Part 2: Abstracts of Posters

Brain and Cognition 49, 194–255 (2002) doi:10.1006/brcg.2001.1463

Part 2: Abstracts of Posters 1. Neuropsychological Impairment in Veterans Who Are HIV-Positive

K. Grohman, K. Donnelly, J. Strang, and J. Kleiner The purpose of this study was to compare the neuropsychological functioning of 12 veterans who were HIV-positive to 21 age-matched veterans who were HIV-negative. Consistent with expectations, the HIV-positive group was found to perform more poorly in areas related to attention and concentration, immediate and delayed verbal recall, immediate and delayed visual recall, visual learning, and tasks requiring psychomotor speed, while a number of language tasks were left intact. This was similar to dysfunction often seen in HIV-related dementia cases. However, this group was also significantly more impaired in confrontation naming, planning, mental calculations, and abstract thought when compared to the HIV-negative group. Comorbid substance abuse found in the majority of our HIV-positive subjects was thought to contribute to the HIV-related dysfunction.  2002 Elsevier Science (USA)

Report

As HIV-positive patients are living longer, and relatively healthier, now than 5 years ago (Hinkin et al., 1998), it becomes even more important to understand the cognitive correlates of this virus before it converts to full-blown AIDS. The literature examining HIV-positive status and progressive neuropsychological impairment is growing but continues to require description (Castellon et al., 2000), especially prior to stages of frank dementia. To date, research findings have variably found a constellation of impairment in later stages, described as HIV-related dementia. This decreased cognitive state is estimated to occur in six to 30% of individuals who are HIV-positive (Snyder & Nussbaum, 1998). The greater level of decline may be especially true for the veteran population, with their elevated rate of comorbid problems. HIV-related dementia impairment includes psychomotor slowing, impaired attention and concentration, and memory deficits, with recognition and language deficits typically spared until later stages (Snyder & Nussbaum, 1998). Research that identifies specific cognitive changes early in the disease process may allow practitioners to flag, with a simple screening, underlying dysfunction related to the HIV disease process. This may serve as a basis for more extensive and expansive neuropsychological testing in a cost-effective manner for the veteran population, as well as for other clinical populations. Benedict et al. (2000) looked at HIV-positive individuals and found that despite disease stage, drug use, depression, or estimated premorbid IQ, the individuals who were cognitively impaired were more likely to be unemployed and fail tasks related to social and medication management. Thus, early identification of HIV-positive individuals with neuropsychological impairment is essential for treatment and quality of life issues. This study represents an early report of ongoing data collection exploring the progressive decline in neuropsychological status of HIV-positive individuals. It was hypothesized there would be a mild, but measurable, decline in the areas identified in HIV-related dementia for 12 HIV-positive veteran patients relative to 21 HIV194 0278-2626/02 $35.00  2002 Elsevier Science (USA) All rights reserved.

ROTMAN INSTITUTE ABSTRACTS

195

negative and neurologically healthy veteran patients. This study included measures of language, attention and concentration, memory and learning, visual–spatial organization, and intellectual and executive functioning. Method

Subjects Subjects were identified from an ongoing archival database collected from the Neuropsychology Clinic at the VA Western New York Healthcare System (VAWNYHS). A total of 33 adult subjects were included in the study. Twelve subjects were HIVpositive. This population included 10 individuals with comorbid substance abuse problems. Twenty-one HIV-negative subjects were matched for age and were without significant neurological, psychiatric, or medical impairment. All subjects over the age of 60 were excluded. The HIV-positive subjects consisted entirely of males and ranged in age from 34 to 51 years (M ⫽ 42.83, SD ⫽ 5.13). Three of the subjects were Caucasian, 8 were African American, and 1 was Hispanic. Their level of education ranged from 10 to 18 years (M ⫽ 13.00, SD ⫽ 5.13). Years of grade retention ranged from zero to one (M ⫽ 0.19, SD ⫽ 0.29). Eleven of the subjects were right-handed and 1 was lefthanded. The HIV-negative subjects consisted of 17 males and 3 females. Twenty of the subjects were Caucasian, and 1 was African American. They ranged in age from 36 to 53 years (M ⫽ 45.14, SD ⫽ 4.91). Their level of education ranged from 9 to 19 years (M ⫽ 13.90, SD ⫽ 2.12). Grade retention ranged from zero to three (M ⫽ 0.24, SD ⫽ 0.70). Nineteen of the subjects were right-handed, and 2 were left-handed. Instruments The measures included in this study are part of a standardized battery used at the VAWNYHS. They include measures of language, attention and concentration, memory, visual–spatial organization, and intellectual and executive functioning. Language. • Categorical generative naming test (animals: 60 in.); • Confrontation naming (modified Boston Naming Test; 18 items); and • Auditory comprehension (8 item analog to the Token Test) Attention and concentration. • Digit span; and • Trail Making Test, part A Memory and learning. • Immediate memory for a paragraph (analog to WMS—III logical memory); • Delayed memory (30 min) for a paragraph; • Immediate memory for a moderately complex figure (simpler analog to Rey– Osterrieth figure; 6-point scoring); and • Delayed memory (30 min) for the figure. Visual–spatial organization. • Design copy; and • Clock drawing and setting.

196

ROTMAN INSTITUTE ABSTRACTS

Intellectual/executive functioning. • Trail Making Test, part B; and • WAIS-R—modified similarities and arithmetic (every other item; prorated).

Procedure All HIV-positive subjects were referred to the Neuropsychology Clinic for cognitive evaluation of baseline functioning. Their test scores were retrieved from a database of all clinical referrals, and their charts were examined to verify information. Control group scores were also retrieved from the archival database.

Analyses Descriptive statistics were examined for all variables. Variables with skewed or kurtotic distributions were transformed by square root to conform to normal distribution assumptions. Zero-order correlations were examined to explore the relationship between demographic variables and neuropsychological test scores. Analyses of variance (ANOVA) were then used to explore group differences for demographic variables. Analyses of covariance, controlling for ethnicity, were used to explore mean differences between groups on the neuropsychological measures.

Results

Correlations Ethnicity was significantly correlated with confrontation naming (p ⬍ .027), digit span (p ⬍ .002), digits backward (p ⬍ .033), incidental memory for design (p ⬍ .048), visual learning ( p ⬍ .041), delayed visual memory (p ⬍ .025), TMT-A time (p ⬍ .030), and errors (p ⬍ .009), TMT-B time ( p ⬍ .001) and errors ( p ⬍ .001), and clock setting ( p ⬍ .01). Highest completed educational grade was significantly correlated with all visual memory and learning measures. Age, gender, number of grade retentions, and handedness were not significantly correlated with any of the neuropsychological outcome measures.

Analyses of Variance Ethnicity was the only demographic variable that significantly differed between the HIV-positive group and the HIV-negative group (F ⫽ 26.437, p ⬍ .001). Therefore, subsequent analyses comparing neuropsychological functioning between the two groups included ethnicity as a covariate. Table 1 provides the means, F values, and significance levels for comparisons of group differences on neuropsychological performance, controlling for race. The HIVpositive group performed significantly worse than the HIV-negative group on most measures including confrontation naming, digits backward, trail making test parts A and B, immediate and delayed verbal memory, immediate and delayed visual memory, visual learning, clock planning, arithmetic, and similarities. The groups did not differ on categorical verbal fluency, auditory comprehension, digits forward, verbal learning, or design copy.

197

ROTMAN INSTITUTE ABSTRACTS

TABLE 1 Mean Comparisons of PTSD and Control Groups on Neuropsychological Measures HIV-positive Area of functioning Language Confront. naming a Fluency animals Auditory comp. Attention and Concentration Digit forward Digit backward Trail Making Test A Time

Mean

SD

Mean

SD

F

p

16.50 17.83 7.17

1.93 4.20 1.47

17.75 23.85 7.75

0.55 5.15 0.55

6.904 1.737 2.148

.014 .198 .154

5.50 3.25

1.09 1.06

6.25 4.30

1.02 0.92

3.865 8.719

.059 .006

56.42 0.011 0.50

29.28

29.29

10.84

7.405

0.67

0.26

0.22

14.649

.001

4.21 5.11 5.60 2.25 1.58 1.87

15.95 17.70 18.70 4.00 4.57 5.00

2.42 2.89 2.00 1.52 1.79 1.38

10.616 2.881 6.784 8.002 7.661 12.121

.000 .072 .004 .011 .012 .002

0.98 1.27 1.48

5.76 3.89 3.86

0.54 0.32 0.82

2.290 8.182 2.962

.147 .008 .096

1.47 1.63

5.10 11.95

0.97 1.76

23.345 26.624

.000 .000

84.57 1.66

62.29 0.68

18.64 0.40

22.404 13.889

.000 .001

Errors a Memory Log. memory—Trial 1 12.50 Log. memory—Trial 2 14.50 Log. memory—Delay 13.58 Design incidental 2.55 Design learning 2.75 Design—delay recall 3.20 Visual–spatial organization Design copy 5.18 2.83 Clock planning a Clock setting 3.00 Intellectual/executive functioning Arithmetic 2.82 Similarities 8.36 Trail Making Test B Time 152.56 1.67 Errors a a

HIV-negative

Variable was transformed by square root due to skewed or kurtotic distribution.

Discussion

The present study examined the neuropsychological functioning of HIV-positive veterans relative to age matched HIV-negative veterans. These two groups differed in a number of ways. In comparison to the HIV-negative group, the HIV-positive group was significantly compromised in tasks requiring attention and concentration, immediate and delayed verbal recall, immediate and delayed visual recall, visual learning, and tasks requiring psychomotor speed, while a number of language tasks were left intact. This was similar to dysfunction often seen in HIV-related dementia cases. However, this group was also significantly more impaired in confrontation naming, planning, mental calculations, and abstract thought when compared to the HIV-negative group. Thus, these findings are not entirely consistent with later stage impairment. These findings may be related to the additive effects of comorbid substance abuse found in the majority of our HIV-positive subjects. The degree of impairment found in this population is similar to reports regarding other HIV-positive and substance abusing individuals. For example, Durvasula et al. (2000) examined the neuropsychological performance of HIV-positive African American, gay, and bisexual men who were abusers of non-IV drugs, such as cocaine. They too found dysfunction related to psychomotor speed and verbal memory and attributed their findings to an interaction of HIV status and substance use.

198

ROTMAN INSTITUTE ABSTRACTS

The findings of this study indicate that individuals in this population are at a significantly greater risk for cognitive impairment. This may impact their medical treatment compliance and skills for daily living. Thus, comprehensive neuropsychological testing of individuals who are HIV-positive is warranted to help maintain their quality of life.

REFERENCES Benedict, R. H. B., Mezhir, J. J., Walsh, K., & Hewitt, R. G. (2000). Impact of human immunodeficiency virus type-1-associated cognitive dysfunction on activities of daily living and quality of life. Archives of Clinical Neuropsychology, 15, 529–534. Castellon, S. A., Hinkin, C. H., & Myers, H. F. (2000). Neuropsychiatric disturbance is associated with executive dysfunction in HIV-1 infection. Journal of the International Neuropsychological Society, 6, 336–347. Durvasula, R. S., Myers, H. F., Satz, P., Miller, E. N., Morganstern, H., Richardson, M. A., Evans, G., & Forney, D. (2000). HIV-1, cocaine, and neuropsychological performance in African American men. Journal of the International Neuropsychological Society, 6, 322–335. Hinken, C. H., Catellon, S. A., van Gorp, W. G., & Satz, P. (1998). Neuropsychological features of HIV disease. In W. van Gorp & S. Buckingham (Eds.), Practitioner’s guide to the neuropsychiatry of HIV/AIDS (pp. 1–41). New York: Guilford Press. Snyder, P. J., & Nussbaum, P. D. (1998). Clinical neuropsychology. Washington, DC: American Psychological Association.

2. Control of Saccades in Parkinson’s Disease

I. T. Armstrong, F. Chan, R. J. Riopelle, and D. P. Munoz Parkinson’s patients (PD) made pro- and antisaccades: In the no-delay condition, the target appeared concurrent with the GO signal. In the delay condition, the target appeared before the signal for movement. Second, we probed spatial working memory in PD. Subjects looked to the remembered locations of sequential targets. In the no-delay prosaccade condition, PD had faster reaction times, made more express saccades, and exhibited hypometria. In the nodelay antisaccade condition, PD had longer reaction times and made more direction errors. In the delay tasks, PD made more direction errors and had more difficulty withholding a movement. PD made more sequencing errors in the spatial working memory task. These findings are consistent with a basal ganglia pathophysiology influencing eye movement processing in the frontal cortex.  2002 Elsevier Science (USA)

Report Saccadic eye movements are controlled by many brain areas including the cerebral cortex, basal ganglia, brain stem, and cerebellum (Leigh & Zee, 1999). Therefore, there is considerable likelihood that patients with neurological or psychiatric disorders will have difficulties controlling aspects of saccadic eye movements. Parkinson’s disease (PD) is a condition of degeneration of dopaminergic neurons in the substantia nigra pars compacta resulting in progressive basal ganglia dysfunction. Because important eye movements pathways travel through the basal ganglia, aspects of saccade control should be impaired by the disease progression. PD characterization includes difficulty initiating movements and in sequencing items (Lezak, 1995). Thus, PD patients would be expected to have slower saccadic

ROTMAN INSTITUTE ABSTRACTS

199

reaction times (SRT) for all eye movement tasks and they would be expected to perform tasks involving eye movements to a series of recalled locations less well than matched controls. Saccadic eye movements can be reflexive (e.g., an eye movement made toward a suddenly appearing eccentric visual target—a prosaccade) or they may be generated voluntarily (e.g., a saccade away from an eccentric target— an antisaccade). We tested PD patients in tasks involving reflexive and voluntary eye movements to visible targets and to recalled targets presented alone and in a sequence. PD patients and controls performed three experiments. In the first study, subjects performed one block of prosaccades followed by two blocks of antisaccades. PD deficits are expected because of the hypothesized role of the frontal cortex and basal ganglia in saccade suppression and the generation of voluntary antisaccades. In the second experiment, subjects performed interleaved pro- and antisaccades in a delayed saccade task. The delay required subjects to maintain fixation stimulus and ignore the eccentric target until a later GO signal to move. In the third study, subjects moved their eyes to the recalled locations of three briefly displayed targets; thus we tested spatial working memory and sustained attention with this task. Methods Ten PD patients and 10 age- and sex-matched controls participated in the experiments. Control subjects had no known neurological, psychiatric, or visual disorder other than refractive errors. Three separate experimental sessions were employed: the immediate pro- and antisaccade tasks, the delayed pro- and antisaccade task, and the sequential memoryguided delay task. Eye movements in the pro- and antisaccade tasks were measured using DC electrooculography (EOG). Subjects sat in complete darkness facing a translucent visual screen located 100 cm away. A red light-emitting diode (LED; 2.0 cd/m 2) was backprojected onto the screen and acted as the central fixation point (FP) in the no-delay task, and a green LED (2.0 cd/m 2) was also used as a central FP in the delayed proand antisaccade task. Two eccentric target LEDs (5.0 cd/m 2) were positioned 20° to the right and left of the FP. Immediate pro- and antisaccade task. Subjects were instructed to look toward an eccentric target in the prosaccade task and away from the eccentric target in the antisaccade task. The stimuli were identical in both tasks. Each trial began with the appearance of the central FP. After 1 s, one of two events occurred. During the overlap condition, the FP remained visible while the eccentric target appeared either to the left or to the right. In the gap condition, the FP was extinguished and following a 200-ms delay; the eccentric target appeared to the left or right. The eccentric target remained visible for 1 s and then all LEDs disappeared. Trials were presented in blocks of 120; each participant performed one block of prosaccade trials followed by two blocks of antisaccade trials with no practice trials. In the delayed pro- and antisaccade task, participants fixated the central FP that was either red or green. After 1 s, the eccentric red target appeared randomly to the right or left. After a variable interval (200 to 1000 ms), the FP then disappeared. Subjects were instructed to wait until the FP disappeared before making an eye movement. Subjects looked toward the target if the FP was red and away from the target if the FP was green. Each subject performed three blocks of 160 trials in which instruction (pro vs anti), target direction (left vs right) and delay period (200, 400, 600, 800, 1000 ms) varied randomly. In the sequential memory-guided task, eye movements were measured using a

200

ROTMAN INSTITUTE ABSTRACTS

video-based infrared eyetracker, sampling horizontal and vertical eye position at 250 Hz. A white circular target stimuli (0.2 cd/m 2) was centred on a black background and was used as the FP. Green circular stimuli (0.2 cd/m 2) were the eccentric targets. Memory-guided sequential delay task trials began with the presentation of a white FP at the center of the computer monitor. Subjects pressed a button to initiate each trial which began with a period of random delay (200 to 1000 ms). After the delay, a series of three green target stimuli appeared in rapid sequence in three of the four quadrants of the visual field. The location of the stimuli varied randomly with the restriction that a quadrant was used only once in any trial. Each target appeared for 100 ms with no interstimulus interval. The FP remained lit for a variable interval after the disappearance of the final target (0, 600, 1200, or 1800 ms). Once the FP disappeared, subjects were instructed to move their eyes to the target locations in the exact sequence of target presentation. Subjects were given up to 2 s to complete the sequence of saccades before the central FP reappeared to indicate the start of the next trial. Two blocks of 96 trials and 20 practice trials were given to each subject. Data analysis. In the immediate pro- and antisaccade task, saccadic reaction time (SRT) was measured as the time from target appearance to the onset of the first saccade. In the delayed pro- and antisaccade task, SRT was measured from the disappearance of the central FP. A saccade was scored as correct if the measured SRT movement was in the correct direction. For the delay tasks, timing errors consisted of saccades initiated after target appearance but before the GO signal (i.e., FP disappearance). In both the immediate and the delayed tasks, reaction latencies between 90 and 1000 ms were used to compute mean SRT, the percentage of express saccades (90 to 140 ms; Fischer et al., 1993), percentage of direction errors, and percentage of timing errors. We measured the mean amplitude of the first correct prosaccade with latencies between 90 and 1000 ms. For saccades with amplitudes between 18 and 21°, the mean peak velocity and duration were also calculated. In the memoryguided sequential delay task, we calculated timing errors, sequence errors, the number of saccades executed during each trial, and the accuracy of each subject to fixate the correct locations of the three flashed targets. Results and Discussion In the immediate prosaccade task, PD subjects had shorter mean SRT and more express saccades. We attribute this reduction in SRT to a facilitation in generation of reflexive saccades due to basal ganglia dysfunction. Consistent with previous findings, PD patients displayed significant hypometria in the amplitude of the first saccade. In the immediate antisaccade task, PD patients had increased SRT and made a greater percentage of direction errors. These findings are consistent with a dysfunction in the basal ganglia affecting the ability of PD patients to suppress reflexive responses and initiate voluntary behaviors. In the delayed pro-/antisaccade tasks, PD subjects showed more timing errors (eye movement initiated before the disappearance of the fixation point) and made several small saccades to reach the target. In the memory-guided sequential delayed task, PD patients displayed greater difficulty in delaying their eye movements until FP disappearance. PD patients were also less successful at moving their eyes in the correct order of target appearance and they were also less accurate in directing their eyes in the correct direction. In conclusion, our results show that PD patients have significant deficits in several eye movement tasks that reveal difficulties in generating voluntary behaviors and using spatial working memory. These deficits are consistent with a pathophysiology in the basal ganglia that can influence frontal lobe function.

ROTMAN INSTITUTE ABSTRACTS

201

REFERENCES Fischer, B., Weber, H., Biscaldi, M., Aiple, F., Otto, P., & Stuhr, V. (1993). Separate populations of visually guided saccades in humans: Reaction times and amplitudes. Experimental Brain Research, 92, 528–541. Leigh, R. J., & Zee, D. S. (1999). The neurology of eye movements (3rd ed.). Oxford: Oxford University Press. Lezak, M. D. (1995). Neuropsychological assessment (3rd ed.). Oxford: Oxford University Press. This is doi:10.1006/brcg.2001.1464.

3. The Relation between APOE Status and Neuropsychological Memory Test Performance: An Analysis of the Canadian Study of Health and Aging

J. Klages and J. D. Fisk This study examined the relationship between two risk factors for dementia, the apolipoprotein (APOE) ε4 allele and poor memory test performance. Participants were from the Canadian Study of Health and Aging, a 4-year longitudinal population-based study. Persons with no cognitive impairment who had an ε4 allele but whose memory was average or better were not at increased risk of developing dementia after five years. Risk was increased for those with below average memory and no ε4 allele, but was particularly increased for those with below average memory and an ε4 allele. While the APOE ε4 allele was associated with slightly lower memory test performance for persons without cognitive impairment at baseline, it only increased their risk of developing dementia if their memory was below average.  2001 Elsevier Science (USA)

Report Identifying Alzheimer disease (AD) and related dementias closer to their onset is increasingly important as the population ages and as new treatments are developed. A number of risk factors have been discovered and disputed over the past few decades, one of which is the apolipoprotein (APOE) ε4 allele. A meta-analysis of community and clinical data reported that, when compared with the ε3/ε3 genotype, the odds ratio for having or developing AD ranges from 2.7 to 3.2 for those with the ε3/ε4 genotype and from 12.5 to 12.9 for those with the ε4/ε4 genotype (APOE and Alzheimer Disease Meta Analysis Consortium, 1997). Despite the importance of APOE ε4 as a risk factor for dementia, its utility in prediction is limited because not everyone with an ε4 allele develops dementia, and the majority of persons with dementia do not have an ε4 allele (Rebeck & Hyman, 1999). Thus, other potential risk factors must be considered. Another risk factor for developing dementia in longitudinal studies is the presence of cognitive impairment at baseline. Since impairment of memory is a core feature of dementia and the most obvious early feature of AD, it is not surprising that impairment on tests of Delayed Recall have shown the most consistent relationship with the development of AD. Moreover, in a 2-year prospective clinic-based study, Tierney and colleagues (1996) found that for persons who did not have dementia, having an APOE ε4 allele was associated with increased risk of developing AD only when relatively poor memory test performance was also considered. The purpose of the current investigation was twofold. First was an examination of the potential relation between the presence of the apoe ε4 allele and neuropsychological memory test performance in a population-based sample of elderly persons

202

ROTMAN INSTITUTE ABSTRACTS

without clinical evidence of cognitive impairment. The second was to examine how the risks associated with poor memory test performance and the risks associated with the apoe ε4 allele are related when one looks at the development of dementia after 5 years. Method All participants were enrolled in the Canadian Study of Health and Aging (CSHA), a 5-year longitudinal study of the prevalence and incidence of dementia (The Canadian Study of Health and Aging Working Group, 2000). The initial sample of 10,263 participants over age 64 was selected from across Canada in a randomized and stratified manner. APOE genotyping was conducted in 1624 participants. Each participant from the community was screened using the Modified MiniMental Examination. Those with scores less than 78/100 received a clinical examination that included neuropsychological testing, as did a random sample with scores greater than 77/100. The memory tests included the Buschke Cued and Selective Reminding Test (BCRT) and the Rey Auditory Verbal Learning Test (RAVLT) (Tuokko, Kristjansson, & Miller, 1995) which were employed in the analyses described below. A consensus diagnosis classified participants with either no cognitive impairment (NCI), cognitive impairment no dementia (CIND), or dementia. All participants in this study had a clinical assessment, neuropsychological testing, a diagnosis of NCI at CSHA-1 and APOE genotyping, and a diagnosis of NCI or dementia 5 years later (CSHA-2). Results Of those who completed the BCRT at CSHA-1, 68 participants had an ε4 allele and 311 did not. Those with an ε4 allele had significantly lower free recall on all 4 trials of the BCRT, including Delayed Recall (F[1, 375] ⫽ 18.5, p ⫽ .000). Age was a significant covariate (F[1, 375] ⫽ 31.1, p ⫽ .000) but education was not (F[1, 375] ⫽ 1.4, p ⫽ .233). For the RAVLT 62 participants had an ε4 allele and 253 did not. Those with an ε4 allele performed less well on all 6 RAVLT recall trials but this was a nonsignificant trend (F[1, 311] ⫽ 3.6, p ⫽ .057). Both age (F[1, 311] ⫽ 31.5, p ⫽ .000) and education (F[1, 311] ⫽ 35.9, p ⫽ .000) were significant covariates. The odds ratio (OR) for the development of dementia after 5 years was determined using logistic regression. The OR for the NCI participants with an APOE ε4 allele was 2.79 (95% CI 1.90–4.10). The OR for participants with below average BCRT Trial 3 scores (the final learning trial) was 5.10 (95% CI 2.45–10.63) while the OR for below average Delayed Recall was 4.37 (95% CI 2.15–8.88). Below average performance on the RAVLT Trial 6 (free recall following a distracter word list) was also a significant risk factor with an OR of 3.98 (95% CI 1.55–10.22). For the next analysis, participants without an APOE ε4 and average or above average memory test performance (i.e., no risk factors) served as the referent group. For the BCRT Trial 3 and Delayed Recall, and for the RAVLT Trial 6, participants with an APOE ε4 allele who had average or above memory were not at increased risk of developing dementia. However, participants without an APOE ε4 allele but below average memory test performance and participants with an APOE ε4 allele and below average memory test performance were at increased risk (see Table 1). When the referent group was changed to participants with an APOE ε4 allele and below average memory test performance (i.e., both risk factors), the ORs of developing dementia for the other three groups were significantly lower in the BCRT Delayed Recall and the RAVLT Trial 6 analyses. The exception to this pattern was the BCRT Trial 3

203

ROTMAN INSTITUTE ABSTRACTS

TABLE 1 Odds Ratio for Persons with Dementia (D) and Persons with No Cognitive Impairment (NCI) for the Development of Dementia after 5 Years BCRT Trial 3 N Rounded mean Score Odds ratio (95% CI) No ε4, high score n (for D and NCI) No ε4, low score n (for D and NCI) ε4, high score n (for D and NCI) ε4, low score n (for D and NCI) Age (years) Education

OR OR OR OR

BCRT Delay

RAVLT Trial 6

259 9

258 10

224 7

D ⫽ 17; NCI ⫽ 154 Referent group D ⫽ 14; NCI ⫽ 22 4.00 (1.63–9.81)† D ⫽ 5; NCI ⫽ 29 2.11 (0.69–6.47) D ⫽ 10; NCI ⫽ 8 11.25 (3.67–34.50)‡ 1.13 (1.06–1.21)‡ 0.98 (0.94–1.03)

D ⫽ 17; NCI ⫽ 134 Referent group D ⫽ 14; NCI ⫽ 42 2.87 (1.24–6.62)* D ⫽ 3; NCI ⫽ 31 1.30 (0.34–5.01) D ⫽ 11; NCI ⫽ 6 17.04 (5.02–57.80)‡ 1.15 (1.08–1.23)‡ 0.99 (0.95–1.03)

D ⫽ 9; NCI ⫽ 102 Referent group D ⫽ 18; NCI ⫽ 48 2.72 (1.07–6.89)* D ⫽ 3; NCI ⫽ 26 1.33 (0.31–5.62) D ⫽ 9; NCI ⫽ 9 12.02 (3.48–41.49)‡ 1.12 (1.05–1.21)† 0.85 (0.75–0.96)†

* p ⬍ .05. † p ⬍ .01. ‡ p ⬍ .001.

in which participants with an APOE ε4 allele and average or above memory test performance did not have significantly lower risk of dementia. Discussion The first purpose of this study was to determine the relation between memory test performance and the presence of the APOE ε4 allele in a population-based sample of persons with no cognitive impairment. For the BCRT, participants with an APOE ε4 allele had significantly lower scores but for the RAVLT there was only a trend toward lower performance for those with the APOE ε4 allele. While these findings support previous reports that nondemented persons with an APOE ε4 allele perform less well on tests of memory, the magnitude of the differences between groups was small and clinically insignificant. Since the APOE ε4 allele is a risk factor for dementia, a population-based sample of persons with the APOE ε4 genotype might contain a higher proportion of persons in the early stages of dementia, thereby decreasing their average memory test scores. However, as all participants with evidence of cognitive impairment (i.e., CIND) were removed before analyses, even those in the early stages of dementia should have been excluded from in our sample. An alternative explanation is the presence of developmental differences in cognitive abilities that reflect APOE genotype. However, developmental differences in cognitive abilities would have been expected to affect attained education levels and this was not the case in the BCRT analyses. While education was a significant covariate in the RAVLT analyses, APOE ε4 had a nonsignificant effect on this measure. Thus, while in cross-sectional analyses APOE genotype is related to memory test performance, further studies will be required to resolve the source of this effect. The second purpose of this study was to examine how the risks associated with poor memory test performance and the presence of the APOE ε4 allele combine to affect the development of dementia after 5 years. Having an APOE ε4 allele was a significant risk factor for the development of dementia in the total study sample as was relatively poor memory test performance at CSHA-1. However, for those partici-

204

ROTMAN INSTITUTE ABSTRACTS

pants whose memory test performance was at or above the mean for the sample, having an APOE ε4 allele was not associated with an increased risk of dementia. These findings replicate those of Tierney et al. (1996) but extend these findings to a population-based sample of elderly persons without clinical evidence of cognitive impairment. Even for participants without an APOE ε4 allele, low memory performance significantly increased the risk of developing dementia, a finding consistent with the fact that most persons in the population with dementia do not carry the APOE ε4 allele. What is most important is the finding that having an APOE ε4 allele and low memory test performance conferred the greatest level of risk. Thus, although the APOE ε4 allele may not be an issue for persons whose memory test performance is average or above, it confers an increased risk of developing dementia to those with low memory test performance. Thus, the APOE ε4 allele increased the risk of developing dementia for persons already at risk of developing dementia by virtue of poor memory test performance. The implications of these findings are twofold. First, in future longitudinal studies a combined approach that employs both genetic and neuropsychological testing seems likely to be both sensitive and specific in identifying persons at risk for the development of dementia. Second, improving the predictive validity of APOE genotyping, by combining it with neuropsychological testing, begins to approach the necessary conditions for using this genetic testing in clinical practice. While the current findings require replication in other population-based samples, they do suggest that it may soon be possible to use genetic and neuropsychological testing in a predictive manner. ACKNOWLEDGMENTS

The data reported in this article were collected as part of the Canadian Study of Health and Aging. The core study was funded by the Seniors’ Independence Research Program, through the National Health Research and Development Program (NHRDP) of Health Canada (Project 6606-3954-MC(S)). Additional funding was provided by Pfizer Canada Incorporated through the Medical Research Council/Pharmaceutical Manufacturers Association of Canada Health Activity Program, NHRDP (Project 6603-1417-302(R)), Bayer Incorporated, and the British Columbia Health Research Foundation (Projects 38 (93-2) and 34 (96-1)). The study was coordinated through the University of Ottawa and the Division of Aging and Seniors, Health Canada. REFERENCES APOE and Alzheimer Disease Meta Analysis Consortium. (1997). Effects of age, sex and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. Journal of the American Medical Association, 278, 1349–1356. Rebeck, G. W., & Hyman, B. T. (1999). Apolipoprotein E and alzheimer disease. In R. D. Terry, R. Katzman, K. L. Bick, & S. S. Sisodia (Eds.), Alzheimer disease (2nd ed.). Philadelphia: Lippincott Williams & Wilkins. Tierney, M. C., Szalai, J. P., Snow, W. G., Fisher, R. H., Tsuda, T., Chi, H., Mclachlan, D. R., & St. George-Hyslop, P. H. (1996). A prospective study of the clinical utility of apoe genotype in the prediction of outcome in patients with memory impairment. Neurology, 46, 149–154. The Canadian study of health and aging working group. (2000). The incidence of dementia in Canada. Neurology, 55, 66–73. Tuokko, H., Kristjansson, E., & Miller, J. (1995). Neuropsychological detection of dementia: An overview of the neuropsychological component of the Canadian study of health and aging. Journal of Clinical and Experimental Neuropsychology, 17, 352–373. This is doi:10.1006/brcg.2001.1465.

ROTMAN INSTITUTE ABSTRACTS

205

4. ‘‘Clockness’’ in the Detection of Dementia: A Semantic–Conceptual Effect

M. M. Saling, C. M. Maccuspie-Moore, V. A. Anderson, and E. Chiu We studied aspects of clock cognition that might underlie the sensitivity of the CDT in screening for dementia of the Alzheimer type (DAT). Two groups, 15 patients with mildmoderate DAT and 15 controls, were assessed with the CDT and specially designed tests of clock-related cognition. Patients were impaired on the CDT, but they did not differ from controls in copying a clock face or selecting the correct representation of a given time. Patients were worse than controls at distinguishing between clock and nonclock objects, detecting anomalies in clocks, and in setting time irrespective of response format. These findings suggest that semantic–conceptual aspects of clock-related cognition are important in discriminating between patients with DAT and controls.  2002 Elsevier Science (USA)

Report The Clock Drawing Test (CDT) has repeatedly proven its efficacy in the detection of dementia of the Alzheimer type (DAT). There is good reason to believe that the CDT is multicomponential. Nevertheless, the particular aspects of clock cognition that are responsible for the CDT’s sensitivity to early dementia have not been identified. We developed a number of measures of clock-related cognition designed to isolate constructional and semantic–conceptual aspects of CDT performance, with the aim of studying their relative effectiveness in discriminating between patients with mild–moderate DAT and normal controls. Participants Two groups were studied. The first (NC) consisted of 15 normal controls, 6 males and 9 females with a mean age of 72.9 years (SD ⫽ 6.5 years; range 61–83 years) and an MMSE score of not less than 26. They were recruited through in-hospital advertisements or by approach to spouses of patients included in the dementia group. The second group (DAT) was recruited through referrals to the Impaired Cognition Clinic for the Elderly (ICCE) at St. George’s Hospital at Kew, Victoria, Australia. This service evaluates patients with suspected cognitive impairment referred by general medical practitioners in the Inner Eastern Region of Melbourne. Prior to inclusion, an ICCE psychogeriatrician, neuropsychologist, and occupational therapist-assessed patients, following a home-based assessment by a community nurse. Patients also underwent standard medical, laboratory, and neuroimaging investigations, which included MRI or CT brain scan, a blood dementia screen, urine analysis, and chest X-ray. Information derived from all assessments, clinical examinations, and investigations was reviewed by the psychogeriatrician and diagnosis was made in terms of DSM-IV and NINCDS-ADRDA criteria. Only potential DAT participants with carers were approached for inclusion in the study. Informed consent was obtained from all participants prior to any testing. In the case of potential DAT participants dual informed consent from patient and carer was obtained. Both groups met the following inclusion and exclusion criteria: age of 60 years or older; visual and auditory function was adequate for performance of all tasks; there was no history of cerebral insult, significant intoxication, or known cerebral pathology; there was no motor impairment; and no participant had a score of 14 or more on the Geriatric Depression Scale (this cutoff has a specificity of 100%, while retaining an acceptable sensitivity of 80%). DAT participants were included only if they were living at home, with a mild to

206

ROTMAN INSTITUTE ABSTRACTS

moderate dementia. They were able to manage personal hygiene and basic activities of daily living independently or with minimal assistance and participate in routine social interaction. DAT subjects were also required to have sufficient cognitive ability to attempt all assessment tasks. The final DAT group consisted of 15 participants, 4 males and 9 females, with a mean age of 76.8 years (SD ⫽ 5.23 years; range 67–84 years). There were no significant group differences on age, gender, level of education, and current level of depression. The NC (M ⫽ 121.2; SD ⫽ 9.88) and DAT (M ⫽ 115.5; SD ⫽ 10.5) groups were also well matched for ‘‘premorbid’’ IQ, estimated by the National Adult Reading Test (t 28 ⫽ 1.54, ns). Current IQ estimated by Reynold’s Short Form of the Wechsler Adult Intelligence Scale-Revised (WAIS-R), however, was lower in the DAT (M ⫽ 89.8; SD ⫽ 10.08) than in the NC (M ⫽ 113.4; SD ⫽ 12.6) participants (t 28 ⫽ 5.66, p ⬍ .0001). DAT participants were also worse at copying line drawings (t 28 ⫽ 2.22, p ⫽ .034) and freehand drawing (t 28 ⫽ 2.51, p ⫽ .18) than the NCs. Clock Tasks Clock drawing test. Participants were given a sheet of paper printed with a circle 4 in. in diameter. They were asked to fill in the numbers that occur around the face of a clock and then to asked to draw the face of a clock, indicating the numbers, and then to draw in the hands at 10 min past 11. Responses were scored according to the Shulman method (Shulman, Gold, Cohen, & Zucchero, 1993). Scores ranged from 1 (perfect performance) to 6 (no reasonable representation of a clock). Clock copy. Participants were presented with a computer-generated line drawing of a clock, 4 in. in diameter, and were asked to draw a copy on a separate sheet of paper. We developed a 10-point scoring protocol, which separately evaluated representation of the shape, central point, numerals, and hands of the copy. Points were deducted for planning difficulties and reduced graphomotor skill. The maximum score was 10. Clock decision. Fifty color photographs (5 ⫻ 7 in.) were presented to participants one at a time. Half of these were pictures of valid clocks, while the remaining half represented nonclock objects that had features in common with clocks (for example, a compass). Order of presentation of the 50 photographs was randomly determined, but the same order was maintained for each participant. Score was the number correct of 50. Clock anomalies. This task consisted of 15 line drawings of clocks, each of which included an anomalous feature (such as repeated or superfluous numbers, additional hands, or shape distortions). Participants were required to identify the anomaly. Score was the number correct of 15. Time matching. This task assessed time reading ability in a forced choice format. Participants were presented with a pair of printed clocks indicating different times and were asked to choose the clock that represented a verbally delivered specified time. There were 20 clock pairs, all showing different times. A new time was specified for each pair. Time setting and drawing. This task consisted of two parts. In the first part, participants were requested to set the time on a real teaching clock by moving the hour and minute hands. Ten times of varying difficulty were presented verbally, half in standard analogue format (for example, half past seven), and half in digital format (for example, nine twenty-five). The second part required participants to set the time on a printed clock template by drawing in the hands. The template had markings to indicate positions of the hours, but numbers were not shown. Ten specified times were given as in the first part.

ROTMAN INSTITUTE ABSTRACTS

207

Results and Discussion As expected, the clock drawing test (CDT) differentiated between the NC (M ⫽ 1.5; SD ⫽ 0.64) and DAT (M ⫽ 2.4; SD ⫽ 0.91) groups (t 28 ⫽ 3.25, p ⫽ .003). Discriminability, or effect size, was expressed as ε 2 and assumed a value of 32%. Clock copy, however, did not differentiate between the groups (NC, M ⫽ 9.07, SD ⫽ 1.10; and DAT, M ⫽ 8.33, SD ⫽ 1.11; ε 2 ⫽ 5%). This pattern suggests that clock construction abilities per se do not explain the difference between patients and controls. Time matching also failed to differentiate between the NC (M ⫽ 20.0; SD ⫽ 0.00) and DAT (M ⫽ 19.4; SD ⫽ 1.59) participants, with ε 2 of less than 10%. Nevertheless, time setting (t 28 ⫽ 4.44; p ⬍ .0001; ε 2 ⫽ 40%) and time drawing (t 28 ⫽ 4.32; p ⬍ .0001; ε 2 ⫽ 40%) were substantially worse in the DAT group. These findings imply that visuoperceptual and visuospatial abilities, as these apply to the configuration of a clock face, are not important in discriminating between patients and controls. Further, the requirement to draw the hands does not appear to be significant in time settings failures. DAT participants (M ⫽ 46.2; SD ⫽ 2.51) were less able than normal controls (M ⫽ 48.2; SD ⫽ 0.78) to decide whether a given photograph was a valid clock or not (t 28 ⫽ 2.72; p ⫽ .011; ε 2 ⫽ 24%). The majority of errors in both groups were false positives (86% for DAT and 95.6% for NC participants). The DAT group (M ⫽ 11.1; SD ⫽ 1.91) was also less able than the NC group (M ⫽ 13.5; SD ⫽ 1.13) to detect anomalies in clocks (t 28 ⫽ 4.2; p ⬍ .0001; ε 2 ⫽ 56%). Thus, the DAT participants are impaired on clock tasks that are not dependent on constructional abilities. The finding that clock anomalies had the largest discriminability is a reflection of the extent to which semantic–conceptual impairment plays a role in CDT performance in DAT. Successful performance is highly dependent on ability to retrieve a clock face template from semantic memory and to evaluate the extent to which the exemplar failed to meet criteria for ‘‘clockness.’’ Conclusion Historically the CDT has been regarded as a measure of visuoconstructional ability. Our findings suggest that semantic–conceptual aspects of clock-related cognition overshadows the role of constructional difficulties in detection of dementia. REFERENCES Shulman, K. I., Gold, D. P., Cohen, C. A., & Zucchero, C. A. (1993). Clock-drawing and dementia in the community: A longitudinal study. International Journal of Geriatric Psychiatry, 8, 487–496. This is doi:10.1006/brcg.2001.1466.

5. A Case Study of Strategic Infarct Dementia Investigated with the Cognitive Assessment System

S. M. McCrea and M. Scott Two subjects with brain lesions who were matched on demographic variables were tested on the Cognitive Assessment System (CAS). AK had been dependent on caregivers after a frontal aneurysm 6 years previously despite intact receptive and expressive language skills and motor functions. GM sustained multiple infarcts although he continued to function well on his own. Nonparametric analysis showed that AK’s T scores on CAS subtests were lower than that of GM’s based on a comparison with a heterogeneous group of brain-damaged pa-

208

ROTMAN INSTITUTE ABSTRACTS

tients ( p ⬍ .003). The CAS’s broad range of complexity of items within subtests, apparent sensitivity in differentiating rates of cognitive decline in dementia, and convergence with dementia rating scales suggests that it could be useful for assessment of strategic infarct dementia.  2002 Elsevier Science (USA)

Report The CAS is a test of performance competencies based on cognitive and neuropsychological theory in addition to its psychometric operationalization (Naglieri, 1999). The CAS incorporates tests of planning and attention specifically designed for use with children and adolescents. Adoption of a cognitive strategy is the most efficient means of completing the planning tasks while the attention tasks purportedly tap behavioral inhibitory and vigilance capacities. The simultaneous scale consists chiefly of visuospatial perceptual and reasoning tasks with the exception of Verbal– Spatial Relations that has both verbal and visuospatial features. In contrast the successive scale consists of verbal memory, repetition, and comprehension tasks. Some of the CAS tasks such as Matching Numbers are unique while other tasks are common derivations of adult neuropsychological tasks well normed for 5- to 18year-olds. Briefly, Matching Numbers is similar to Teuber’s visual search tasks used for assessing symptoms of frontal dysfunction. Planned Codes is similar to the Wechsler scale’s coding subtests (e.g., symbol search) but in the CAS version adoption of a specific strategy optimizes efficiency of completion. Planned Connections is almost identical to the commonly employed Trail Making Tests A and B. The Attention Scale consists of Expressive Attention that is virtually indistinguishable from the classic Stroop Test; Number Detection that is similar to many Line Cancellation tests; and Receptive Attention that is modeled on Posner’s lexical and physical identity match tasks. The simultaneous scale consists of Nonverbal Matrices that is modeled on the Raven’s Progressive Matrices; Verbal–Spatial Relations that is similar to the Tokens Test; and Figure Memory that is similar to Memory-for-Designs test. The successive scale consists of Word Series which is similar to other digit or word span tasks; the Sentence Repetition test which is like other sentence repetition tasks except that in the CAS version the sentences are semantically meaningless so as to emphasize syntactic processing. In Sentence Questions the examinee must rely on the serial positions of words within sentences to correctly answer the prompts since as with the Sentence Repetition test the cue sentences are semantically meaningless (see Naglieri, 1999, for a complete description of these tasks). In the context of neuropsychology the CAS’s planning and attention tasks have been shown to be effective for assessing executive dysfunction in a sample of traumatic brain injured children (see Naglieri, 1999). Also relevant is an earlier study by Das et al. (1995) comparing the performance of young and old groups of Down syndrome (DS) subjects to that of young and old groups of IQ-matched non-DS subjects. Subjects were tested on the Peabody Picture Vocabulary Test—Revised (PPVT-R), Mattis Dementia Rating Scale (MDRS), and the CAS (see Spreen & Strauss, 1998). Older individuals with DS (age range 50 to 62 years) performed worse than the other three groups on the PPVT-R, MDRS, and the CAS when using Wechsler or Stanford-Binet IQ as a covariate. Moreover, in the DS old individuals in whom there is a propensity for development of disease of the Alzheimer’s type (DAT) baseline IQ scores and MDRS scores were substantially correlated whereas in the young DS group these scores were not correlated. Together these results suggest that cognitive changes associated with DAT common in older persons with DS appear to be readily detectable with the CAS scales and individual subtests. Almkvist et al. (1996,

209

ROTMAN INSTITUTE ABSTRACTS

p. 47) have concluded that there is a ‘‘. . . need for multifunctional assessment of moderate to severe dementia free from both floor and ceiling effects . . .’’ for the purposes of diagnosis, rehabilitation, and assessment of effects of pharmacological treatment of dementia. This descriptive case study sought to determine whether the CAS might also be useful in the context of the neuropsychological assessment of vascular dementia, specifically strategic infarct dementia. Single infarct stroke patients (N ⫽ 14) ranging in age from 50 to 64 years (mean 55) were the comparative group from which were derived T scores for AK and GM. All patients were tested between 2 weeks and 1 month postadmission to the stroke unit. Patients AK and GM were matched for age, sex, education, ethnic grouping, occupational attainment, and region using the Barona Index of premorbid ability (Spreen & Strauss, 1998, p. 47). Both patients’ level of premorbid ability (VIQ) was estimated at between 87 and 111 at 15% confidence intervals. AK and GM were each assessed on the Mini-Mental Status Exam (MMSE) within days of the CAS assessment. AK’s MMSE score of 20 was in the impaired range compared to noncognitively impaired elderly patients with 8 or less years of education between the ages of 65 and 79 (5th percentile). The positive predictive power of the MMSE in discriminating Alzheimer patients from age-matched normal elderly patients at a criterion MMSE score of 20 has been previously shown to be 67% (Spreen & Strauss, 1998, pp. 71–72). GM’s MMSE score of 28 was in the average range for his age grouping. AK. Extensive cortical and white matter disease involving the superior and middle gyri of the left frontal lobe as well as the orbital gyrus and gyrus rectus of the right frontal lobe is noted. Cortical disease also extends to the inferior gyrus of the left temporal lobe. There is subcortical white matter disease involving the inferior

TABLE 1 Raw Scores for the Comparison Group (N ⫽ 14); Raw Scores for the Heterogeneous BrainDamaged Group (N ⫽ 16) Including the Two Case Studies (Age Range ⫽ 50 to 67, x ⫽ 57); T Scores for Two Case Studies Based on the Comparison Group; and Difference Values in T Scores between the Two Cases CAS subtests & COMPOSITE SCALE

Mean (SD) raw scores (N ⫽ 14)

Mean (SD) raw scores (N ⫽ 16)

Patient AK MMSE ⫽ 20 (Age 65 years)

Patient GM MMSE ⫽ 28 (Age 67 years)

Matching Numbers Planned Codes Planned Connections PLANNING

8 (5) 41 (23) 349 (221) —

8 (5) 37 (24) 355 (209) —

43 35 43 40

58 38 52 50

14 4 9 9

Expressive Attention Number Detection Receptive Attention ATTENTION

35 (16) 38 (28) 32 (18) —

32 (17) 36 (27) 30 (18) —

31 47 38 38

37 45 46 43

7 ⫺1 8 4

Nonverbal Matrices Verbal Spatial Relations Figure Memory SIMULTANEOUS

15 (9) 17 (5) 11 (6) —

14 (9) 16 (5) 11 (6) —

40 40 37 39

45 45 46 45

6 6 9 7

Word Series Sentence Repetition Sentence Questions SUCCESSIVE

12 (3) 10 (3) 12 (3) —

11 (3) 9 (3) 11 (3) —

24 29 29 27

43 40 33 39

19 12 4 11

—

—

36

44

8

FULL-SCALE

Difference values in T scores

210

ROTMAN INSTITUTE ABSTRACTS

gyrus of the left frontal lobe. As well, there is disease of the head of the left caudate nucleus and anterior portions of the left globus pallidus and putamen. There is deep white matter disease involving the left internal and external capsules. These gliotic changes are the result of an intracranial aneurysm 6 years previously. GM. There is subcortical white matter disease involving the left centrum semiovale and subcortical white matter of the left parietooccipital lobe due to an oligodendroma tumor excision 5 years prior. There is a new finding of subtle cortical and white matter disease involving the right parietal lobe (see Table 1). A nonparametric Wilcoxon signed-rank test demonstrated GM’s performance on the 12 CAS subtests was significantly higher overall compared to AK (z ⫽ 2.981, p ⬍ .003), for a two-tailed test. Although assumptions regarding normality of distributions of scores within subtests were violated, there are a number of trends worthy of comment between AK and GM. GM’s performance on the verbal scale of Word Series appeared to be significantly better than that of AK. In addition, of note were the consistently lower performances on the successive scale in AK compared to GM. Matching Numbers also appeared to discriminate between these two subjects. The MMSE, neurologists’ reports, observational data, history, and neuroradiological findings are consistent with strategic infarct dementia as a sequalae of left caudate and globus pallidus infarction and extensive damage to the left frontal lobe in AK (McPherson & Cummings, 1996). The results suggest the CAS is useful in the context of neuropsychological assessment of strategic infarct dementia. It remains to be determined whether the CAS also demonstrates neuropsychological sensitivity and specificity at the subtest level (e.g., localizing properties) and if it would provide useful information in this respect beyond that offered by existing neuropsychological batteries. ACKNOWLEDGMENT

S.M.M. was supported by SSHRC Grant 752-2000-1344. REFERENCES Almkvist, O., Brane, G., & Johanson, A. (1996). Neuropsychological assessment of dementia: State of the art. Acta Neurologica Scandinavia, Suppl. 168, 45–49. Das, J. P., Davis, B., Alexander, J., Parrila, R. K., & Naglieri, J. A. (1995). Cognitive decline due to aging among persons with Down syndrome. Research in Developmental Disabilities, 16, 461–478. McPherson, S. E., & Cummings, J. L. (1996). Neuropsychological aspects of vascular dementia. Brain and Cognition, 31, 269–282. Naglieri, J. A. (1999). Essentials of CAS assessment. New York: Wiley. Spreen, O., & Strauss, E. (1998). A compendium of neuropsychological tests: Administration, norms and commentary (2nd ed.). New York: Oxford University Press. This is doi:10.1006/brcg.2001.1467.

6. Predictors of Cognitive Change from Preclinical to Clinical Alzheimer’s Disease

S. Jones, B. J. Small, L. Fratiglioni, and L. Ba¨ckman We examined individual-difference variables in relation to the rate of change in global cognitive performance, measured by the MMSE, from 3 years prior to diagnosis of Alzheimer’s

ROTMAN INSTITUTE ABSTRACTS

211

disease (AD) to the time of diagnosis. The population-based sample consisted of 230 incident AD persons who were followed over a 3-year interval. The average annual decline in MMSE was 1.81 points. Being older and acquiring additional diseases during the 3 years preceding diagnosis predicted a faster rate of decline in global cognitive functioning. However, other individual difference variables such as sex, education, depression, vitamin levels (vitamin B12 and folic acid), apolipoprotein status, and social network did not precipitate the rate of decline in the preclinical phase of AD.  2002 Elsevier Science (USA)

Report Normal cognitive aging is characterized by large interindividual differences in performance. Although many cognitive functions decline in normal aging, there is systematic variation in several domains, such as demographics (e.g., age, sex, education), health (e.g., circulatory factors, vitamin, and thyroid status), genetics (e.g., apolipoprotein E), and social factors (e.g., social network), with regard to the size of the age-related cognitive decline (Ba¨ckman, Small, Wahlin, & Larsson, 1999). In Alzheimer’s disease (AD), the effect of these individual-difference variables for cognitive functioning is reduced. Although onset age has been extensively investigated as a potential predictor of the rate of cognitive decline in AD, the research efforts have yielded mixed results. Some studies demonstrate a faster decline with increasing age, whereas others show the opposite pattern, and still some others have found negligible effects of onset age (Ba¨ckman et al., 1999). This suggests that onset age acts like many other individual-difference variables in AD. That is, variables known to affect cognitive performance in normal aging have a limited influence on cognitive performance in this disease. A possible reason thereof may be that the influence of various participant characteristics is overshadowed by the pathogenesis itself. Studies have shown that there is a long preclinical phase in AD, during which cognitive deficits can be detected. The length of the preclinical phase is not yet established but cognitive deficits can be observed at least 7 years before a clinical diagnosis of AD can be rendered (e.g., Small, Fratiglioni, Viitanen, Winblad, & Ba¨ckman, 2000). During the preclinical phase cognitive functioning is relatively stable until a few years before diagnosis when there is a precipitous decline. The purpose of this research was to examine whether the variability in rate of decline during the last 3 years before the diagnosis of AD could be meaningfully linked to participant characteristics within demographic, health-related, genetic, and social domains. Method The study sample was taken from a population-based study, including all inhabitants aged 75 years and older in the Kungsholmen Parish of Stockholm, Sweden (2368 inhabitants). The participants have been examined longitudinally at approximately 3-year intervals. More detailed information about the original study and methods used can be found in Fratiglioni et al. (1991). In the present study data from all incident AD persons from the first (n ⫽ 150) and second (n ⫽ 80) follow-up were aggregated to maximize sample size and variability in the predictor variables. The outcome measure was the Mini Mental State Examination (MMSE), a brief test that measures various aspects of cognitive functioning, such as orientation to time and place, episodic memory, verbal ability, and attention, with a maximum score of 30. A Swedish version of the MMSE was administered according to the same standardized procedure at baseline and follow-up. Predictor variables included demographic factors (age, sex, and education), disease measures (history of disease, additional diseases acquired between baseline and

212

ROTMAN INSTITUTE ABSTRACTS

follow-up, and high blood pressure), depression, and vitamin status (vitamin B12 and folic acid). The disease measures were calculated using data obtained from a registry of all in-patients in the hospitals in the Stockholm area. These predictors were selected as they are known to affect cognitive functioning in normal aging. Apolipoprotein genotype and social network, which have been implicated as risk factors for AD, were also used as predictor variables. Results and Discussion The results show considerable decline in the MMSE during the 3 years preceding diagnosis of AD. The mean annual decline in the present sample was 1.81 points. This figure may be compared with the decline that has been reported for clinical AD, where an annual decline of around 3 points on the MMSE is typically found (e.g., Clark et al., 1999). There may be at least two explanations for the slower rate of decline during the 3 years preceding AD diagnosis compared to the rate of decline in clinical AD. First, it is conceivable that the decline is not as rapid preclinically as after clinical diagnosis, when the disease has manifested itself. Another explanation is that once the decline starts it is as rapid as in clinical AD but in this sample all participants did not decline over the whole retest interval. As could be expected, MMSE score at baseline was related to MMSE score at follow-up (R 2 ⫽ .16, p ⬍ .001). However, the main findings in this study were that only age (β ⫽ ⫺.16, p ⫽ .01) and additional diseases between baseline and followup (β ⫽ ⫺.12, p ⫽ .05) were independently related to rate of cognitive change from preclinical to clinical AD. Older persons had a faster rate of decline. Considering that cognitive decline in normal aging is an inevitable fact, it is not unrealistic to expect that both being very old and in a preclinical phase of AD would exacerbate the rate of cognitive decline. Also having greater number of additional diseases between baseline and follow-up was related to a faster rate of cognitive decline. A critical question is whether this association is due to that individuals develop additional diseases in the late preclinical phase because of their impending dementia or if the cognitive decline accelerates because of the individuals’ health status. History of diseases, that is, health status before the start of the neurodegenerative process, appears to have no relation to the rate of decline. This would seem to suggest that poor health is a result of the neurodegenerative process rather than the opposite (Table 1). Conclusions Being older and having poorer health during the final years preceding diagnosis of AD are associated with a faster cognitive decline. Other individual-difference variables known to affect cognitive functioning in normal aging (e.g., sex, education, TABLE 1 Regression Analysis Predicting MMSE Change Predictor

R

∆R 2

Cum R 2

β

p

MMSE Age Additional disease Age ⫻ Additional disease

.400 .422 .436 .442

.16 .02 .01 .006

.16 .18 .19 .196

0.40 ⫺0.14 ⫺0.11 ⫺1.31

⬍.001 .03 .07 ns

Note. The results of a final regression analysis where only predictors that were reliable when entered alone were included are shown. This model explains about 19% of the variance, most of the variance being associated with cognitive functioning at baseline.

ROTMAN INSTITUTE ABSTRACTS

213

high BP, depression, vitamin status) or risk factors for AD (e.g., apolipoprotein Eε4, social network) do not precipitate the decline. Thus, even preclinically, the disease process appears to reduce the impact of factors that are known to affect cognitive functioning in nondemented individuals. REFERENCES ˚ ., & Larsson, M. (1999).Cognitive functioning in very old age. In Ba¨ckman, L., Small, B. J., Wahlin, A F. I. M. Craik & T. A. Salthouse (Eds.), The handbook of cognitive aging (2nd ed.) (pp. 499–558). Mahwah, NJ: Erlbaum. Clark, C. M., Sheppard, L., Fillenbaum, G. G., Galasko, D., Morris, J. C., Koss, E., Mohs, R., & Heyman, A. (1999). Variability in annual Mini-Mental State Examination score in patients with probable Alzheimer disease: A clinical perspective of data from the Consortium to Establish a Registry for Alzheimer’s Disease. Archives of Neurology, 56, 857–862. Fratiglioni, L., Grut, M., Forsell, Y., Viitanen, M., Grafstro¨m, M., Holmen, K., Ericsson, K., Ba¨ckman, L., Ahlbom, A., & Winblad, B. (1991). Prevalence of Alzheimer’s disease and other dementias in an elderly urban population: Relationship with sex and education. Neurology, 41, 1886–1892. Small, B. J., Fratiglioni, L., Viitanen, M., Winblad, B., & Ba¨ckman, L. (2000). The course of cognitive impairment in preclinical Alzheimer’s disease: Three- and 6-year follow-up of a population-based sample. Archives of Neurology, 57, 839–844. This is doi:10.1006/brcg.2001.1468.

7. Changes to the Object Recognition System in Patients with Dementia of the Alzheimer’s Type

K. S. Purdy, P. A. McMullen, and M. Freedman Do DAT patients show category-specific deficits in object identification, and do they arise from semantic or visual damage? Participants decided whether line drawings of living and nonliving objects matched names at superordinate, basic, or subordinate levels. Patients were most impaired with superordinate decisions. Controls had most difficulty with subordinate decisions. No category-specific deficit was found with patients. Impaired superordinate decisions by the patients support semantic damage. If category-specific deficits arise from damaged semantics, they should have been found. Since they were not, and since patients performed subordinate decisions the best, a visual basis to category specificity is supported. Finally, a living advantage was found with normal observers which cannot be spurious due to differences in concept familiarity since living and nonliving objects were matched for this variable.  2002 Elsevier Science (USA)

Report The present study addresses two questions with respect to object recognition in patients with dementia of the Alzheimer’s type (DAT). First, do DAT patients show category-specific deficits in recognizing objects, and second, if category-specific deficits in object naming exist, are these deficits due to damage to semantic or visual representations? Alzheimer’s patients often have difficulty naming common objects as evidenced by their poor performance on the Boston naming test. The ability to name common objects is usually compromised at early stages of DAT and becomes progressively worse over the course of the disease (Hodges, Patterson, Graham, & Dawson, 1996). Patients with DAT are thought to have damage to their semantic systems as a result of their disease which contributes to their difficulty identifying objects (Patterson &

214

ROTMAN INSTITUTE ABSTRACTS

Hodges, 1995). There has been some suggestion in recent literature that DAT patients show category-specific deficits, that is, that they have more difficulty naming either living or nonliving objects. Garrard, Patterson, Watson, and Hodges (1998) found that some DAT patients showed a significant advantage for naming nonliving objects, while others showed an advantage for naming living objects. Whether or not DAT patients actually show category-specific deficits in object naming, or whether this effect is due to some other confounding variable, such as the frequency with which the objects have been encountered, has been the topic of much debate in recent literature. There are two major organizational principles to the object recognition system. The first is that the system is organized in a hierarchy of superordinate (e.g., animal), basic (e.g., dog), and subordinate (e.g., collie) representations. Rosch, Mervis, Gray, Johnson, and Boyes-Braem (1976) found that in object verification tasks, basic representations are accessed fastest, and so, basic level identification occurs first. Beyond basic level identification, superordinate identification requires more semantic processing, and subordinate identification requires more visual processing (Jolicoeur, Gluck, & Kosslyn, 1984). The other organizational principle of the object recognition system is that objects are generally part of one of two broad categories: living or nonliving (artifacts). Some researchers support the notion that there is no difference in the way that living and nonliving objects are represented in the brain, contending that there really are no category-specific deficits. Others support the idea that there is a difference in the way living and nonliving objects are represented in the brain and that this difference may exist at the level of visual or semantic representations. Thus, with brain damage, category-specific deficits may appear. If DAT patients show category-specific deficits and if they are due to semantic damage then DAT patients should show these deficits particularly at superordinate levels of object identification because this level of identification is most dependent on semantic processing. If these category-specific deficits are due to damage to visual representations, then DAT patients should show these deficits particularly at subordinate levels of object identification because identification at this level requires the most visual processing. Methods Subject groups. Subject groups were 16 college-aged observers with normal or corrected-to-normal vision participated for class credit (age range 18–23 years); 3 normal elderly volunteers with normal or corrected-to-normal vision participated (age range 74–82 years); and 3 DAT patients (ages 60, 65, and 81 years). Elderly controls were included if they had no previous diagnosis of neurological or psychiatric illness. DAT inclusion criteria. Patients were diagnosed by neurologists with probable DAT based on the criteria set forth by NINCDS-ADRDA (National Institute of Neurology and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association) as well as by the criteria of the DSM-IV. DAT patients were also tested on the Mini Mental Status Examination (with a score above 20 for inclusion), and adequate scores on both the Boston Naming Test and Test No. 48 of the PALPA (Psycholinguistic Assessments of Language Processing in Aphasia). DAT patients were excluded if they had visual impairments that could affect their performance during the experimental task or if they had a concurrent diagnosis of psychiatric or additional neurological illness. Experimental task. Participants viewed stimuli on a computer screen and made key-press responses on a response box. On each trial, a word and line drawing were simultaneously presented. Participants were asked to respond with one key on the

ROTMAN INSTITUTE ABSTRACTS

215

response box if the word and picture matched and to respond with another key if the word and picture did not match. Each participant responded to three blocks of 72 stimuli. The same 72 pictures were used in each block, but in each block, drawings were presented with either matching or mismatching subordinate names, basic names, or superordinate names. All of the subjects saw all of the conditions. The 72 stimuli included 36 living objects and 36 nonliving objects (matched for concept familiarity). The blocks of trials were presented in random order, and stimuli in each block were randomized as well. Reaction time and error rates were recorded in an output file. Results Inverse efficiency scores were calculated as Mean reaction time/proportion correct and were used in the statistical analyses. This measure was used as it corrects for any speed–accuracy trade-offs. The data were analyzed using an analysis of variance (ANOVA), for a 3 ⫻ 3 ⫻ 2 mixed measures design comparing subject group, level of object recognition (superordinate, basic, subordinate), and living vs nonliving objects. College-aged controls. There was a main effect of level of object recognition ( p ⬍ .006). Subordinate representations were accessed the least efficiently, and basic and superordinate were not significantly different from each other. There was also a marginal main effect of living and nonliving (p ⬍ .054). Living responses were more efficiently processed than nonliving. There was a main effect for match/mismatch (p ⬍ .004), with responses to matches being more efficient than responses to mismatches. There were no significant interactions. Elderly controls. There was a main effect of age, in that older controls were less efficient than younger controls. The pattern of the other main effects was the same as those of the college-aged controls. There was a main effect of level of object recognition, with subordinate representations being accessed least efficiently, and basic and superordinate not significantly different from each other. There was also a main effect of living/nonliving, with an advantage for living objects, and a main effect for match/mismatch, with responses to matches being more efficient than responses to mismatches. Again, there were no significant interactions. DAT patients. There was a main effect of dementia, with elderly control subjects responding much more efficiently than DAT patients. There was a main effect of level of object recognition. Interestingly, in a pattern opposite to that of the control subjects, superordinate representations were accessed least efficiently, and basic and subordinate were not significantly different from each other (with subordinate being most efficient). There was also a main effect of living/nonliving, with the pattern looking similar to controls—a living advantage. There was a main effect for match/ mismatch also, with responses to matches being much more efficient than responses to mismatches. There were no significant interactions. Conclusions Results indicated that DAT patients were most impaired with superordinate decisions. This effect is in contrast to results from the control observers who had most difficulty with subordinate decisions (see Fig. 1). We did not find a category-specific deficit in DAT patients, as has been previously reported. The impaired performance of superordinate decisions demonstrated by DAT patients does support damage at the level of semantic representations. Consequently, category-specific deficits should have been found if they arise from damaged semantics. Since we did not find a category-specific deficit in the patients studied, this supports a visual basis to category-specific representations. This conclusion is further supported by the finding

216

ROTMAN INSTITUTE ABSTRACTS

FIG. 1. Inverse efficiency graphs comparing (L–R) college-aged controls, elderly controls, and DAT patients.

that DAT patients performed subordinate decisions relatively better than decisions at other levels. Finally, the living advantage found here with normal observers challenges reports of a nonliving advantage. This living advantage cannot be spurious because of differences in concept familiarity since the living and nonliving objects that were used were matched for concept familiarity. Note added in proof. This research is still ongoing, and the projected sample sizes for the entire study are 32 normal college-age controls and 20 DAT patients with 20 elderly age-matched controls.

REFERENCES Garrard, P., Patterson, K., Watson, P. C., & Hodges, J. R. (1998). Category specific semantic loss in dementia of the Alzheimer’s type: Functional–anatomical correlations from cross-sectional analyses. Brain, 121, 633–646. Hodges, J. R., Patterson, K., Graham, N., & Dawson, K. (1996). Naming and knowing in dementia of Alzheimer’s type. Brain and Language, 54, 302–325. Jolicoeur, P., Gluck, M. A., & Kosslyn, S. M. (1984). Pictures and names: Making the connection. Cognitive Psychology, 16, 243–275. Patterson, K., & Hodges, J. R. (1995). Disorders of semantic memory. In A. D. Baddeley, B. A. Wilson, & F. N. Watts (Eds.), Handbook of memory disorders. Chichester: Wiley. Rosch, E., Mervis, C. B., Gray, W. D., Johnson, D. M., & BoyesBraem, P. (1976). Basic objects in natural categories. Cognitive Psychology, 8, 382–439. This is doi:10.1006/brcg.2001.1469.

8. Verbal Learning and Visuomotor Attention in Alzheimer’s Disease and Geriatric Depression

J. M. Strang, K. Z. Donnelly, K. Grohman, and J. Kleiner We compared the verbal learning and visuomotor attention of 34 Alzheimer’s patients and 18 depressive patients. Verbal learning was assessed using The Hopkins Verbal Learning Test—Revised (HVLT—R); visuomotor attention was assessed using the Trail Making Test (TMT). The Alzheimer’s patients had significantly lower scores on immediate and delayed recall of a word list. There was a nonsignificant trend in this group toward a fewer number of true positives and a greater number of false positives. Alzheimer’s patients were significantly slower on Trails A, with a nonsignificant trend toward slower performance on Trails B. No difference was observed in accuracy of attentional processing. The results are discussed

ROTMAN INSTITUTE ABSTRACTS

217

in terms of other factors, such as stage of cognitive decline, which might have influenced the findings.  2002 Elsevier Science (USA)

Report Elderly patients with symptoms of both depression and cognitive impairment present a diagnostic challenge. Impaired attention and memory are common symptoms of depression, particularly in elderly patients (King, Cox, Lyness, Conwell, & Caine, 1998). Likewise, depression is one of the first signs of a dementing illness, such as Alzheimer’s disease (Lezak, 1995). Symptoms that are most likely to confound differential diagnosis of depression and dementia are depressed mood, agitation, psychomotor retardation, impaired immediate learning and memory, defective attentional processing, impaired orientation, and apathy (Lezak, 1995). However, accurate diagnosis is crucial, as depression is reversible and may be effectively treated (Whall, 1986). The term ‘‘depressive pseudodementia’’ is often used to describe patients who present as demented but whose depressive symptoms predominate and do not fulfill the criteria for a diagnosis of dementia (Lezak, 1995; Whall, 1986). Several differences have been observed between demented patients and psychiatrically depressed patients. For example, the onset of dementia is slow and insidious with a progressive course, compared to the more sudden onset of depression, which often occurs in response to an identifiable event or stressor. Additionally, depressed patients are more likely to be aware of their cognitive difficulties, and they may subjectively complain of deficits that do not appear when they are formally assessed (Lezak, 1995). Depressed patients are also more likely to respond to questions with ‘‘I don’t know’’ or ‘‘I don’t care.’’ Dementia patients, on the other hand, attempt, often unsuccessfully, to provide the correct answer, and they are often unaware of their mistakes (Whall, 1986). Certain cognitive patterns also distinguish dementia of the Alzheimer’s type (DAT) from depressive pseudodementia (DPD). In the area of verbal memory, Gainotti and Marra (1994) found that the most sensitive diagnostic differentiator was the presence in the DAT group of a higher rate of false positive responses on a delayed recognition trial. Other patterns characteristic of Alzheimer’s patients include slowed performance on timed visuomotor tasks and deficits in attention, orientation, and reasoning (Lezak, 1995; Nebes, Halligan, Rosen, & Reynolds, 1998). In the present study, we compared the verbal learning and visuomotor attention of Alzheimer’s patients versus depressive patients. The goals of the study were (1) to confirm the previously explored aspects of verbal learning and visuomotor attention that differentiate Alzheimer’s patients from depressive patients and (2) to explore the utility of the Hopkins Verbal Learning Test—R in measuring differences in verbal learning. We predicted that the Alzheimer’s patients would be significantly more impaired on immediate and delayed recall and recognition of a word list compared to patients with a primary diagnosis of major depression. We also predicted that the Alzheimer’s patients would show significantly slowed attentional processing speed on the Trail Making Test, Parts A and B. Methods Research participants. Fifty-two veterans (51 men, 1 woman) diagnosed with either Alzheimer’s disease (AD) or unipolar major depression (DEP) were selected from a larger database with a total of 676 cases, 43 different medical and psychiatric diagnoses, and a comprehensive neuropsychological battery. Participants in the AD group (N ⫽ 34) ranged in age from 70 to 85, with a mean of 76.09 (SD ⫽ 5.17).

218

ROTMAN INSTITUTE ABSTRACTS

Years of education for this group ranged from 4 to 16, with a mean educational level of 10.79 years (SD ⫽ 2.9). The mean number of grade retentions (number of years held back in school) was 0.09 (SD ⫽ 0.39). Participants in the DEP group (N ⫽ 18) did not differ significantly on any of the demographic variables. Age ranged from 68 to 83, with a mean of 75.22 (SD ⫽ 4.97). Years of education ranged from 8 to 16, with a mean educational level of 11.39 years (SD ⫽ 2.43). The mean number of grade retentions was .29 (SD ⫽ .85). Measures. Verbal learning and memory were assessed using The Hopkins Verbal Learning Test—Revised (HVLT—R). The HVLT was administered as part of a full neuropsychological battery that included tests of language ability, attention and concentration, memory, visuospatial functioning, and intellectual efficiency. The HVLT involves reading a list of 12 words from three semantic categories. Three learning trials are followed by a 20-min delayed free recall and a 24-word recognition list containing all 12 target words plus six semantically related distracters and six unrelated ones. The following variables were of primary interest for this study: immediate recall (total number of words recalled on each of the three learning trials), longdelayed recall (free recall after 20 minutes), true positives (number of target words recognized from the recognition list), and false positives (number of distracter words chosen from the recognition list). Visuomotor attention was assessed using the Trail Making Test (TMT). This test is given in two parts, A and B. On Part A, the subject must draw lines connecting consecutively numbered circles. On Part B, the subject must connect the same number of consecutively numbered and lettered circles by alternating between number and letter. The subject is urged to connect the circles as quickly as possible without lifting the pencil from the paper. The following variables were of primary interest: speed of attentional processing (number of seconds required to complete each part) and accuracy of attentional processing (number of errors on each part). Statistical analysis. In order to compare the memory and attentional performance of participants in the AD and DEP groups, analysis of variance was performed on the aforementioned variables of interest. In the analyses, plots of standardized residuals were used to check the assumptions of normally distributed errors with constant variance. Examination of the residuals revealed some outliers. However, the same individuals were not outliers across the different variables, and no pattern could be determined. Thus, none of the outliers were removed, as no single individual was a consistent outlier across variables.

Results The means and standard deviations of the AD and DEP groups on each of the dependent variables are displayed in Table 1. The data are displayed without removing the outliers. Participants in the AD group had significantly lower scores on HVLT trial one learning [F(1, 50) ⫽ 13.83, p ⬍ .0001], HVLT trial two learning [F(1, 50) ⫽ 17.71, p ⬍ .0001], HVLT trial three learning [F(1, 47) ⫽ 11.97, p ⬍ .001], and HVLT delayed recall [F(1, 46) ⫽ 22.48, p ⬍ .0001]. There was a nonsignificant trend in the AD group toward a fewer number of true positives [F(1, 46) ⫽ 2.72, p ⫽ .11] and a greater number of false positives [F(1, 46) ⫽ 2.73, p ⫽ .11] on the HVLT recognition trial. Of the 52 participants, 42 completed Trail Making Test A and 21 completed Trail Making Test B. Participants in the AD group were significantly slower on Trails A [F(1,40) ⫽ 6.26, p ⬍ .05]. There was also a nonsignificant trend in the AD group

219

ROTMAN INSTITUTE ABSTRACTS

TABLE 1 Scores of Participants in the AD and DEP Groups on the HVLT—R and TMT AD (N ⫽ 34)

DEP (N ⫽ 18)

Measures

M

SD

M

SD

HVLT1*** HVLT2*** HVLT3** HVLT4*** HVLTTP HVLTFP TRAILSAS* TRAILSAE TRAILSBS TRAILSBE

2.53 3.5 4.55 .71 7.9 3.27 75.8 .20 256 1.71

1.02 1.42 1.31 1.64 3.36 2.92 28.32 .50 204.1 .93

4.28 5.72 6.56 3.88 9.5 2.0 56.12 .06 141 1.80

2.37 2.4 2.75 3.02 3.07 1.81 19.03 .24 54.8 1.00

F ( p) 13.83 17.71 11.97 22.48 2.72 2.73 6.26 1.16 4.06 1.69

(.0005) (.0001) (.0012) (.0001) (.106) (.105) (.017) (.29) (.058) (.209)

Note. HVLT1, immediate recall 1; HVLT2, immediate recall 2; HVLT3, immediate recall 3; HVLT4, delayed free recall; HVLTTP, true positives; HVLTFP, false positives; TRAILSAS, A, processing speed; TRAILSAE, A, number of errors; TRAILSBS, B, processing speed; TRAILSBE, B, number of errors. * p ⬍ .05; ** p ⬍ .001; *** p ⬍ .0001.

toward slower performance on Trails B [F(1, 20) ⫽ 4.06, p ⫽ .06]. There was no difference between groups on the accuracy of attentional processing. Discussion The results confirmed the presence of a differential pattern of verbal learning and visuomotor attention in Alzheimer’s versus depressive patients. As predicted, the Alzheimer’s patients had significantly lower scores on immediate and delayed recall of a word list. The trend toward a greater number of false positive responses in the AD group, in contrast to Gainotti and Marra’s (1994) findings, was not significant. In comparison, the DEP group in the present study had a higher level of false positive responses. Gainotti and Marra used a false positive cutoff score of two to differentiate between normal controls and patients. Only 4 of 26 depressed patients exceeded the cutoff, compared to 34 of 42 Alzheimer’s patients. In contrast, the false positive mean for the DEP group in the present study was two, indicating that a significant number of depressed patients made two or more false positive errors. This difference may be accounted for by differences in the samples and in the verbal learning measures utilized. Nonetheless, the results demonstrate the utility of the HVLT—R in measuring differences in verbal learning in Alzheimer’s versus depressed geriatric patients. The Alzheimer’s patients were also significantly slower on Trails A. This trend was evident on Trails B, but it was not significant. As discussed by Lezak (1995), the within group variance on Trails B is typically very large, which may account for this finding. These results must be interpreted in light of the study’s limitations. The sample, composed of male veterans (with the exception of one female), is not representative of the population at large. Additionally, the effects of psychoactive medication on performance are unknown, as this information was not gathered. The stage of cognitive decline in the Alzheimer’s patients may also have an impact on the results. Once again, this information was unavailable. Nevertheless, the results of this study confirm previously explored differences in verbal learning and visuomotor attention, as well as the utility of the HVLT—R in measuring differences in verbal learning.

220

ROTMAN INSTITUTE ABSTRACTS

REFERENCES King, D. A., Cox, C., Lyness, J. M., Conwell, Y., & Caine, E. D. (1998). Quantitative and qualitative differences in the verbal learning performance of elderly depressives and healthy controls. Journal of the International Neuropsychological Society, 4,115–126. Gainotti, G., & Marra, C. (1994). Some aspects of memory disorders clearly distinguish dementia of the Alzheimer’s type from depressive pseudo-dementia. Journal of Clinical and Experimental Neuropsychology, 16, 65–78. Lezak, M. D. (1995). Neuropsychological assessment (3rd ed.). New York: Oxford University Press. Nebes, R. D., Halligan, E. M., Rosen, J., & Reynolds, C. F. (1998). Cognitive and motor slowing in Alzheimer’s disease and geriatric depression. Journal of the International Neuropsychological Society, 4, 426–434. Whall, A. L. (1986). Identifying the characteristics of pseudodementia. Journal of Gerontological Nursing, 12, 34–35. This is doi:10.1006/brcg.2001.1470.

9. In Vivo Neuroanatomy of Alzheimer’s Disease: Evidence from Structural and Functional Brain Imaging

P. Poulin and K. K. Zakzanis In vivo structural (CT, MRI) and functional (SPECT, PET) brain imaging techniques have been widely used to study the neuroanatomy and neurophysiology of Alzheimer’s disease (AD) and to identify definite biological markers of the disease. We used meta-analytic methods to synthesize this literature to determine what neuroanatomical structures best differentiate patients with AD from healthy normal controls. A total of 125 studies published between 1984 and 2000 that included 3543 patients with AD and 1698 normal healthy controls met inclusion criteria. We found that measures of the temporal cortices, including the amygdala, hippocampus, and inferior temporal lobes, along with the anterior cingulate cortex, associated with the largest magnitudes of effects and, hence, could serve as the most useful structures to help clinicians differentiate AD from healthy normal aging.  2002 Elsevier Science (USA)

Report In vivo structural and functional brain imaging techniques have been widely used to study the neuroanatomy and neurophysiology of Alzheimer’s disease (AD) and to identify definite biological markers of the disease. The purpose of anatomical markers in AD can broadly be characterized as follows: (1) to diagnose the disease in individual patients, (2) to follow the course of the disease, and (3) to assess the response to therapeutic intervention both in individuals and in groups (i.e., drug trials). Structural imaging (CT and MRI) provides measures of normal age-related changes in the brain as well as pathological brain atrophy. In the study of AD it has been used to delineate the topographic distribution of pathological brain atrophy and to understand its longitudinal progression (e.g., Kidron et al., 1997). Functional imaging provides information on metabolic brain function and dysfunction. In the study of AD, it has sought to identify a physiologic ‘‘signature’’ or functional neuroanatomy that corresponds to the clinical phenomenology of the dementia and permits its positive identification. Such a signature image feature could be the foundation for rational therapy as well as early differential diagnosis. We used meta-analytic methods to review, synthesize, and evaluate neuroimaging studies of AD. In addition to solving problems with traditional narrative reviews, meta-analysis provides tools for the analysis of magnitude of effects and may be

ROTMAN INSTITUTE ABSTRACTS

221

used to identify potential clinical markers for a disease and aid in the differential diagnosis of neurologic and psychiatric disease (see Zakzanis et al., 1999). We report effect sizes and associated measures for 113 areas of the AD brain. Method Literature search. We conducted a manual search through the volumes of pertinent journals. This was done with every issue for the following journals: American Journal of Psychiatry; Annals of Neurology; Archives of Clinical Neuropsychology; Archives of General Psychiatry; Archives of Neurology; Biological Psychiatry; Brain; British Journal of Psychiatry; Dementia; Journal of Clinical and Experimental Neuropsychology; Journal of Neuropsychiatry and Clinical Neurosciences, Journal of Nervous and Mental Disease; Journal of the International Neuropsychological Society; Neurology; Neuropsychology; Neuropsychiatry, Neuropsychology, and Behavioral Neurology; Neuropsychopharmacology; and Psychiatry Research. To reduce the likelihood that bias was involved in the manual search outcome, we also conducted a computer based search using the PsychInfo and Medline databases. The key words used were ‘‘Alzheimer’s’’ with independent matched searches with the key word(s) ‘‘PET,’’ ‘‘positron emission tomography,’’ ‘‘MRI,’’ ‘‘magnetic resonance,’’ ‘‘CT,’’ ‘‘computed tomography,’’ ‘‘CAT,’’ ‘‘computed axial tomography,’’ ‘‘NMRI,’’ ‘‘nuclear magnetic resonance,’’ ‘‘brain metabolism,’’ ‘‘blood flow, ‘‘neuroimaging,’’ and ‘‘imaging.’’ The studies located by the computer search were limited to those published English written studies and dissertations. They were obtained at two large Canadian universities and through interlibrary loan. Study inclusion criteria. Articles were included if they met the following criteria: (1) publication between 1984 and 2000 (cutoff date for use of systematic and reliable diagnostic criteria, i.e., NINCDS-ADRDA, see McKhann et al., 1984); (2) presence of a control group comprising healthy participants; (3) study statistics convertible to effect size d (e.g., means, standard deviations, F, t, X); and (4) blind reading of the scans. Statistical manipulations. The d statistic (Cohen, 1988) was calculated for each comparison as the difference between Alzheimer and control group means normalized by the pooled standard deviation whenever these measures where reported. We also calculated effect sizes from inferential statistics. We transformed all effect sizes into a nonoverlap percentage (U ) using Cohen’s (1988) idealized distributions. As suggested by Zakzanis et al. (1999), we then converted the U statistic to represent the degree of overlap by subtracting the nonoverlap from 100. When there is less than 5% overlap (d ⬎ 3) between the patient and control distribution on a given measure, it can be trusted as a potential marker of the disease. The OL% statistic used here represents the degree of overlap between patients with AD and normal control participants in the distributions of structural and physiological measures of the cerebrum. Recorded variables. Recorded variables for each article used in our meta-analysis included the full study reference, any moderator variables reported (e.g., age, onset age, duration of illness, percent male, and total score on the Mini Mental Status Examination (MMSE); also the type of neuroimaging equipment along with the procedural outline—e.g., T1/T2, length of cut, number of cuts, angle of cut (i.e., transaxial, sagittal, or coronal). These study characteristics were used to describe the study set retrieved and treated uniformly for moderator variable analysis. Results One-hundred twenty-five studies—60 structural and 65 functional—published between 1984 and 2000 met criteria for inclusion in the present analysis. In total, neuro-

222

ROTMAN INSTITUTE ABSTRACTS

TABLE 1 Effect Size and Overlap Percentages for Structural and Functional Brain Imaging in Alzheimer’s Disease Structural imaging

Whole brain Whole brain L. hemisphere R. hemisphere Frontal lobes Frontal lobes L. frontal lobe R. frontal lobe Prefrontal Orbital frontal L. orbital frontal R. orbital frontal L. primary motor R. primary motor Primary motor Anterior cingulate Temporal lobes Temporal lobes Sylvian fissure L. sylvian fissure R. Sylvian fissure L. temporal R. temporal Superior temporal L. superior temporal R. superior temporal Medial temporal L. medial temporal R. medial temporal Inferior temporal L. inferior temporal R. inferior temporal Posterior temporal L. posterior temporal R. posterior temporal Temporal pole L. temporal pole R. temporal pole Hippocampus L. hippocampus R. hippocampus Temporal horns L. temporal horn R. temporal horn Amygdala L. amygdala R. amygdala L. enthorhinal R. enthorhinal Parahippocampus L. Subiculum R. Subiculum

Functional imaging

N

Mean d (SD)

OL%

N

Mean d (SD)

OL%

11 1 1

⫺1.02 (0.50) ⫺0.44 ⫺0.49

44.6 69 66

16 1 1

⫺1.28 (0.78) ⫺0.15 ⫺0.39

34.7 85.3 72.6

4 8 7 1

⫺0.65 (0.21) ⫺0.94 (0.54) ⫺0.82 (0.59) ⫺1.28

61.8 48.4 52.6 37.8

23 19 20 14 8 5 5 7 7 8 10

⫺1.21 ⫺0.96 ⫺0.80 ⫺1.03 ⫺1.02 ⫺1.62 ⫺1.54 ⫺1.10 ⫺1.04 ⫺0.95 ⫺1.93

(0.86) (0.74) (0.76) (0.61) (0.69) (1.12) (1.02) (1.14) (1.04) (0.60) (1.21)

37.8 44.6 52.6 44.6 44.6 26.9 29.3 41.1 44.6 44.6 20.6

4 5 5 5 5 5 1

⫺1.32 1.51 1.12 0.74 ⫺1.33 ⫺0.79 ⫺4.35

(0.54) (0.25) (0.27) (0.30) (0.75) (0.45)

34.7 29.3 41.1 57.0 34.7 52.6 —

19

⫺1.37 (0.83)

31.9

1 2 2

⫺0.98 ⫺0.97 (1.24) ⫺1.66 (1.78)

44.6 44.6 26.9

⫺1.66 ⫺1.58 ⫺1.71 1.31 1.95 1.63 ⫺2.17 ⫺1.84 ⫺1.82 ⫺1.19 ⫺1.46 ⫺0.71 ⫺0.84 ⫺1.01

24.6 26.9 24.6 34.7 18.9 26.9 15.7 22.6 22.6 37.8 29.3 57.0 52.6 44.6

5 6 10 5 5 9 6 7 6 2 2 3 3 3 2 7 7 5 5 5

⫺1.62 ⫺1.56 ⫺1.50 ⫺0.59 ⫺0.69 ⫺1.73 ⫺1.53 ⫺1.54 ⫺1.96 ⫺0.58 ⫺1.04 ⫺1.30 ⫺1.01 ⫺0.85 ⫺1.65 ⫺1.18 ⫺0.93 ⫺1.56 ⫺1.92 ⫺1.76

(0.68) (0.96) (0.56) (0.76) (0.59) (0.72) (0.93) (1.06) (0.70) (0.77) (0.61) (1.42) (0.86) (0.74) (0.10) (0.91) (0.88) (1.02) (1.34) (1.26)

26.9 26.9 29.3 48.4 57.0 24.6 29.3 29.3 20.6 61.8 92.3 34.7 44.6 52.6 26.9 37.8 48.4 26.9 20.6 22.6

⫺0.42 ⫺2.17 (0.92) ⫺1.69 (0.49) ⫺1.94 ⫺1.36

72.6 15.7 24.6 20.6 31.9

6 15 15 3 2 2 3 9 9 4 4 3 1 1

(0.42) (0.58) (0.17) (1.18) (0.87) (0.82) (0.68) (0.46) (0.34)

1 5 5 1 1

223

ROTMAN INSTITUTE ABSTRACTS

TABLE 1—Continued Structural imaging N Parietal lobes Parietal lobes L. parietal R. parietal Anterior parietal L. anterior parietal R. anterior parietal Sensorimotor L. sensorimotor R. sensorimotor Medial parietal L. medial parietal R. medial parietal Posterior parietal R. posterior parietal L. posterior parietal Superior parietal R. superior parietal L. superior parietal Inferior parietal L. inferior parietal R. inferior parietal Posterior cingulate Angular gyrus Occipital lobes Occipital lobes L. occipital R. occipital L. Calcarine fissure R. Calcarine fissure Subcortical anatomy Corpus callosum Basal forebrain Ventricles L. ventricule R. ventricule Third ventricule Putamen Basal ganglia L. basal ganglia R. basal ganglia Striatum Internal capsule Lenticular nucleus Caudate R. caudate L. caudate Globus Pallidus Thalamus L. thalamus R. thalamus L. insula R. insula Pons Cerebellum L. Cerebellum

2 5 5 2

Mean d (SD) ⫺1.56 ⫺0.55 ⫺0.46 ⫺1.57

(0.97) (0.53) (0.39) (0.17)

Functional imaging OL%

N

Mean d (SD)

OL%

26.9 61.8 66.6 26.9

22 13 13

⫺1.77 (1.07) ⫺1.25 (0.89) ⫺1.16 (0.84)

24.6 37.8 37.8

1 1 17 3 3 3 2 2 2 2 2 4 3 3 5 4 4

⫺0.59 ⫺0.24 ⫺0.86 ⫺1.67 ⫺2.16 ⫺1.37 ⫺1.79 ⫺3.29 ⫺1.92 ⫺1.21 ⫺1.34 ⫺1.44 ⫺2.25 ⫺1.76 ⫺1.34 ⫺1.68 ⫺1.12

(1.01) (1.47) (1.42) (0.43) (0.25) (2.83) (0.17) (0.42) (0.36) (0.50) (0.69) (0.75) (0.48) (1.33) (0.71)

61.8 85.3 48.4 57.0 15.7 31.9 22.6 5.8 20.6 37.8 34.7 31.9 10.7 22.6 34.7 24.6 41.1

25 13 12 2 2

⫺0.70 ⫺0.98 ⫺0.89 ⫺0.74 ⫺0.73

(0.46) (1.00) (0.98) (0.50) (0.53)

57.0 44.6 48.4 57.0 57.0

5 13 2 2 2

⫺2.10 ⫺1.16 ⫺0.70 ⫺0.66 ⫺1.19

(2.06) (1.44) (1.02) (0.54) (1.61)

18.9 37.8 57.0 57.0 37.8

⫺0.79 ⫺1.18 (0.20) ⫺0.98 (0.98) ⫺0.68 (0.31)

52.6 37.8 44.6 57.0

⫺0.85 ⫺0.91 ⫺1.04 ⫺1.03 ⫺0.69 ⫺0.45 ⫺0.17 ⫺0.39

48.4 48.4 44.6 44.6 57.0 66.6 85.3 72.6

1

⫺0.64

61.8

1

⫺1.31

34.7

7 1

⫺1.61 (0.73) ⫺1.22

26.9 37.8

1 1 1

⫺0.49 ⫺0.17 ⫺0.21

66.6 85.3 85.3

8 1 18 4 4 16 3 2

⫺1.01 ⫺0.83 1.24 0.88 1.31 1.19 ⫺0.49 ⫺0.92

1 1 2 4 4 1 3 5 5

⫺0.20 ⫺0.53 ⫺0.91 ⫺0.66 ⫺0.60 ⫺1.26 ⫺1.73 ⫺0.28 ⫺0.34

1 1 1

⫺0.39 ⫺0.82 ⫺0.27

(0.43) (0.35) (0.52) (0.77) (0.53) (0.47) (0.22)

(0.47) (0.63) (0.53) (0.77) (0.26) (0.32)

44.6 52.6 37.8 48.4 34.7 37.8 66.6 48.4

85.3 66.6 48.4 57.0 61.8 34.7 24.6 78.7 78.7

72.6 52.6 78.7

1 4 3 3 18 4 4 2 2 2 16 2

(1.01) (0.85) (1.27) (0.41) (0.52) (0.10) (0.90) (0.12)

224

ROTMAN INSTITUTE ABSTRACTS

TABLE 1—Continued Structural imaging

R. Cerebellum Vermis Brain stem Hypothalamus Multiple cortices Temporal-parietal L. temporal–parietal R. temporal–parietal Fronto–parietal Fronto–temporal Parietal–occipital

Functional imaging

N

Mean d (SD)

OL%

N

Mean d (SD)

OL%

1 1 4 1

⫺1.46 ⫺0.02 0.00 (0.88) ⫺1.04

29.3 100 100 44.6

2

⫺0.37 (0.08)

72.6

1 1 1

⫺0.47 ⫺1.99 ⫺1.42

66.6 18.9 31.9

2

⫺1.06 (0.38)

44.6

6 1 1 1 1 1

⫺1.33 (0.34) ⫺1.31 ⫺1.19 ⫺0.94 ⫺0.74 ⫺0.14

34.7 34.7 37.8 48.4 57.0 92.3

imaging results from 3543 patients with AD and 1698 normal healthy controls were recorded across meta-analyses. In the structural meta-analysis as in the functional meta-analysis, both genders were equally represented, with males constituting, respectively, 45.5 and 49.4% of the sample. The average age of patients in the structural studies was approximately 70 years old; they had shown symptoms of dementia for 4.3 years, had been diagnosed for 2.9 years, and were moderately impaired cognitively (MMSE mean ⫽ 18.3). The average age of patients with AD in the functional studies was 67 years old; they had shown symptoms of dementia for 4.2 years, had been diagnosed 3.5 years earlier, and were mildly impaired (MMSE mean ⫽ 19.4). Table 1 provides effect sizes and overlap percentages for structural and functional brain imaging in Alzheimer’s disease. Discussion We used meta-analytic methods to review, synthesize, and evaluate neuroimaging studies of AD. Although almost every study in which imaging measures of global or specific anatomical atrophy have identified a statistically significant difference between the mean value found in patients with AD and control subjects, we found that substantial overlap exists between groups that in turn limits the clinical utility of this approach for diagnosis in individual patients. However, we did find that structural and functional imaging of the hippocampus and the amygdala as well as functional imaging of the inferior temporal lobes and of the anterior cingulate cortex are associated with the greatest sensitivities to overlap. Accordingly, imaging these structures might serve the clinician best in their diagnostic work-up of AD. REFERENCES Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). New York: Academic Press. Kidron, D., Black, S. E., Stanchev, P., Buck, B., Szalai, J. P., Parker, J., Szekely, C., & Bronskill, M. J. (1997). Quantitative MR volumetry in Alzheimer’s disease: Topographic markers and the effects of sex and education. Neurology, 49, 1504–1512. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., & Stadlan, E. M. (1984). Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology, 34, 939–944.

ROTMAN INSTITUTE ABSTRACTS

225

Zakzanis, K. K., Leach, L., & Kaplan, E. (1999). Neuropsychological differential diagnosis. Lisse, The Netherlands: Swets and Zeitlinger. This is doi:10.1006/brcg.2001.1471.

10. Encoding and Retrieval in Aging and Memory Loss, a fMRI Study

J. Mandzia, S. Black, C. Grady, M. P. McAndrews, and S. Graham Mild cognitive impairment (MCI) is a term for nondemented individuals with memory complaints and deficits greater than age-adjusted normal performance. Functional MRI (f MRI) may be a more sensitive method than other techniques to reveal functional abnormalities in individuals with MCI, only a proportion of whom progress to Alzheimer’s disease (AD). f MRI was carried out while subjects (four MCI, five age-matched normal controls, and one AD) performed incidental encoding (deep and shallow) and recognition tasks for colored and black and white photographs contrasted to baseline fixation. f MRI revealed interesting dissociations between activation patterns and behavioral performance when comparing the MCI and AD to the NC.  2001 Elsevier Science (USA)

Report Physicians are faced with a growing number of elderly individuals who have memory complaints and show memory deficits greater than age-adjusted normal performance, but who do not meet the criteria for dementia. Mild cognitive impairment (MCI) is one term given to these individuals, only a proportion of whom progress to Alzheimer’s disease (AD). The risk of converting to AD is estimated to be 10– 15% per year, but follow-up in published studies so far has been limited to 3 to 5 years (Petersen et al., 1999). As new treatment opportunities emerge, it will be important to identify those individuals who are at risk for progression to dementia, as appropriate treatment could delay the onset of disabling symptoms. Functional magnetic resonance imaging (fMRI) measures hemodynamic variations associated with task-induced changes in neuronal activity with good spatial and temporal resolution (Ogawa et al., 1990). Most functional imaging studies in MCI and AD have examined resting cerebral blood flow or metabolism, but recently activation states have begun to be investigated (e.g., Small et al., 1999). Analogous to a cardiac stress test, these ‘‘cognitive stress tests’’ can measure the brain’s capacity to deal with increasing functional demands and may be a more sensitive index of neuronal compromise. As AD is characterized by an insidious course with a cascade of pathological changes occurring over a period of four to five decades (Ohm et al., 1995), impairments undetected by more conventional diagnostic techniques and neuropsychological testing may be revealed with fMRI.

Methods Participants. Five MCI, one mild AD, and five normal controls (NC) matched for age and education underwent fMRI scanning. An additional three NC and five AD underwent behavioral testing. The patients were recruited from a memory clinic at Sunnybrook and Women’s College HSC and the normal controls were recruited from the community. Individuals were given a diagnosis of MCI if they did not meet the NINCDS-ADRDA criteria for dementia, but suffered from an isolated subjective and objective memory impairment (⬍1 SD below on memory tests), had a global

226

ROTMAN INSTITUTE ABSTRACTS

CDR of 0.5, and had a MMSE of 24. AD patients were diagnosed with possible or probable AD. f MRI tasks. Subjects performed incidental encoding and recognition tasks for colored and black and white photographs. The incidental encoding task consisted of a deep and shallow encoding run. During the deep encoding run subjects decided whether the photograph represented a natural (e.g., butterfly) vs a man-made (e.g., car) item. During the shallow encoding task subjects decided whether the photograph was in color or black and white. Following encoding, subjects underwent a structural scan for 12 min which provided a delay before the recognition task. Recognition testing followed which consisted of a deep and shallow retrieval run where subjects decided whether the photograph was old or new. The deep recognition task contained photographs from the deep encoding condition and the shallow encoding task contained photographs from the shallow encoding condition. All tasks were contrasted to a baseline condition consisting of a colored scrambled picture. Subjects indicated their decision with a response device. Each run was 5 min 30 s and consisted of six blocks of five photographs presented for 4 s each with a 1-s interstimulus interval (ISI). This was interleaved with five blocks of baseline fixation in which a scrambled color picture was presented repeatedly four times for 4 ⫹ 1 s ISI. Before each block subjects viewed a set of instructions for 6 s reminding them which buttons to press. Accuracy and reaction time were recorded for each trial. The order of the encoding and retrieval tasks was counterbalanced across subjects. f MRI scanning and analysis. Participants were placed in the MRI scanner and their head was immobilized using a vacuum pillow. Single shot spiral scanning on a 1.5-T GE scanner was performed (TE/TR ⫽ 40/3000 ms; τ ⫽ 80; acquisition matrix, 64 ⫻ 64; coronal orientation, 5 mm/slice, 30–34 slices). Following 3D registration for residual head motion, statistical analyses were performed on a voxel by voxel basis using the orthogonalized correlation method (Bandettini et al., 1993). All analyses were performed using AFNI (Analysis for Functional Neuroimages, 2000). Voxels were deemed significant if they exceeded a correlation coefficient of r ⫽ 0.4 (p ⬍ 1 ⫻ 10 ⫺4) and formed a cluster of ⬎4 contiguous voxels. After thresholding and clustering, the corrected p value was (p ⬍ 0.01). Individual activation maps were transformed to standard Talairach space and average activation maps were created for the MCI and AD group. Group activation maps were overlaid on averaged anatomical maps produced by averaging the individual structural brain scans for each subject within each group which were spatially smoothed with a 3-mm FWHM filter. Behavioral analysis. Median accuracy performance and reaction time were calculated for correct trials for each subject. A MANOVA was performed to test for differences in performance between the three groups on the four memory conditions. A Paired t test analysis, corrected for multiple comparisons, was conducted to test for a level of processing effect by examining differences in accuracy and reaction time between the encoding conditions (deep vs shallow) and the retrieval conditions (deep vs shallow) within each group. Results Behavioral Results. AD (N ⫽ 4 M/2 F, mean age 78.3 ⫾ 6.2 years; mean MMSE 25.3 ⫾ 1.5) differed significantly from the NC (N ⫽ 6 F/2 M, mean age 69.1 ⫾ 7.7 years; mean MMSE 28.7 ⫾ 1.1) on deep encoding with the AD subjects showing a reduction in accuracy ( p ⫽ 0.039). Both MCI (N ⫽ 2 F/3 M, mean age 69.8 ⫾ 5.8 years; mean MMSE 28.2 ⫾ 0.8) and AD showed significant reductions in accuracy during deep retrieval ( p ⫽ .008, p ⫽ .003). MCI and AD subjects demonstrated a

ROTMAN INSTITUTE ABSTRACTS

227

significant increase in reaction time during deep retrieval ( p ⫽ .007, p ⫽ .026). Normal controls performed significantly better during deep retrieval compared to shallow retrieval (p ⫽ .01) suggesting a levels of processing effect. f MRI results. One MCI subject was excluded from fMRI analysis because of movement related artifacts. Activation maps were compared between the MCI and NC groups and the AD patient in a preliminary analysis. When deep encoding was contrasted to baseline fixation, the NC group activated a network of regions which included bilateral occipitotemporal area (BA 37), the anterior cingulate (BA 32), left parietal (BA 40/7), left (BA 9/10) and right (BA 46) prefrontal, left middle temporal gyrus (BA 21), and the left parahippocampal gyrus. The MCI group activated a similar network of regions, but with no apparent activation in the left PHG, the right prefrontal region, and the medial prefrontal region (BA 10). Activation in the occipitotemporal region was reduced, but not absent while the left parietal region was increased relative to the NC. The AD subject showed similar activations compared to the MCI group. The shallow encoding vs baseline contrast demonstrated a reduction in activation in the occipitotemporal regions compared to deep encoding in both the NC and the MCI groups and a lack of significant prefrontal activation. The NC group relative to the MCI group demonstrated bilateral prefrontal and posterior cingulate activation during deep retrieval compared to the MCI group who activated predominantly left prefrontal regions. The NC group also activated the left ⬎ right hippocampal region during shallow retrieval (Fig. 1). The MCI group demonstrated a reduction in prefrontal activation during both retrieval conditions, but showed increased parietal activation. The AD subject demonstrated greater bilateral parietal and inferior frontal (BA 44) activity in both deep and shallow retrieval conditions. Discussion It would be useful to be able to characterize individuals with MCI who are more likely to progress to AD. The purpose of this early work, therefore, was to demonstrate that we could show differences in activations in our MCI group compared to the NC group. Our MCI group relative to our NC group did not perform significantly different on the encoding tasks. However, the MCI group failed to activate the medial frontal

FIG. 1. Increased left hippocampal activation (white arrow) during shallow retrieval in the NC group (left image) compared to the MCI group (right image).

228

ROTMAN INSTITUTE ABSTRACTS

and parahippocampal regions and showed reduced occipitotemporal activation during the deep encoding condition. The occipitotemporal, PHG, and frontal regions may be important in establishing new memories. Compensatory activation of the parietal region in the AD subject and the MCI group may have been necessary to carry out the deep encoding task at an adequate level. During the shallow encoding task, lack of prefrontal activity in the NC and MCI group may have resulted from less semantic elaboration or task difficulty required by the task. During retrieval, the NC activated to a greater extent the prefrontal regions compared to the MCI group. The prefrontal region may be recruited for successful execution of the task and also may be involved in retrieval success. In addition, the MCI group and AD subject failed to activate the hippocampus in both conditions. Hippocampal activation may have only reached significance during shallow retrieval in the NC group, as it may be required for satisfactory task performance during this more difficult condition. Future work will be required to quantify these differences in activation patterns and to examine more closely individual differences in the MCI group. By comparing the activation patterns of individuals with MCI to more AD subjects, we hope to identify prodromal cases that are more likely to progress to AD.

REFERENCES Bandettini, P. A., Jesmanowicz, A., Wong, E. C., & Hyde, J. S. (1993). Processing strategies for timecourse data sets in functional MRI of the human brain. Magnetic Resonance in Medicine, 30, 161– 173. Ohm, T. G., Muller, H., Braak, H., & Bohl, J. (1995). Close-meshed prevalence rates of different stages as a tool to uncover the rate of Alzheimer’s disease-related neurofibrillary changes. Neuroscience, 64, 209–217. Ogawa, S, Lee, T. M., Kay, A. R., & Tank, D. W. (1990). Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proceedings National Academy of Sciences USA, 87, 9868–9872. Petersen, R. C., Smith, G. E., Waring, S. C., Ivnik, J., Tangalos, E. G., & Kokmen, E. (1999). Mild cognitive impairment: Clinical characterization and outcome. Archives of Neurology, 56, 303–308. 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. Annals of Neurology, 45, 466–472. This is doi:10.1006/brcg.2001.1472.

11. Cognitive Impairment in Dementia: Correlations with Atrophy and Cerebrovascular Disease Quantified by Magnetic Resonance Imaging

R. H. Swartz, S. E. Black, G. Sela, and M. J. Bronskill This project assessed the contributions of atrophy and cerebrovascular disease (CVD) to cognitive impairment in dementia. Ten individuals with clinically diagnosed pure VaD were age-, sex-, and education-matched to individuals with AD. All participants underwent neuropsychological testing and MRI which were processed to generate quantitative indices of atrophy and CVD. A linear regression, including thalamic lesion and vCSF volumes, predicted cognitive status (R 2 ⫽ .74; p ⬍ .0005). Three VaD subgroups were identified: thalamic lesion (n ⫽ 4), hippocampal infarcts (n ⫽ 3), and other (n ⫽ 3). In participants without thalamic lesion, vCSF predicted general cognition (R 2 ⫽ .48), hippocampal atrophy predicted memory

ROTMAN INSTITUTE ABSTRACTS

229

impairment (R 2 ⫽ .33), and white matter lesions predicted executive dysfunction (R 2 ⫽ .48). Both atrophy and CVD burden correlated highly with cognitive impairment and should be simultaneously assessed in studies of brain–behaviour relations in dementia.  2001 Elsevier Science (USA)

Background Alzheimer’s disease (AD) and vascular dementia (VaD) are two common causes of dementia which frequently co-occur. Cerebrovascular disease (CVD), in addition to causing VaD, has been shown to increase the prevalence and severity of dementia in individuals with co-occurring AD. Fewer pathological AD changes are required to cause clinical dementia in those with infarcts to the thalamus, basal ganglia, or deep white matter (Snowdon et al., 1997). This relationship has not been demonstrated in vivo. Most studies of progression and treatment examine ‘‘pure’’ populations of either AD or VaD. However, current clinical diagnostic criteria cannot reliably distinguish VaD from AD. Many individuals with AD have cerebrovascular brain changes and many individuals with clinically diagnosed VaD are found to have co-occurring AD pathology at autopsy. In the current climate of experimental treatments for dementia, it is becoming increasingly important to understand how AD and CVD interact in vivo. Purpose This project set out to assess the simultaneous contributions of atrophy and cerebrovascular disease to the type and severity of cognitive impairments in relatively pure groups of individuals with AD and VaD. Method The Sunnybrook & Women’s College Health Sciences Centre Dementia database was queried for individuals who had magnetic resonance imaging (MRI), single photon emission computed tomography (SPECT) scanning, and detailed neuropsychological testing within 10 weeks. The subset of cognitive tests used in this study included: the Mattis Dementia Rating Scale (DRS) (1976) as an index of general cognitive function, the California Verbal Learning Test as a measure of verbal memory, and the Wisconsin Card Sort Test (WCST) as an index of executive function. From the set of 125 individuals with dementia who completed all tests, 10 individuals were identified who met NINDS-AIREN criteria (Roman et al., 1993) for probable VaD. These individuals were age-, sex-, and education-matched with 10 individuals judged clinically to have pure AD (McKhann et al., 1984). MR images were acquired using a 1.5-T Signa MR imager (GE Medical Systems, Milwaukee, WI) at Sunnybrook and Women’s CHSC in Toronto. A standard interleaved spin-echo acquisition was performed in the axial plane covering the whole brain including cerebellum. T2- and proton density-weighted MR images were acquired without gaps using 3-mm-thick slices (TE ⫽ 30, 80 ms; TR ⫽ 3000 ms, 0.5 excitations, field of view 20 ⫻ 20 cm, matrix 256 ⫻ 192). Images were transferred to a Sun workstation (Sun Microsystems Inc., Mountain View, CA) and all processing was performed blind to patient demographic, clinical, and neuropsychological information. The MR scans were processed using a set of newly developed and validated, semiautomated algorithms to generate quantitative indices of atrophy (ventricular (v-) and sulcal (s-)CSF) and cerebrovascular disease (thalamic, basal ganglia, periventricular,

230

ROTMAN INSTITUTE ABSTRACTS

and deep white matter hyperintensity volumes). The initial segmentation was performed using a two dimensional k-NN approach described elsewhere (Byrum et al., 1996), with modifications to include hyperintensities. Briefly, a structured ‘‘seeding’’ protocol was applied over three slices to select a total of 30 training points each for gray matter, white matter, and CSF. Lesion voxels were selected by a standardized protocol as follows: one seed for each periventricular cap or rim in each hemisphere and one seed for each white matter hyperintensity, to a maximum of 10 points per slice, over three slices for a total maximum of 30 points. Segmented images were generated from k-NN extrapolation to all voxels in the image. Using a three-dimensional floodfill tool, subtypes of hyperintensities were separated from the general deep white matter lesions. Hyperintensities in the deep gray matter structures were outlined and renamed as either basal ganglia or thalamic hyperintensities and periventricular hyperintensities were defined automatically by connection to the ventricles. In addition, hippocampal and amygdala/parahippocampal volumes were quantified using user-defined intensity threshold traces of both left and right sides. The average volumes were used except in three cases where posterior cerebral artery strokes had ablated one hippocampus. In this case, the contralateral hippocampal volume was used as an estimate of hippocampal atrophy. All volumes could be collected reliably, with inter- and intraclass correlation coefficients over 0.8. In order to account for individual variations in brain size, all volumes were expressed as a percentage of total intracranial capacity (TIC). Analysis of variance was performed to evaluate differences between the AD and VaD groups. In order to determine which brain measures best predicted general neuropsychological status as measured by the DRS (Mattis, 1976), a two-block linear regression model was performed (block one entered age and education, block two allowed brain measures to enter stepwise if significant). A logistic regression was performed to determine which measures, if any, could separate those with pure AD from those with VaD. Results As expected, since presence of significant lesion was used to exclude cases of mixed disease, there were significant differences between the two groups on hyperintensity measures. However, after accounting for multiple comparisons, there were no significant differences between the groups on any measures of atrophy or cognitive function. Regardless of diagnosis, a linear regression model was found that predicted cognitive status as measured by the DRS (R 2 ⫽ .74; p ⬍ .0005; see Fig. 1). Thalamic lesion and vCSF volumes were the only two variables that contributed significantly to this model. Logistic regression analysis separated the 20 individuals into two groups with 100% accuracy (p ⬍ .0005) with a model that included hippocampal, vCSF, and thalamic lesion volumes. Based on these observations, three subgroups in the VaD sample were identified. One group of four individuals had thalamic lesions, another group of three individuals had infarcts involving the hippocampus, and a final group of three individuals had small vessel disease (periventricular and deep white matter hyperintensities) with no strategic infarcts detected. These individuals were most similar to the AD group in their neuropsychological functioning. Thalamic infarcts were highly correlated with cognitive function. Thus, in order to explore the contributions of other brain lesions to cognitive status, we performed a series of regression models in participants without thalamic infarcts. In this group, increased vCSF volume predicted lower DRS score (R 2 ⫽ .48; p ⬍ .01); smaller hippocampal volumes predicted lower memory performance (R 2 ⫽ .33; p ⬍ .01); and more white matter lesion predicted poorer frontal lobe function measured by the Wisconsin Card Sort Test (R 2 ⫽ .48; p ⬍ .01).

ROTMAN INSTITUTE ABSTRACTS

231

FIG. 1. Plot of the observed Dementia Rating Scale (DRS) versus the score predicted by the regression function including vCSF and thalamic lesion volumes (R 2 ⫽ .74; p ⬍ .0005). Three VaD subgroups (with hippocampal, thalamic infarcts, or subcortical deep white and periventricular hyperintensities) were identified. (Note: two individuals with AD are superimposed at observed DRS ⫽ 113.)

Discussion While those with AD and VaD differed on measures of cerebrovascular disease, there were no differences on measures of atrophy. This may be due to global and hippocampal atrophy caused by lesions to the thalamus and deep white matter. Alternatively, the same pathological processes that place individuals at risk for cerebrovascular disease may also cause global cell and neurite loss. In either case, measures of atrophy were not specific to AD and cannot be used as an index of AD pathology. However, the group of VaD participants who segregated with the AD volunteers and had no strategic infarcts may be the most likely to have underlying AD pathology unmasked by their cerebrovascular disease, rather than vascular pathology as their sole cause of dementia. While thalamic lesions were very important determinants of global cognitive status in those that had them, global atrophy as indexed by vCSF volume was also a significant predictor of cognition, in both the VaD and the AD populations. Either atrophy or lesions to different brain regions may contribute to the expression of dementia in different ways. For example, hippocampal atrophy predicted memory function, while deep white matter lesions predicted executive dysfunction. Measures of atrophy and cerebrovascular disease can be used to generate an index of pathological severity that correlates highly with cognitive impairment in dementia. This approach may be useful for understanding the simultaneous contribution of vascular disease and AD, especially in studies of the diagnosis, progression, and treatment of dementia. A larger study to examine the relationships between atrophy, cerebrovascular disease and cognitive status in a heterogeneous memory clinic sample would be valuable. REFERENCES Byrum, C. E., Macfall, J. R., Charles, H. C., Chitilla, V. R., Boyko, O. B., Upchurch, L., Smith, J. S., Rajagopalan, P., Passe, T., Kim, D., Xanthakos, S., Ranga, K., & Krishan, R. (1996). Accuracy

232

ROTMAN INSTITUTE ABSTRACTS

and reproducibility of brain and tissue volumes using a magnetic resonance segmentation method. Psychiatry Research, 67, 215–234. Mattis S. (1976). Mental status examination for organic mental syndrome in the elderly patient. In L. Bellak and T. B. Karasu (Eds.). Geriatric psychiatry. New York: Grune and Stratton. McKhann, G., Drachman, D. A., Folstein, M. F., Katzman, R., Price, D. L., & Stadlan, E. M. (1984). Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Workgroup under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology, 34, 939–944. Roman, G. C., Tatemichi, T. K., Erkinjuntti, T., Cummings, J. L., Masdeu, J. C., Garcia, J. H., Amaducci, L., Orgogozo, J. M., Brun, A., & Hofman, A. (1993). Vascular dementia: Diagnostic criteria for research studies. Neurology, 43, 250–260. Snowdon, D. A., Greiner, L. H., Mortimer, J. A., Riley, K. P., Greiner, P. A., & Markesbery, W. R. (1997). Brain infarction and the clinical expression of Alzheimer disease: The Nun Study. Journal of the American Medical Association, 277, 813–817. This is doi:10.1006/brcg.2001.1473.

12. Nicotine and Sensory Memory in Alzheimer’s Disease: An Event-Related Potential Study

C. Engeland, C. Mahoney, E. Mohr, V. Ilivitsky, and V. Knott The auditory mismatch negativity (MMN) event-related brain potential (ERP) reflects the storage of information in acoustic sensory memory. Thirteen patients with probable Alzheimer’s disease (AD), 6 receiving treatment with the cholinesterase inhibitor (ChEI) tacrine (tetrahydroaminoacridine, THA) and 7 receiving no treatment, were administered 2 mg of nicotine polacrilex and placebo. MMNs were recorded with 1- and 3-s interstimulus intervals pre- and postplacebo/nicotine administration. In nontreated patients, amplitudes were decreased from pre- to postplacebo recordings but remained stable in THA-treated patients. Comparison of pre- and postnicotine MMNs found amplitude increases with nicotine in nontreated but not THA-treated patients. MMN latencies were shortened by nicotine in both treatment groups. These exploratory findings suggest that nicotine-improved strength of acoustic sensory memory traces and speed of acoustic sensory discrimination in AD are differentially affected by chronic ChEI treatment.  2001 Elsevier Science (USA)

Rationale Physical features of repetitive (‘‘standard’’) auditory stimuli are purported to be fully analysed and encoded as neural traces in short-term echoic memory (i.e., the preattentive sensory register). The electroencephalographically (EEG)-derived mismatch negativity (MMN) component of the event-related potential (ERP) is believed to be elicited (via an automatic comparator process) each time any afferent auditory input fails to match (i.e., ‘‘deviates’’ from) features encoded in the prevailing neuronal representation. Under certain stimulus conditions, MMN recordings have shown altered auditory trace decay in the normal elderly and in patients with dementia of the Alzheimer type (AD) (Gene´-Cos, Ring, Pottinger, & Barett, 1999). Cholinesterase inhibitor (ChEI) treatment of AD with single doses of tacrine (tetrahydroaminoacridine, THA) disrupts MMNs (Riekkinen, Pa¨a¨kko¨nen, Karhu, Partanen, Soininen, Luakso, & Riekkinen, 1997) but chronic ChEI treatment effects have yet to be examined. As selective activation of neuronal nicotinic receptors have been advocated as a promising approach to (a) ameliorating cholinergic deficits in attention deficit disorder and (b) augmenting ChEI treatment in AD (Sjo¨berg, Svensson, Zhang, & Nordberg, 1998), this pilot study explored the acute effects of nicotine on MMN in THAtreated and nontreated AD patients.

ROTMAN INSTITUTE ABSTRACTS

233

Subjects. Thirteen (7 male) volunteers (mean age 71.6 years; range 53–82), assessed via laboratory (including EKG and blood and urine analysis), neurological (including CAT scans), and psychiatric screens, and meeting DSM-IV criteria for probable AD, were included in the study. Seven (5 male) patients had not received any pharmacological treatments with CNS agents for at least a 1-month period (nontreated group). The other 6 (5 male) patients were receiving tacrine treatment (THAtreated group) and had been at their maximally tolerated dose (mean 140.0 mg/day; range 80–160 mg/day) for 6 months or longer (mean 11.6 months; range 6–15 months). The mean and range of Mini-Mental State Examination (MMSE) rating scores for the patients were 23.8 and 4–30, respectively. The average MMSE score (25.5) in the 6 THA-treated patients at the time of this study had not substantially changed from the average MMSE score (26.3) assessed at the beginning of their tacrine treatment. MMSE scores of THA-treated patients did not differ from scores of nontreated patients. Design. Patients attended the laboratory for one ‘‘orientation’’ session for familiarization with study procedures and for two additional ‘‘test’’ sessions in which they received either placebo and nicotine within a repeated measures, pseudo-randomized, double-blind, crossover design. MMNs and vital signs were assessed pre- and postplacebo/nicotine administration. Treatments. Nicotine was administered orally in the form of nicotine polacrilex (Nicorette), 2 mg. Both the nicotine and the placebo gum were chewed in compliance with Nicorette instructions (two bites/minute for 20 min) and in so doing, the nicotine gum produced an estimated blood nicotine level of 5.0 ng/ml approximately 25 min from the initiation of chewing. Stimuli. ERPs were elicited with the presentation of 400 sinusoidal tone (60 dB SL) stimuli (70-ms duration; 7-ms rise/decay) delivered to the right tear in an ‘‘oddball’’ format in three separate blocks: 200 tones in one block using an ISI of 1 s and 100 tones in each of the other two blocks using an ISI of 3 s. Each block consisted of standard (85%) and deviant (15%) tones with frequencies of 800 and 550 Hz, respectively. To divert attention away from the tones, patients watched a silent video and were instructed not to attend to the auditory stimuli. ERPs. Frontal (Fz, F3, F4) EEG and vertical/horizontal electrooculographic (EOG) signals were amplified using a 1.0-s time constant and a 30-Hz filter. Analogto-digital sampling time locked to each stimulus was carried out at 512 Hz for 350 ms beginning 50 ms pre-stimulus onset. Digital values of ERP averages (corrected for EOG artifact) elicited by standard stimuli were subtracted from digital values of ERP averages elicited by deviant stimuli, and the resulting MMN component was scored for peak amplitude (maximum negative values between 50 and 250 ms poststimulus onset; measured with respect to prestimulus average) and latency (i.e., time to peak amplitude). Parametric, univariate split-plot repeated measures analysis of variance (ANOVA) procedures and post hoc t tests were used to statistically analyze the data.

Results In general, MMN amplitudes appeared to diminish postplacebo but this was significant only in the nontreated group (p ⬍ .01). Similarly, only the nontreated group evidenced significant nicotine-induced MMN amplitude changes, with amplitudes increasing postnicotine administration compared to prenicotine administration (p ⬍ .05). Analysis of MMN latency values found, in both nontreated and THA-treated groups, postplacebo latencies to be significantly slower than both preplacebo latencies

234

ROTMAN INSTITUTE ABSTRACTS

(p ⬍ .01) and postnicotine latencies (p ⬍ .05). No significant latency or amplitude differences were observed between groups or between ISI conditions. Conclusions These exploratory results, achieved only with a single, low dose of nicotine and in relatively small samples, must be treated cautiously especially as they were obtained in the absence of behavioral assessments which may have provided useful insight into the relationship between MMN alterations and clinical/cognitive deficits in AD. The postplacebo reduction in MMN amplitudes seen in nontreated, but not in THA-treated, patients suggests that sustained ChEI treatment may act to stabilize acoustic traces over short time intervals, an effect which appears to be at odds with that observed with ChEI administered acutely. The apparent strengthening of acoustic sensory memory traces, seen with amplitude increments postnicotine, was limited to nontreated patients, suggesting that chronic treatment using ChEI may modulate neuronal nicotinic receptor sensitivity in AD. The general shortening of MMN latencies induced by nicotine, reflecting faster sensory discrimination, may be of relevance to subsequent and/or parallel cognitive processing of the sensory input as AD is frequently associated with mental and motoric slowness. REFERENCES Gene´-Cos, N., Ring, H., Pottinger, R., & Barett, G. (1999). Possible roles for mismatch negativity in neuropsychiatry. Neuropsychiatry, Neuropsychology and Behavioural Neurology, 12, 17–27. Riekkinen, P., Jr., Pa¨a¨kko¨nen, A., Karhu, J., Partanen, J., Soininen, H., Laakso, M., & Riekkinen, P., Sr. (1997). THA disrupts mismatch negativity. Psychopharmacology 133, 203–206. Sjo¨berg, R., Svensson, A., Zhang, X., & Nordberg, A. (1998). Neuronal nicotinic receptor activation: A processing strategy for treatment of Alzheimer’s disease. International Journal of Geriatric Psychopharmacology, 1, 145–149. This is doi:10.1006/brcg.2001.1474.

13. rCBF/SPECT in the Evaluation of Inner-City Minority Patients with a History of Impaired Memory: A Pilot Blind Read Pre- and Poststudy

S. Jonas, R. Van Heertum, R. Tikofsky, E. Millman, D. Singh, J. Brust, H. McCurtis, and R. Locko Eight patients (seven women), mean ⫾ SD T1 age 68.57 ⫾ 12.43 years, average educational level 5.83 ⫾ 3.70 years, had two Tc-99m ECD SPECT examinations separated by an average 8.49 ⫾ 5.59 months. Patients were imaged using standard Harlem Hospital acquisition and processing protocols with approximately 30 mCi of ECD on a Prism 3000 triple head gamma camera. Images were interpreted by an independent reader blinded to the patients’ clinical history and imaging date. T1 psychiatric diagnosis was seven Alzheimer’s dementia (AD) and one depression. Eight T1 images were interpreted as abnormal, six indicative of AD. Binomial 95% two-tail confidence interval for T1 agreement between diagnosis and interpretation was 0.25 0.63 0.92. T2 diagnosis was seven AD and one none. Seven T2 images were abnormal and indicative of AD, and one was normal. T2 confidence interval was 0.34 0.75 0.97. These findings suggest SPECT’s value in assessing AD in uneducated socioeconomically disadvantaged geriatric patients.  2002 Elsevier Science (USA)

Introduction Inner-city minority geriatric patients are often socioeconomically disadvantaged with low levels of formal education. These patients frequently obtain scores on the

ROTMAN INSTITUTE ABSTRACTS

235

MMSE suggestive of Alzheimer’s disease. However, establishing a firm diagnosis is clouded by their socioeconomic, educational, and linguistic background. This pilot study estimates the extent to which Tc-99m ECD SPECT images of geriatric innercity minority patients obtained before and after a period of geriatric psychiatric treatment corroborate psychiatrists’ clinical diagnosis of dementia of the Alzheimer’s type (AD). Images were interpreted by an expert nuclear medicine physician blind to clinical history and the imaging dates. Background Difficulties in assessing geriatric patients with little formal education using psychiatric and psychometric assessment suspected of being demented is well known. Anthony et al. (1982) studied 91 patients in a hospital general medical unit with the MMSE. They found 87% sensitivity and 82% specificity in detecting dementia and delirium as clinically diagnosed by a psychiatrist. False positives (39%) had less than 9 years of education and many were 60 years or older. Clinical experience with rCBF/ SPECT demonstrates that there are specific patterns of radiotracer uptake indicative of Alzheimer’s dementia (Van Heertum et al., 2000). Brain SPECT images of patients classified by neurologists as presumed or confirmed Alzheimer’s disease, multiinfarct dementia, HIV-related dementia, ‘‘mixed,’’ and normal are concordant across institutions and readers (Hellman et al., 1994). Blinded visual interpretation of 220 SPECT scans was reported. One-hundred and four patients who met NIDA-ADRDA criteria for probable AD (43 with early onset, age ⬍65 years; 56 with late onset) showed the classic pattern of bilateral parietotemporal (PT) hypoperfusion. It was most often seen in severe AD, early onset, and in men independent of duration of symptoms, tracer, or camera (Nitrini et al., 2000). Similar results are reported by Masterman et al. (1997). Based on a blinded analysis of SPECT scans on 139 patients who met NIDA-ARDA criteria for dementia they report that all patients whose SPECT scans showed bilateral hypoperfusion in the PT regions had dementia. This pattern was most frequent in AD patients, but was seen in other patients, and thus cannot be considered as specific to AD. However, they conclude that demented and nondemented persons can be differentiated by reduced PT perfusion. Patient Characteristics Eight ethnic minority group inner-city socioeconomically disadvantaged patients (seven women, one man) presenting to the Geriatric Psychiatry service of Harlem Hospital Center with complaints of memory problems were evaluated by two psychiatrists boarded in Geriatric Psychiatry. Based on neuropsychological and psychosocial assessments, physical health history and examination, and clinical interviews with the patients and available family members or close associates, five patients were initially diagnosed (T1) as having Alzheimer’s dementia of the (AD), one AD with depression, one AD with alcohol dependence, and one depressed. On reassessment (T2) the psychiatric diagnosis remained the same for all but one patient who originally was diagnosed as depressed and T2 as not demented. Patient age at T1 ranged from 41.20 to 83.40 years, with mean 68.57 years and standard deviation ⫾ 12.43 years. At T2 patient age ranged from 41.80 to 83.50 years, with mean 69.25 and standard deviation ⫾ 12.29 years. Their highest completed grade of education ranged from 1 to 10 with mean 5.38 and standard deviation ⫾ 3.70. SPECT Acquisition Protocols SPECT imaging was performed in the Nuclear Medicine Laboratory, Department of Radiology, Harlem Hospital Center, using standard data acquisition protocols.

236

ROTMAN INSTITUTE ABSTRACTS

Subjects were placed in a quiet room with the lights on, and an intravenous line was placed in the antecubital vein. Subjects remained in the room for about 5 min. Between 20 and 30 mCi of Tc-99m-ECD was injected through the existing line. Time between injection and image acquisition varied from 30 min to 1 h. Image acquisition was performed with Picker 3000 triple-head gamma camera equipped with ultra high resolution fan beam collimators. Patients were positioned on the camera gantry in a foam padded head holder, and the collimators were adjusted to a minimal radius of rotation as function of the subjects’ size. Four rapid acquisition sequences were performed. The acquired data were then processed and reconstructed using the Harlem Hospital Nuclear Medicine Laboratory protocols. Resultant images were displayed in all three planes, transaxial, coronal, and saggital, for interpretation. The interval from first to second SPECT imaging (T1 to T2 interval) ranged from 1.00 to 16.30 months with mean 8.49 and standard deviation ⫾ 5.59. SPECT Interpretation Method A nuclear medicine physician experienced in interpreting brain SPECT studies viewed the images using a lightbox presentation. The reader was blinded to the patients’ clinical history and the date of imaging. He rated all 16 images [T1 and T2 combined, randomized with respect to both patient and time of imaging (T1 vs T2)]. Images were rated on the basis of: (1) image quality—good, adequate, poor, or uninterpretable; (2) overall impression—normal or abnormal; (3) type of abnormality— global, focal, or combined; (4) severity of abnormality—minimal, moderate, or severe; (5) laterality of abnormality—bilateral of hemisphere (right or left); (6) primary perfusion deficit—absent, decreased, or increased; and (7) diagnostic impression— high probability AD, low probability AD, vascular dementia, frontal dementia, cardiovascular dementia, depression, trauma, other psychiatric, NPH, and/or atrophy appropriate for age. Findings Eight T1 images were interpreted as abnormal, six indicative of AD. Of the eight T1 scans six were read as having both focal and global abnormalities, one as global only, and one as focal only. Five of eight T1 images were read as showing severe abnormality, one as moderate, one as minimal, and one as none. The binomial 95% two-tail confidence interval for T1 agreement between geriatric psychiatric diagnosis of AD and SPECT interpretation indicative of AD was 0.25 ⋅ 0.63 ⋅ 0.92. At T2 seven images were read as abnormal and indicative of AD, and one was normal. Four T2 images were read as having both focal and global abnormalities, three as focal only, one as global only, and one as none. At T2 no images were read as showing severe abnormality (a decrease of 5), six as moderate (an increase of 5), one as minimal, and one as none. The T2 confidence interval for agreement of AD diagnosis and impression was 0.34 ⋅ 0.75 ⋅ 0.97. The decrease in severity of perfusion abnormalities from T1 to T2 is clear. When dichotomized ‘‘severe’’ versus ‘‘less than severe’’ five of eight patients changed from severe at T1 to less than severe at T2, whereas three patients were less than severe at both T1 and T2. All five T1 scans which showed severe abnormality showed less than severe abnormality at T2 (Wilcoxon Matched-pairs signed-ranks test Z ⫽ ⫺2.02, two-tailed p ⫽ .04). Discussion Agreement between scan diagnostic impressions and geriatric psychiatric diagnosis was significant (p ⬍ .05) for T1 and T2. Seven of eight patients had an AD geriatric

ROTMAN INSTITUTE ABSTRACTS

237

psychiatric diagnosis. Of these seven patients, at T1 the reader reported a diagnostic impression of AD for five scans. Of the other three T1 scans, one was interpreted as AD but with no geriatric psychiatry dementia diagnosis, and the other two scans had diagnostic impressions other than AD with a geriatric psychiatric dementia diagnosis. At T2 the reader and psychiatrists designated AD in six of eight patients. There was a pair of one-case discrepancies. The reader’s diagnostic impressions at T1 and T2 had a prevalence of AD equivalent to that of the psychiatrists (two-tailed binomial p ⫽ 1.000). However the T1 scan interpretations were distinctly different from those of T2. The patient whose T1 scan showed no perfusion abnormality showed a moderate perfusion abnormality at T2. However, the scan interpreted as severe perfusion abnormality at T1 was reported to show no perfusion abnormality on the T2 scan. The predominant SPECT scan abnormality at both T1 and T2 was bilaterally decreased perfusion in the TP regions. At T1 all eight were scans read as having bilateral abnormalities. At T2 seven of eight scans were read as showing bilateral abnormalities, and one as none. Conclusion Results of this pilot study suggest brain SPECT’s value in the assessment for AD in inner-city ethnic minority socioeconomically disadvantaged geriatric patients with low levels of formal education whose cognitive functions may be compromised. The results of the blindly read SPECT identified AD with findings specific to T2 versus T1 images. This also suggests that SPECT may be valuable in detecting change as a result of treatment. REFERENCES Anthony, J. C., LeResche, L., Niaz, U., Von Korff, M. R., & Folstein, M. F. (1982). Limits of the ‘MiniMental State’ as a screening test for dementia and delirium among hospital patients. Psychological Medicine, 12, 397–408. Hellman, R. S., Tikofsky, R. S., Van Heertum R. L., Coade, G., Carretta, R., & Hoffmann, R. G. (1994). A multi-institutional study of interobserver agreement in the evaluation Of dementia with rCBF/ SPET technetium-99m exametazime (HMPAO). European Journal of Nuclear Medicine, 21, 306– 313. Masterman, D. L., Mendez, M. F., Fairbanks, L. A., & Cummings, J. L. (1997). Sensitivity, specificity, and positive predictive value of technetium 99-HMPAO SPECT in discriminating Alzheimer’s disease from other dementias. Journal of Geriatric Psychiatry and Neurology, 10, 15–21. Nitrini, R., Buchpiguel, C. A., Caramelli, P., Bahia, V. S., Mathias, S. C., Nascimento, C. M., Degenszajn, J., & Caixeta, L. (2000). SPECT in Alzheimer’s disease: Features associated with bilateral parietotemporal hypoperfusion. Acta Neurologica Scandinavica, 101, 172-–176. Van Heertum, R. L., Tikofsky, R. S., & Rubens, A. B. (2000). Dementia. In R. L. Van Heertum & R. S. Tikofsky (Eds.). Functional cerebral SPECT and PET imaging (3rd ed.) (pp. 127–188). Philadelphia, PA: Lippincott Williams & Wilkins. This is doi:10.1006/brcg.2001.1475.

14. Evaluation of Prospective Memory Training for Individuals with Mild Alzheimer’s Disease

J. S. Kixmiller Prospective memory (PM) refers to the timely execution of a previously formed intention. PM deficits are one of the earliest measurable deficits in Alzheimer’s disease (AD; Huppert &

238

ROTMAN INSTITUTE ABSTRACTS

Beardsall, 1993) and yet, there has been little systematic research devoted to identifying ways in which these deficits can be compensated for or minimized. In this pilot study, individuals with mild AD participated in a new training program aimed at improving PM performance on an experimental PM task. PM training consisted of errorless learning and spaced retrieval techniques taught in six sessions in subjects’ homes. Results showed that compared to untrained AD controls, AD patients who received PM training performed a PM task successfully across 7 weeks posttreatment (over 90% accuracy; controls 16–33% accuracy).  2002 Elsevier Science (USA)

Report

Prospective memory (PM) refers to remembering to perform an intention at a particular time in the future or as the timely execution of a previously formed intention (Kvavilashvili & Ellis, 1996). PM deficits are one of the earliest measurable deficits in the progression of Alzheimer’s disease (AD; Huppert & Beardsall, 1993). Despite the adverse impact that poor PM performance has on mild AD patients’ lives and the burden that is experienced by the caregivers, there has been little systematic research devoted to identifying ways in which these deficits can be compensated for or minimized. Several studies have examined the effect of PM training in AD patients (McKitrick, Camp, & Black, 1992; Camp, Foss, Stevens, & O’Hanlon, 1996). Both studies trained AD patients to perform simple PM tasks using spaced retrieval, which requires subjects to recall newly acquired information across increasingly longer retention periods. In one, subjects were trained to redeem a target-colored coupon when they next saw the researcher each week for several weeks following training. In the other task, AD subjects were trained to learn to look at a calendar on which was listed a PM task to be completed that day and which changed daily. Results were positive in terms of improved PM performance for both studies. In redeeming coupons following only two training sessions, 40% of mild AD patients redeemed one coupon in the first posttraining week, and 33% redeemed a second coupon in the second successive posttraining week. However, 27% failed to redeem any coupons. Results from the second experiment taught patients to perform a PM task that was written on the calendar each day as the to-be-performed activity. Results indicated that posttraining, over 61% of subjects had learned the correct verbal/motor responses to use the calendar to perform daily PM tasks within three training sessions or less (26% more learned after four to seven sessions; 13% never learned the strategy). Of the subjects who learned the correct verbal/motor calendar use, 35% were able to meet criteria for effective calendar PM task performance independently. However, 25% of patients who had initially learned the correct verbal–motor responses in training never learned to carry out the strategy effectively. Together, these results show that many mild AD patients are able to learn to carry out future intentions with PM training. However, given that 25% of the patients in both studies failed to perform PM tasks independently, augmented training with additional memory rehabilitation techniques and more extensive training may benefit the one-quarter of patients who failed PM tasks in posttraining periods. In this pilot study, individuals with mild AD participated in a new training program aimed at improving PM performance. PM training consisted of errorless learning and spaced retrieval learning techniques. The errorless learning approach optimized learning through training that avoids (or at least limits) errors during the acquisition phase of learning. Spaced retrieval relied on having subjects repeatedly recall learned information across increasingly longer retention periods.

ROTMAN INSTITUTE ABSTRACTS

239

Methods

Subjects Seven elderly individuals with mild Alzheimer’s disease (AD) were recruited from the University of California, Davis–Alzheimer’s Disease Research Center (UCDADC) and met the diagnostic criteria for probable AD developed by the National NINCDS and the ADRDA. Average MMSE scores were 20/30. Subjects were screened for any neurological events, history of alcoholism, or major psychiatric disturbance. Additionally, all individuals were screened for being able to physically carrying out simple household tasks such as answering and dialing the phone. Five of the patients were trained on the experimental tasks and two served as no-training control subjects. Experimental Tasks Over the course of 2 weeks (six training sessions total), subjects in the experimental group (n ⫽ 5) received training on one of two home-based, naturalistic PM tasks, an event-based task, or time-based task. The tasks are listed below. Event-based task. The event that initiated this task sequence was a telephone call from the research assistant (RA) to subjects at home. Subjects were trained to follow these steps in order: 1. Answer the phone; 2. Write down the date and time of the phone call; 3. Write down three pieces of information given by the experimenter (while still on the phone); 4. After hanging up, transfer the date, time, and three pieces of information to a blank postcard; 5. Address the postcard to the experimenter; 6. Stamp the postcard; and 7. Immediately put the postcard in the mailbox to go out. Time-based task. In this task, subjects were trained the following task steps: 1. Pick up the newspaper; 2. Read the newspaper and find three ‘‘target’’ pieces of information (one of the day’s front page headlines, description of the content of one of the front pages pictures, forecasted weather description); 3. Write down the three pieces of information onto a notepad; 4. At the SPECIFIED TIME (_ _:_ _), call the RA (phone number is provided to subjects); 5. Leave a message on the RA’s voicemail, containing the following information: • Your name, the date, and time of the call and • The three pieces of information from the newspaper; 6. Hang up. Experimental Training Procedure RA’s conducted all training in the subjects’ homes. The same RA worked with the subject throughout both time- and event-based task conditions so that no experimenter based artifacts contributed to performance differences. Training/test timeline. Training for a task was carried out for a 2-week period, during which the RA met with the subject 3⫻/week in his/her home. Training ses-

240

ROTMAN INSTITUTE ABSTRACTS

sions ranged between 30 and 120 min. Immediately following the training period, posttest evaluations began. Posttest periods were week-long intervals in which subjects were instructed to perform their PM task independently. RA’s called the subjects prior to each postevaluation week the night before to remind them to perform the PM tasks. The schedule of posttest evaluations was as follows: the week immediately following the completion of training, after 3 weeks, after 4 weeks, and after 7 weeks. Training content. During the 2-week/six session training period, the RA and subject will progress through an ordered series of stages. One of the goals of the training procedure was to reduce the possibility that the subject made mistakes as he/she learned the task. Therefore, early in the training process the subject’s actions were substantially under the control of the RA. As training progressed, the subject gradually became more independent until he/she was able to carry out the task independently. The errorless learning approach was as follows: 1. Observation/context setting: The subject watched videotaped examples of individuals carrying out the target task. 2. Walk through: The RA walked the subject through the task steps in his/her own home. 3. Verbal prompt: The RA gave the subject a verbal prompt to carry out each task step, one at a time, in order. 4. Self-instruction: The subject initiated verbal descriptions of each step prior to carrying it out. RA’s provided necessary correction before subject acted out the step, thereby preventing any motor errors. 5. Independent with feedback: The subject walked through steps of task and RA provided immediate correction for any mistakes. 6. Drilling/self-instruction: Subject practiced tasks using self-instruction and external cueing (in the form of a posted list of the task steps). RA supervised and provided feedback. 7. Independent: Subject carried out tasks independently and RA provided feedback only upon completion.

Control Group Procedure A RA visited subjects in the control group in their homes. Subjects were given tasks and questionnaires to complete so that they received an equivalent amount of training time to that received by the experimental subjects. During each experimental posttest period the control subjects also carried out the target task on the same schedule as the experimental subjects and their performance was evaluated the same way. They received a verbal description of the task, such as ‘‘When the RA calls you, write down the information you are given on a postcard along with the date and time and mail it back to us.’’ However, no instruction or training concerning task steps or order was given to the control subjects.

Results and Interpretations

Results from this PM training for mild AD patients are positive and suggest that PM training using errorless learning and spaced retrieval is efficacious for assisting mild AD patients with remembering to perform PM tasks in home settings. Compared to untrained AD controls, AD patients who received training using errorless learning and spaced retrieval performed a PM task well across 7 weeks posttreatment (see Table 1). Additional training process comparisons, comparisons of event-based and

241

ROTMAN INSTITUTE ABSTRACTS

TABLE 1

Mild AD trained subjects Mild AD controls (no treatment)

Posttest 1 (lag 0 weeks) (%)

Posttest 2 (lag 3 weeks) (%)

Posttest 3 (lag 4 weeks) (%)

Posttest 4 (lag 7 weeks) (%)

100

93

93

93

33

36

26

16

Note. Lag, No. weeks elapsed between training and posttests of subjects’ recall of prospective memory tasks.

time-based tasks, and qualitative performance measures will be discussed in the poster. REFERENCES Camp, C. J., Foss, J. W., Stevens, A. B., & O’Hanlon, A. M. (1996). Improving prospective memory task performance in persons with Alzheimer’s Disease. In M. Brandimonte, G. O. Einstein, & M. A. McDaniel (Eds.). Prospective memory: Theory and applications. Hillsdale, NJ: Erlbaum. Huppert, F. A., & Beardsall, L. (1993). Prospective memory impairment as an early indicator of dementia. Journal of Clinical and Experimental Neuropsychology, 15, 805–821. Kvavilashvili, L., & Ellis, J. (1996). Varieties of intention: Some distinctions and classifications. In M. Brandimonte, G. O. Einstein, & M. A. McDaniel (Eds.). Prospective memory: Theory and applications. pp. 23–51, Hillsdale, NJ: Erlbaum. McKitrick, L. A., Camp, C. J., & Black, W. (1992). Prospective memory intervention in Alzheimer’s disease. Journal of Gerontology, 47, 337–343. This is doi:10.1006/brcg.2001.1476.

15. Therapy for Anomia in Semantic Dementia

R. Jokel, E. Rochon, and C. Leonard Semantic dementia is a commonly accepted term for a language disorder resulting from neurodegenerative changes due to frontotemporal dementia. An intervention program was designed to halt and/or decelerate the effects of progressive anomia in AK, a 63-year-old female with semantic dementia. Pictorial stimuli were selected and labeled with their respective names and a description most relevant to AK’s experience. Daily homework assignments were carried out, during which AK looked at the picture, read the label aloud, and read the description. Positive short-term effects of treatment were observed on treated items.  2001 Elsevier Science (USA)

Introduction ‘‘Semantic dementia (SD) is a clinical designation applied to a multimodal breakdown of meaning. It is characterized by impaired word comprehension and naming in the presence of effortless, fluent and often rapid speech and preserved repetition, and coexistent or subsequent development of impaired face and object recognition in the presence of well preserved non-semantic perceptual skills’’ (Snowden et al., 1996, p. 92). Functionally, the most devastating and immediate feature of SD is the loss of expressive and receptive vocabulary. Traditionally, therapy for anomia has been carried out in cases where a certain degree of recovery can be predicted. Indeed,

242

ROTMAN INSTITUTE ABSTRACTS

approaches to remediate anomia arising from impairments of the semantic system have proven beneficial in patients recovering from a stroke (Hillis & Caramazza, 1994). We describe a treatment program for progressive loss of expressive and receptive vocabulary in AK, a patient with SD. We used a multimodality treatment approach in which a picture was paired with both a written word and a brief description of the pictured item, to facilitate retrieval of the semantic representation for that word. We predicted maintenance of the words AK knew at the start of the study, improvements on items she could not name and/or comprehend, and no generalization to the untreated items. Method Case study. AK is a 63-year-old right-handed woman with a 7-year history of progressive changes in her ability to retrieve and comprehend words. Her initial complaints were those of memory decline and a diagnosis of possible Alzheimer’s disease was made at that time. However, with time, it became clear that her memory decline was limited to words. As an avid concert goer, she was devastated by her inability to remember the names of musical instruments she used to know so well. At the time of this project, she presented with a classical profile of semantic dementia. Her spontaneous speech was fluent and anomic. Confrontation naming skills were severely impaired (8/60, BNT), while responsive naming (e.g., what do we cut paper with?) was intact (10/10, BDAE). Repetition was intact, as were other aspects of expressive speech, such as syntax, phonology, and prosody. AK’s auditory comprehension of sentences and complex morphosyntactic structures was also relatively normal, while her comprehension of single words was impaired (78/175, PPVT). Reading was characterized by surface dyslexia on those irregularly spelled words which she did not understand. Writing single words to dictation yielded a similar pattern of surface dysgraphia. All nonverbal cognitive skills were intact. A CT scan of the head was normal and her SPECT showed left temporal atrophy. A pilot treatment program was designed to investigate the potential for decelerating her progressive anomia. Stimuli A large set of pictures from the Peabody Picture Collection was administered to AK. The pictures were presented twice, once for confrontation naming and then for auditory comprehension, in order to establish the status of AK’s semantic knowledge for each picture. A final set of 180 pictures was selected. The set consisted of three subsets/conditions: 1. Sixty items that AK could name and comprehend (⫹N⫹C); 2. Sixty items she could not name but could comprehend (⫺N⫹C); and 3. Sixty items she could not name and could not comprehend (⫺N⫺C). Words from each of the three conditions were subsequently incorporated into a group of 90 words to be used as therapy items and 90 control, untreated items. Therapy and control items within each condition were balanced for word frequency and semantic category. The categories included food items, household items, musical instruments, clothing, personal care items, and accessories. The mean frequency count (Kucera & Francis, 1967) for all three conditions was as follows: treatment

243

ROTMAN INSTITUTE ABSTRACTS

items ⫹N⫹C 224.4, ⫺N⫹C 16.8, ⫺N⫺C 3.1; and control items ⫹N⫹C 192.9, ⫺N⫹C 13.3, ⫺N⫺C 4.4. Each treatment stimulus was labeled to provide orthographic and phonological information about the corresponding word. To evoke a stronger semantic representation, a description of the pictured item, most relevant to AK’s experience, was written on the back of each picture. Treatment program and design. Therapy was limited to 3 weeks in duration, during which time AK practiced the therapy items at home. The 90 items were divided into three groups of 30 (10 ⫹N⫹C, 10 ⫺N⫹C, 10 ⫺N⫺C). Each group of items was devoted to a separate week of practice. Items were practiced daily for approximately 30 min. Practice involved looking at the picture, reading the label aloud, and then reading the description on the back of the picture. For each set of 30 pictures, AK practiced for six consecutive days and came for testing on the seventh day. At that point she was presented with pictures she practiced that week (without the labels) interspersed with 30 control items, and the results were recorded as positive only when she was able to name the picture spontaneously. All 180 words were tested again 4 weeks and 6 months after the last therapy session to assess potential shortterm and long-term effects, respectively. Results A Fisher’s Exact Test was used to assess a treatment effect on naming ability at the three intervals (posttreatment, short term, long term). α was set at 0.006 using the Bonferroni adjustment. Within the ⫹N⫹C condition, no significant effects emerged. Within the ⫺N⫹C condition, an effect of therapy was found immediately posttreatment and marginally at short term ( p ⫽ .007). Within the ⫺N⫺C condition, a significant effect of treatment was found only immediately posttreatment. As evidenced in Table 1, AK showed immediate improvements in both the items she could not name prior to therapy (18) and those items she could neither name nor comprehend prior to therapy (11). Although without statistical significance, at 6 months posttherapy she could still retrieve practised words that she could not name (9/30) and comprehend (4/30) prior to therapy. Discussion The results of this pilot project are as predicted in suggesting that improvements in word retrieval are possible, even when there is a progressive loss of lexical knowledge. They demonstrate that it is possible to relearn items for which semantic knowledge has been partially or totally lost. Improvement in naming performance was seen immediately posttreatment for those items that our patient both could and could not comprehend. However, the results do suggest that comprehension may affect the ability to relearn items as, overall, more items were retrieved posttreatment under TABLE 1 Number of Pictures AK named Correctly (N ⫽ 30) Pretreatment

Short-term

Posttreatment

Long-term

Tx Items Controls Tx Controls Fisher Tx Controls Fisher Tx Controls Fisher ⫹N⫹C ⫺N⫹C ⫺N⫺C

30 0 0

30 0 0

26 18 11

22 4 1

.33 .0004 .0025

30 13 7

23 3 0

.01 .007 .01

24 9 4

18 1 0

.16 .01 .11

244

ROTMAN INSTITUTE ABSTRACTS

the ⫺N⫹C condition (n ⫽ 18) than under the ⫺N⫺C condition (n ⫽ 11) and the effect was more enduring. Generalization to untreated items was not observed. This is not unexpected since the treatment was item-specific and included semantic information about treated items only. Regarding the effect of therapy on the maintenance of semantic knowledge (i.e., under the ⫹N⫹C condition), the results are equivocal. While there was no significant difference in naming ability to therapy and control items at the different intervals, there does appear to be a trend toward a greater loss of control than therapy items long term (24 vs 18). The rate of decay of lexical/semantic knowledge is unknown and it is possible that a greater difference between therapy and control items would be seen at a longer interval posttherapy. While this possibility is intriguing and potentially useful for maintaining lexical/semantic knowledge in SD patients, it is speculative and awaits more rigorous testing. Consistent with the results of Graham et al. (1999), AK subsequently forgot some words on which she had shown improvement, but at a slightly slower rate than they reported. Their patient, DM retained only 40% of exemplars at 10 weeks posttreatment. These results are closer to the long-term effects (6 months posttreatment) observed in AK. Direct comparisons are difficult to perform, as AK’s testing was carried out at different intervals than DM’s. In addition to that, it also appears that DM’s comprehension deficits were milder than those of AK’s. Graham et al. commented on the fact that DM was their only patient who benefited from repeated practice. AK’s improvements and slower forgetting of the practiced items are encouraging with regard to the potential for treating naming deficits in SD. Ongoing work will incorporate these findings into a subsequent treatment program. REFERENCES Graham, K. S., Patterson, K., Pratt, K. H., & Hodges, J. R. (1999). Relearning and subsequent forgetting of semantic category exemplars in a case of semantic dementia. Neuropsychology, 13, 359–380. Hillis, A., & Caramazza, A. (1994). Theories of lexical processing and rehabilitation of lexical deficits. In M. Ridoch & G. Humpreys (Eds.). Cognitive neuropsychology and cognitive rehabilitation. Hillsdale, NJ: Erlbaum. Snowden, J. S. (1999). Semantic dysfunction in frontotemporal lobar degeneration. Dementia and Geriatric Cognitive Disorders, 10, 33–36. Snowden, J. S., Neary, D., & Mann, D. M. A. (1996). Fronto-temporal lobar degeneration. In Frontotemporal dementia, progressive aphasia, semantic dementia. New York: Churchill Livingstone. This is doi:10.1006/brcg.2001.1477.

16. Differential Memory Impairment in Dementia with Lewy Bodies and Alzheimer’s Disease

M. Simard, R. van Reekum, D. Myran, M. Panisset, T. Cohen, M. Freedman, S. Black, and B. Suvajac This study was conducted in order to elucidate the functioning of the Central Executive System of Working Memory (WM) and to clarify the status of other cognitive functions in Dementia with Lewy bodies (DLB) and Alzheimer’s disease (AD). Fourteen DLB, 22 AD, and 23 control subjects were assessed with the dual task paradigm and other cognitive tests. When compared with controls, DLB subjects performed more poorly in concurrent conditions on semantic WM tasks, and AD subjects performed more poorly on the spatial WM task. The DLB subjects had an inferior verbal span and AD subjects, an inferior recall on the CVLT.

ROTMAN INSTITUTE ABSTRACTS

245

These data suggest relative impairments of verbal and semantic WM in DLB and relative impairments of spatial WM and verbal episodic memory in AD.  2002 Elsevier Science (USA)

INTRODUCTION

The literature often reports more severe attentional and visuospatial deficits in subjects with dementia with Lewy bodies (DLB) than in subjects with Alzheimer’s disease (AD). One of the most salient findings is a more severe impairment of spatial working memory (a subsystem of working memory), in subjects with DLB compared to subjects with AD, in the mild and moderate stages of dementia. However, some discrepancies between the results exist. In addition, there are no data regarding the differential profile between subjects with DLB and AD on tests of the dual task paradigm measuring the functioning of the central executive system (CES) of working memory (WM) (Simard et al., 2000). These tasks have been shown to be sensitive to the dysexecutive syndrome and to cognitive deterioration overtime in Alzheimer’s disease (Baddeley et al., 1997). The primary goal of this study was to establish a differential cognitive profile between dementia with Lewy bodies and Alzheimer’s disease particularly with reference to tasks of the CES of WM that have not been studied between these two disorders. The secondary goal was to replicate the results of previous studies on other cognitive measures. METHOD Subjects

Two groups of subjects with dementia and one control group were recruited at Baycrest Centre for Geriatric Care and Sunnybrook Health Sciences Centre. Groups with Dementia

Inclusion Criteria The subjects with DLB met the criteria for probable or possible dementia with Lewy bodies of the consortium on dementia with Lewy bodies (CDLB). The subjects with AD met the NINCDS-ADRDA criteria for probable Alzheimer’s disease. Exclusion Criteria Exclusion criteria included CVA or other focal neurological disease; vascular dementia; and/or history of alcoholism or drug abuse. Supplementary Exclusion Criteria for Subjects with AD Supplementary exclusion criteria included history of hallucinations or presence of hallucinations and/or repeated falls. Healthy Control (HC) Group

Inclusion Criteria Patients were healthy male or female subjects, aged 50 to 85 years. Their total score on the Hamilton Depression Scale was ⬍17. Their total score on the Mattis Dementia Rating Scale (MDRS) was ⱖ136 (possible maximum 144).

246

ROTMAN INSTITUTE ABSTRACTS

MATERIAL AND PROCEDURES (PRIMARY VARIABLES) 1. The Dual Task (Baddeley et al., 1997)

Unitary Conditions (UC) a. Verbal part. (1) The examiner determined the length of the verbal span. The subject had to succeed in repeating successively three series of digits of the same length before continuing with longer series. The digit span (DS) task stopped after the subject had failed to repeat three successive digit series of the same length. The number of digits included in the last series repeated without mistake was recorded as the verbal span (secondary variable). (2) The subject had to repeat, after the examiner, as many digit series as he could for 2 min. To avoid the length effect of the digit span on the results of the dual task, the score of this phase (repetition of digits) gave the percentage of series correctly repeated in the unitary condition (UC). b. Visuospatial part. On the practice sheets each containing 40 boxes related to each other by arrows to indicate directions, the subject had to practice, for 1 min, the skill to properly fill each box with a cross. Then on the testing sheets (each containing 80 boxes related to each other by simple lines), the subject had, for 2 min, to fill as many boxes as he could with crosses, following the order of the lines. The score of this phase gave the number of boxes correctly filled by crosses in the unitary condition (UC). Concurrent Condition For 2 min, the subject had to simultaneously repeat series of digits and to execute the visual-tracking task. This phase gave two scores: the percentage of digit series correctly repeated in the concurrent condition (CC) and the difference between the number of boxes correctly filled by crosses in the unitary condition and the number of boxes correctly filled by crosses in the concurrent condition (UC-CC). 2. The Talking/Walking Test

Unitary Condition (UC) (1) The subject had to walk down a hallway (15 ft long, indicated with tape on the floor) and back. Time to walk the distance (s) was recorded. (2) The animal fluency task (words/s) was used for the unitary condition of the verbal task. Concurrent Condition (CC) The subject generated male names while walking the route. The same procedure was repeated with female names. This phase gave two verbal concurrent conditions scores, the difference between the number of words per second generated in the unitary condition and the number of words per second generated in the concurrent conditions (the UC-CC/fluency/animals—male names and UC-CC/fluency/animals— female names scores, in words per second); and two walking concurrent conditions scores, time to walk the distance (the CC/walking/female names and the CC/ walking/male names scores in seconds). Tests (Secondary Variables) Tests of verbal (digit span) and spatial working memory (visuospatial span), together with tests of verbal episodic memory [short-term free recall, short-term cued

247

ROTMAN INSTITUTE ABSTRACTS

recall, long-term free recall, long-term cued recall, etc., of the California Verbal Learning Test (CVLT)], visual perception, visuospatial construction abilities, mental flexibility, and lexical and semantic fluency were administered. Statistical Analyses

An ANOVA was performed between the three groups on the following variables: age, education, and the total score on the MDRS. Since the control group was younger and more educated than the patient groups, age and level of education were computed as covariates in the MANCOVA analyses. The sequential Holm correction procedure was applied to the results of the secondary variables. Student–Newman–Keuls post hoc analyses (level of significance p ⬍ .05) were performed on the significant variables. RESULTS

Subjects Seven subjects with possible DLB, 7 subjects with probable DLB, 22 subjects with probable AD, and 23 healthy controls were recruited. An ANOVA showed no difference between the two patient groups on their age (DLB 76.7 ⫾ 5.6; AD 78.5 ⫾ 8.1 years), level of education (DLB 9.4 ⫾ 4.4; AD ⫽ 10.6 ⫾ 3 years), duration of their illness (DLB 2.7 ⫾ 1.6; AD 2.6 ⫾ 1.8 years). and total score on the MDRS (DLB 100.1 ⫾ 21.1; AD 101 ⫾ 13.9). However, the control subjects were significantly (p ⬍ .05) younger (71.04 ⫾ 9.1 years), were more educated (14.9 ⫾ 3.1 years), and had a superior score on the MDRS (140 ⫾ 2.1) when compared with the patient groups (see Table 1). TABLE 1 Results of the Student–Newman–Keuls Post Hoc Analyses Following the MANCOVA Analyses Tests Dual task UC/digit repetition UC/number boxes CC/digit repetition UC-CC/number boxes Talking/walking test UC/walking (s) UC/fluency/animals (words/s) CC/walking/female names (s) CC/walking/male names (seconds) UC-CC/fluency/animals—male (words/s) UC-CC/fluency/animals—female (words/s) DS CVLT Short-term/free recall Short-term/cued recall Long-term/free recall Long-term/cued recall

DLB Mean (SD)

AD Mean (SD)

HC Mean (SD)

58.4 79.6* 50.7* 34.9

62.1 93.3* 46.5* 54.8*

(20.9) (40.8) (23.6) (43.8)

71.9 157.4 68.6 23.7

(18.1) (18.6) (21.5) (28.4) (2.90) (0.092) (20.44) (16.34) (0.233)

(27.2) (50.1) (23.2) (32.5)

15.60 0.133* 31.66 24.11 ⫺0.039*

(5.46) (0.082 (16.19) (6.05) (0.108)

16.30 0.112* 46.00* 35.70 ⫺0.130*

(5.62) (0.042) (35.19) (22.02) (0.141)

12.00 0.338 22.65 20.60 ⫺0.301

⫺0.030*

(0.115)

⫺0.094

(0.136)

⫺0.175 (0.170)

5.076*¶ (1.115)

5.863* (0.990)

6.783 (0.998)

4.00*¶ 4.90*¶ 4.30*¶ 4.50*¶

0.72* 2.67* 1.33* 2.56*

(3.83) (3.60) (3.49) (3.59)

(1.36) (1.49) (1.45) (1.62)

* Significant difference ( p ⬍ .05) when compared with the HC group. ¶ Significant difference ( p ⬍ .05) when compared with the subjects with AD.

10.71 11.76 11.00 12.05

(1.79) (1.92) (2.49) (2.13)

248

ROTMAN INSTITUTE ABSTRACTS

Other Results Patient groups had inferior performances (p ⬍ .05) compared to those of control subjects on other components of the CVLT, the visuospatial span, other fluency, and mental flexibility tasks. There was no significant difference between the three groups on the visuoperceptual and visuoconstruction tests. Age and education had no effect on the variables. DISCUSSION

The CES of WM The dual task paradigm showed, on the dual task and the talking/walking test, an alteration of processes of the CES of WM in both DLB and AD, when compared to control subjects. However, a differential profile emerged between DLB and AD when their performances were compared to those of control subjects. On the visuospatial part of the dual task, only the performance of subjects with AD was significantly inferior to that of the healthy control subjects, in the concurrent condition. The alteration on the dual task was then more severe in subjects with AD than in subjects with DLB, when compared to the control subjects. In the concurrent conditions of the talking/walking test, the performance of subjects with DLB on the two fluency tasks was inferior to that of control subjects, whereas only the performance of subjects with AD was inferior to that of control subjects on the walking test. Verbal and Spatial WM The DLB subjects had an inferior verbal span compared to that of the AD and control subjects. The patient groups were equally impaired on the spatial working memory (span) tasks. Verbal Episodic Memory The AD subjects had poorer verbal recall when compared with DLB and control subjects. Limitations of the Study The limitations of the study included the following: (1) Small sample size of the DLB group: Subjects with DLB are difficult to recruit due to the severity of their clinical problems and to the fact that paranoid delusions are very common in this population. The effect of the small sample size is to potentially limit statistical power. (2) This study employed multiple comparisons which may lead to false associations; this issue was addressed through the use of the sequential Holm correction procedure for multiple comparisons. (3)The clinical nature of the study (the absence of pathological confirmation for the diagnosis of DLB and AD): The sensitivity values of the CDLB criteria range from 22 to 57% whereas their specificity values range from 90 to 100% (Luis et al., 1999; Holmes et al., 1999). The sensitivity values of the NINCDSADRDA criteria range from 66 to 98% whereas their specificity values range from 40 to 75% (Nagy et al., 1998; Holmes et al., 1999). The current criteria for DLB and AD are therefore useful for recruiting patients for research. Despite the limitations of this study, the results suggest an overall impairment of the CES in subjects with DLB and AD, with a greater vulnerability to the effects of

ROTMAN INSTITUTE ABSTRACTS

249

the dual task paradigm for semantic working memory in DLB than in AD, and a greater vulnerability for spatial working memory in AD than in DLB. The subsystem of verbal working memory is more impaired in DLB than in AD; whereas verbal episodic memory is more impaired in AD than in DLB. REFERENCES Baddeley, A. D., Della Sala, S., Papagno, C., & Spinnler H. (1997). Dual task performance in dysexecutive and nondysexecutive patients with frontal lesion. Neuropsychology, 11, 187–194. Holmes, C., Cairns, N., Lantos, P., & Mann, A. (1999). Validity of current clinical criteria for Alzheimer’s disease, vascular dementia and dementia with Lewy bodies. British Journal of Psychiatry, 174, 45–50. Luis, C. A., Barker, W. W., Gajaraj, K., Harwood, D., Petersen, R., Kashuba, A., Waters, C., Jimison, P., Pearl, G., Petito, C., Dickson, D., & Duara, R. (1999). Sensitivity and specificity of three clinical criteria for dementia with Lewy bodies in an autopsy-verified sample. International Journal of Geriatric Psychiatry, 14, 526–533. Nagy, Z., Esiri, M. M., Hindley, N. J., Joachim, C., Morris, J. H., King, E. M., McDonald, B, Litchfield, S., Barnetson, L., Jobst, K. A., & Smith, A. D. (1998). Accuracy of clinical operational diagnostic criteria for Alzheimer’s disease in relation to different pathological diagnosis protocols. Dementia and Geriatric Cognitive Disorders, 9, 219–226. Simard, M., van Reekum, R., & Cohen, T. (2000). A review of cognitive and behavioral symptoms in dementia with Lewy bodies. The Journal of Neuropsychiatry and Clinical Neurosciences, 12, 425–450. This is doi:10.1006/brcg.2001.1478.

17. Evidence for a Shrinking Span of Personal and Present Existence in Dementia of the Alzheimer’s Type

K. K. Zakzanis and L. Leach Severely impaired episodic memory deprives patients with Alzheimer’s disease (AD) of a sense of personal continuity in their daily lives, yet there are no tests that accurately measure this impairment. Recently, Zakzanis, Leach, and Moscovitch (1999) examined the integrity of memory function in terms of temporal continuity in a way that would engage the patient in everyday behavior, such as informal conversation, but still allow memory function to be quantified. The task allowed the measurement of the duration of continuous, conscious experience of the present and was therefore termed ‘‘span of temporal continuity (STC).’’ Given that we were able to document static and growing STCs, we wanted to know whether our measure could track progressive memory loss. Accordingly, we followed a patient we believed was in the very early stages of AD to measure the change of his STC longitudinally. Along with his STC, we present our neuropsychological and brain imaging findings over the course of the investigation.  2002 Elsevier Science (USA)

Report Severely impaired episodic memory deprives patients with Alzheimer’s disease (AD) of a sense of personal continuity in their daily lives, yet there are no tests that accurately measure this impairment. Several neuropsychological tasks, such as the Brown–Peterson technique (Peterson & Peterson, 1959), have been developed to measure the loss of information from primary memory. These tests measure memory in terms of how much information could be held in working or short-term memory, not memory span in the sense of conscious continuity of existence. Zakzanis, Leach, and Moscovitch (1999) examined the integrity of memory function in terms of temporal continuity in a way that would engage the patient in everyday behavior, such as

250

ROTMAN INSTITUTE ABSTRACTS

informal conversation, but still allow memory function to be quantified. The task allowed the measurement of the duration of continuous, conscious experience of the present. Zakzanis, Leach, and Moscovitch (1999) referred to this measure as the ‘‘span of continuity,’’ or ‘‘personal and present span of existence.’’ In order to estimate span of existence Zakzanis, Leach, and Moscovitch, (1999) first obtained a ‘‘fact’’ from both the patient and family that could be reliably recalled by the patient. That fact concerned either the patient’s occupation or spouse’s name. Having asked the patient a question regarding that fact (e.g., ‘‘What did you do for a living?’’) the patient was engaged in conversation for a few minutes and then the examiner repeated the question 5 min later. If the patient answered the question without any indication of having answered it previously, the patient’s span of temporal continuity was assumed to be less than 5 min. If the patient gave some indication of having being asked the question (e.g., if the patient said, ‘‘You just asked me that’’ or ‘‘I just told you’’ or answered the question with ‘‘Were you not listening?’’ or anything else that indicated examiner oversight or incompetence), span of temporal continuity was assumed to be greater than 5 min. If the span of temporal continuity was less than 5 min, the question was repeated in decrements of 1 min until the person gave some indication that the question was repeated. That is, the question (‘‘What did you do for a living?’’) was asked again 4 min later. If the threshold for recognition had not been achieved after a 1-min delay then the question was repeated in 10-s intervals until the patient indicated that he or she had been asked the question previously. Once this span was approximated (e.g., at 60 s), the question was posed at decrements of 1 s (that is, the question was repeated 69 s later, 68 s later, and so on) until the patient indicated that they had answered the question previously (which should be within 60 and 70 s in keeping with the example). This decrement span was used as our approximation of temporal continuity. To ensure reliability of the measure, however, we also used an increment rather than decrement procedure with the same span intervals only in reverse order. The final span of temporal continuity was determined by averaging the decrement and increment span measures. Using our measure we found a reliable span of temporal continuity across eight weeks in two patients with herpes simplex encephalitis (SEP) and streptococcal meningioencephalitis (SM) who presented to our memory disorder clinic with severe amnesia (span of continuity was found to be 22 s in the patient with SEP and 3 min 34 s in the patient with SM). We also found a ‘‘growing’’ span of temporal continuity in a patient recovering from an anterior communicating artery aneurysm rupture during her 8-week stay in hospital (span of temporal continuity at week 1 was 45 s; span of temporal continuity at week 8 was greater than 10 min). Given that Zakzanis, Leach, and Moscovitch (1999) were able to document static and growing span of temporal continuities, we wanted to know whether our measure could track progressive memory loss. Accordingly, we followed a patient we believed was in the very early stages of AD to measure the change of his span of temporal continuity over 3 years. Along with his span of temporal continuity, we present our neuropsychological and brain imaging findings over the course of the investigation. Case History VB is a 71-year-old, right-handed married man who presented initially with a 6to 8-month history of memory failures. VB is a well-educated, medical specialist who retired in 1996. After his retirement, his wife encouraged him to stay intellectually involved, but she reported that VB now prefers watching television to reading. VB’s wife brought him to the attention of their family physician when she began to notice VB’s frequent repetitive questioning and newly found difficulty of recalling

ROTMAN INSTITUTE ABSTRACTS

251

people’s names. VB was subsequently referred to the Memory Disorders Clinic at Baycrest Centre for Geriatric Care in April of 1997. On their initial visit to our department (also in April of 1997), VB’s wife also reported that her husband’s memory decline had been gradual. He had no history of psychiatric disturbance or alcohol abuse. His mother and brother, both deceased, had coronary heart disease, and his father died of a brain tumor. Neuropsychological Testing April 1997. VB was alert, obliging, and very pleasant. He understood the purpose of the assessment and cooperated fully. He exhibited no unusual difficulty following instructions. His speech was generally fluent and articulate with occasional circumlocution and word-finding pauses. His voice was normal in prosody with reduced volume and rate. We noted at the time of the first assessment that VB repeated the same comment or story within a 10- to 15-min span. VB admitted to having memory problems, and although he did not appear unduly distressed by this knowledge, he did not try to minimize or excuse his poor memory. There were no behavioral signs of a depressed mood. He was oriented to time and place and was fully independent in activities of daily living (e.g., came to his appointment independently and on time). The first neuropsychological examination found evidence of a very mild impairment in cognitive functioning (see Table 1). Although VB’s general intellectual functioning remained in the high average to superior range, mild impairments were noted on tests of episodic memory for visually as well as verbally presented information. In addition, a moderate impairment was noted on a test of naming to visual confrontation. Word list generation was within normal limits for first letter cues and semantic category. Abstract categorical reasoning was high average. Drawing and visualconstructive ability were unimpaired and rated as high average for his age. Brain imaging using single photon emission computed tomography (SPECT) revealed no focal or diffuse perfusion abnormalities. An unenhanced head computed tomography (CT) was interpreted as being normal. March 1998. On this occasion VB admitted that he was having memory failures although he was unable to recount specific examples of such failures. His wife reported that VB’s memory had indeed worsened in keeping with her observations of repeated questioning. She noted, however, that there were no transient episodes of extreme confusion, that there was an absence of hallucinations and delusions, and that progression of memory loss remained gradual. The second neuropsychological examination (see Table 1) revealed a noticeable increase in the total number of intrusion errors made on the California Verbal Learning Test (CVLT). VB’s output on tests of verbal fluency to both phonemic and semantic cues also had declined. A brain imaging examination at this time using SPECT now revealed mild global hypoperfusion. May 2000. At the time of the third assessment, VB’s wife reported that her husband’s memory had declined precipitously. She reported that he frequently retold the same stories and repeated questions after a matter of minutes. The results of the third assessment are given in Table 1. His scores on the DRS subtests of Attention, Initiation/Perseveration, and Memory had declined. He was unable to state the correct weekday and date. On the CVLT, there was a marked decline on measures of list learning as well as on delayed recall and recognition measures. He could recall no more than four items from the CVLT during learning trials; he could not recall any items after a short delay and only a single item after a 20-min delay. Furthermore, the number of intrusions made during recall was extraordinarily high. His recognition

252

ROTMAN INSTITUTE ABSTRACTS

TABLE 1 Neuropsychological Status Neuropsychological measure

April 1997 raw score

March 1998 raw score

Dementia rating scale Total 138 135 Attention 37 37 Initiation/perseveration 37 37 Construction 6 6 Conceptualization 39 39 Memory 19 16 California Verbal Learning Test List A Trials 1–5 Total 29 26 List A Trial 1 2 4 List A Trial 5 8 4 List B 4 2 List A short-delay free recall 6 4 List A short-delay cued 6 6 recall List A long-delay free recall 6 7 List A long-delay cued recall 5 7 Discriminability 86% 82% Intrusions (free and cued 16 34 recall total) Rey–Osterieth Complex Figure Copy 32 32 Immediate recall 4.5 4.5 Delayed recall 2.5 2.5 Wechsler Adult Intelligence Scale–Revised (Age Corrected Scales Scores) Digit span 19 19 Vocabulary 13 16 Block design 11 12 Boston Naming Test 40 40 Verbal fluency F, A, S 61 46 Animals 17 12 a 4.5 min Span of Temporal Continuity SPECT findings Within normal mild global limits hypoperfusion

May 2000 raw score 127 34 34 6 39 14 17 3 4 2 0 3 1 1 64% 59

30 4 0 19 16 11 40 38 14 50 s bilateral frontal hypoperfusion

a Not formally assessed at this time. An estimate span of temporal continuity of 10–15 min can be approximated based on clinical notes made at the initial assessment where the examiner noted the repetition of stories after about 10–15 min without conscious recollection of having told the story.

discriminability approached chance. With respect to visual memory, his immediate reproduction from memory of the Rey–Osterrieth complex figure was deemed severely impaired and after a 40-min delay he was unable to recall copying the figure. Although his total score on the Boston Naming Test remained unchanged, he made many more phonemic and semantic paraphasic errors and circumlocutory responses. In terms of verbal fluency, his ability to generate words using a phonemic cue had dramatically declined as opposed to his milder decline in semantic fluency. A SPECT investigation revealed evidence of a mild bilateral frontal hypoperfusion. Span of Continuity Patient VB was seen on three occasions in our memory disorders clinic between April 1997 and May 2000. In addition to completing a number of neuropsychological

ROTMAN INSTITUTE ABSTRACTS

253

measures on each occasion (see Table 1), we estimated his span of temporal continuity in March 1998 and May 2000 (see Table 1). Span of temporal continuity was estimated in keeping with the description noted above and in Zakzanis, Leach, and Moscovitch (1999). In April 1997, his span of temporal continuity was not formally assessed. An estimate span of temporal continuity of 10–15 was approximated based on clinical notes made at the initial assessment where the examiner noted the repetition of stories after about 10–15 min without conscious recollection of having told the story. REFERENCES Peterson, L. R., & Peterson, M. J. (1959). Short-term retention of individual verbal items. Journal of Experimental Psychology, 58, 193–198. Zakzanis, K. K., Leach, L., & Moscovitch, M. (1999). Span of temporal continuity as a measure of personal and present existence. Journal of the International Neuropsychological Society, 5, 85–86. This is doi:10.1006/brcg.2001.1479.

18. Noncognitive Symptoms in Alzheimer Disease and Caregivers Distress in Chile

G. Rohde, P. Quiroga, M. Fasce, and F. Fasce This study describes the profile of noncognitive symptoms in Chilean AD patients and its effects on the caregiver’s mental health. In a sample of 26 urban dwelling AD patients, 21 of them women, age range 63–90 years, diagnosed using NINCDS-ADRDA, the intensity and frequency of noncognitive symptoms and caregivers distress was assessed using NPI. Seventeen caregivers were first degree relatives. The most prevalent symptoms were apathy, irritability, and anxiety. The least were euphoria, hallucinations, and disinhibition. The most stressful conditions were disinhibition and agitation; the best well-tolerated were euphoria and anxiety. Caregivers responded most frequently with feelings of concern and depression. Few of them reported to be ashamed. These results may reflect cultural differences and represents the first description of this reality in Chile.  2002 Elsevier Science (USA)

Report Noncognitive symptoms in AD patients are among the most common and disturbing conditions in dementia (Santulli, 2000). It has been demonstrated that these symptoms, more than the cognitive ones, lead to early institutionalization and cause serious caregivers distress, generating mental and even physical diseases in them (Wilz, Adler, Gunzelmann, & Brahler, 1999). The impact of noncognitive symptoms in the burden of caregivers has been associated with different variables among which gender, social and cultural background, and availability of support systems seem to be important (Donaldson & Burns, 1999). The aim of this study is to describe the profile of noncognitive symptoms in Chilean patients and its effects on caregiver’s mental health. Methodology In a sample of consecutively assessed 26 urban dwelling AD patients diagnosed using NINCDS-ADRDA criteria, the intensity and frequency of noncognitive symptoms and caregiver distress, referred to the month prior to the interview, were assessed using the neuropsychiatric inventory (NPI) by trained neurologists and psychiatrists.

254

ROTMAN INSTITUTE ABSTRACTS

Severity of the disease was rated according to Clinical Dementia Rating Scale (CDR). Descriptive data analysis was performed using the EPI-Info statistical computer package. The sample included 21 women. The age range was 63–90 years. According to severity 6 patients had mild AD, 19 had moderate AD, and 1 had severe AD. Twentyfive of the caregivers were female, 17 were first degree relatives, 5 were professional, and 5 were other relatives (nieces and daughters-in-law). Results The most common reported noncognitive symptoms were apathy and indifference present in 19 cases (73%) followed by anxiety and irritability in 15 subjects each (58%) and depression and dysphoria in 14 subjects (54%). The least informed symptoms were euphoria with 3 cases (12%), hallucinations in 7 subjects (27%), and disinhibition in 8 cases (31%). Delirium and aberrant motor activity were reported in 14 (54%) and 13 cases (50%), respectively. In relation to frequency of appearance, the symptoms followed a different pattern: aberrant motor activity occurred as frequently as one or more times a day in 54% of the cases (very frequently) and in the rest of the cases was several times a week (frequently). Apathy and indifference were the following most frequent symptoms when present with 53% in the very frequently range and 21% occurring several times a week. Depression, anxiety, and euphoria were occasional (present less than once a week) in at least 33% of the cases. Disinhibition was the most occasional symptom occurring in 50% of the subjects less than once a week. According to severity, only apathy was rated as severe in 37% of the subjects and euphoria was moderate in 67% of the cases All the other symptoms were reported as mild: depression and dysphoria 86%,disinhibition 75%, anxiety and irritability 73%, agitation in 64%, hallucinations in 57%, delusions 54%, and finally aberrant motor activity 46%. The most distressful conditions for caregivers were disinhibition generating moderate to severe distress, with a mean of 3.87 in the NPI scale, followed by agitation with 3.45 and by aberrant motor activity causing mild to moderate (2.76). The most well tolerated were euphoria with none distress, anxiety with mild to moderate (1.46), and hallucinations with moderate discomfort (2). The most important stress for caregivers was the feeling of being worried by accidents occurred to the patients reported as moderate with a mean of 2.08 in the NPI scale. Second, they responded to the burden of caring with feelings of depression (2) and of not foreseeing solution to the problem (1.73). The least felt stressing conditions rated as none or scarce were being forbidden from social life (0), being ashamed by the patient (0.12), and the inability to manage the situation (0.85). Conclusions As in other reported results in different populations the presence of noncognitive symptoms is high in our patients (Chen, Borson, & Scanlan, 2000). According to our results it seems that for our caregivers the most stressing factor is frequency more than severity of the symptoms, except for disinhibited behavior, which could be explained for the social impact of this symptom. Chilean caregivers, being mostly women and daughters, tend to assume their burden with a fatalistic determinism which could be due either to the fact that they are used by social requirements to assume this kind of role as part of their gender and

ROTMAN INSTITUTE ABSTRACTS

255

familial role or to the fact that in Chile there are scarce social support systems for these patients. These results may reflect differences due to the social structure of Latin American populations and represent the first description of this increasing problem in Chile. REFERENCES Chen, J. C., Borson, S., & Scanlan, J. M. (2000). Stage-specific prevalence of behavioral symptoms in Alzheimer’s disease in a multi-ethnic community sample. American Journal of Geriatric Psychiatry, 8, 123–133. Donaldson, C., & Burns, A. (1999). Burden of Alzheimer’s disease: Helping the patient and the caregiver. Journal of Geriatric Psychiatry and Neurology, 12, 21–28 Santulli, R. (2000). Treating of mood and behavioral symptoms of Alzheimer disease in bridging research and care: Alzheimer’s Disease and Related Disorders Association In World Alzheimer Congress 2000 (pp. 55–63). Washington, DC. Wilz, G., Adler, C., Gunzelmann, T., & Brahler, E. (1999). Effects of chronic stress on physical or psychological health—An analysis of family caregivers of demented patients. Zeitung fur Gerontologie und Geriatrie, 32, 255–265. This is doi:10.1006/brcg.2001.1480.