- Email: [email protected]
Part 2: Abstracts of Posters
Brain and Cognition 39, 48–57 (1999) Article ID brcg.1998.1056, available online at http://www.idealibrary.com on
9. Event-Related Brain Potential Correlates of Short-Term Memory in the Healthy Elderly, Patients with Aging-Associated Cognitive Decline, and Alzheimer’s Disease
N. A. Phillips, M. Leblanc, and H. Chertkow Department of Psychology, Concordia University, Quebec, Canada It is difficult to distinguish normal age-related memory changes from the early pathological changes of dementia of the Alzheimer’s type. A late positive eventrelated brain potential (ERP) elicited in a short-term memory search paradigm may reflect memory dysfunction in DAT. We investigated patients with DAT (n ⫽ 14; mean age ⫽ 70.6), patients with mild memory loss at risk for developing DAT (aging-associated cognitive decline, AACD; n ⫽ 16; mean age ⫽ 75.8) and healthy controls (n ⫽ 15; mean age ⫽ 75.2). ERP amplitudes of DAT subjects (6.1 µV) were significantly smaller (F(2, 42) ⫽ 4.73, p ⬍ .01) than those of AACD (9.2 µV) and normal subjects (9.9 µV), which did not differ from each other. The sensitivity of this ERP measure to early memory dysfunction is discussed. 1999 Academic Press
Rationale A significant diagnostic challenge facing us today is the need to distinguish between benign mild changes in cognition associated with normal aging and the pathologic, albeit initially mild, deficits associated with the onset of dementia of the Alzheimer’s type (DAT). Although a deficit in learning and memory are the most obvious early symptoms of DAT, initial symptoms and disease course can be variable. Moreover, changes in memory are a common consequence of normal aging. Some aged individuals present with verifiable evidence of mild cognitive impairment, but which is not of a severity to warrant a diagnosis of dementia. To meet WHO diagnostic criteria for such aging-associated cognitive decline (AACD; Levy, 1994), a patient must exhibit a decline in cognitive function which is gradual and which must be present for a minimum of 6 months. Although typically memory is most affected, the decline can be in any cognitive domain; however, the impairment must not be of sufficient magnitude to diagnosis dementia. This population includes individuals considered to be at very high risk for the development of DAT over long-term follow-up. A significant proportion (50–60%) of AACD individuals will develop DAT during 2–3 year follow-up. This proportion could be considered to be in a prodromal stage of DAT, yet currently it is often necessary to wait a signifi48 0278-2626/99 $30.00
Copyright 1999 by Academic Press All rights of reproduction in any form reserved.
ROTMAN INSTITUTE ABSTRACTS
49
cant period of time to document further cognitive decline in order to attach a diagnosis. There is evidence that a particular event-related brain potential (ERP) component may provide a marker of memory/hippocampal dysfunction. The late positive component (LPC) is recorded during a Sternberg memory scanning paradigm. LPC amplitude has been shown to be differentially abnormal in DAT patients relative to controls (deToledo-Morrell et al., 1991). Further research suggested that its amplitude is correlated with magnetic resonance image (MRI)-derived volume measures of the hippocampus, a structure intimately involved in memory function (deToledo-Morrell et al., 1992). The objective of the research reported here is to determine whether the memory scanning LPC is an early marker of memory dysfunction in subjects with DAT and those at risk (i.e., AACD patients), and thus can be used to distinguish between normal and pathological changes in memory. Methodology Participants. Healthy elderly adults (n ⫽ 15). Healthy elderly subjects (mean age 75 years) without cognitive impairment (verified through neuropsychological assessment) were recruited from the outpatient clinics of the Jewish General Hospital (JGH), Montreal. These subjects were matched in terms of age and education with the patients described below. AACD subjects (n ⫽ 16). Patients with AACD were recruited from the JGH Memory Clinic (mean age 76 years). All met diagnostic criteria for AACD (Levy, 1993) and had Clinical Dementia Ratings of 0.5. Briefly, there was a reported decline (by either the individual or family) in cognitive function which was gradual and of least 6 months duration. This was substantiated by impaired performance on objective neuropsychological assessments. DAT Subjects (n ⫽ 14). Patients with probable DAT (mean age 71) were recruited from the JGH Memory Clinic. All met NINCDS–ADRDA research criteria for probable Alzheimer’s disease (McKhann et al., 1984) and had Clinical Dementia Ratings of 1.0 or greater. Procedure. The memory scanning LPC was elicited during a fixed-set Sternberg memory scanning paradigm. Stimuli consisted of single consonant (yellow) letters presented on a black-background computer screen. Memory set size ranged from 1 to 5 letters. Stimuli comprising each memory set were presented in a horizontal array for 60 s during which the subject was instructed to memorize the stimuli. Following each set, 100 probe trials (single letters) were presented; 24% had been in the memory set. Each probe appeared for 700 ms (stimulus onset asynchrony ⫽ 4200 ms). Subjects indicated (YES/NO manual button press) whether or not each probe had been a member of the memory set. EEG data was amplified in a 0.1–100 Hz bandwidth, sampled at 200 Hz for 1000 ms (100 ms prestimulus baseline), and time-locked to the presentation of each probe stimulus. Trials containing
50
ROTMAN INSTITUTE ABSTRACTS
EOG artifact were corrected off-line using a computerized regression algorithm. EEG was recorded from 3 midline sites (Fz, Cz, Pz) and six lateral electrode placements (frontal: F7, F8; temporal: T3, T4; parietal: P3, P4). LPC was defined as the largest positive peak occurring between 250 and 750 ms poststimulus. Results The groups did not differ in terms of LPC latency (F[2, 42] ⫽ 0.83, p ⬎ .05). For all groups, LPC latency was significantly earlier when there was only one item to hold in memory (mean ⫽ 485 ms) than when the set size was larger (Set Sizes 2–5: mean ⫽ 550 ms). However, LPC amplitude did differ between groups (F[2, 42] ⫽ 3.63, p ⫽ .035). LPC amplitudes of DAT subjects (mean ⫽ 6.8 µV) were significantly smaller than those of AACD (mean ⫽ 9.1 µV) and normal elderly subjects (mean ⫽ 9.9 µV), which did not differ from each other. There was a significant effect of Set Size on LPC amplitude (F[3, 137] ⫽ 5.28, p ⫽ .0007), in which LPC amplitude was significantly smaller in Set Sizes 2–5 (mean ⫽ 8.2 µV) compared to Set Size 1 (mean ⫽ 9.6 µV). There was a significant positive relationship between mean LPC amplitude and performance on the WMS-R Logical Memory Immediate recall (r ⫽ .315, p ⫽ .035, df ⫽ 43) and a trend toward a relationship between LPC amplitude and delayed recall on the Logical Memory subtest (r ⫽ .289, p ⫽ .055, df ⫽ 43). Accuracy to respond to the probe stimuli differed significantly between the groups (F[2, 42] ⫽ 23.5, p ⬍ .0001), as did reaction time (F[2, 42] ⫽ 10.8, p ⫽ .0002). DAT subjects were less accurate and slower (mean ⫽ 80%, mean ⫽ 1050 ms) than the AACD (mean ⫽ 92%, mean ⫽ 720 ms) or control subjects (mean ⫽ 96%, mean ⫽ 710 ms), who did not differ from each other. Discussion LPC amplitude was significantly reduced in DAT subjects relative to the Controls and AACD subjects, but failed to distinguish between the AACD and Control subjects. It is premature to conclude that LPC amplitude will not distinguish between AACD subjects and normal controls. It is expected that only a proportion of our AACD subjects tested (⬃50%) will decline to dementia in 3 to 5 years. We are following these subjects to determine who of our original cohort are destined to develop DAT. It is hypothesized that those subjects who do decline will have manifested significantly reduced LPC amplitudes (similar to those of the DAT subjects) during this initial testing period. It is also possible that the current experimental paradigm was too easy to differentiate the AACD subjects (who do have mild memory dysfunction) from the Control subjects. Finally, we have measured the volume of our subjects’ hippocampi from MRI scans obtained from our three subject groups. Our preliminary results
ROTMAN INSTITUTE ABSTRACTS
51
show that there is a significant positive relationship between LPC amplitude and mean hippocampal volume (r ⫽ 0.45, p ⫽ .006, df ⫽ 34). Further analyses will be conducted to determine whether the combination of functional (LPC amplitude) and anatomical (hippocampal volumes) measures will better discriminate AACD, DAT, and Control subjects. REFERENCES deToledo-Morrell, L., et al. 1991. Archives of Neurology, 48, 605–609. Levy, R. (1994). Age-associated cognitive decline. International Psychogeriatrics, 6, 63–68. Semlitsch, H. V., et al. 1986. Psychophysiology, 23, 695–703.
10. Effects of Smoking History on Aging-Associated Cognitive Decline: An Event-Related Potential Study
V. Knott, A. Harr, and C. Mahoney Royal Ottawa Hospital, Ontario, Canada Cigarette smoking has been shown to reduce the risk for dementia of the Alzheimer type but little is known of the effects of a lengthy smoking history on cognitive processes in the normal elderly. This investigation utilized event-related (P300) brain potentials (ERPs) in a secondary (auditory discrimination) task to examine cognitive involvement in a primary (visual memory) task in groups of young and elderly smoker and nonsmoker adults. Reduced performance efficiency in the elderly was associated with attenuated P300 amplitudes (an index of attentional resources) and delayed P300 latencies (an index of processing speed). Neither performance nor ERP indices were influenced by smoker status. 1999 Academic Press
Rationale Cigarette smoking has been shown to be a variable risk factor for a number of neurological disorders in that it appears to potentiate the risk for stroke and vascular dementia but reduces the risk for dementia of the Alzheimer type. Neurobiological research into the smoking/nicotine habit has focused almost exclusively on young and middle-aged adults and it is yet unknown as to whether a long-term smoking history alters the cognitive status of the aging brain. As scalp-recorded event-related potentials (ERPs) such as the P300 are sensitive indices of cognition, this investigation compared young and elderly adults with smoking and nonsmoking histories using the socalled dual-task paradigm of mental workload assessment: ERP and performance measures of a secondary task are used to test cognitive involvement in an ongoing primary task.
52
ROTMAN INSTITUTE ABSTRACTS
Methodology Participants. Forty healthy volunteers, 20 young (10 M) adults, aged 18– 39 years, and 20 elderly (10 M) adults, aged 65–81 years, participated in this study. Half of the young and elderly were nonsmokers with no previous smoking history and the remaining half of the young and elderly were current smokers (15 or more cigarettes/day) with average smoking histories of 8.3 and 52.5 years, respectively. Procedure. Subjects attended the laboratory for one orientation/practice session and one morning test session which followed 12 hr of overnight tobacco abstinence. During the test session subjects were assessed (in counterbalanced order) separately and concurrently on a ‘‘primary’’ visual word recognition task and a ‘‘secondary’’ auditory tone discrimination task. The primary task required overt discrimination of words presented for the first versus the second time. A list of words were presented (at 2 s interstimulus intervals) and subjects were required to respond (dominant hand) on every trial, pressing one microswitch button when they saw a word displayed for the first time (a ‘‘new’’ word) and responding with another button when a word was presented for the second time (an ‘‘old’’ word: p ⫽ .25). The secondary task was a standard auditory oddball task used to elicit the P300 ERP and consisted of a regular series (2 s interstimulus intervals of high-pitch (2000 Hz) and low-pitch (1000 Hz) tones (70 dB [SPL]) differing in occurrences probability (p ⫽ .80 vs p ⫽ .20). The rare low-pitch tone had to be responded to as quickly and accurately as possible by a button press (nondominant hand). During the dual-task condition the two stimulus series were intermixed, but not overlapping, resulting in a 1-s interstimulus interval. In the dual-task condition, subjects were instructed to maximize their attention and effort to the word recognition task. ERPs extracted only for secondary task stimuli, were collected from midline (Fz, Cz, Pz) sites (referenced to linked earlobes) with amplifier bandpass settings and sampling frequency set to 0.1–30.0 and 256 Hz, respectively. Ocular-corrected ERP averages to currently identified rare tones were scored for P300 peak latency and amplitude, with P300 being defined as the maximum positive voltage within 250–600 ms poststimulus onset. Performance measures in both tasks included the frequency of correct detections, commission errors, and reaction time. Results Performance analysis on the primary task showed elderly subjects to exhibit reduced accuracy (i.e., fewer correct detections and more false alarms) and slower reaction times in both single- and dual-task conditions. Performance analysis on the secondary task also found elderly subjects to exhibit more false alarms and slower reaction times. Performance efficiency (i.e.,
ROTMAN INSTITUTE ABSTRACTS
53
accuracy and speed) was less efficient in the secondary task when performed under dual-task conditions and the dual-task condition induced significant increases in false alarms in elderly but not young adult subjects. Elderly adults exhibited smaller parietal and central P300 amplitudes and longer parietal P300 latencies than younger adults in both single and dual-task conditions. Neither performance nor ERP measures were influenced by smoker– nonsmoker status. The results will be discussed in relation to cognition in normal and pathological aging.
11. Age-Related Differences in Brain Activation during Encoding and Retrieval under Divided Attention: A Positron Emission Tomography (PET) Study
T. Iidaka,* N. Anderson,*,† S. Kapur,†,‡ R. Cabeza,§ C. Okamoto,* and F. I. M. Craik*,† *Department of Psychology, University of Toronto, Canada; †Rotman Research Institute, Baycrest Centre for Geriatric Care, Toronto, Ontario, Canada; ‡Clarke Institute of Psychiatry, Toronto, Ontario, Canada; §Department of Psychology, University of Alberta, Canada A PET study was conducted to examine the neural correlates of encoding and retrieval under divided attention (DA) and full attention (FA) in young and old subjects. The subjects performed both the memory and the pitch discrimination task concurrently. PET scans were obtained using H215O as a tracer and the data were analyzed by SPM96. In the young subjects during encoding, the right inferior parietal lobule was activated under DA, whereas the left inferior temporal and left inferior frontal cortex were activated under FA. The left anterior cingulate and the right cerebellum were related to DA. In the comparison between the young and old, the neural network involved in the age-related decrements in memory and attention included the right cerebellum, left thalamus, left inferior frontal, right medial temporal, and right medial frontal cortex. 1999 Academic Press
Rationale Behavioral studies have shown that divided attention (DA) at encoding is associated with a substantial drop in later memory performance, but in contrast, DA at retrieval resulted in comparatively slight drops in memory performance. The first aim of the present study is to examine the effects of DA on encoding and retrieval processes in young subjects, and to reveal the nature of differences between them by using PET. A PET study concerning age-related differences in memory has revealed that old subjects failed to activate hippocampal and other cortical regions during encoding. The second aim of the present study is to examine the effect of aging on encoding and retrieval processes in a DA paradigm.
54
ROTMAN INSTITUTE ABSTRACTS
Methodology Participants. Twelve young (M/F; 3/9, aged from 21 to 31, mean; 24 years) and eight old (M/F; 3/5, aged from 63 to 76, mean; 69 years) subjects participated in the study after giving informed consent. All subjects spoke English as their first language and were right-handed. No subject had a history of psychiatric or neurological diseases. None of the subjects were taking any drug that affected cerebral blood flow. In the experiment, subjects were instructed to perform both the memory task and the secondary task concurrently. The experimental conditions were (1) encoding under full attention (EF), (2) encoding under divided attention (ED), (3) retrieval under full attention (RF), and (4) retrieval under divided attention (RD). In the memory task, 20 word pairs were presented on display. During encoding, subjects read aloud the second word of the pair and memorized the pair for subsequent retrieval. During retrieval, only the first word of each pair was presented in a random order and subjects were instructed to recall the associated word or say ‘‘pass.’’ As a secondary task we employed a pitch discrimination task. In DA conditions subjects were instructed to listen to randomly presented high and low tones, discriminate the level of the pitch, and press the corresponding button of the mouse. In full attention (FA) conditions the low tone was repeatedly presented and subjects pressed a left mouse button each time in response to the tone. In the control condition, a simple word-reading task plus the FA secondary task were employed. PET scans were obtained with a GE PC2048–15B scanner. After taking a transmission scan, 60-s emission scans were performed following a bolus injection of 35 mCi of H215O. Nine PET scans were conducted 11 min apart (twice for each of the four experimental conditions and once for the control condition). The order of the conditions was counterbalanced across subjects. The data were analyzed by SPM96. We set the statistical significance at a p value of 0.05 (corrected for height and uncorrected for extent) for the first aim of the study in young subjects, and at a p value of 0.001 (uncorrected for height) for the comparison between young and old subjects. Results Behavioral data. Mean memory performance of the young subjects in EF, ED, RF, and RD conditions was 0.79, 0.58, 0.78, and 0.75 respectively. In the old subjects, memory performance in each condition was 0.52, 0.34, 0.52, and 0.44, respectively. A two-way ANOVA revealed significant main effects of condition (performance in ED was lower than the other three conditions) and age (performance in the young subjects was lower than in the old subjects), but there was no interaction.
ROTMAN INSTITUTE ABSTRACTS
55
PET data. In the young subjects, the ED–EF contrast revealed three activated clusters in right cerebellum, left anterior cingulate gyrus (BA32), and right inferior parietal lobule (BA40). In the RD–RF contrast, the activated clusters were located in left anterior cingulate gyrus (BA32) and right premotor area (BA6). In the EF–ED contrast, two activated clusters were found in left inferior temporal gyrus (BA37) and left inferior frontal gyrus (BA44). There was no significant activation in the RF–RD contrast. In the comparison between the young and old subjects, the young subjects demonstrated significant activation in the right cerebellum in the ED–EF contrast, left thalamus in the RD–RF contrast, left inferior frontal gyrus (BA9/44) and right medial temporal cortex in the EF–ED contrast, and right medial frontal cortex (BA9/11) in the RF–RD contrast. Discussion For young subjects our study showed that in the DA condition the right cerebellum and the right inferior parietal lobule were more activated in encoding than in retrieval, whereas activation in the left anterior cingulate was common to both of them. In FA, the inferior temporal cortex and inferior frontal cortex in the left hemisphere were activated during encoding. These results may imply that a trade-off exists between the auditory and visual systems during encoding; that is, greater attention was paid to auditory information in DA, whereas greater attention was paid to visual information in FA. The involvement of the cerebellum in episodic memory encoding may relate to its modulatory function in human cognition. The comparison between the young and old indicates that the old subjects failed to activate several regions of the brain. The neural network involved in the age-related decrements in memory and attention included the right cerebellum, left thalamus, left inferior frontal, right medial temporal, and right medial frontal cortex.
12. The Neurophysiological Effects of Aging on the Ability of the Visual Cortex to Process Temporal Information
J. R. Mendelson and E. F. Wells Department of Speech–Language Pathology, Faculty of Medicine, University of Toronto, 6 Queen’s Park Crescent West, Toronto, Ontario M5S 3H2, Canada Aging can produce a decline in the ability to process visual motion. It has been suggested that this decline is due in part to a slowing down in temporal processing, the locus of which may reside in central, as opposed to peripheral, visual system. We used critical flicker fusion (CFF), or the frequency of flickering light (Hz) required to
56
ROTMAN INSTITUTE ABSTRACTS
produce an appearance of steady light, to compare the temporal processing abilities of visual cortical neurons in young (2–4 months) and old (24–28 months) Long Evans Hooded rats. CFF thresholds were significantly reduced in neurons recorded from old animals in both primary and secondary visual cortex. This study provides the first indication of a difference in temporal processing speed in the aging visual cortex. 1999 Academic Press
Rationale It is well known that as we age, many people experience a deterioration in visual function. This deterioration is commonly associated with diseases of the visual system such as cataract, glaucoma, and macular degeneration, or with changes in the peripheral structures such as an increase in lenticular density of yellowing. However, it now appears that not all age-related visual deficits can be attributed to optical changes, but instead may also be a result of changes to more central visual pathways (Spear, 1993). This problem is exemplified by elderly individuals who demonstrate no signs of optical pathology and perform well on clinical tests of visual acuity, yet still complain of poor vision. These visual ‘‘underachievers’’ indicate that central visual structures may also contribute to age-related deficits. However, little information as to the neural bases of these deficits currently exists. One aspect of vision that appears to decline with age is the ability to detect temporal change (Sekuler & Sekuler, 1991). The ability to process temporally contiguous visual stimuli is important in everyday vision for the perception of moving targets (Kline, Kline, Fozard, Kosnik, Schieber, & Sekuler, 1991). One way in which a deficit in this process can manifest itself, is in an increase in vehicle accidents involving elderly drivers who complain of being unable to perceive the speed of surrounding vehicles, as well as their own speed (Kline et al., 1991). The tremendous social impact, as well as psychological effect on the self-esteem of elderly individuals who are forced to surrender their driver’s license and cope with subsequent lifestyle changes (Salvage, 1995), indicates the importance of this deficit and need for study. Unfortunately, there is currently little information on how changes in neuronal information processing and integration contribute to this apparent ageassociated decrease in the rat of temporal processing. However, work on the retina (Spear et al., 1996) and lateral geniculate nucleus (Spear et al., 1993, 1994) of monkeys, indicates that the changes in temporal processing that accompany aging may occur in primary or association cortex. To examine the neural bases underlying age-related temporal processing deficits in the visual system, we compared the response of visual cortical neurons in young and old Long Evans Hooded rats to critical flicker fusion (CFF). The ability to perceive a certain number of flashes per unit time is an index of the efficacy of the visual system’s temporal processing capabilities (Brozek & Keys, 1945). Since the ability to detect CFF declines with age, the hypothesis was that neurons recorded in the old visual cortex would have depressed CFF threshold values compared to those in young visual cortex.
ROTMAN INSTITUTE ABSTRACTS
57
Methodology Subjects. Subjects were 36 male Long Evans Hooded rats. Twenty-eight rats between two and four months of age served as young subjects with the eight old subjects ranged from 24 to 28 months of age. The optics of the eye were checked regularly with an ophthalmoscope in order to exclude any subjects with signs of optical pathology such as glaucoma or cataract. Procedure. Extracellular single-cell recordings were made from both primary and secondary visual cortex. Once a visually responsive cell was isolated, both its critical flicker fusion threshold and receptive field properties were determined. CFF was examined by stimulating the contralateral eye with light pulses delivered by a stimulator lamp at frequencies ranging from 1 and 60 Hz. To determine threshold, frequency was varied pseudo-randomly until the cell was no longer able to entrain, or respond to each of the light flashes within the presentation. The cell was also characterized as simple, complex, or hypercomplex according to its response to a bar of light moved across the visual hemifield at various speeds, orientations, and directions. After the characteristics of each cell was determined, lesions were made for histological verification. Results Extracellular recordings were made from 113 neurons in rat visual cortex: 65 were from young animals and 45 were from old animals. The results showed that there was a significant difference in the way cells recorded from the visual cortex of young and old animals responded to a flashing light. The range of CFF thresholds obtained from young rat visual cortex was between 10 and 35 Hz with a mean of 21.54 ⫾ 6.0 Hz, while the range recorded in old animals was 6 to 15 Hz with a mean of 11.89 ⫾ 2.79 Hz. The difference was statistically significant (t ⫽ 10.10; p ⬍.001). When CFF was analyzed with respect to receptive field properties, a significant reduction in CFF thresholds was found for both complex and hypercomplex cells in old animals as compared to young animals (F ⫽ 5.12; p ⬍ 0.05). There was no significant difference between young and old rats for simple cells. In addition, simple cells in both young and old rats exhibited lower CFF thresholds than complex of hypercomplex cells. When we compared CFF thresholds between cells recorded from primary and secondary visual cortex within each age group, no significant difference was found. However, significant differences were observed when the responses of primary and secondary visual cortical units in young animals were compared with those from old animals (F ⫽ 60.77; p ⬍ 0.001).
9. Event-Related Brain Potential Correlates of Short-Term Memory in the Healthy Elderly, Patients with Aging-Associated Cognitive Decline, and Alzheimer’s Disease
N. A. Phillips, M. Leblanc, and H. Chertkow Department of Psychology, Concordia University, Quebec, Canada It is difficult to distinguish normal age-related memory changes from the early pathological changes of dementia of the Alzheimer’s type. A late positive eventrelated brain potential (ERP) elicited in a short-term memory search paradigm may reflect memory dysfunction in DAT. We investigated patients with DAT (n ⫽ 14; mean age ⫽ 70.6), patients with mild memory loss at risk for developing DAT (aging-associated cognitive decline, AACD; n ⫽ 16; mean age ⫽ 75.8) and healthy controls (n ⫽ 15; mean age ⫽ 75.2). ERP amplitudes of DAT subjects (6.1 µV) were significantly smaller (F(2, 42) ⫽ 4.73, p ⬍ .01) than those of AACD (9.2 µV) and normal subjects (9.9 µV), which did not differ from each other. The sensitivity of this ERP measure to early memory dysfunction is discussed. 1999 Academic Press
Rationale A significant diagnostic challenge facing us today is the need to distinguish between benign mild changes in cognition associated with normal aging and the pathologic, albeit initially mild, deficits associated with the onset of dementia of the Alzheimer’s type (DAT). Although a deficit in learning and memory are the most obvious early symptoms of DAT, initial symptoms and disease course can be variable. Moreover, changes in memory are a common consequence of normal aging. Some aged individuals present with verifiable evidence of mild cognitive impairment, but which is not of a severity to warrant a diagnosis of dementia. To meet WHO diagnostic criteria for such aging-associated cognitive decline (AACD; Levy, 1994), a patient must exhibit a decline in cognitive function which is gradual and which must be present for a minimum of 6 months. Although typically memory is most affected, the decline can be in any cognitive domain; however, the impairment must not be of sufficient magnitude to diagnosis dementia. This population includes individuals considered to be at very high risk for the development of DAT over long-term follow-up. A significant proportion (50–60%) of AACD individuals will develop DAT during 2–3 year follow-up. This proportion could be considered to be in a prodromal stage of DAT, yet currently it is often necessary to wait a signifi48 0278-2626/99 $30.00
Copyright 1999 by Academic Press All rights of reproduction in any form reserved.
ROTMAN INSTITUTE ABSTRACTS
49
cant period of time to document further cognitive decline in order to attach a diagnosis. There is evidence that a particular event-related brain potential (ERP) component may provide a marker of memory/hippocampal dysfunction. The late positive component (LPC) is recorded during a Sternberg memory scanning paradigm. LPC amplitude has been shown to be differentially abnormal in DAT patients relative to controls (deToledo-Morrell et al., 1991). Further research suggested that its amplitude is correlated with magnetic resonance image (MRI)-derived volume measures of the hippocampus, a structure intimately involved in memory function (deToledo-Morrell et al., 1992). The objective of the research reported here is to determine whether the memory scanning LPC is an early marker of memory dysfunction in subjects with DAT and those at risk (i.e., AACD patients), and thus can be used to distinguish between normal and pathological changes in memory. Methodology Participants. Healthy elderly adults (n ⫽ 15). Healthy elderly subjects (mean age 75 years) without cognitive impairment (verified through neuropsychological assessment) were recruited from the outpatient clinics of the Jewish General Hospital (JGH), Montreal. These subjects were matched in terms of age and education with the patients described below. AACD subjects (n ⫽ 16). Patients with AACD were recruited from the JGH Memory Clinic (mean age 76 years). All met diagnostic criteria for AACD (Levy, 1993) and had Clinical Dementia Ratings of 0.5. Briefly, there was a reported decline (by either the individual or family) in cognitive function which was gradual and of least 6 months duration. This was substantiated by impaired performance on objective neuropsychological assessments. DAT Subjects (n ⫽ 14). Patients with probable DAT (mean age 71) were recruited from the JGH Memory Clinic. All met NINCDS–ADRDA research criteria for probable Alzheimer’s disease (McKhann et al., 1984) and had Clinical Dementia Ratings of 1.0 or greater. Procedure. The memory scanning LPC was elicited during a fixed-set Sternberg memory scanning paradigm. Stimuli consisted of single consonant (yellow) letters presented on a black-background computer screen. Memory set size ranged from 1 to 5 letters. Stimuli comprising each memory set were presented in a horizontal array for 60 s during which the subject was instructed to memorize the stimuli. Following each set, 100 probe trials (single letters) were presented; 24% had been in the memory set. Each probe appeared for 700 ms (stimulus onset asynchrony ⫽ 4200 ms). Subjects indicated (YES/NO manual button press) whether or not each probe had been a member of the memory set. EEG data was amplified in a 0.1–100 Hz bandwidth, sampled at 200 Hz for 1000 ms (100 ms prestimulus baseline), and time-locked to the presentation of each probe stimulus. Trials containing
50
ROTMAN INSTITUTE ABSTRACTS
EOG artifact were corrected off-line using a computerized regression algorithm. EEG was recorded from 3 midline sites (Fz, Cz, Pz) and six lateral electrode placements (frontal: F7, F8; temporal: T3, T4; parietal: P3, P4). LPC was defined as the largest positive peak occurring between 250 and 750 ms poststimulus. Results The groups did not differ in terms of LPC latency (F[2, 42] ⫽ 0.83, p ⬎ .05). For all groups, LPC latency was significantly earlier when there was only one item to hold in memory (mean ⫽ 485 ms) than when the set size was larger (Set Sizes 2–5: mean ⫽ 550 ms). However, LPC amplitude did differ between groups (F[2, 42] ⫽ 3.63, p ⫽ .035). LPC amplitudes of DAT subjects (mean ⫽ 6.8 µV) were significantly smaller than those of AACD (mean ⫽ 9.1 µV) and normal elderly subjects (mean ⫽ 9.9 µV), which did not differ from each other. There was a significant effect of Set Size on LPC amplitude (F[3, 137] ⫽ 5.28, p ⫽ .0007), in which LPC amplitude was significantly smaller in Set Sizes 2–5 (mean ⫽ 8.2 µV) compared to Set Size 1 (mean ⫽ 9.6 µV). There was a significant positive relationship between mean LPC amplitude and performance on the WMS-R Logical Memory Immediate recall (r ⫽ .315, p ⫽ .035, df ⫽ 43) and a trend toward a relationship between LPC amplitude and delayed recall on the Logical Memory subtest (r ⫽ .289, p ⫽ .055, df ⫽ 43). Accuracy to respond to the probe stimuli differed significantly between the groups (F[2, 42] ⫽ 23.5, p ⬍ .0001), as did reaction time (F[2, 42] ⫽ 10.8, p ⫽ .0002). DAT subjects were less accurate and slower (mean ⫽ 80%, mean ⫽ 1050 ms) than the AACD (mean ⫽ 92%, mean ⫽ 720 ms) or control subjects (mean ⫽ 96%, mean ⫽ 710 ms), who did not differ from each other. Discussion LPC amplitude was significantly reduced in DAT subjects relative to the Controls and AACD subjects, but failed to distinguish between the AACD and Control subjects. It is premature to conclude that LPC amplitude will not distinguish between AACD subjects and normal controls. It is expected that only a proportion of our AACD subjects tested (⬃50%) will decline to dementia in 3 to 5 years. We are following these subjects to determine who of our original cohort are destined to develop DAT. It is hypothesized that those subjects who do decline will have manifested significantly reduced LPC amplitudes (similar to those of the DAT subjects) during this initial testing period. It is also possible that the current experimental paradigm was too easy to differentiate the AACD subjects (who do have mild memory dysfunction) from the Control subjects. Finally, we have measured the volume of our subjects’ hippocampi from MRI scans obtained from our three subject groups. Our preliminary results
ROTMAN INSTITUTE ABSTRACTS
51
show that there is a significant positive relationship between LPC amplitude and mean hippocampal volume (r ⫽ 0.45, p ⫽ .006, df ⫽ 34). Further analyses will be conducted to determine whether the combination of functional (LPC amplitude) and anatomical (hippocampal volumes) measures will better discriminate AACD, DAT, and Control subjects. REFERENCES deToledo-Morrell, L., et al. 1991. Archives of Neurology, 48, 605–609. Levy, R. (1994). Age-associated cognitive decline. International Psychogeriatrics, 6, 63–68. Semlitsch, H. V., et al. 1986. Psychophysiology, 23, 695–703.
10. Effects of Smoking History on Aging-Associated Cognitive Decline: An Event-Related Potential Study
V. Knott, A. Harr, and C. Mahoney Royal Ottawa Hospital, Ontario, Canada Cigarette smoking has been shown to reduce the risk for dementia of the Alzheimer type but little is known of the effects of a lengthy smoking history on cognitive processes in the normal elderly. This investigation utilized event-related (P300) brain potentials (ERPs) in a secondary (auditory discrimination) task to examine cognitive involvement in a primary (visual memory) task in groups of young and elderly smoker and nonsmoker adults. Reduced performance efficiency in the elderly was associated with attenuated P300 amplitudes (an index of attentional resources) and delayed P300 latencies (an index of processing speed). Neither performance nor ERP indices were influenced by smoker status. 1999 Academic Press
Rationale Cigarette smoking has been shown to be a variable risk factor for a number of neurological disorders in that it appears to potentiate the risk for stroke and vascular dementia but reduces the risk for dementia of the Alzheimer type. Neurobiological research into the smoking/nicotine habit has focused almost exclusively on young and middle-aged adults and it is yet unknown as to whether a long-term smoking history alters the cognitive status of the aging brain. As scalp-recorded event-related potentials (ERPs) such as the P300 are sensitive indices of cognition, this investigation compared young and elderly adults with smoking and nonsmoking histories using the socalled dual-task paradigm of mental workload assessment: ERP and performance measures of a secondary task are used to test cognitive involvement in an ongoing primary task.
52
ROTMAN INSTITUTE ABSTRACTS
Methodology Participants. Forty healthy volunteers, 20 young (10 M) adults, aged 18– 39 years, and 20 elderly (10 M) adults, aged 65–81 years, participated in this study. Half of the young and elderly were nonsmokers with no previous smoking history and the remaining half of the young and elderly were current smokers (15 or more cigarettes/day) with average smoking histories of 8.3 and 52.5 years, respectively. Procedure. Subjects attended the laboratory for one orientation/practice session and one morning test session which followed 12 hr of overnight tobacco abstinence. During the test session subjects were assessed (in counterbalanced order) separately and concurrently on a ‘‘primary’’ visual word recognition task and a ‘‘secondary’’ auditory tone discrimination task. The primary task required overt discrimination of words presented for the first versus the second time. A list of words were presented (at 2 s interstimulus intervals) and subjects were required to respond (dominant hand) on every trial, pressing one microswitch button when they saw a word displayed for the first time (a ‘‘new’’ word) and responding with another button when a word was presented for the second time (an ‘‘old’’ word: p ⫽ .25). The secondary task was a standard auditory oddball task used to elicit the P300 ERP and consisted of a regular series (2 s interstimulus intervals of high-pitch (2000 Hz) and low-pitch (1000 Hz) tones (70 dB [SPL]) differing in occurrences probability (p ⫽ .80 vs p ⫽ .20). The rare low-pitch tone had to be responded to as quickly and accurately as possible by a button press (nondominant hand). During the dual-task condition the two stimulus series were intermixed, but not overlapping, resulting in a 1-s interstimulus interval. In the dual-task condition, subjects were instructed to maximize their attention and effort to the word recognition task. ERPs extracted only for secondary task stimuli, were collected from midline (Fz, Cz, Pz) sites (referenced to linked earlobes) with amplifier bandpass settings and sampling frequency set to 0.1–30.0 and 256 Hz, respectively. Ocular-corrected ERP averages to currently identified rare tones were scored for P300 peak latency and amplitude, with P300 being defined as the maximum positive voltage within 250–600 ms poststimulus onset. Performance measures in both tasks included the frequency of correct detections, commission errors, and reaction time. Results Performance analysis on the primary task showed elderly subjects to exhibit reduced accuracy (i.e., fewer correct detections and more false alarms) and slower reaction times in both single- and dual-task conditions. Performance analysis on the secondary task also found elderly subjects to exhibit more false alarms and slower reaction times. Performance efficiency (i.e.,
ROTMAN INSTITUTE ABSTRACTS
53
accuracy and speed) was less efficient in the secondary task when performed under dual-task conditions and the dual-task condition induced significant increases in false alarms in elderly but not young adult subjects. Elderly adults exhibited smaller parietal and central P300 amplitudes and longer parietal P300 latencies than younger adults in both single and dual-task conditions. Neither performance nor ERP measures were influenced by smoker– nonsmoker status. The results will be discussed in relation to cognition in normal and pathological aging.
11. Age-Related Differences in Brain Activation during Encoding and Retrieval under Divided Attention: A Positron Emission Tomography (PET) Study
T. Iidaka,* N. Anderson,*,† S. Kapur,†,‡ R. Cabeza,§ C. Okamoto,* and F. I. M. Craik*,† *Department of Psychology, University of Toronto, Canada; †Rotman Research Institute, Baycrest Centre for Geriatric Care, Toronto, Ontario, Canada; ‡Clarke Institute of Psychiatry, Toronto, Ontario, Canada; §Department of Psychology, University of Alberta, Canada A PET study was conducted to examine the neural correlates of encoding and retrieval under divided attention (DA) and full attention (FA) in young and old subjects. The subjects performed both the memory and the pitch discrimination task concurrently. PET scans were obtained using H215O as a tracer and the data were analyzed by SPM96. In the young subjects during encoding, the right inferior parietal lobule was activated under DA, whereas the left inferior temporal and left inferior frontal cortex were activated under FA. The left anterior cingulate and the right cerebellum were related to DA. In the comparison between the young and old, the neural network involved in the age-related decrements in memory and attention included the right cerebellum, left thalamus, left inferior frontal, right medial temporal, and right medial frontal cortex. 1999 Academic Press
Rationale Behavioral studies have shown that divided attention (DA) at encoding is associated with a substantial drop in later memory performance, but in contrast, DA at retrieval resulted in comparatively slight drops in memory performance. The first aim of the present study is to examine the effects of DA on encoding and retrieval processes in young subjects, and to reveal the nature of differences between them by using PET. A PET study concerning age-related differences in memory has revealed that old subjects failed to activate hippocampal and other cortical regions during encoding. The second aim of the present study is to examine the effect of aging on encoding and retrieval processes in a DA paradigm.
54
ROTMAN INSTITUTE ABSTRACTS
Methodology Participants. Twelve young (M/F; 3/9, aged from 21 to 31, mean; 24 years) and eight old (M/F; 3/5, aged from 63 to 76, mean; 69 years) subjects participated in the study after giving informed consent. All subjects spoke English as their first language and were right-handed. No subject had a history of psychiatric or neurological diseases. None of the subjects were taking any drug that affected cerebral blood flow. In the experiment, subjects were instructed to perform both the memory task and the secondary task concurrently. The experimental conditions were (1) encoding under full attention (EF), (2) encoding under divided attention (ED), (3) retrieval under full attention (RF), and (4) retrieval under divided attention (RD). In the memory task, 20 word pairs were presented on display. During encoding, subjects read aloud the second word of the pair and memorized the pair for subsequent retrieval. During retrieval, only the first word of each pair was presented in a random order and subjects were instructed to recall the associated word or say ‘‘pass.’’ As a secondary task we employed a pitch discrimination task. In DA conditions subjects were instructed to listen to randomly presented high and low tones, discriminate the level of the pitch, and press the corresponding button of the mouse. In full attention (FA) conditions the low tone was repeatedly presented and subjects pressed a left mouse button each time in response to the tone. In the control condition, a simple word-reading task plus the FA secondary task were employed. PET scans were obtained with a GE PC2048–15B scanner. After taking a transmission scan, 60-s emission scans were performed following a bolus injection of 35 mCi of H215O. Nine PET scans were conducted 11 min apart (twice for each of the four experimental conditions and once for the control condition). The order of the conditions was counterbalanced across subjects. The data were analyzed by SPM96. We set the statistical significance at a p value of 0.05 (corrected for height and uncorrected for extent) for the first aim of the study in young subjects, and at a p value of 0.001 (uncorrected for height) for the comparison between young and old subjects. Results Behavioral data. Mean memory performance of the young subjects in EF, ED, RF, and RD conditions was 0.79, 0.58, 0.78, and 0.75 respectively. In the old subjects, memory performance in each condition was 0.52, 0.34, 0.52, and 0.44, respectively. A two-way ANOVA revealed significant main effects of condition (performance in ED was lower than the other three conditions) and age (performance in the young subjects was lower than in the old subjects), but there was no interaction.
ROTMAN INSTITUTE ABSTRACTS
55
PET data. In the young subjects, the ED–EF contrast revealed three activated clusters in right cerebellum, left anterior cingulate gyrus (BA32), and right inferior parietal lobule (BA40). In the RD–RF contrast, the activated clusters were located in left anterior cingulate gyrus (BA32) and right premotor area (BA6). In the EF–ED contrast, two activated clusters were found in left inferior temporal gyrus (BA37) and left inferior frontal gyrus (BA44). There was no significant activation in the RF–RD contrast. In the comparison between the young and old subjects, the young subjects demonstrated significant activation in the right cerebellum in the ED–EF contrast, left thalamus in the RD–RF contrast, left inferior frontal gyrus (BA9/44) and right medial temporal cortex in the EF–ED contrast, and right medial frontal cortex (BA9/11) in the RF–RD contrast. Discussion For young subjects our study showed that in the DA condition the right cerebellum and the right inferior parietal lobule were more activated in encoding than in retrieval, whereas activation in the left anterior cingulate was common to both of them. In FA, the inferior temporal cortex and inferior frontal cortex in the left hemisphere were activated during encoding. These results may imply that a trade-off exists between the auditory and visual systems during encoding; that is, greater attention was paid to auditory information in DA, whereas greater attention was paid to visual information in FA. The involvement of the cerebellum in episodic memory encoding may relate to its modulatory function in human cognition. The comparison between the young and old indicates that the old subjects failed to activate several regions of the brain. The neural network involved in the age-related decrements in memory and attention included the right cerebellum, left thalamus, left inferior frontal, right medial temporal, and right medial frontal cortex.
12. The Neurophysiological Effects of Aging on the Ability of the Visual Cortex to Process Temporal Information
J. R. Mendelson and E. F. Wells Department of Speech–Language Pathology, Faculty of Medicine, University of Toronto, 6 Queen’s Park Crescent West, Toronto, Ontario M5S 3H2, Canada Aging can produce a decline in the ability to process visual motion. It has been suggested that this decline is due in part to a slowing down in temporal processing, the locus of which may reside in central, as opposed to peripheral, visual system. We used critical flicker fusion (CFF), or the frequency of flickering light (Hz) required to
56
ROTMAN INSTITUTE ABSTRACTS
produce an appearance of steady light, to compare the temporal processing abilities of visual cortical neurons in young (2–4 months) and old (24–28 months) Long Evans Hooded rats. CFF thresholds were significantly reduced in neurons recorded from old animals in both primary and secondary visual cortex. This study provides the first indication of a difference in temporal processing speed in the aging visual cortex. 1999 Academic Press
Rationale It is well known that as we age, many people experience a deterioration in visual function. This deterioration is commonly associated with diseases of the visual system such as cataract, glaucoma, and macular degeneration, or with changes in the peripheral structures such as an increase in lenticular density of yellowing. However, it now appears that not all age-related visual deficits can be attributed to optical changes, but instead may also be a result of changes to more central visual pathways (Spear, 1993). This problem is exemplified by elderly individuals who demonstrate no signs of optical pathology and perform well on clinical tests of visual acuity, yet still complain of poor vision. These visual ‘‘underachievers’’ indicate that central visual structures may also contribute to age-related deficits. However, little information as to the neural bases of these deficits currently exists. One aspect of vision that appears to decline with age is the ability to detect temporal change (Sekuler & Sekuler, 1991). The ability to process temporally contiguous visual stimuli is important in everyday vision for the perception of moving targets (Kline, Kline, Fozard, Kosnik, Schieber, & Sekuler, 1991). One way in which a deficit in this process can manifest itself, is in an increase in vehicle accidents involving elderly drivers who complain of being unable to perceive the speed of surrounding vehicles, as well as their own speed (Kline et al., 1991). The tremendous social impact, as well as psychological effect on the self-esteem of elderly individuals who are forced to surrender their driver’s license and cope with subsequent lifestyle changes (Salvage, 1995), indicates the importance of this deficit and need for study. Unfortunately, there is currently little information on how changes in neuronal information processing and integration contribute to this apparent ageassociated decrease in the rat of temporal processing. However, work on the retina (Spear et al., 1996) and lateral geniculate nucleus (Spear et al., 1993, 1994) of monkeys, indicates that the changes in temporal processing that accompany aging may occur in primary or association cortex. To examine the neural bases underlying age-related temporal processing deficits in the visual system, we compared the response of visual cortical neurons in young and old Long Evans Hooded rats to critical flicker fusion (CFF). The ability to perceive a certain number of flashes per unit time is an index of the efficacy of the visual system’s temporal processing capabilities (Brozek & Keys, 1945). Since the ability to detect CFF declines with age, the hypothesis was that neurons recorded in the old visual cortex would have depressed CFF threshold values compared to those in young visual cortex.
ROTMAN INSTITUTE ABSTRACTS
57
Methodology Subjects. Subjects were 36 male Long Evans Hooded rats. Twenty-eight rats between two and four months of age served as young subjects with the eight old subjects ranged from 24 to 28 months of age. The optics of the eye were checked regularly with an ophthalmoscope in order to exclude any subjects with signs of optical pathology such as glaucoma or cataract. Procedure. Extracellular single-cell recordings were made from both primary and secondary visual cortex. Once a visually responsive cell was isolated, both its critical flicker fusion threshold and receptive field properties were determined. CFF was examined by stimulating the contralateral eye with light pulses delivered by a stimulator lamp at frequencies ranging from 1 and 60 Hz. To determine threshold, frequency was varied pseudo-randomly until the cell was no longer able to entrain, or respond to each of the light flashes within the presentation. The cell was also characterized as simple, complex, or hypercomplex according to its response to a bar of light moved across the visual hemifield at various speeds, orientations, and directions. After the characteristics of each cell was determined, lesions were made for histological verification. Results Extracellular recordings were made from 113 neurons in rat visual cortex: 65 were from young animals and 45 were from old animals. The results showed that there was a significant difference in the way cells recorded from the visual cortex of young and old animals responded to a flashing light. The range of CFF thresholds obtained from young rat visual cortex was between 10 and 35 Hz with a mean of 21.54 ⫾ 6.0 Hz, while the range recorded in old animals was 6 to 15 Hz with a mean of 11.89 ⫾ 2.79 Hz. The difference was statistically significant (t ⫽ 10.10; p ⬍.001). When CFF was analyzed with respect to receptive field properties, a significant reduction in CFF thresholds was found for both complex and hypercomplex cells in old animals as compared to young animals (F ⫽ 5.12; p ⬍ 0.05). There was no significant difference between young and old rats for simple cells. In addition, simple cells in both young and old rats exhibited lower CFF thresholds than complex of hypercomplex cells. When we compared CFF thresholds between cells recorded from primary and secondary visual cortex within each age group, no significant difference was found. However, significant differences were observed when the responses of primary and secondary visual cortical units in young animals were compared with those from old animals (F ⫽ 60.77; p ⬍ 0.001).