Vascular dementia (VaD) and Alzheimer disease (AD) are

Preclinical Vascular Cognitive Impairment and Alzheimer Disease Neuropsychological Test Performance 5 Years Before Diagnosis Janet L. Ingles, PhD; Denise C. Boulton, MSc; John D. Fisk, PhD; Kenneth Rockwood, MD Background and Purpose Neuropsychological changes that precede a diagnosis of vascular cognitive impairment (VCI) and the differences between preclinical VCI and Alzheimer disease (AD) are not well understood. We compared the neuropsychological performances of people with incident VCI, incident AD, and no cognitive impairment (NCI) 5 years before their clinical diagnoses. Methods The Canadian Study of Health and Aging is a prospective, cohort study of 10 263 randomly selected persons age 65 years or older. We studied 332 individuals who had completed a battery of neuropsychological tests and were diagnosed with NCI at baseline. After 5 years, 41 were diagnosed with VCI, 25 with AD, and 266 with NCI. Results At baseline, the incident-vci group performed worse on a wide range of neuropsychological tests compared with the NCI group. A test of abstract reasoning was selectively low in the incident-vci group, relative to both the incident-ad and NCI groups. The incident-ad group performed worse at baseline on memory tests compared with incident-vci and NCI groups. Conclusions This study suggests a preclinical phase may exist in VCI that differs from that in AD. Neuropsychological measures may aid the design of preventive strategies for VCI. (Stroke. 2007;38:1148-1153.) Key Words: Alzheimer disease cerebrovascular disorders dementia neuropsychology Vascular dementia (VaD) and Alzheimer disease (AD) are the 2 most prevalent forms of dementia. 1 Knowing prodromal features for each is important for the design of preventive strategies. Neuropsychological deficits, particularly memory problems, precede, often by many years, the diagnosis of AD. 2 7 However, the cognitive changes that may precede VaD and the differences between preclinical VaD and AD have been less well studied. Previously, we found that among elderly people who had vascular cognitive impairment (VCI) who did not meet dementia criteria (ie, vascular cognitive impairment, no dementia [VCI-ND]), low scores on tests of memory and semantic category fluency predicted incident cases of dementia 5 years later. 8 Most of these incident cases had VaD. Other groups have also shown that reductions in memory, as part of a more global pattern of impairment, precede diagnoses of both VaD and AD by up to 4 years. 9 13 Although few differences in the preclinical cognitive profiles of these 2 conditions have been observed, 9,10,12,13 Jones et al 14 found that semantic category fluency was more impaired in preclinical AD, whereas deficits in letter category fluency were present in both preclinical AD and VaD. Still, findings of preclinical memory deficits common to both VaD and AD might simply reflect that dementia diagnoses traditionally require deficits in memory. People with preclinical cognitive impairment due to cerebrovascular disease that progresses, eg, in the domain of executive functioning, and that causes significant functional impairment even without the memory loss typical of AD have not been included in these investigations. The term VCI incorporates VaD and AD with a vascular component (referred to as mixed AD/VCI), as well as VCI-ND. 15 17 The advantage of VCI as an outcome not restricted to the requirements of VaD is that it allows for inclusion of people who develop cognitive impairment that does not meet the criteria for dementia and who nevertheless have increased risks of institutionalization and death. 18 In this secondary analysis of data from the Canadian Study of Health and Aging (CSHA), we compared the neuropsychological performances of people with incident VCI, incident AD, and no cognitive impairment (NCI) 5 years before their clinical diagnoses. Subjects and Methods The CSHA is a prospective, cohort study of 10 263 randomly selected, community-dwelling (n 9008) and institution-dwelling (n 1255) persons age 65 years and older. 1,19 In 1991 to 1992 (CSHA-1), all institutionalized and community participants who scored 78 of 100 on the Modified Mini-Mental State Examination Received October 11, 2006; accepted October 31, 2006. From the School of Human Communication Disorders (J.L.I., D.C.B.), the Department of Psychology (J.L.I., J.D.F.), the Department of Medicine (J.D.F., K.R.), and the Department of Psychiatry (J.D.F.), Dalhousie University, Halifax, Canada. Correspondence to Dr Janet Ingles, School of Human Communication Disorders, Dalhousie University, 5599 Fenwick St, Halifax, Nova Scotia, B3H 1R2, Canada. E-mail janet.ingles@dal.ca 2007 American Heart Association, Inc. Stroke is available at http://www.strokeaha.org DOI: 10.1161/01.STR.0000259716.04739.60 1148

Ingles et al Neuropsychological Predictors of Dementia 1149 (3MS) 20 or who scored 78 of 100 and were part of a randomly selected comparison group received a comprehensive clinical assessment (n 2914). The assessments involved evaluations of participants medical and neurological status, informants accounts of functional ability, laboratory blood work, and neuropsychological examination. The neuropsychological component of the assessment was completed by 1879 participants (789 were deemed untestable because of 3MS scores 50, and 246 refused or did not complete testing). The neuropsychological test battery assessed constructs described in the DSM-III-R criteria for dementia, as detailed elsewhere. 21 Short- and long-term memory was assessed with Buschke s Cued Recall Test, 22,23 the Rey Auditory Verbal Learning Test, 24 26 the Benton Visual Retention Test, 27 and the Information and Digit Span subtests of the Wechsler Memory Scale. 28 Abstract thinking was assessed with the Wechsler Adult Intelligence Scale, Revised (WAIS-R) Similarities subtest. 29,30 Judgment was assessed with the WAIS-R Comprehension subtest. 29,30 Other higher cortical functions were assessed with the Token Test, 31 the Controlled Oral Word Association Test (ie, letter category fluency), 32 Animal Naming (ie, semantic category fluency), 33 visual identification of items from Buschke s Cued Recall Test, 22,23 the WAIS-R Digit Symbol subtest, 29 and the WAIS-R Block Design subtest. 29,30 In 1995 to 1996 (CSHA-2), follow-up data were collected and surviving participants were reassessed. The clinical assessment procedure was the same as in CSHA-1. All participants provided written, informed consent, and the protocol was approved at each participating institution. Separate approval for secondary analyses was also obtained from the research ethics committee of Capital Health, Halifax, Nova Scotia. Participants in both CSHA-1 and CSHA-2 were diagnosed by the same criteria at a consensus conference of the examining physician, neuropsychologist, and nurse. From a DSM-III-R dementia diagnosis, etiologic diagnoses were based on NINCDS-ADRDA criteria for AD 34 and ICD-10 criteria for VaD. 35 A diagnosis of cognitive impairment but no dementia (CIND) was made if they had cognitive impairment that did not meet the DSM-III-R criteria for dementia (ie, they did not have both memory impairment and other cognitive impairment that caused functional deficits). 36 Participants who did not meet the criteria for dementia or CIND were classified as having NCI. In CSHA-2, VCI was diagnosed by a staged process. A checklist that comprised every criterion from each set of VaD criteria then in use was completed. These criteria did not include the presence of vascular risk factors but required clinical signs of vascular involvement (eg, onset after stroke, stepwise progression, focal cognitive deficits, focal neurologic findings, or neuroimaging data when available). 37 From this information, VCI was diagnosed and subtyped (ie, VaD, VCI-ND, and mixed AD/VCI). The construct and predictive validity of this subtype classification scheme was subsequently demonstrated in a clinic-based cohort study with the use of clinical features, neuroimaging, and outcome data (ie, time to death and institutionalization). 38 Of the 1879 participants who completed the neuropsychological assessment at CSHA-1, 786 were classified as NCI. Four hundred fifty-five of these individuals were alive and participated in the CSHA-2 assessment process. Of these 455 individuals, 266 were diagnosed with NCI, 25 with probable AD, and 41 with VCI (13 with VaD, 18 with VCI-ND, and 10 with mixed AD/VCI). An additional 123 subjects had diagnoses that were not of interest in the current study. Thus, a total of 332 participants constituted our sample. We examined differences in baseline neuropsychological test scores between the incident-vci, incident-ad, and NCI groups by ANCOVAs with age, years of education, and 3MS test scores (as a baseline measure of severity of cognitive impairment) as covariates. To facilitate comparisons between the neuropsychological variables, raw scores were transformed into Z scores. Bonferroni post hoc tests adjusted for multiple between-group comparisons were used to compare the specific groups. We examined the ability of the neuropsychological variables to predict outcomes by conducting 3 sets of logistic-regression analyses (ie, incident VCI vs NCI; incident TABLE 1. Demographic Characteristics of the 3 Groups at CSHA-1 Incident VCI (n 41) Incident AD (n 25) NCI (n 266) Mean age, y (SD) 79.2 (6.4) 81.8 (5.0) 76.3 (6.1) % Female 56.0 76.0 61.3 Mean years of education (SD) 9.2 (3.7) 8.2 (3.1) 10.4 (4.1) % Institutionalized 17.1 16.0 13.5 Mean 3MS Score (SD) 86.4 (7.3) 80.3 (10.6) 89.2 (7.9) AD vs NCI; incident VCI vs incident AD), with age, sex, years of education, and baseline 3MS test scores as covariates. Results The NCI group was younger than the incident-vci (P 0.02) and incident-ad groups (P 0.001), although the latter 2 groups did not differ significantly in age (P 0.2, Table 1). The NCI group had more years of education than the incident-ad group (P 0.03) but did not differ from the incident-vci group (P 0.1); the incident-vci and AD groups did not differ in education (P 0.5). The NCI and incident-vci groups had higher baseline 3MS scores than the incident-ad group (P 0.01 for both comparisons), although the scores of the NCI and incident-vci groups did not differ significantly (P 0.08). There were no significant group differences in sex or residence distributions (P 0.1 for all comparisons). At baseline, those with incident VCI performed worse than those with incident AD and NCI on the WAIS-R Similarities subtest (Table 2). The incident-vci group also performed worse than the NCI group on the semantic category fluency test and the WAIS-R Digit Symbol and Block Design subtests. The incident-ad group performed worse than the incident-vci group on the Cued Recall measure of Buschke s Cued Recall Test (ie, total words recalled on cued recall trials/total number of cues provided) and worse than both the incident-vci and NCI groups on the Wechsler Memory Scale Information subtest. Both the incident-ad and incident-vci groups performed worse than the NCI group on the Free Recall measure of Buschke s Cued Recall Test but did not differ significantly from each other. No significant baseline differences were found between the 3 groups on the remaining tests. The baseline analyses were repeated after excluding those with a diagnosis of mixed AD/VCI (n 10) from the incident- VCI group. The pattern of results was the same, with the exception that the NCI versus VCI comparison was statistically significant for letter category fluency (P 0.05) but did not reach significance for the WAIS-R Block Design subtest (P 0.08). Table 3 presents the neuropsychological tests that contributed independently to the logistic-regression analyses. For incident VCI versus NCI, low baseline scores on the Free Recall measure of Buschke s Cued Recall Test; the WAIS-R Similarities, Comprehension, Digit Symbol, and Block Design subtests; and the letter and semantic fluency tests were associated with incident VCI. For incident AD versus NCI, low baseline scores on the Free and Cued Recall measures of

1150 Stroke April 2007 TABLE 2. CSHA-1 Neuropsychological Test Results in Incident-VCI, Incident-AD, and NCI Groups Test Incident VCI Incident AD NCI Post Hoc Group Comparisons (P Value) Buschke s Cued Recall Test: 33.49 (0.8) 30.92 (1.1) 38.10 (0.3) NCI VCI ( 0.001) Free Recall Total (36) NCI AD ( 0.001) VCI AD (0.17) Buschke s Cued Recall Test: 0.99 (0.01) 0.94 (0.02) 0.98 (0.005) NCI VCI (1.0) Cued Recall Measure (1.00) NCI AD (0.08) VCI AD (0.04) Rey Auditory Verbal Learning Test: 37.17 (1.4) 35.62 (1.9) 39.38 (0.6) NS Total Trials 1 5 (75) Benton Visual Retention Test (16) 11.44 (0.3) 11.95 (0.4) 11.79 (0.1) NS Wechsler Memory Scale: Information Subtest (6) 5.24 (0.1) 4.69 (0.2) 5.36 (0.1) NCI VCI (1.0) NCI AD (0.001) VCI AD (0.03) Wechsler Memory Scale: Digit Span Subtest (12) 5.68 (0.2) 5.99 (0.2) 5.70 (0.1) NS WAIS-R Similarities (14) 5.75 (0.5) 8.03 (0.7) 7.57 (0.2) NCI VCI (0.02) NCI AD (1.0) AD VCI (0.003) WAIS-R Comprehension (16) 8.69 (0.4) 8.78 (0.6) 9.5 (0.2) NS Token Test (44) 38.68 (0.9) 39.16 (1.2) 37.70 (0.4) NS Letter Category Fluency 24.72 (1.5) 25.74 (2.0) 28.37 (0.6) NS Semantic Category Fluency 12.85 (0.6) 14.25 (0.8) 14.89 (0.2) NCI VCI (0.004) NCI AD (1.0) VCI AD (0.44) Buschke s Cued Recall Test: Visual Identification (12) 11.77 (0.1) 11.65 (0.1) 11.82 (0.03) NS WAIS-R Digit Symbol (93) 26.67 (1.3) 28.08 (1.7) 30.95 (0.5) NCI VCI (0.007) NCI AD (0.36) VCI AD (1.0) WAIS-R Block Design (29) 9.19 (0.7) 10.27 (0.9) 11.29 (0.3) NCI VCI (0.01) NCI AD (0.84) VCI AD (0.95) Values are means adjusted for baseline age, education, and 3MS scores (SE). Maximum scores are given in brackets after the name of each test. P values were adjusted by Bonferroni correction. Buschke s Cued Recall Test, the Rey Auditory Verbal Learning Test, the Wechsler Memory Scale Information subtest, letter category fluency, and the WAIS-R Digit Symbol subtest were associated with incident AD. For incident VCI versus incident AD, there was a trend toward an association between low baseline scores on the WAIS-R Similarities subtest and incident VCI, although this did not reach statistical significance (P 0.058). Discussion In this population-based study, individuals with NCI who progressed to VCI after 5 years performed worse at baseline on a number of neuropsychological tests (ie, Free Recall measure of Buschke s Cued Recall Test; WAIS-R Similarities, Block Design, and Digit Symbol subtests; and semantic category fluency) relative to those who remained cognitively unimpaired. Each of these neuropsychological variables, as well as the WAIS-R Comprehension subtest and letter category fluency, also contributed to a predictive model for VCI. Those who progressed to AD performed worse at baseline on tests of memory (ie, Free Recall measure of Buschke s Cued Recall Test and Wechsler Memory Scale Information subtest) relative to those who remained cognitively unimpaired. Additional memory tests (ie, Cued Recall measure of Buschke s Cued Recall Test and Rey Auditory Verbal Learning Test) and tests of executive abilities (ie, letter category fluency and WAIS-R Digit Symbol subtest) contributed to the predictive model for AD. At baseline, tests of memory (ie, Cued Recall measure of Buschke s Cued Recall Test and Wechsler Memory Scale Information subtest) were selectively low in the incident-ad group relative to the incident-vci group. Only a test of abstract reasoning, the WAIS-R Similarities subtest, was selectively low at baseline in the incident-vci group relative to the incident-ad group. This test was also the only 1 included in a predictive model distinguishing incident-vci and incident-ad groups. The results suggest the existence of a preclinical phase of VCI during which widespread but subtle reductions in cognitive ability may be detected some years before a clinical diagnosis can be rendered. Our findings extend those of

Ingles et al Neuropsychological Predictors of Dementia 1151 TABLE 3. Results of Logistic-Regression Analyses Predicting Diagnostic Category at CSHA-2 Odds Ratio 95% CI P Incident VCI vs NCI* Buschke s Cued Recall Test: Free Recall Total 3.11 1.99 4.86 0.001 WAIS-R Similarities 2.04 1.32 3.14 0.001 WAIS-R Comprehension 1.55 1.01 2.39 0.05 Letter Category Fluency 1.84 1.11 3.05 0.02 Semantic Category Fluency 2.14 1.34 3.42 0.001 WAIS-R Digit Symbol 2.54 1.41 4.59 0.002 WAIS-R Block Design 1.92 1.26 2.94 0.003 Incident AD vs NCI Buschke s Cued Recall Test: Free Recall Total 3.24 1.81 5.81 0.001 Buschke s Cued Recall Test: Cued Recall Measure 1.33 1.00 1.78 0.05 Rey Auditory Verbal Learning Test: Total Trials 1 5 3.02 1.41 6.50 0.005 Wechsler Memory Scale: Information Subtest 1.67 1.10 2.53 0.02 Letter Category Fluency 2.20 1.03 4.70 0.05 WAIS-R Digit Symbol 2.85 1.21 6.71 0.02 Incident VCI vs Incident AD WAIS-R Similarities 2.41 0.97 5.98 0.06 All analyses were adjusted for baseline age, sex, years of education, and 3MS scores. *Low test scores were associated with incident VCI. Low test scores were associated with incident AD. Low test scores were associated with incident VCI. previous studies that have shown a relatively global pattern of cognitive decline in preclinical VaD 9 13 by demonstrating similarly widespread lowering of cognitive test performance in preclinical VCI, a diagnosis that does not require clinical evidence of memory deficits. Using this broader diagnosis may have facilitated our identification of different patterns of neuropsychological performance in preclinical VCI and AD. In those studies that have found preclinical cognitive performance to be similar in VaD and AD, it must be acknowledged that the diagnostic outcomes are also similar (ie, both VaD and AD require memory deficits). 9,10,12,13 Although we did not replicate the findings of Jones et al 14 of poorer semantic category fluency in preclinical AD relative to preclinical VaD, we note that VCI is a heterogeneous condition and that different forms may well present with different prodromal cognitive profiles. For example, people with VaD who have significant subcortical pathology may be more impaired on tests of executive functioning and visual construction, whereas those with little subcortical pathology may be more impaired on tests of language and memory. 39 Our findings are also consistent with the frequently reported executive dysfunction (including reduced abstract reasoning) of VaD compared with the memory impairment that characterizes AD. 40 43 We show that this differential pattern of cognitive performance may be evident in a population-based sample even 5 years before a clinical diagnosis is evident. Therefore, measuring such neuropsychological variables might aid the design of preventive strategies for both VCI and AD through appropriate selection of people at risk in the population. Neuropsychological test performance in relation to VCI has been investigated in the CSHA before. The importance of memory and semantic fluency test performance for predicting risk of progression to dementia (VaD and AD combined) in patients with VCI-ND has been demonstrated. 8 Similarly, Klages et al 44 evaluated apolipoprotein E genotype, memory test performance, and other clinical variables in a subsample (n 223) of the 332 studied here. They found that apolipoprotein E status was no longer predictive of VCI or AD once memory test performance was taken into account and that free recall was an important predictor of both VCI and AD, as in the current results. Owing to their more limited sample size, however, they did not attempt to examine potential differences in neuropsychological risk factors for VCI and AD. All of our participants were classified by clinical consensus at baseline as having NCI. Although people who later progressed to VCI and AD performed worse on a number of neuropsychological tests relative to those who maintained their NCI diagnosis, these reductions were subtle and fell 1SD below the performance of the NCI group. We have therefore referred to our participants who later progressed to VCI and AD as being in preclinical phases at baseline. It could be argued, however, that our terminology is inconsistent with a prime tenet of the VCI concept: ie, that VCI includes people with wide-ranging levels of cognitive ability, from the brain-at-risk stage in which patients have no frank cognitive deficit, to severe VaD. 45 Thus, the relatively lower levels of neuropsychological test performance in our incident-vci group may indicate that these participants in fact had VCI at baseline, albeit at an early stage. Establishing when the preclinical phase of a disorder ends and the actual disorder begins is a difficult task. This is particularly true for

1152 Stroke April 2007 neuropsychological tests in which a wide range of scores can be generated by individuals who are, and will remain, cognitively healthy. Identification of cognitive measures that are specific and sensitive to the progression of deficits in VCI will be a challenge for future research. Our results must be interpreted with caution, given the relatively small number of cases in the incident-vci and -AD groups. Because of sample size limitations, neuropsychological predictors of the different VCI subgroups (ie, VaD, VCI-ND, and mixed AD/VCI) could not be evaluated. Additional differences between the predictors of VCI and AD might also have been detected with a larger sample. Limited sample size also prevented us from using the CSHA-2 neuropsychological data as a baseline in an attempt to replicate our findings by using the third phase of the CSHA (conducted in 2001 to 2002) to identify incident cases (14 subjects with incident AD, 18 with incident VCI, and 29 with NCI). Similarly, we were unable to examine neuropsychological predictors 10 years before diagnosis by using CSHA-1 as the baseline and CSHA-3 as the follow-up (6 subjects with incident AD, 8 with incident VCI, and 11 with NCI). The lack of neuroimaging data is an additional weakness of our study and prevented further subtyping of the VCI cases. Different subtypes of VCI may well present with different clinical courses. For example, subcortical small-vessel dementia appears to have an insidious onset and an extended prodromal period, in contrast to the abrupt onset of the multi-infarct, strategic infarct, and intracranial hemorrhage subtypes of VCI. 10 The clinical course of mixed AD/VCI may be more similar to that of AD and may be characterized by progressive memory impairment in addition to executive dysfunction. 45 Although the pattern of baseline neuropsychological results changed little when the mixed AD/VCI cases were excluded from our incident-vci group, as noted earlier, sample size limitations meant that we were unable to examine VCI subgroups in detail. Mixed AD/VCI cases in which the vascular component was present only on neuroimaging would also have been missed in our study. However, the effect of this is conservative: ie, it would have made our AD cases look more like VCI cases, a factor that might account for some of the negative findings. On the other hand, this should also make the finding of poor abstract reasoning in preclinical VCI robust. We note, too, that although the CSHA battery included tests of abstract thinking, judgment, verbal fluency, and psychomotor speed, other classic measures of executive functions (eg, Trail Making, Card Sorting), were not included. Given the predominance of executive dysfunction in VCI, additional differences between the neuropsychological profiles of preclinical VCI and AD may have been detected with better measurement of this domain. In addition, the 3MS screening tool is a relatively weak measure of executive function, and some incident VCI cases may have been excluded from the baseline neuropsychological assessment by virtue of having been screened negative. Comprehensive assessment of executive function in preclinical VCI and AD will be particularly important in future population-based and clinic-based studies because of the strong association between executive dysfunction and functional outcomes in dementia. 46 50 Replication of our findings, especially in studies with routine baseline and follow-up that combines neuropsychological and neuroimaging studies, could help identify groups of individuals to whom preventive maneuvers might be targeted. Acknowledgments The data reported in this article were collected as part of the Canadian Study of Health and Aging. The study was coordinated through the University of Ottawa and the federal government s Laboratory Centre for Disease Control. K.R. and J.F. were funded by an operating grant from the Canadian Institutes of Health Research. Source of Funding This was funded by the Seniors Independence Research Program, administered by the National Health Research and Development Program of Health and Welfare Canada (project No. 6606-3954- MC [S]). None. Disclosures References 1. Canadian Study of Health and Aging Working Group. Canadian study of health and aging: study methods and prevalence of dementia. Can Med Assoc J. 1994;150:899 913. 2. Backman L, Jones S, Berger AK, Laukka EJ, Small BJ. Cognitive impairment in preclinical Alzheimer s disease: a meta-analysis. Neuropsychology. 2005;19:520 531. 3. Tierney MC, Yao C, Kiss A, McDowell I. Neuropsychological tests accurately predict incident Alzheimer disease after 5 and 10 years. Neurology. 2005;64:1853 1859. 4. Bozoki A, Giordani, B, Heidebrink, JL, Berent, S. Foster, NL. Mild cognitive impairments predict dementia in nondemented elderly patients with memory loss. Arch Neurol. 2001;58:411 416. 5. Galvin JE, Powlishta KK, Wilkins K, McKeel DW, Xiong C, Grant E, Storandt M, Morris JC. Predictors of preclinical Alzheimer disease and dementia: a clinicopathologic study. Arch Neurol. 2005;62:758 765. 6. Amieva H, Jacqmin-Gadda H, Orgogozo JM, Le Carret N, Helmer C, Letenneur L, Barberger-Gateau P, Fabrigoule C, Dartigues JF. The 9 year cognitive decline before dementia of the Alzheimer type: a prospective population-based study. Brain. 2005;128:1093 1101. 7. Saxton J, Lopez OL, Ratcliff G, Dulberg C, Fried LP, Carlson MC, Newman AB, Kuller L. Preclinical Alzheimer disease: neuropsychological test performance 1.5 to 8 years prior to onset. Neurology. 2004; 63:2341 2347. 8. Ingles JL, Wentzel C, Fisk JD, Rockwood K. Neuropsychological predictors of incident dementia in patients with vascular cognitive impairment, without dementia. Stroke. 2002;33:1999 2002. 9. Sacuiu S, Sjogren M, Johansson B, Gustafson D, Skoog I. Prodromal cognitive signs of dementia in 85-year-olds using four sources of information. Neurology. 2005;65:1894 1900. 10. Meyer JS, Xu G, Thornby J, Chowdhury MH, Quach M. Is mild cognitive impairment prodromal for vascular dementia like Alzheimer s disease? Stroke. 2002;33:1981 1985. 11. Laukka EJ, Jones S, Fratiglioni L, Backman L. Cognitive functioning in preclinical vascular dementia: a 6-year follow-up. Stroke. 2004;35: 1805 1809. 12. Laukka EJ, Jones S, Small BJ, Fratiglioni L, Backman L. Similar patterns of cognitive deficits in the preclinical phases of vascular dementia and Alzheimer s disease. J Int Neuropsychol Soc. 2004;10:382 391. 13. Jones S, Jonsson Laukka E, Small BJ, Fratiglioni L, Backman L. A preclinical phase in vascular dementia: cognitive impairment three years before diagnosis. Dement Geriatr Cogn Disord. 2004;18:233 239. 14. Jones S, Laukka EJ, Backman L. Differential verbal fluency deficits in the preclinical stages of Alzheimer s disease and vascular dementia. Cortex. 2006;42:347 355. 15. O Brien JT, Erkinjuntti T, Reisberg B, Roman G, Sawada T, Pantoni L, Bowler JV, Ballard C, DeCarli C, Gorelick PB, Rockwood K, Burns A, Gauthier S, DeKosky ST. Vascular cognitive impairment. Lancet Neurol. 2003;2:89 98. 16. Bowler JV, Steenhuis R, Hachinski V. Conceptual background to vascular cognitive impairment. Alzheimer Dis Assoc Disord. 1999; 13(suppl 3):S30 S37.

Ingles et al Neuropsychological Predictors of Dementia 1153 17. Rockwood K, Bowler J, Erkinjuntti T, Hachinski V, Wallin A. Subtypes of vascular dementia. Alzheimer Dis Assoc Disord. 1999;13(suppl 3): S59 S65. 18. Rockwood K, Wentzel C, Hachinski V, Hogan DB, MacKnight C, McDowell I. Prevalence and outcomes of vascular cognitive impairment. Neurology. 2000;54:447 451. 19. The Canadian Study of Health and Aging Working Group. The incidence of dementia in Canada. Neurology. 2000;55:66 73. 20. Teng EL, Chui HC. The modified Mini-Mental State (3MS) examination. J Clin Psychiatry. 1989;48:314 318. 21. Tuokko H, Kristjansson E, Miller J. Neuropsychological detection of dementia: an overview of the neuropsychological component of the Canadian study of health and aging. J Clin Exp Neuropsychol. 1995;17: 352 373. 22. Buschke H. Cued recall in amnesia. J Clin Neuropsychol. 1984;6: 433 440. 23. Tuokko H, Vernon-Wilkinson R, Weir J, Beattie BL. Cued recall and early identification of dementia. J Clin Exp Neuropsychol. 1991;13: 871 879. 24. Bleecker ML, Bolla-Wilson K. Age-related sex differences in verbal memory. J Clin Psychol. 1988;44:403 411. 25. Taylor EM. The Appraisal of Children With Cerebral Deficits. Cambridge, Mass: Harvard University Press; 1959. 26. Rey A. L examen Clinique en Psychologie. Paris: Presses Universitaires de France; 1964. 27. Benton AL. Revised Visual Retention Test: Clinical and Experimental Applications. New York: Psychological Corp; 1974. 28. Wechsler D. Wechsler Memory Scale. New York: Psychological Corp; 1975. 29. Wechsler D. Wechsler Adult Intelligence Scale Revised. New York: Psychological Corp; 1981. 30. Satz P, Mogel S. An abbreviation of the WAIS for clinical use. J Clin Psychol. 1962;18:77 79. 31. Benton AL, Hamsher K. Multilingual Aphasia Examination: Manual of Instructions. Iowa City, Iowa: AJA Associates; 1989. 32. Spreen O, Benton AL. Neurosensory Center Comprehensive Examination for Aphasia: Manual of Instructions. Victoria, Canada: University of Victoria; 1977. 33. Rosen WG. Verbal fluency in aging and dementia. J Clin Neuropsychol. 1980;2:135 146. 34. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadian EM. Clinical diagnosis of AD: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on AD. Neurology. 1984;34:939 944. 35. World Health Organization. International Statistical Classification of Diseases and Related Health Problems, 10th Revision. Geneva: World Health Organization; 1992. 36. Graham JE, Rockwood K, Beattie BL, Eastwood R, Gauthier S, Tuokko H, McDowell I. Prevalence and severity of cognitive impairment with and without dementia in an elderly population. Lancet. 1997;349:1793 1796. 37. Rockwood K, Davis H, MacKnight C, Vandorpe R, Gauthier S, Guzman A, Montgomery P, Black S, Hogan DB, Kertesz A, Bouchard R, Feldman H. The consortium to investigate vascular impairment of cognition: methods and first findings. Can J Neurol Sci. 2003;30:237 243. 38. Rockwood K, Black SE, Song X, Hogan DB, Gauthier S, MacKnight C, Vandorpe R, Guzman A, Montgomery P, Kertesz A, Bouchard RW, Feldman H. Clinical and radiographic subtypes of vascular cognitive impairment in a clinic-based cohort study. J Neurol Sci. 2006;240:7 14. 39. Price CC, Jefferson AL, Merino JG, Heilman KM, Libon DJ. Subcortical vascular dementia: integrating neuropsychological and neuroradiologic data. Neurology. 2005;65:376 382. 40. Looi JC, Sachdev PS. Differentiation of vascular dementia from AD on neuropsychological tests. Neurology. 1999;53:670 678. 41. Lafosse JM, Reed BR, Mungas D, Sterling SB, Wahbeh H, Jagust WJ. Fluency and memory differences between ischemic vascular dementia and Alzheimer s disease. Neuropsychology. 1997;11:514 522. 42. Tierney MC, Black SE, Szalai JP, Snow WG, Fisher RH, Nadon G, Chui HC. Recognition memory and verbal fluency differentiate probable Alzheimer disease from subcortical ischemic vascular dementia. Arch Neurol. 2001;58:1654 1659. 43. Graham NL, Emery T, Hodges JR. Distinctive cognitive profiles in Alzheimer s disease and subcortical vascular dementia. J Neurol Neurosurg Psychiatry. 2004;75:61 71. 44. Klages JD, Fisk JD, Rockwood K. APOE genotype, vascular risk factors, memory test performance and the five-year risk of vascular cognitive impairment or Alzheimer s disease. Dement Geriatr Cogn Disord. 2005; 20:292 297. 45. Desmond DW. The neuropsychology of vascular cognitive impairment: is there a specific cognitive deficit? J Neurol Sci. 2004;226:3 7. 46. Boyle PA, Paul RH, Moser DJ, Cohen RA. Executive impairments predict functional declines in vascular dementia. Clin Neuropsychol. 2004;18:75 82. 47. Boyle PA, Paul R, Moser D, Zawacki T, Gordon N, Cohen R. Cognitive and neurologic predictors of functional impairment in vascular dementia. Am J Geriatr Psychiatry. 2003;11:103 106. 48. Royall DR, Lauterbach EC, Cummings JL, Reeve A, Rummans TA, Kaufer DI, LaFrance WC, Coffey CE. Executive control function: a review of its promise and challenges for clinical research. J Neuropsychiatry Clin Neurosci. 2002;14:377 405. 49. Royall DR. Executive cognitive impairment: a novel perspective on dementia. Neuroepidemiology. 2000;19:293 299. 50. Royall DR. Precis of executive dyscontrol as a cause of problem behavior in dementia. Exp Aging Res. 1994;20:73 94.