated orthogonal to each subject’s individual motion parameters

720

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

to further reduce the potential for artifacts related to head motion. Subjects were excluded if head motion was greater than 2 mm in all three planes (x, y, z) measured. A mask was created for each fMRI slice location to remove noise outside the brain and regions of low signal intensity due to susceptibility artifact. Each individual data set was transformed to Talairach space and smoothed using a 6 mm full width half maximum (FWHM) kernel to reduce the influence of individual anatomical and functional variability. Our main goal was to compare differences in fMRI activation between the NC and MCI groups on the four separate tasks. To examine group differences two-sample t-tests were performed for each memory task between NC and MCI groups. All group comparisons were displayed on a representative brain that was transformed into Talairach space. Activations were deemed significant if the threshold based on the correlation coefficient exceeded p ≤ 0.005 for an individual voxel and were within a minimum cluster volume of 164 ␮l (equivalent to three voxels in original space and to 164 voxels in Talairach space). This approach of combining thresholding and clustering minimizes false positive activations and numerical simulations indicate that for the parameters chosen the estimated overall corrected alpha corresponds to p < 0.05 (Forman et al., 1995). 2.5. Region of interest (ROI) analysis Four ROIs were drawn in Talairach space for each fMRI participant in each hemisphere: the hippocampus (all sub-regions in the hippocampal formation); and the entire parahippocampal gyrus by the same rater (JM). The subsequent ROI analysis enabled examination of group differences in these regions, and correlation between BOLD percent signal during deep encoding, as measured by percent signal change vs. fixation, and recognition performance for items which were deeply encoded. We focused on this condition for the ROI analysis, as we predicted that it would elicit the most successful encoding and subsequent retrieval performance and potentially the largest separation between groups. Regions were automatically drawn using the Talairach Daemon template in AFNI. The ROIs were also edited by creating a mask for each individual to correct for variability in brain anatomy, degree of atrophy and differences in signal dropout using both the spiral and anatomical scans. Only voxels within the ROI mask within 0.5–3.5% signal change were selected to eliminate sampling voxels which potentially were corrupted by noise. On the basis of this selection, the percentage of activated voxels within an ROI (number of voxels which met percent signal criteria/total number of voxels in the region) was calculated for each individual and this is identified as the ROI measure for the rest of our analyses. First, a multivariate general linear model (GLM) was performed to compare differences on the percentage of activated voxels and volumes between NC and MCI groups for each region in each hemisphere. Then an interaction term ROI measure × diagnosis was created and entered into each model

to examine whether the relationship was different between hit rate and the ROI measure in the NC and MCI groups. Because there was a significant interaction effect, the relationship between hit rate and ROI measure was examined separately for each group; eight models (one for each region and for each group) were tested using simple regression to determine whether our ROI measure calculated separately in each ROI could predict hit rate on the recognition task for deeply encoded items. The alpha level was corrected for multiple comparisons (0.05/8 = 0.006). 2.6. fMRI behavioural performance analysis To examine differences in behavioural performance between groups (NC vs. MCI) across tasks, a multivariate GLM analysis was conducted separately for encoding and retrieval. The dependent measures for encoding were percent correct and reaction time (RT) and for recognition were d , hit and false alarm rates and RT. d prime (d ) is a bias-free measure of recognition sensitivity.

3. Results 3.1. Neuropsychological testing results The degree of memory impairment on the CVLT in the MCI group varied from 1 S.D. below normal on long delay (n = 4) to ≥2 S.D. in nine MCI. One individual had CVLT performance within normal range for age, but performed abnormally on paragraph recall. Although not all individuals met the Petersen criteria of 1.5 S.D. below normal for age on memory testing, the intent was to cover a range of memory difficulties. General cognition and functional performance were sufficiently preserved to preclude a diagnosis of possible or probable AD (see Table 1). All MCI subjects were followed up annually for the duration of the project to monitor decline and progression to Alzheimer’s disease. Five MCI participated in a randomized clinical trial of cholinesterase inhibitors, but their treatment status is still unknown. MCI Table 2 fMRI Behavioural differences NC encoding*

Percentage of correct deep Percentage of correct shallow encoding* Reaction time (RT) deep encoding (ms)* RT shallow encoding (ms) Hit rate deep retrieval* Hit rate shallow retrieval False alarm rate deep retrieval* False alarm rate shallow retrieval* d deep retrieval* d shallow retrieval* RT deep retrieval (ms)* RT shallow retrieval (ms)* *

99.7 95.8 1022 1068 0.85 0.70 0.1 0.06 4.2 3.2 1217 1221

MCI ± ± ± ± ± ± ± ± ± ± ± ±

Significant differences between groups p ≤ 0.05.

0.96 3.80 178 278 0.16 0.2 0.09 0.07 1.3 1.6 205 189

95.1 88.1 1269 1128 0.67 0.60 0.26 0.24 1.4 1.6 1567 1634

± ± ± ± ± ± ± ± ± ± ± ±

6.8 9 203 146 0.22 0.21 0.24 0.21 1.7 1.7 205 255

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

participants have now been followed up for a mean of 6.7 [3.9–12.8] years and 11/14 (78%) have progressed to meet criteria for Alzheimer’s dementia. 3.2. Behavioural performance Multivariate GLM analyses revealed significant differences (Table 2) between NC and MCI groups on percent correct for deep (F(26) = 6, p = 0.02) and shallow (F(26) = 7.4, p = 0.01) encoding and on reaction time for deep encoding (F(26) = 11.8, p = 0.002). A separate multivariate analysis for retrieval tasks revealed group differences in hit rate only for the retrieval of deeply encoded items (F(26) = 7.2, p = 0.013) and d for both deep (F(26) = 28.1, p = 0.0001) and shallow (F(26) = 6.8, p = 0.02) conditions. The MCI group was significantly slower than NC participants for both deep (F(26) = 13, p = 0.001) and shallow retrieval (F(26) = 23.8, p = 0.0001). False alarm rate was significantly higher for both conditions in the MCI group (deep-F(26) = 23.7, p = 0.0001 and shallow-F(26) = 9 p = 0.006). A paired t-test was conducted to

721

examine levels of processing effects between groups for MCI and NC separately. Only NC participants showed a significant difference in hit rate (t = 4.32, p = 0.001) and d (t = 2.7, p = 0.02) between the two retrieval conditions, indicating superior recognition of deeply encoded stimuli. 3.3. fMRI activation results 3.3.1. Deep encoding The MCI group showed no areas of increased activation for this condition relative to the NC group. Greater temporal lobe activity in NC compared to MCI was localized to the left superior temporal gyrus extending into the left middle temporal gyrus, and also to the right middle temporal gyrus. There was no significant difference in MTL activity between groups. Relative increases in frontal activity were located in the left superior and inferior frontal (BA 47) gyri (Fig. 1). Parietal activation in NC was also greater bilaterally in the precuneus. The left cuneus and anterior cingulate, as well as the right lentiform nucleus, caudate and putamen also

Fig. 1. Regions of increased brain activity (% BOLD signal change) for the NC relative to the MCI group (p ≤ 0.05, corrected, minimum cluster volume = 164 ␮l = 3 voxels): increased L PHG—left parahippocampal gyrus (BA 35; −18 −35 −10) activity and R IFG—right inferior frontal (BA 46; 44 30 10) during shallow encoding; L IFG—left inferior frontal (BA 47; −34 34 −8) activity during deep encoding; Increased L HC—left hippocampal (−33 −23 −10) activation during deep retrieval.

722

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

Table 3 Voxel locations and volumes of areas more active in normal controls vs. MCI during encoding tasks Brain regions

Volume (␮l)

Maximum % signal intensity difference

Talairach coordinates

Temporal Left superior temporal gyrusa Left BA 34 Right middle temporal gyrus

548 146 473

0.53 0.62 0.30

−42 6 −21 −27 5 −9 43 3 −37

Frontal Left superior frontal gyrus Left superior frontal gyrus Left inferior frontal gyrus (BA 47) Left BA 47b

222 201 346 511

0.78 0.32 0.55 0.35

0 21 60 −20 25 51 −46 16 −9 −34 34 −8

Parietal Left precuneus Right precuneus

501 146

0.26 0.23

−16 −47 32 9 −50 35

Occipital Left cuneus (BA 18)

245

0.71

−1 −88 17

224 1441 254

0.22 0.46 0.33

−12 20 17 27 8 −10 11 10 −1

Shallow encoding Temporal Left parahippocampal gyrus (BA 35) Left fusiform gyrus (BA 37) Right middle temporal gyrus

263 262 150

0.32 0.79 0.30

−18 −35 −10 −47 −64 −12 53 4 −25

Frontal Right inferior frontal gyrus (BA 44) Right inferior frontal gyrus (BA 46)

809 294

0.30 0.36

44 15 14 44 30 10

Parietal Right inferior parietal lobule

478

0.23

47 −30 36

Occipital Left middle occipital gyrus Left cuneus Left cerebellum Right caudate and caudate body

794 176 176 155

0.49 0.20 0.32 0.19

−30 −90 16 −18 −83 12 −2 −54 −37 24 −21 23

Other Left anterior cingulate gyrus Right lentiform nucleus and putamen Right caudate and caudate head

Results displayed for regions of maximum % signal intensity difference between NC and MCI groups, p < 0.05 (corrected), minimum cluster volume = 164 ␮l = 3 voxels. a Includes left middle temporal gyrus (BA 21). b Includes left middle frontal gyrus (BA 10).

showed significantly greater activity in NC compared to MCI (Table 3). 3.3.2. Shallow encoding Clusters of increased activation in the NC compared to MCI group were located in the left parahippocampal gyrus (Fig. 1), fusiform and right middle temporal gyrus. There was also increased activation in the right inferior (BA 44 and 46) frontal regions (Fig. 1). Other regions of greater activation in NC included the right inferior parietal lobule (BA 40), caudate and the left cerebellum, middle occipital gyrus and cuneus (Table 3). 3.3.3. Recognition for deeply encoded items Compared to the MCI group, normal controls activated to a greater extent the left hippocampus (Fig. 1), bilateral superior and middle frontal gyri (left BA 9 and right BA 6), and

right lateral inferior and medial frontal gyri. Other regions significantly activated to a greater extent in NC included cingulate gyrus bilaterally, left thalamus and middle occipital gyrus (Table 4). 3.3.4. Recognition for shallowly encoded items NC activated to a greater extent the left lentiform nucleus and putamen (Table 4). In addition, this was the only comparison in which MCI showed more activation in any region than did the controls. The MCI group had greater activity in the left fusiform gyrus, superior frontal gyrus (BA 10) and the right cingulate gyrus (Table 4). 3.4. ROI analysis GLM multivariate analysis revealed no significant differences in percent of activated voxels between NC and MCI

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

723

Table 4 Voxel locations and volumes of areas more active in normal controls vs. MCI during deep retrieval and in MCI vs. NC in shallow retrieval Brain regions

Volume (␮l)

Maximum % signal intensity difference

Talairach coordinates

Temporal Left hippocampus

366

0.28

−33 −23 −10

Frontal Left middle frontal gyrus (BA 9) Left middle frontal gyrus Left superior frontal gyrus Right middle frontal gyrus (BA 6) Right medial frontal gyrus (BA 6) Right superior frontal gyrus Right inferior frontal gyrus

153 156 194 163 224 350 393

0.46 0.39 0.41 0.33 0.31 0.56 0.51

−52 7 37 −33 36 −5 −18 28 51 21 17 54 3 37 36 29 56 −11 48 26 12

Occipital Left middle occipital gyrus (BA 19)

298

0.36

−41 −79 22

Other Right cingulate gyrus Left cingulate gyrus

186 148

0.21 0.22

13 −9 28 −12 −13 31

Left thalamus Left lentiform nucleus and putamen

506 147

0.35 0.44

−20 −22 17 −17 4 −10

164

0.38

−22 −56 −9

223 164

0.48 0.29

−39 48 23 −12 16 48

721

0.34

20 −7 45

Deep retrieval, NC > MCI

Shallow retrieval, MCI > NC Temporal Left fusiform gyrus Frontal Left superior frontal gyrus (BA 10) Left superior frontal gyrus (medial) Other Right cingulate gyrus (within 3 mm)

Results displayed for regions of maximum % signal intensity difference between NC and MCI groups, p < 0.05 (corrected), minimum cluster volume = 164 ␮l = 3 voxels.

groups in the four regions of interest (Table 5). There was also no significant difference in the ROI volumes between NC and MCI. Recognition accuracy for deeply encoded items in NC was predicted by the percent of activated voxels separately in the left parahippocampal gyrus (PHG) (F = 14.5, p = 0.003, R2 = 0.55), left hippocampus and right PHG (F = 17.6, p = 0.001, R2 = 0.59) (Fig. 2). However, the relationship between the recognition of deeply encoded items and activation in the left hippocampus did not survive correction for multiple comparisons (F = 6.6, p = 0.024, R2 = 0.36) and as well there was no significant relationship between right hippocampal activation and recognition performance

Table 5 ROI analysis: % of activated voxels in NC and MCI during deep encoding condition NC Right hippocampus Left hippocampus Right PHG Left PHG

0.19 0.22 0.24 0.28

MCI ± ± ± ±

0.15 0.16 0.1 0.1

0.22 0.34 0.26 0.29

± ± ± ±

0.16 0.22 0.1 0.1

ROI, region of interest; PHG, parahippocampal gyrus. No significant differences between groups were found.

(F = 3.1, p = 0.1, R2 = 0.22) (Fig. 2). Individuals with MCI did not demonstrate this same predictive relationship between activation at encoding and subsequent recognition accuracy in any of these regions.

4. Discussion Individuals with MCI are at risk for progression to AD, and yet cross-sectionally MCI can be difficult to distinguish from normal aging. Functional MRI can help characterize differences in brain activity between MCI and normal aging and ultimately may have some utility in predicting conversion to AD. At this point, however, the relationship between functional activation and cognitive performance in individuals with neurologic compromise is relatively poorly understood. In the current study we characterized group and ROI differences in fMRI activation between MCI and NC groups during incidental encoding and yes/no recognition tasks by manipulating depth of encoding. As expected, the MCI group demonstrated poorer recognition memory and functional MRI revealed a pattern of reduced activation in some key brain regions that may underlie this performance difference.

724

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

Fig. 2. Regression of the percent of activated voxels during deep encoding in the right (p = 0.001) and left PHG (parahippocampal gyrus) (p = 0.003) and the right (p = 0.1) and left (p = 0.024) hippocampus with hit rate for the retrieval of deeply encoded items. Results were considered significant at p ≤ 0.006 after correction for multiple comparisons.

4.1. Encoding Comparisons revealed reduced activation in the MCI group compared to normal controls in key regions: the left parahippocampal and right inferior frontal (BA 44 and 46) gyri during shallow encoding and in the left inferior frontal gyrus (BA 47) during deep encoding. Of interest, there was no significant difference in medial temporal activation between groups for the deep encoding condition, indicating that the MCI patients can still adequately engage this region (hippocampus and PHG), as seen in the group activation patterns (not shown here) and ROI group results (Table 5). The MCI group did not show increases in either magnitude or extent of activation nor did they activate compensatory brain regions. Others have reported such increases during episodic memory tasks, but most of these studies have been done in AD patients (Grady et al., 2003; Sperling et al., 2003) or genetically at risk, asymptomatic individuals (Bookheimer et al., 2000; Fleisher et al., 2005), rather than MCI. Previous studies in MCI have shown both reduced and increased activation in comparison to the normal controls in the MTL during encoding (Dickerson et al., 2004, 2005; Johnson et al., 2004, 2006; Machulda et al., 2003; Small et al., 1999). Our study does support previous reports in individuals with comparable MCI who showed reduced activation in the

MTL region during encoding compared to normal controls, as demonstrated in our shallow encoding condition (Johnson et al., 2004, 2006; Machulda et al., 2003; Small et al., 1999). The discordance in MTL activation patterns across studies is most likely due to differences in the heterogeneity of MCI, as nicely demonstrated by a recent study which showed the spectrum of MTL activity in relation to the degree of impairment across MCI individuals (Celone et al., 2006). Furthermore, we expected the extent of activation during encoding to be correlated with retrieval success as has been shown in event-related studies of encoding (Brewer et al., 1998). Dickerson and colleagues (Dickerson et al., 2004) recently reported that greater extent of activation within the hippocampal formation and PHG correlated with better memory performance in individuals with MCI which converges with similar results found in the hippocampus from their most recent fMRI study during associative encoding (Dickerson et al., 2005). In the present study this relationship only held for controls and not for the MCI patients; we will discuss this association further, below. Together with the lack of a significant group difference in MTL activation during deep encoding despite obvious performance differences, these findings suggest that a straight-forward relationship between neuronal activity (as reflected in BOLD signal) and effectiveness of encoding operations may not exist in the damaged brain.

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

The other main difference in encoding was greater activation in prefrontal regions in the NC group. During shallow encoding this was seen in right BA 44, 46. Activity in this region has been reported in older individuals during encoding (Rosen et al., 2002) in contrast to younger adults who tend to show a left-lateralized pattern of activity (Cabeza, 2002; Grady et al., 1999). Activity in the right inferior frontal region has also been shown during encoding of pictorial stimuli in young subjects (Brewer et al., 1998; Kirchhoff et al., 2000) leading authors to suggest hemispheric specific activation of the material depending on the type of material encoded. A recent 4T fMRI study by Petrella and colleagues, who also conducted full brain analyses, also found MCI to show a reduction in bilateral prefrontal activation during the encoding of face–name pairs (Petrella et al., 2006). Individuals with MCI may not be able to mount a sufficient neural response in these regions leading to ineffective encoding; we can speculate that there may be a ‘functional disconnection’ between the MTL and inferior frontal regions (Salat et al., 2001). This is in contrast to older adults who have been found to demonstrate an increased, ‘compensatory’ response in prefrontal regions during memory processing (Cabeza et al., 2002). During deep encoding, we found increased left inferior frontal activation for NC compared to MCI. This region (BA 47) is important for semantic elaboration and effective encoding (Bookheimer, 2003; Wagner et al., 1998). Other studies have reported a reduction in inferior frontal activation during encoding tasks in AD patients (Corkin, 1998; Rombouts et al., 1998; Sperling et al., 2003). In contrast, Bookheimer et al. (2000) reported hyperactivation in the prefrontal region in asymptomatic APOE e4 carriers during the encoding of word pairs, but that may be a capacity that symptomatic individuals can no longer engage. The inferior, medial and orbital frontal cortices may be more vulnerable in terms of AD pathology as they have strong connections to the MTL and limbic regions, resulting in both anterograde and retrograde degeneration (Salat et al., 2001). These regions may be affected then in MCI, as pathology in the MTL is probably already underway at least in some MCI patients (DeKosky et al., 2002; Kordower et al., 2001; Van Der Flier et al., 2002a). While no gross atrophy was present, structural or functional damage may have begun as shown in individuals with MCI who have changes in magnetic transfer ratios (MTR) in the frontal lobes compared to a NC group (Kabani et al., 2002; Van Der Flier et al., 2002b). Difficulty during encoding in activating this important region, which is involved in the process of retrieving semantic information and drawing comparisons between semantic concepts (Martin and Chao, 2001), also may have contributed to the inability to form a stable memory trace. Other left-lateralized regions, which are part of the semantic network (Vandenberghe et al., 1996), such as the middle and superior temporal regions, as well as the left fusiform gyrus were activated to a greater extent in the NC vs. MCI group. The homologous right middle temporal gyrus was also more activated. These regions are important for visual perceptual processing (e.g. left fusiform) (Murtha et al., 1999)

725

and contribute to higher order semantic processing, such as semantic integration, which is believed to take place in more anterior temporal regions (Martin and Chao, 2001). Structural damage in these regions has been found to be a significant predictor of progression of MCI to AD (Convit et al., 2000; Killiany et al., 2000). These temporal regions may in fact be the first isocortical regions to be affected early in the AD degenerative process and may already be showing some AD pathology in MCI subjects (Braak and Braak, 1995). Our MCI group performed worse than NC on naming and fluency tests, but their scores were not sufficient to be considered impaired relative to normal limits (see Table 1). At this stage, MCI can recruit these critical regions for semantic processing, but perhaps not as effectively as NC, as demonstrated by a reduction in activity in these regions in MCI. Longitudinal studies, which have tracked neuropsychological performance in MCI over time, have found that language and visuospatial functions do become affected after memory (Lambon et al., 2003). Specifically, semantic memory, in terms of person naming, seems to be quite vulnerable in the pre-dementia stage (Dudas et al., 2005), which was not examined in our study. 4.2. Does activity in the MTL predict subsequent recognition? Several previous studies have shown that activity in the parahippocampal gyrus during encoding is predictive of recognition performance (hit rate) for words (Wagner et al., 1998) and pictures (Brewer et al., 1998). The optimal way to assess this relationship is using an event-related fMRI design, which allows one to examine encoding activity partitioned on the basis of subsequent recognition success (i.e., hits vs. correct rejections or misses). Although this was not possible given our blocked design, a between-subjects correlation such as has been used in PET studies does afford an examination of the relationship between regional activation magnitude and overall recognition success. Using this strategy we found that the percentage of activated voxels during deep encoding in a group of MTL regions predicted hit rate during the retrieval of deeply encoded items, but only for the NC group. The association was strongest in the right PHG, followed by the left PHG, and weakly in the left hippocampus (i.e. not significant after correction for multiple comparisons). Thus, successful encoding in NC appears to be related to bilateral activity in the PHG region. The PHG is important in the initial formation of memories, as it receives multiple neorcortical inputs, is involved in integrating this information, and relaying it to the hippocampus (Lavenex and Amaral, 2000). Our PHG ROI did not distinguish between the various sub-regions of this structure and therefore we cannot comment on the role of the entorhinal or perirhinal cortices per se in memory encoding. Bilateral PHG activity here may be related to the complexity of the pictures and the fact that they were also highly verbalizable, as studies have shown different lateralization in the MTL depending on the type of material (Kelley et al., 1998).

726

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

The hippocampus is believed to play a conjunctive role in memory encoding whereby it binds together the different facets and features that make up an episodic experience (Eichenbaum, 2001; Suzuki and Eichenbaum, 2000). Neuroimaging studies support a role for hippocampal activity for successful encoding (Davachi and Wagner, 2002; Fernandez et al., 1998; Haxby et al., 1996; Kirchhoff et al., 2000; Morcom et al., 2003; Otten et al., 2001; Stern et al., 1996; Strange et al., 2002). We found that the percent of activated voxels in the left hippocampus only weakly predicted subsequent recognition for deeply encoded items in NC. This may be related to the variability in activation of this region whereby a larger sample would have increased our ability to show significance. It is surprising that the MCI group did not show a parallel relationship between ROI activity and recognition success for deeply encoded items; nor did they differ from NCs in terms of the percent of activated voxels. This may be due to variability in the ROI measures and/or hit rate in the MCI group. Nonetheless, it suggests that activity in these regions in MCI is not a sufficient condition for successful recognition. Dickerson et al. (2004) found a significant relationship between extent of activation in the PHG and hippocampus in MCI subjects, similar to the one we found in our normal control group, but their sample size was more than double of ours (n = 32), and their group of patients had milder memory deficits (Dickerson et al., 2004). In addition their MCI group was recruited through the community, whereas our MCI subjects came to our attention through our memory clinic. Their methodology was also different than our study where their ROI measure selected voxels based on a predetermined threshold, and they did not divide by the total number of voxels within the ROI, as we did. Rather they included the ROI volume as a separate variable in their analysis. In contrast, another group of investigators suggested that reduced activity in ERC during encoding may be prognostic for conversion, but this has yet to be replicated (Small et al., 1999). MTL activity may represent a response to brain pathology that is presumably occurring in some individuals with MCI in specific regions, for example the ERC (DeKosky et al., 2002), and therefore may be related more broadly to clinical features than to the memory task at hand. As there are so few studies addressing this issue, coupled with the presumed heterogeneity of MCI samples within and between studies, the precise relationship between MTL activity and memory performance in affected individuals and its prognostic value with regards to conversion will require considerable further study. 4.3. Retrieval task comparisons The MCI group performed significantly worse than NC on both deep retrieval and shallow retrieval based on hit rate and d measures. Furthermore, the NC group benefited from the deeper encoding whereas the MCI group did not. MCI were also slower than NC in both retrieval conditions.

We had previously shown increases in right prefrontal cortex and left hippocampus during deep vs. shallow recognition in healthy controls (Mandzia et al., 2004) and suggested that activity in those regions was likely to be a signature of successful retrieval. Thus, we expected to see reduced activation in these areas for MCI relative to NC. That was clearly the pattern for the contrast involving recognition of deeply encoded items where NC showed greater activation in the left hippocampus, right inferior and bilateral prefrontal region. Of note, performance was considerably worse in the MCI group for this condition, whereas when recognition was more equivalent (as in shallow recognition), there was no significant difference in activity in these regions. Petrella et al’s (2006) recent fMRI study also showed reduced left hippocampal and bilateral prefrontal activation during retrieval of face–name association pairs and they report that this difference remained when they used covariance analyses to control for group differences in recognition accuracy (Petrella et al., 2006). Although the overall pattern of findings is compatible with the concept that hippocampal activation at recognition is associated with retrieval success (Eldridge et al., 2000; Nyberg et al., 1996; Stark and Squire, 2000, 2001), the design of our recognition task precludes analysis of the specific processes involved. As old and new items were intermixed in our task blocks, we cannot ascertain whether any differences between conditions or groups reflects one or a combination of processes such as familiarity, recollection or novelty detection (Daselaar et al., 2006) or in some more non-specific aspect such as retrieval ‘mode’ or ‘set’. Nonetheless, there is clearly a relationship between MTL activation effects and recognition accuracy across groups in our data that parallels what has been established in other studies and in our previous study contrasting deep and shallow recognition conditions in controls (Mandzia et al., 2004). Increased prefrontal activation has also been shown to be correlated with enhanced retrieval in AD patients (Corkin, 1998; Grady et al., 2003) and asymptomatic genetically at risk individuals (Bookheimer et al., 2000). In our case enhanced prefrontal activation appears to have benefited the normal control group in the deep retrieval condition. Fewer studies have examined retrieval differences in MCI (Johnson et al., 2006; Petrella et al., 2006; Ries et al., 2006). Another looked more broadly at specific regions that were selected based on findings from a young normative sample, and there did not appear to be any differences between MCI and controls in prefrontal activation differences between the two groups (Johnson et al., 2006). Although performance was similar for the two groups in that study, Petrella’s findings of reduced prefrontal activation were apparent irregardless of control for performance (Petrella et al., 2006). In our study recognition was superior in the NC group during deep retrieval and therefore we cannot exclude the possibility that differences in prefrontal activity may reflect non-mnemonic processes such as effort and attention which may contribute to successful memory performance.

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

Of interest, the shallow retrieval condition was the only one in which we found certain regions activated to a greater extent in MCI than in NC. These regions included the left fusiform gyrus and superior frontal gyrus and the right cingulate gyrus. The significance of greater activation of these regions in MCI is unclear, but may be a reflection of increased allocation of attention (right cingulate gyrus), and executive resources (left superior frontal) to carry out the task. Greater visual processing or reactivation of the object representation may also be suggested by the increased left fusiform gyrus activation relative to the NC group. It is possible that such increased activation reflected a compensatory function that yielded more equivalent performance between the groups, although we do not have direct evidence for this. 4.4. Limitations and caveats Functional MRI may offer promise as a novel technique to elucidate neurocognitive dysfunction, but there are considerable limitations to use and interpretation of findings in patient populations. For one, the effect of neurodegeneration on neurovascular coupling that underlies the BOLD response remains unclear (D’Esposito et al., 2003). Rombouts et al. (2005) recently demonstrated the complexity of hemodynamics in patient populations such as MCI where it may be the case of a delay in the BOLD response rather than a reduction in comparison to normal controls during activation tasks. Post-mortem analyses in AD brains reveal capillary degeneration, which does not correlate with Braak and Braak pathology staging, suggesting these changes may not be the consequence of neuronal pathology, but rather part of the pathology spectrum (de la Torre, 2002). If the cerebral vasculature is affected in neurodegenerative diseases such as AD and MCI (when it is prodromal AD), then the coupling between neurons and capillaries may be compromised raising questions about the basis of the fMRI technique, as neuronal activity is inferred through changes in cerebral blood flow and oxygenation. Functional MRI studies are also extremely sensitive to head motion, which is a real challenge in patient populations. In our sample, head motion was more problematic in the MCI than NC group, necessitating that data be discarded in 15% of our original sample. For our study, an event-related analysis may have been more appropriate to examine brain–behaviour correlations and differences especially at retrieval, as trials could have been sorted based on recognition success, whereby only accurate trials could be used to characterize regional activation, as well activation related to the familiarity (old vs. new) of the events could have been separated. However, event-related designs can be more difficult to implement in patient populations due to longer scanning sessions and because the hemodynamic response requires more careful characterization, which itself may be affected by the disease. Also, there were performance differences between two groups, and this may have confounded our activation analysis. The inclusion of RT and/or a measure of accuracy in our fMRI analysis

727

would have allowed us to tease out the effects of performance of brain activation and may have revealed greater differences between the groups during retrieval conditions. Differences in performance at time of encoding may have also influenced accuracy at retrieval. We do not have a good explanation why MCI subjects performed significantly worse on the shallow encoding task, but this was also found in our normal control subjects (Mandzia et al., 2004). Another possible limitation associated with fMRI scanning and analysis was the ROI method used to detect the brain–behaviour relationship between hit rate and the number of activated voxels in this region. The ROI, drawn using the semi-quantitative program from AFNI, and edited individually to account for atrophy and variability in anatomy, may not have had sufficient precision. Further error may have been introduced by tracing on brains which had been transformed into Talairach space. Another published study defined ROIs in Talairach space which were generated by using average anatomical data (Fleisher et al., 2005), whereas our ROIs were drawn on each individual brain. Krishnan et al. (2006) demonstrated that normalization is less accurate in the hippocampus in individuals such as MCI who have MTL atrophy relative to normal controls, as compared to ROI manual tracing of this region. A compromise was made between reliability and efficiency and such procedures can be better developed and implemented if this method does prove to have clinical utility. Last in our NC group and a sub-set of MCI (n = 7) the CDR was administered by telephone. We felt this was only a mild concern, as our NC and MCI were carefully screened by an experienced neurologist (SEB) at baseline and all subsequent evaluations. Even though fMRI in the past decade has become more available across centers, an experienced support system must be in place for an fMRI study to be properly conducted and analyzed. At the current state of development, this technique will be challenging to implement as a standard screening technique across even academic centers. Our sample size did not permit inferences concerning the utility of fMRI in identifying those at risk of progression to AD. Nevertheless, fMRI can play an important role in understanding the neural correlates underlying the memory deficits that affect individuals with dementing disorders and their relationship to normal memory.

5. Conclusion To our knowledge this is one of the first studies to characterize fMRI activation and behavioural differences using a combination of tasks which manipulated the depth of encoding in MCI and normal aging. During encoding MCI participants can activate the MTL but may not do so unless directed toward the meaning of stimuli. Even when so directed, and with adequate MTL activation during encoding, they do not show a strong relationship between regional activity and subsequent retrieval success unlike controls.

728

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

Nonetheless, MTL activation during recognition does show a rough parallel with success in that group differences in signal tracked group differences in performance. Clearly, the relationship between activation in a damaged region and on-line performance is complex, especially given the heterogeneous nature of MCI presentation even when patients are selected for ‘pure’ memory impairment. Another important result from this study is that patients tended to show weaker activation of inferior prefrontal regions which are known to play important roles in facilitating the commitment of information to memory and its subsequent retrieval. Reduction in inferior frontal activation in MCI may be the result of pathological changes beginning to occur in this region or may be due to the degeneration of connections between the MTL and the inferior frontal lobe. Examining effective connectivity between these regions will likely provide important clues toward assessing the deterioration in the memory encoding/retrieval system in MCI and predicting progression or conversion to AD.

Disclosure statement All authors have identified no actual or potential conflicts of interest including financial, personal or other relationships with other people or organizations within three years of beginning the work submitted that could inappropriately influence (bias) their work.

Acknowledgements This study was supported by MRC/CIHR grant # 13129 (SEB) and Alzheimer Society of Canada Training award and OSOTF Scace Graduate Scholarship in Alzheimer’s disease (JLM).

References Bookheimer, S., 2003. Functional MRI of language: new approaches to understanding the cortical organization of semantic processing. Annu. Rev. Neurosci. 25, 151–188. Bookheimer, S.Y., Strojwas, M.H., Cohen, M.S., Saunders, A.M., PericakVance, M.A., Mazziotta, J.C., Small, G.W., 2000. Patterns of brain activation in people at risk for Alzheimer’s disease. New Engl. J. Med. 343, 450–456. Braak, H., Braak, E., 1995. Staging of Alzheimer’s disease-related neurofibrillary changes. Neurobiol. Aging 16, 271–278. Brewer, J.B., Zhao, Z., Desmond, J.E., Glover, G.H., Gabrielli, J.D., 1998. Making memories: brain activity that predicts how well visual experience will be remembered. Science 281, 1185–1187. Cabeza, R., 2002. Hemispheric asymmetry reduction in older adults: the HAROLD model. Psychol. Aging 17, 85–100. Cabeza, R., Anderson, N.D., Locantore, J.K., McIntosh, A.R., 2002. Aging gracefully: compensatory brain activity in high-performing older adults. Neuroimage 17, 1394–1402. Celone, K.A., Calhoun, V.D., Dickerson, B.C., Atri, A., Chua, E.F., Miller, S.L., DePeau, K., Rentz, D.M., Selkoe, D.J., Blacker, D., Albert, M.S.,

Sperling, R.A., 2006. Alterations in memory networks in mild cognitive impairment and Alzheimer’s disease: an independent component analysis. J. Neurosci. 26, 10222–10231. Convit, A., de Asis, J., de Leon, M.J., Tarshish, C.Y., De Santi, S., Rusinek, H., 2000. Atrophy of the medial occipitotemporal, inferior, and middle temporal gyri in non-demented elderly predict decline to Alzheimer’s disease. Neurobiol. Aging 21, 19–26. Corkin, S., 1998. Functional MRI for studying episodic memory in aging and Alzheimer’s disease. Geriatrics 53, S13–S15. Cox, R.W., 1996. AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput. Biomed. Res. 29, 162–173. Cox, R.W., Jesmanowicz, A., 1999. Real-time 3D image registration for functional MRI. Magn. Reson. Med. 42, 1014–1018. Craik, F.I.M., Lockhart, R.S., 1972. Levels of processing: a framework for memory research. J. Verbal Learn. Verbal Behav. 11, 671–684. D’Esposito, M., Deouell, L.Y., Gazzaley, A., 2003. Alterations in the BOLD fMRI signal with ageing and disease: a challenge for neuroimaging. Nat. Rev. Neurosci. 4, 863–872. Daselaar, S.M., Fleck, M.S., Cabeza, R., 2006. Triple dissociation in the medial temporal lobes: recollection, familiarity, and novelty. J. Neurophysiol. 96, 1902–1911. Davachi, L., Wagner, A.D., 2002. Hippocampal contributions to episodic encoding: insights from relational and item-based learning. J. Neurophysiol. 88, 982–990. de la Torre, J.C., 2002. Alzheimer disease as a vascular disorder: nosological evidence. Stroke 33, 1152–1162. DeKosky, S.T., Ikonomovic, M.D., Styren, S.D., Beckett, L., Wisniewski, S., Bennett, D.A., Cochran, E.J., Kordower, J.H., Mufson, E.J., 2002. Upregulation of choline acetyltransferase activity in hippocampus and frontal cortex of elderly subjects with mild cognitive impairment. Ann. Neurol. 51, 145–155. Delis, D.C., 1987. California Verbal Learning Test. The Psychological Corporation, New York. Dickerson, B.C., Salat, D.H., Bates, J.F., Atiya, M., Killiany, R.J., Greve, D.N., Dale, A.M., Stern, C.E., Blacker, D., Albert, M.S., Sperling, R.A., 2004. Medial temporal lobe function and structure in mild cognitive impairment. Ann. Neurol. 56, 27–35. Dickerson, B.C., Salat, D.H., Greve, D.N., Chua, E.F., Rand-Giovannetti, E., Rentz, D.M., Bertram, L., Mullin, K., Tanzi, R.E., Blacker, D., Albert, M.S., Sperling, R.A., 2005. Increased hippocampal activation in mild cognitive impairment compared to normal aging and AD. Neurology 65, 404–411. Dudas, R.B., Clague, F., Thompson, S.A., Graham, K.S., Hodges, J.R., 2005. Episodic and semantic memory in mild cognitive impairment. Neuropsychologia 43, 1266–1276. Eichenbaum, H., 2001. The hippocampus and declarative memory: cognitive mechanisms and neural codes. Behav. Brain Res. 127, 199–207. Eldridge, L.L., Knowlton, B.J., Furmanski, C.S., Bookheimer, S.Y., Engel, S.A., 2000. Remembering episodes: a selective role for the hippocampus during retrieval. Nat. Neurosci. 3, 1149–1152. Fernandez, G., Weyerts, H., Schrader-Bolsche, M., Tendolkar, I., Smid, H., Tempelman, C., Hinrichs, H., Scheich, H., Elger, C., Mangun, G.R., Heinze, H.J., 1998. Successful verbal encoding into episodic memory engages posterior hippocampus. A parametrically analyzed functional magnetic resonance imaging study. J. Neurosci. 18, 1841–1847. Fleisher, A.S., Houston, W.S., Eyler, L.T., Frye, S., Jenkins, C., Thal, L.J., Bondi, M.W., 2005. Identification of Alzheimer disease risk by functional magnetic resonance imaging. Arch. Neurol. 62, 1881–1888. Forman, S.D., Cohen, J.D., Fitzgerald, M., Eddy, W.F., Mintun, M.A., Noll, D.C., 1995. Improved assessment of significant activation in functional magnetic resonance imaging (fMRI): use of a cluster-size threshold. Magn. Reson. Med. 33, 636–647. Glover, G.H., Lai, S., 1998. Self-navigated spiral fMRI: interleaved versus single-shot. Magn. Reson. Med. 39, 361–368. Grady, C.L., McIntosh, A.R., Beig, S., Keightley, M.L., Burian, H., Black, S.E., 2003. Evidence from functional neuroimaging of a compensatory prefrontal network in Alzheimer’s disease. J. Neurosci. 23, 986–993.

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730 Grady, C.L., McIntosh, A.R., Rajah, M.N., Beig, S., Craik, F.I., 1999. The effects of age on the neural correlates of episodic encoding. Cereb. Cortex 9, 805–814. Haxby, J.V., Ungerleider, L.G., Hortwitz, B., Maisog, J.M., Rapoport, S.I., Grady, C.L., 1996. Face encoding and recognition in the human brain. Proc. Natl. Acad. Sci. U.S.A. 93, 922–927. Johnson, S.C., Baxter, L.C., Susskind-Wilder, L., Connor, D.J., Sabbagh, M.N., Caselli, R.J., 2004. Hippocampal adaptation to face repetition in healthy elderly and mild cognitive impairment. Neuropsychologia 42, 980–989. Johnson, S.C., Schmitz, T.W., Moritz, C.H., Meyerand, M.E., Rowley, H.A., Alexander, A.L., Hansen, K.W., Gleason, C.E., Carlsson, C.M., Ries, M.L., Asthana, S., Chen, K., Reiman, E.M., Alexander, G.E., 2006. Activation of brain regions vulnerable to Alzheimer’s disease: the effect of mild cognitive impairment. Neurobiol. Aging 27, 1604–1612. Kabani, N.J., Sled, J.G., Shuper, A., Chertkow, H., 2002. Regional magnetization transfer ratio changes in mild cognitive impairment. Magn. Reson. Med. 47, 143–148. Kelley, W.M., Miezin, F.M., McDermott, K.B., Buckner, R.L., Raichle, M.E., Cohen, N.J., Ollinger, J.M., Akbudak, E., Conturo, T.E., Snyder, A.Z., Petersen, S.E., 1998. Hemispheric specialization in human dorsal frontal cortex and medial temporal lobe for verbal and nonverbal memory encoding. Neuron 20, 927–936. Killiany, R.J., Gomez-Isla, T., Moss, M., Kikinis, R., Sandor, T., Jolesz, F., Tanzi, R., Jones, K., Hyman, B.T., Albert, M.S., 2000. Use of structural magnetic resonance imaging to predict who will get Alzheimer’s disease. Ann. Neurol. 47, 430–439. Kirchhoff, B.A., Wagner, A.D., Maril, A., Stern, C.E., 2000. Prefrontal–temporal circuitry for episodic encoding and subsequent memory. J. Neurosci. 20, 6173–6180. Kordower, J.H., Chu, Y., Stebbins, G.T., DeKosky, S.T., Cochran, E.J., Bennett, D., Mufson, E.J., 2001. Loss and atrophy of layer II entorhinal cortex neurons in elderly people with mild cognitive impairment. Ann. Neurol. 49, 202–213. Krishnan, S., Slavin, M.J., Tran, T.T., Doraiswamy, P.M., Petrella, J.R., 2006. Accuracy of spatial normalization of the hippocampus: implications for fMRI research in memory disorders. Neuroimage 31, 560– 571. Lambon, R.M., Patterson, K., Graham, N., Dawson, K., Hodges, J.R., 2003. Homogeneity and heterogeneity in mild cognitive impairment and Alzheimer’s disease: a cross-sectional and longitudinal study of 55 cases. Brain 126, 2350–2362. Lavenex, P., Amaral, D.G., 2000. Hippocampal–neocortical interaction: a hierarchy of associativity. Hippocampus 10, 420–430. Machulda, M.M., Ward, H.A., Borowski, B., Gunter, J.L., Cha, R.H., O’Brien, P.C., Petersen, R.C., Boeve, B.F., Knopman, D., Tang-Wai, D.F., Ivnik, R.J., Smith, G.E., Tangalos, E.G., Jack, C.R.J., 2003. Comparison of memory fMRI response among normal, MCI, and Alzheimer’s patients. Neurology 61, 500–506. Mandzia, J.L., Black, S.E., McAndrews, M.P., Grady, C., Graham, S., 2004. fMRI differences in encoding and retrieval of pictures due to encoding strategy in the elderly. Hum. Brain Mapp. 21, 1–14. Martin, A., Chao, L.L., 2001. Semantic memory and the brain: structure and processes. Curr. Opin. Neurobiol. 11, 194–201. Morcom, A.M., Good, C.D., Frackowiak, R.S., Rugg, M.D., 2003. Age effects on the neural correlates of successful memory encoding. Brain 126, 213–229. Morris, J.C., 1993. The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology 43, 2412–2414. Morris, J.C., Storandt, M., Miller, J.P., McKeel, D.W., Price, J.L., Rubin, E.H., Berg, L., 2001. Mild cognitive impairment represents early-stage Alzheimer disease. Arch. Neurol. 58, 397–405. Murtha, S., Chertkow, H., Beauregard, M., Evans, A., 1999. The neural substrate of picture naming. J. Cogn. Neurosci. 11, 399–423. Nyberg, L., McIntosh, A.R., Houle, S., Kilsson, L.G., Tulving, E., 1996. Activation of the medial temporal structures during episodic memory retrieval. Nature 380, 715–717.

729

Ogawa, S., Lee, T.M., Kay, A.R., Tank, D.W., 1990. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc. Natl. Acad. Sci. U.S.A. 87, 9868–9872. Otten, L.J., Henson, R.N., Rugg, M.D., 2001. Depth of processing effects on neural correlates of memory encoding: relationship between findings from across- and within-task comparisons. Brain 124, 399– 412. Petersen, R.C., Doody, R., Kurz, A., Mohs, R.C., Morris, J.C., Rabins, P.V., Ritchie, K., Rossor, M., Thal, L., Winblad, B., 2001. Current concepts in mild cognitive impairment. Arch. Neurol. 58, 1985– 1992. Petersen, R.C., Thomas, R.G., Grundman, M., Bennett, D., Doody, R., Ferris, S., Galasko, D., Jin, S., Kaye, J., Levey, A., Pfeiffer, E., Sano, M., van Dyck, C.H., Thal, L.J., 2005. Vitamin E and donepezil for the treatment of mild cognitive impairment. New Engl. J. Med. 352, 2379–2388. Petrella, J.R., Krishnan, S., Slavin, M.J., Tran, T.T., Murty, L., Doraiswamy, P.M., 2006. Mild cognitive impairment: evaluation with 4-T functional MR imaging. Radiology 240, 177–186. Ries, M.L., Schmitz, T.W., Kawahara, T.N., Torgerson, B.M., Trivedi, M.A., Johnson, S.C., 2006. Task-dependent posterior cingulate activation in mild cognitive impairment. Neuroimage 29, 485–492. Rombouts, S.A., Barkhof, F., Goekoop, R., Stam, C.J., Scheltens, P., 2005. Altered resting state networks in mild cognitive impairment and mild Alzheimer’s disease: an fMRI study. Hum. Brain Mapp. 26, 1075–1085. Rombouts, S.A.R.B., Lazeron, R.H.C., Barkhof, F., Machielsen, W.C.M., Witter, M.P., Sprenger, J., Scheltens, Ph., 1998. Comparison of medial temporal lobe activation in young and elderly subjects and patients with Alzheimer’s disease. Neuroimage, S520. Rosen, A.C., Prull, M.W., O’Hara, R., Race, E.A., Desmond, J.E., Glover, G.H., Yesavage, J.A., Gabrieli, J.D., 2002. Variable effects of aging on frontal lobe contributions to memory. Neuroreport 13, 2425–2428. Salat, D.H., Kaye, J.A., Janowsky, J.S., 2001. Selective preservation and degeneration within the prefrontal cortex in aging and Alzheimer disease. Arch. Neurol. 58, 1403–1408. Saykin, A.J., Wishart, H.A., Rabin, L.A., Flashman, L.A., McHugh, T.L., Mamourian, A.C., Santulli, R.B., 2004. Cholinergic enhancement of frontal lobe activity in mild cognitive impairment. Brain 127, 1574–1583. Small, S.A., Perera, G.M., Delapaz, R., Mayeux, R., Stern, Y., 1999. Differential regional dysfunction of the hippocampal formation among elderly with memory decline and Alzheimer’s disease. Ann. Neurol. 45, 466–472. Sperling, R.A., Bates, J.F., Chua, E.F., Cocchiarella, A.J., Rentz, D.M., Rosen, B.R., Schacter, D.L., Albert, M.S., 2003. fMRI studies of associative encoding in young and elderly controls and mild Alzheimer’s disease. J. Neurol. Neurosurg. Psychiatry 74, 44–50. Stark, C.E., Squire, L.R., 2001. Simple and associative recognition memory in the hippocampal region. Learn. Memory 8, 190–197. Stark, C.E., Squire, L.R., 2000. Functional magnetic resonance imaging (fMRI) activity in the hippocampal region during recognition memory. J. Neurosci. 20, 7776–7781. Stern, C.E., Corkin, S., Gonzalez, R.G., Guimaraes, A.R., Baker, J.R., Jennings, P.J., Carr, C.A., Sugiura, R.M., Vedantham, V., Rosen, B.R., 1996. The hippocampal formation participates in novel picture encoding: evidence from functional magnetic resonance imaging. Proc. Natl. Acad. Sci. U.S.A. 93, 8660–8665. Strange, B.A., Otten, L.J., Josephs, O., Rugg, M.D., Dolan, R.J., 2002. Dissociable human perirhinal, hippocampal, and parahippocampal roles during verbal encoding. J. Neurosci. 22, 523–528. Suzuki, W.A., Eichenbaum, H., 2000. The neurophysiology of memory. Ann. N.Y. Acad. Sci. 911, 175–191. Van Der Flier, W.M., Van Den Heuvel, D.M., Weverling-Rijnsburger, A.W., Bollen, E.L., Westendorp, R.G., Van Buchem, M.A., Middelkoop, H.A., 2002a. Magnetization transfer imaging in normal aging, mild cognitive impairment, and Alzheimer’s disease. Ann. Neurol. 52, 62–67. Van Der Flier, W.M., Van Den Heuvel, D.M., Weverling-Rijnsburger, A.W., Spilt, A., Bollen, E.L., Westendorp, R.G., Middelkoop, H.A.,

730

J.L. Mandzia et al. / Neurobiology of Aging 30 (2009) 717–730

Van Buchem, M.A., 2002b. Cognitive decline in AD and mild cognitive impairment is associated with global brain damage. Neurology 59, 874–879. Vandenberghe, R., Price, C., Wise, R., Josephs, O., Frackowiak, R.S., 1996. Functional anatomy of a common semantic system for words and pictures. Nature 383, 254–256.

Wagner, A.D., Shacter, D.L., Rotte, M., Koutstaal, W., Maril, A., Dale, A.M., Rosen, B.R., Buckner, R.L., 1998. Building memories: remembering and forgetting of verbal experiences as predicted by brain activity. Science 281, 1188–1191. Wechsler, D., 1987. Wechsler Memory Scale-Revised. The Psychological Corporation, New York.

Neural correlates of incidental memory in mild cognitive ...

Available online 25 October 2007. Abstract. Behaviour ... +1 416 480 4551; fax: +1 416 480 4552. ..... hit and false alarm rates and RT. d prime (d ) is a bias-free.

602KB Sizes 29 Downloads 305 Views

Recommend Documents

Severity of mild cognitive impairment in early ...
E-mail address: [email protected] (R.A. Lawson). Contents lists available at ... Participants were recruited from community and outpatient clinics through general practitioners ... advertising to provide normative data. 2.2. Scales and ...

Neural correlates of symbolic number processing in children and ...
Neural correlates of symbolic number processing in children and adults.pdf. Neural correlates of symbolic number processing in children and adults.pdf. Open.

Orthographic dependency in the neural correlates of reading-evidence ...
Orthographic dependency in the neural correlates of rea ... nce from audiovisual integration in English readers.pdf. Orthographic dependency in the neural ...

Neural correlates of symbolic number processing in children and ...
Neural correlates of symbolic number processing in children and adults.pdf. Neural correlates of symbolic number processing in children and adults.pdf. Open.

Predictors of progression from mild cognitive impairment ... - Neurology
May 8, 2007 - Stockholm, Sweden; and Unit of Neurology and Rehabilitation (R.M.), Laboratory of Epidemiology and Psychology of Aging and Dementia,. DiNOOP, University of Palermo, Italy. Supported by grants from the Swedish Council for Working Life an

Processing complex pseudo-words in mild cognitive ...
Nov 20, 2015 - Submit your article to this journal. Article views: 2. View related articles · View Crossmark data ...... 1843 (360). 1713 (330). 1282 (211). Control.

THE AMBiGUiTiES OF MiLD COGNiTivE IMPAiRMENT
Thus if MCI does turn out to have a predictively valid connection to Alzheimer's, there is a way of resolving the question of its pathologic status. It can inherit a ...

Generation of novel motor sequences-the neural correlates of musical ...
Generation of novel motor sequences-the neural correlates of musical improvisation.pdf. Generation of novel motor sequences-the neural correlates of musical ...

Neural correlates of observing joint actions with ... - Floris de Lange
Jun 5, 2015 - (>5 mm) or other technical difficulties at the time of data collection. .... of paired t-tests of contrast images to analyze the statistical difference ...

Neural correlates of risk prediction error during ...
12 May 2009 - decreases (because it is likely that the second card will be below 9). So there is a positive reward prediction error ... condition, payoffs were displayed on each card so participants knew the outcome in advance (Fig. 1, right). Blocks

The neural correlates of verb and noun processing
formed into a standard stereotactic space; global differences in cerebral blood flow were covaried out for all voxels and comparisons across conditions were made ..... the monkey ventral area 6 where 'mirror' neurons were recorded (Rizzolatti et al.,

Neural correlates of the chronic fatigue syndrome—an ...
2560 ms; 32 axial slices, slice thickness = 3.5 mm; FOV (field of view). = 224 mm]. .... in error detection and the online monitoring of performance. (Carter et al.

Neural correlates of risk prediction error during reinforcement ... - Lobes
May 12, 2009 - 2. Mean payoff and standard error (SE) by deck for the four versions of the Iowa Gambling Task. Each deck consisted of 60 cards. Good decks are in green and bad .... p=0.04 (one-sided test).2 The risk preference parameter was not .....

Neural Correlates of Envy and Schadenfreude
Feb 13, 2009 - The following resources related to this article are available online at ... results (28) that intraperitoneal (ip) administration of DMT (2 mg per kilogram ..... Bioinformatics, Graduate School of Health Sciences, Tokyo. Medical and ..

Neural correlates of the chronic fatigue syndrome—an ...
education-matched healthy controls participated in the study after giving written informed .... nificantly higher levels of fatigue, experienced more concentration ...

The Neural Basis of Relational Memory Deficits in ...
2006;63:356-365 ... gions previously found to support transitive inference in .... participants were shown pairs of visual items on a computer screen and asked to ...

Searching for an Oddball: Neural Correlates of ...
1Institute of Experimental Psychology, Heinrich-Heine-University Düsseldorf, 40225 Düsseldorf, Germany, and 2Systems Neurobiology Laboratory, Salk. Institute for Biological Studies, ... gle red apple among a pile of lemons, as depicted in Figure 1A

Neural Correlates of Integral and Separable Processing ...
the emotion expressed by the faces (integral priming). Neural Correlates of Integral and Separable Processing During Category Learning. Catherine Hanson Stephen José Hanson Tara Schweighardt. Rutgers University, Newark, NJ. Categorization Tasks. Fol

Neural correlates of task and source switching - Semantic Scholar
Jan 20, 2010 - them wait for a fixed duration between each letter/response. Furthermore ...... is of course possible that the neural activity that may have ..... Ridderinkhof, K.R., van den Wildenberg, W.P., Segalowitz, S.J., Carter, C.S., 2004.

Neural correlates of symbolic and nonsymbolic arithmetic (2005).pdf ...
Imaging and image analysis ... each subject were prepro- cessed and analyzed using BrainVoyager 2000 software ... Fixed-effect analysis at the group level.

Neural Correlates of Envy and Schadenfreude
Feb 13, 2009 - mobility in rodents concurrently treated with the monoamine oxidase ..... A list of misfortunes is provided in table S1, and a schematic depiction ...

The Neural Correlates of Third-Party Punishment
Dec 10, 2008 - set (n = 20) included actions of comparable gravity to those de- ..... or threat of injury to the subject or a close friend/family member. ...... DD. Victim Compensation. Community Service Short Jail Sentence Long Jail Sentence ...

Dissociating Neural Correlates of Action Monitoring and ...
wise error corrected). RESULTS. Behavioral Results. Performance. The results of the hit rate analysis revealed a significant main effect of Turbulence (F(1, 10) ...