Incidence of dementia after ischemic stroke: results of a longitudinal study

Incidence of Dementia After Ischemic Stroke Results of a Longitudinal Study David W. Desmond, PhD; Joan T. Moroney, MD, MRCPI; Mary Sano, PhD; Yaakov Stern, PhD Background and Purpose—A number of cross-sectional epidemiological studies have reported that one fourth of elderly patients meet criteria for dementia 3 months after ischemic stroke, but few longitudinal studies of the incidence of dementia after stroke have been performed. We conducted the present study to investigate the incidence and clinical predictors of dementia after ischemic stroke. Methods—We administered neurological, neuropsychological, and functional assessments annually to 334 ischemic stroke patients (age, 70.4⫾7.5 years) and 241 stroke-free control subjects (age, 70.6⫾6.5 years), all of whom were nondemented in baseline examinations. We diagnosed incident dementia using modified Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition criteria requiring deficits in memory and ⱖ2 additional cognitive domains, as well as functional impairment. Results—The crude incidence rate of dementia was 8.49 cases per 100 person-years in the stroke cohort and 1.37 cases per 100 person-years in the control cohort. A Cox proportional-hazards analysis found that the relative risk (RR) of incident dementia associated with stroke was 3.83 (95% CI, 2.14 to 6.84), adjusting for demographic variables and baseline Mini-Mental State Examination score. Within the stroke cohort, intercurrent medical illnesses associated with cerebral hypoxia or ischemia were independently related to incident dementia (RR, 4.40; 95% CI, 2.20 to 8.85), adjusting for recurrent stroke, demographic variables, and baseline Mini-Mental State Examination score. Conclusions—The risk of incident dementia is high among patients with ischemic stroke, particularly in association with intercurrent medical illnesses that might cause cerebral hypoxia or ischemia, suggesting that cerebral hypoperfusion may serve as a basis for some cases of dementia after stroke. (Stroke. 2002;33:2254-2262.) Key Words: Alzheimer disease 䡲 cerebrovascular disorders 䡲 dementia 䡲 dementia, vascular 䡲 stroke

A

number of cross-sectional epidemiological studies have suggested that ischemic stroke is a potent risk factor for dementia.1–3 In our own work,3 we diagnosed dementia in one fourth of a large cohort of elderly patients 3 months after ischemic stroke. The clinical determinants of dementia included features of the presenting stroke such as its size and location, vascular risk factors such as diabetes mellitus and prior stroke, and host characteristics such as older age. Because of the increased frequency of adverse outcomes among patients with stroke and dementia,4 which results in early patient attrition and the underestimation of the true frequency of dementia after stroke in prevalence surveys,5 however, studies of the incidence of dementia would be likely to provide a more accurate estimate of the magnitude of the association between ischemic stroke and dementia.6 Such studies could also permit the recognition of risk factors for incident dementia, which might be responsive to targeted interventions to slow or arrest the course of cognitive decline, but few such studies have been performed.

See Editorial Comment, page 2261 We conducted the present study to investigate the frequency and clinical determinants of incident dementia after ischemic stroke. Following our previous report on this topic,7 we recruited a second stroke cohort and extended the length of follow-up of the stroke and control cohorts that we had recruited previously. We administered comprehensive annual assessments to all subjects who were found to be free of dementia in baseline assessments in an effort to answer 2 specific questions. First, what is the risk of incident dementia associated with ischemic stroke? Second, consistent with studies that have suggested that cerebral hypoperfusion may serve as a basis for vascular dementia,8,9 what role do intercurrent medical illnesses associated with cerebral hypoxia or ischemia play as determinants of incident dementia after ischemic stroke?

Methods Subjects As part of a longitudinal study of stroke and dementia,3 we recruited 585 subjects among patients consecutively admitted to Columbia-

Received November 16, 2001; final revision received March 7, 2002; accepted May 13, 2002. From the Departments of Neurology and Pathology, SUNY Downstate Medical Center, Brooklyn, NY (D.W.D.); Department of Clinical Neurosciences, Royal College of Surgeons in Ireland, Beaumont Hospital, Dublin, Ireland (J.T.M.); and Departments of Neurology and Psychiatry and the Gertrude H. Sergievsky Center, Columbia University, College of Physicians and Surgeons, New York, NY (M.S., Y.S.). Correspondence to Dr David W. Desmond, SUNY Downstate Medical Center, 450 Clarkson Ave, Box 25, Brooklyn, NY 11203. E-mail [email protected] © 2002 American Heart Association, Inc. Stroke is available at http://www.strokeaha.org

DOI: 10.1161/01.STR.0000028235.91778.95

2254 Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015

Desmond et al Presbyterian Medical Center for ischemic stroke. We recruited 297 of those patients from 1988 to 1990 and the remaining 288 patients from 1994 to 1997; those 2 recruitment phases corresponded to 2 funding cycles. Eligibility requirements included age ⱖ60 years and a diagnosis of ischemic stroke within the previous 30 days confirmed by brain imaging (relevant infarct or normal). Patients were excluded when certain clinical features precluded reliable assessment of cognitive function, including a Boston Diagnostic Aphasia Examination10 severity rating ⬍3 (lower scores represent greater severities), persistent impairment of consciousness, or a primary language other than English or Spanish. Additional exclusion criteria included the presence of a concomitant neurological disorder potentially affecting cognitive function (eg, Parkinson’s disease) or a severe comorbid medical illness (eg, terminal cancer) that would preclude follow-up throughout the course of this longitudinal study. Using neuropsychological and functional assessments performed 3 months after stroke and modified criteria from the Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition (DSM-III-R),11 all of which are described in detail below, we diagnosed dementia in 119 of the 453 ischemic stroke patients available for examination.3 The remaining 334 nondemented patients (age, 70.4⫾7.5 years; education, 10.7⫾4.9 years) constitute the incidence patient cohort for this study. Regarding race and ethnicity, 126 of the 334 stroke patients were black (37.7%), 98 patients were Hispanic (29.3%), 104 patients were white (31.2%), and 6 patients (1.8%) were of other race/ethnicity. Regarding sex, 167 of the 334 stroke patients (50.0%) were women. We also recruited a control cohort of 249 subjects who were ⱖ60 years of age and who were free of any history of stroke or evidence of stroke in neurological examination. Most of the control cohort was randomly selected from the surrounding community from a Medicare list (53.0%); the remaining subjects were spouses of stroke patients also enrolled in our study (17.3%) or neighborhood volunteers who came to our attention through advertisements or referrals by friends (29.7%). As a group, control subjects were matched to the stroke cohort by age. In baseline examinations, 8 of those 249 subjects were found to be demented, and the remaining 241 subjects (age, 70.6⫾6.5 years; education, 12.4⫾4.5 years) constitute the incidence control cohort for this study. Regarding race and ethnicity, 77 of the 241 control subjects were black (32.0%), 36 subjects were Hispanic (14.9%), and 128 subjects were white (53.1%). Regarding sex, 159 of the 241 control subjects (66.0%) were women. This study was approved by the Institutional Review Board of Columbia-Presbyterian Medical Center, and all subjects provided informed consent.

Assessment Procedures and Follow-Up Seven to 10 days after stroke onset, neurologists specializing in stroke administered a structured neurological examination and documented any history of stroke, transient ischemic attack, or exposure to risk factors for cerebrovascular disease on the basis of a review of medical records and a structured interview administered to all patients and knowledgeable informants. A comprehensive medical history was also recorded. Patients were classified by infarct location and stroke syndrome using modified methods of the Stroke Data Bank12 on the basis of a review of clinical features and brain imaging performed immediately after stroke. We performed our baseline comprehensive assessment of the stroke cohort 3 months after stroke. We then examined all subjects annually on the basis of the date of stroke onset for patients and the date of the baseline examination for control subjects using the same assessment protocol. During the baseline assessment and all annual examinations, all subjects were administered a comprehensive battery of neuropsychological tests developed for use in epidemiological studies of dementia,13 which is described in detail elsewhere,3 with testing performed in either English or Spanish, whichever was spoken in the subject’s home; the Mini-Mental State Examination (MMSE),14 which was not part of our dementia diagnosis paradigm; the Barthel Index,15 which taps the physical aspects of activities of daily living; and the Structured Interview Guide for the Hamilton Depression Rating Scale (SIGH-D),16 with depression defined as an

Incidence of Dementia After Ischemic Stroke

2255

SIGH-D total score ⬎11 with the acknowledgment of depressed mood. Knowledgeable informants were administered the Blessed Functional Activity Scale (BFAS),17 which taps the cognitive aspects of activities of daily living. Neurologists specializing in stroke administered a structured neurological examination and documented any recurrent strokes and intercurrent illnesses that may have occurred. Stroke patients were also rated on the Stroke Data Bank Stroke Severity Scale.12 To maximize follow-up rates, we visited subjects’ homes or healthcare facilities if they were unable or unwilling to be examined in our clinic, and we did not consider subjects to be “refusals” for a particular interval until at least 4 attempts had been made to contact and examine them. When in-person examinations were not possible, we obtained information by telephone to ascertain vital status and the occurrence of major clinical events.

Diagnosis of Hypoxic-Ischemic Disorders Consistent with our previous work,7 intercurrent illnesses and hospitalizations were reviewed to identify all disorders that could result in cerebral hypoxia or ischemia, including cardiopulmonary arrest, cardiac arrhythmias, congestive heart failure, myocardial infarction, syncope, seizures, sepsis, pneumonia, respiratory failure, drug overdose, burns, hypotension with general anesthesia, profound hypoglycemia, hanging, and strangulation. Some of those disorders were not present in our sample (eg, strangulation), and others were always associated with patient death (eg, cardiopulmonary arrest). We did not rely on specific diagnostic criteria for each hypoxic-ischemic (HI) disorder but instead relied on the clinical judgment of the treating physician. Exact dates for all HI disorders were recorded when available; otherwise, the date of occurrence was assigned to the midpoint of the corresponding follow-up interval. For those patients who experienced ⬎1 HI disorder, we used the date of the event closest to that of the next scheduled follow-up examination. The diagnosis of HI disorders was performed by researchers blinded to incident dementia status.

Dementia Diagnosis Dementia was diagnosed using criteria modified from the DSM-IIIR.11 We required deficits in memory and ⱖ2 additional cognitive domains as determined in the neuropsychological evaluation, as well as functional impairment not solely related to physical disability documented with the BFAS. When patients were aphasic, we required that they exhibit evidence of nonverbal memory impairment. We defined impairment within any cognitive domain as any neuropsychological test score within that domain falling below a predetermined cutoff that was selected in a pilot study. We consider those cutoffs to be conservative. We used this same paradigm to diagnose incident dementia at every interval. We did not require a specific decrement in performance for that diagnosis. Instead, we required only performance falling below our cutoff scores combined with functional impairment documented with the BFAS. These procedures were intended to maximize consistency and reliability in diagnosis over time.

Statistical Analyses We calculated survival time from the date of the baseline examination. The date of onset of new dementia was defined as the date of the examination during which a subject’s performance first met diagnostic criteria for that disorder. For patients who never met criteria for incident dementia, the date of censoring was considered to be the date of the final completed annual examination. Reasons for censoring included death, subject dropout or loss to follow-up, or the end of formal study follow-up on May 31, 1999. The crude incidence rate of new dementia, stratified by stroke status, was calculated using life-table methods. To compute the relative risk (RR) of incident dementia associated with stroke versus control status, we performed Cox proportional-hazards analyses, first unadjusted and then adjusted for demographic variables and baseline MMSE score, which served to represent the severity of cognitive impairment 3 months after stroke for patients and at baseline for control subjects.

Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015

2256

Stroke

September 2002

We then investigated the clinical predictors of incident dementia within our stroke cohort. We performed log-rank tests to investigate the location and severity of the presenting stroke, vascular risk factors, and demographic variables as potential predictors. We also performed unadjusted Cox proportional-hazards analyses to investigate the risk associated with HI disorders and recurrent stroke, which were entered as time-dependent covariates. HI disorders and recurrent stroke were eligible for inclusion in our analyses only if they occurred between the baseline examination and either the date of diagnosis of incident dementia or the date of censoring. We then performed Cox proportional-hazards analyses to determine whether any of the variables found to be related to incident dementia in the univariate analyses (P⬍0.10) would be independently related to the incidence of dementia.

Results Incidence of Dementia in the Stroke and Control Cohorts

In the stroke cohort, 290 of the 334 patients (86.8%) had ⱖ1 follow-ups, 184 patients (55.1%) had ⱖ2 follow-ups, and 125 patients (37.4%) had ⱖ3 follow-ups. Death occurred in 63 patients, and 17 patients were lost to follow-up. Overall, members of the stroke cohort completed a median of 64.6% of all possible follow-up visits before a diagnosis of incident dementia, death, or the formal end of follow-up. The median follow-up was 21.1 months, with a maximum follow-up of 120.0 months. It is important to note that a proportion of our second stroke cohort was never eligible for second- or third-year follow-up examinations because of the restricted interval between their baseline examinations and the formal end of follow-up. New dementia was diagnosed in 72 stroke patients (21.6%) during 848.1 person-years of follow-up, yielding a crude incidence rate of 8.49 cases per 100 person-years. In the control cohort, 209 of the 241 subjects (86.7%) had ⱖ1 follow-ups, 196 subjects (81.3%) had ⱖ2 follow-ups, and 164 subjects (68.0%) had ⱖ3 follow-ups. Death occurred in 35 control subjects, and 16 subjects were lost to follow-up. Overall, members of the control cohort completed a median of 62.5% of all possible follow-up visits before a diagnosis of incident dementia, death, or the formal end of follow-up. The median follow-up was 62.2 months, with a maximum follow-up of 119.9 months. New dementia was diagnosed in 17 control subjects (7.0%) during 1243.7 person-years of follow-up, yielding a crude incidence rate of 1.37 cases per 100 person-years. A log-rank test determined that the survival curves of the stroke and control cohorts, which are shown in the Figure, were significantly different (P⬍0.0001). An unadjusted Cox proportional-hazards analysis found that stroke status was a significant predictor of incident dementia (RR, 6.12; 95% CI, 3.57 to 10.50). That RR suggests that patients with ischemic stroke are at a 6-foldincreased risk of dementia during long-term follow-up, whereas the CI suggests that there is a 95% probability that the risk of dementia is increased by at least 4-fold and possibly by as much as 10-fold among those patients. When demographic variables were added to that model, the RR associated with stroke status was 5.20 (95% CI, 2.97 to 9.12), adjusting for age ⱖ80 years (RR, 6.30; 95% CI, 3.50 to 11.33) and age 70 to 79 years (RR, 2.70; 95% CI, 1.62 to 4.48) versus age 60 to 69, education of 0 to 8 years (RR, 3.57;

Kaplan-Meier analysis showing cumulative proportion of subjects surviving free of dementia stratified by stroke status during follow-up of up to 120 months. Numbers of subjects remaining active at the end of each interval for the stroke group (bottom line) and control group (top line) were 211 and 206 at year 1, 144 and 195 at year 2, 92 and 170 at year 3, 60 and 150 at year 4, 48 and 125 at year 5, 40 and 112 at year 6, 30 and 97 at year 7, 20 and 74 at year 8, and 7 and 21 at year 9, respectively.

95% CI, 1.85 to 6.87) and 9 to 12 years (RR, 1.61; 95% CI, 0.84 to 3.07) versus ⱖ13 years, black race (RR, 1.56; 95% CI, 0.90 to 2.72) and Hispanic ethnicity (RR, 1.21; 95% CI, 0.63 to 2.31) versus white race, and female sex (RR, 1.08; 95% CI, 0.68 to 1.70). As shown in Table 1 (model 1), when a baseline MMSE score ⬍24, which was not part of our dementia diagnosis paradigm, was added to that model, the RR associated with stroke status was 3.83 (95% CI, 2.14 to 6.84), adjusting for age ⱖ80 years (RR, 5.88; 95% CI, 3.22 to 10.76) and age 70 to 79 years (RR, 2.15; 95% CI, 1.27 to TABLE 1. Primary Cox Proportional-Hazards Models of the Predictors of Incident Dementia RR (95% CI) Model 1, Stroke and Control Cohorts

Model 2, Stroke Cohort

3.83 (2.14–6.84)

䡠䡠䡠

ⱖ80

5.88 (3.22–10.76)

4.17 (2.03–8.56)

70–79

2.15 (1.27–3.62)

1.95 (1.08–3.52)

0–8

2.69 (1.34–5.38)

2.02 (0.88–4.64)

9–12

1.49 (0.76–2.92)

1.53 (0.70–3.35)

Black

1.72 (0.98–3.02)

1.28 (0.67–2.44)

Variable Stroke vs control status Age (vs 60–69 y), y

Education (vs ⱖ13 y), y

Race/ethnicity (vs white) Hispanic

1.47 (0.76–2.84)

1.34 (0.64–2.81)

Sex (female)

0.93 (0.58–1.49)

0.89 (0.51–1.54)

MMSE total score⬍24

3.11 (1.90–5.09)

3.45 (2.02–5.88)

Recurrent stroke

䡠䡠䡠

2.71 (1.44–5.10)

HI disorders

䡠䡠䡠

4.40 (2.20–8.85)

Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015

Desmond et al TABLE 2. Demographic Variables, Vascular Risk Factors, and HI Disorders by Incident Dementia Status in the Stroke Cohort

Incidence of Dementia After Ischemic Stroke

TABLE 3. Index Stroke Characteristics and Recurrent Stroke by Incident Dementia Status in the Stroke Cohort

Incident Dementia, n (%) Variable

Incident Dementia, n (%)

Yes (n⫽72)

No (n⫽262)

P

ⱖ80

15 (20.8)

27 (10.3)

⬍0.001

70–79

30 (41.7)

85 (32.4)

60–69

27 (37.5)

150 (57.3)

Age, y

Variable

Yes (n⫽72)

No (n⫽262)

Major dominant hemispheral

8 (11.1)

12 (4.6)

Major nondominant hemispheral

6 (8.3)

17 (6.5)

Minor dominant hemispheral

9 (12.5)

39 (14.9)

Minor nondominant hemispheral

8 (11.1)

30 (11.5)

Lacunar/deep hemispheral

28 (38.9)

103 (39.3)

Brainstem/cerebellar

13 (18.1)

61 (23.3)

Left hemisphere

22 (30.6)

80 (30.7)

Right hemisphere

30 (41.7)

91 (34.9)

Brainstem/cerebellum

20 (27.8)

90 (34.5)

P

Stroke syndrome

Education, y 0–8

31 (43.1)

75 (28.6)

9–12

30 (41.7)

104 (39.7)

ⱖ13

11 (15.3)

83 (31.7)

Black

33 (46.5)

93 (36.2)

Hispanic

20 (28.2)

78 (30.4)

0.010

0.323

Stroke location

Race/ethnicity 0.515

White

18 (25.4)

86 (33.5)

Sex (female)

47 (65.3)

120 (45.8)

0.050

ICA

2 (2.8)

11 (4.2)

Hypertension

52 (72.2)

194 (74.0)

0.758

ACA

3 (4.2)

5 (1.9)

Diabetes mellitus

25 (34.7)

82 (31.3)

0.286

MCA

36 (50.0)

124 (47.3)

Myocardial infarction

16 (22.2)

41 (15.6)

0.117

PCA

10 (13.9)

34 (13.0)

Angina

16 (22.2)

54 (20.8)

0.662

Vertebrobasilar

21 (29.2)

88 (33.6)

Atrial fibrillation

12 (16.7)

34 (13.1)

0.200

Stroke mechanism

7 (9.9)

24 (9.2)

0.626

Large-artery atherosclerosis

12 (16.7)

48 (18.3)

Congestive heart failure

2257

0.344

Vascular territory 0.558

0.463

Hypercholesterolemia

13 (18.1)

67 (25.9)

0.485

Cardiac embolism

16 (22.2)

52 (19.8)

Consistent cigarette use

42 (58.3)

159 (61.6)

0.624

Lacunar

26 (36.1)

100 (38.2)

Consistent alcohol use

32 (45.1)

137 (53.1)

0.332

Unknown/other cause

18 (25.0)

62 (23.7)

Prior stroke

20 (27.8)

52 (19.8)

0.217

9 (13.8)

13 (5.1)

0.086

Prior transient ischemic attack

10 (14.3)

47 (18.1)

HI disorders

13 (18.1)

26 (9.9)

Depression (SIGH-D total score ⬎11 with depressed mood)

17 (23.6)

22 (8.4)

⬍0.001

0.429 ⬍0.001

Significance levels are based on log-rank tests for all variables except HI disorders, which was entered into an unadjusted Cox proportional-hazards analysis as a time-dependent covariate.

3.62) versus age 60 to 69 years, education of 0 to 8 years (RR, 2.69; 95% CI, 1.34 to 5.38) and 9 to 12 years (RR, 1.49; 95% CI, 0.76 to 2.92) versus ⱖ13 years, black race (RR, 1.72; 95% CI, 0.98 to 3.02) and Hispanic ethnicity (RR, 1.47; 95% CI, 0.76 to 2.84) versus white race, female sex (RR, 0.93; 95% CI, 0.58 to 1.49), and an MMSE score ⬍24 (RR, 3.11; 95% CI, 1.90 to 5.09). The results of this analysis were essentially unchanged after further adjustment for vascular risk factors such as hypertension, diabetes mellitus, and cardiac disease.

Risk Factors for Incident Dementia in the Stroke Cohort Demographic variables, vascular risk factors, and the occurrence of HI disorders by incident dementia status in the stroke cohort are shown in Table 2. Index stroke characteristics and recurrent stroke by incident dementia status are shown in Table 3. Log-rank tests demonstrated that patients with incident dementia were older, had received fewer years of education, tended to be female, and more often met our operationalized criteria for depression than patients who did not develop incident dementia. Because of the numerous categories included in the stroke syndrome variable, we also

Recurrent stroke

ICA indicates internal carotid artery; ACA, anterior cerebral artery; MCA, middle cerebral artery; and PCA, posterior cerebral artery. Significance levels are based on log-rank tests for all variables except recurrent stroke, which was entered into an unadjusted Cox proportional-hazards analysis as a timedependent covariate.

examined a variable representing a major hemispheral stroke syndrome versus all other stroke syndromes, and that variable was significantly related to incident dementia status by a log-rank test (P⫽0.032). Dichotomous recodings of the vascular territory and stroke mechanism variables were not related. Unadjusted Cox proportional-hazards analyses, with HI disorders and recurrent stroke entered as time-dependent covariates, demonstrated that each of those variables was significantly associated with incident dementia. An unadjusted Cox proportional-hazards analysis determined that the RR of incident dementia associated with HI disorders, which was entered as a time-dependent covariate, was 3.20 (95% CI, 1.66 to 6.14). That RR suggests that stroke patients with HI disorders are at a 3-fold-increased risk of dementia during long-term follow-up, whereas the CI suggests that there is a 95% probability that the risk of dementia is increased by at least 2-fold and possibly by as much as 6-fold among those patients. Adding demographic variables to that model showed that the risk of incident dementia

Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015

2258

Stroke

September 2002

remained elevated in association with HI disorders (RR, 4.70; 95% CI, 2.37 to 9.35), with adjustment for age ⱖ80 years (RR, 4.66; 95% CI, 2.36 to 9.22) and age 70 to 79 years (RR, 2.68; 95% CI, 1.52 to 4.74) versus age 60 to 69 years, education of 0 to 8 years (RR, 3.06; 95% CI, 1.43 to 6.58) and 9 to 12 years (RR, 1.68; 95% CI, 0.80 to 3.50) versus ⱖ13 years, black race (RR, 1.45; 95% CI, 0.78 to 2.71) and Hispanic ethnicity (RR, 1.21; 95% CI, 0.59 to 2.49) versus white race, and female sex (RR, 1.19; 95% CI, 0.71 to 1.99). Finally, as shown in Table 1 (model 2), our primary Cox proportional-hazards analysis demonstrated that the risk of incident dementia was elevated in association with HI disorders (RR, 4.40; 95% CI, 2.20 to 8.85), while adjusting for recurrent stroke (RR, 2.71; 95% CI, 1.44 to 5.10), age ⱖ80 (RR, 4.17; 95% CI, 2.03 to 8.56) and age 70 to 79 (RR⫽1.95; 95% CI, 1.08 to 3.52) versus age 60 to 69, education of 0 to 8 years (RR, 2.02; 95% CI, 0.88 to 4.64) and 9 to 12 years (RR, 1.53; 95% CI, 0.70 to 3.35) versus ⱖ13 years, black race (RR, 1.28; 95% CI, 0.67 to 2.44) and Hispanic ethnicity (RR, 1.34; 95% CI, 0.64 to 2.81) versus white race, female sex (RR, 0.89; 95% CI, 0.51 to 1.54), and an MMSE score ⬍24 (RR, 3.45; 95% CI, 2.02 to 5.88). Depression and a major hemispheral stroke syndrome were not significantly related to incident dementia status and were excluded from this multivariate model. Within the group of 39 stroke patients who had experienced intercurrent HI disorders, 27 patients experienced cardiac HI disorders (myocardial infarction in 1 incident and 10 nonincident patients, congestive heart failure in 2 incident and 6 nonincident patients, atrial fibrillation or other arrhythmias in 2 incident and 3 nonincident patients, and syncope in 1 incident and 2 nonincident patients), and 12 patients experienced noncardiac HI disorders (pneumonia in 4 incident and 1 nonincident patients, seizures in 3 incident and 2 nonincident patients, and sepsis in 2 nonincident patients). Although the cell sizes are relatively small, noncardiac HI disorders, particularly pneumonia and seizures, were associated with a significantly higher frequency of incident dementia (58.3%) than cardiac HI disorders (22.2%; P⫽0.027 by ␹2). The results of that analysis were unchanged after adjustment for age.

Discussion We found that the risk of incident dementia was increased 4-fold among ischemic stroke patients who were initially nondemented relative to clinically stroke-free elderly control subjects after adjustment for demographic factors and baseline level of cognitive function. Among stroke patients, the risk of incident dementia was elevated in association with intercurrent illnesses that might cause cerebral hypoxia or ischemia after adjustment for those same variables and recurrent stroke, suggesting that cerebral hypoperfusion may serve as a basis for some cases of dementia after stroke. Although the cell sizes were relatively small, noncardiac HI disorders, particularly pneumonia and seizures, were associated with a significantly higher frequency of incident dementia than cardiac HI disorders. In addition to the risk factors that we identified in this study, it is also likely that a proportion of our cohort was affected by concomitant Alz-

heimer’s disease, which may have served as the primary basis for their cognitive decline, and certain patients whose baseline neuropsychological test scores were slightly higher than our operationalized cutoffs may have crossed over those cutoffs without exhibiting clinically meaningful decline, causing them to be identified as incident cases of dementia when they might have been more accurately characterized as prevalent cases. The findings of the few previous longitudinal studies that have been performed based on hospitalized stroke series are consistent with our own in suggesting that the risk of incident dementia associated with stroke is high. Tatemichi et al18 reported that the incidence of dementia was 6.7% among patients 60 to 64 years of age and 26.5% among patients ⱖ85 years of age after 1 year of follow-up in a sample of 610 patients who were initially nondemented after stroke, but they did not report an overall frequency of incident dementia. Bornstein et al19 reported that 56 of 175 patients who were initially nondemented after stroke (32.0%) developed incident dementia during 5 years of follow-up after first ischemic stroke. Hénon et al20 examined a cohort of 169 patients who had been nondemented before stroke onset and reported that the cumulative proportion of patients with incident dementia was 21.3% after 3 years of follow-up. The onset of new dementia occurred immediately after the index stroke in most cases, however, and only 7% of patients who were nondemented 6 months after that index stroke developed incident dementia during the remainder of the 3 years of follow-up. In 2 studies based on patients presenting with a lacunar infarction as their first stroke, Loeb et al21 found that 25 of 108 patients (23.2%) developed incident dementia during an average of 4 years of follow-up, and Samuelsson et al22 found that 4 (4.9%) and 8 (9.9%) of 81 patients developed incident dementia after 1 and 3 years of follow-up, respectively. In population-based studies of stroke and incident dementia, Kokmen et al23 reviewed the medical records of a sample of 971 patients who were free of dementia before first stroke. The cumulative incidence of dementia, which includes prevalent cases with an onset immediately after stroke, was 7% at 1 year, 10% (ie, an additional 3% of new cases) at 3 years, 15% (ie, a further 5% of new cases) at 5 years, and 23% (ie, a further 8% of new cases) at 10 years. Zhu et al24 studied 1301 initially nondemented subjects ⱖ75 years of age, 7.1% of whom had a history of stroke, and diagnosed incident dementia in 224 subjects (17.2%) after 3 years of follow-up. The RR of incident dementia associated with prior stroke was 1.7 (95% CI, 1.1 to 2.6) after adjustment for potential confounders, and prior stroke was particularly potent when it had occurred within the preceding 3 years. In addition, the RR of incident dementia associated with incident stroke, or a first stroke occurring during follow-up, was 2.4 (95% CI, 1.6 to 3.5). Our study and certain of the studies cited earlier have similarly recognized the importance of recurrent stroke occurring during the study period,21,23 and the central role of recurrent stroke as a risk factor for incident dementia has received a great deal of attention in studies of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL).25,26

Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015

Desmond et al Although numerous studies have focused on the clinical consequences of catastrophic HI events, particularly cardiopulmonary arrest,27,28 and some studies have reported an association between HI events and memory disorders,29,30 only a few studies have investigated the association between HI disorders and dementia. Using a sample of 133 patients with vascular dementia, Sulkava and Erkinjuntti8 identified 6 patients (4.5%) who exhibited the acute onset of dementia in temporal association with cardiac arrhythmias and systemic arterial hypotension that were judged to be responsible for cerebral hypoperfusion. Similarly, Skoog et al31 found that 4.1% of 147 elderly (85 years of age) patients with dementia had cerebral hypoperfusion as the primary cause of their dementia syndrome. In the study described above, Hénon et al20 found that 13.9% of 36 stroke patients experienced intercurrent HI disorders before the onset of incident dementia, but it should also be noted that 12.0% of the 133 patients who did not develop incident dementia also experienced those disorders. In the Cardiovascular Health Study,32 orthostatic hypotension was significantly associated with whitematter lesions on MRI of the brain, and those lesions were associated with poorer performance on a mental status test. Cooper and Mungas33 compared 502 patients with vascular dementia with 810 patients with Alzheimer’s disease and found that patients with vascular dementia more frequently had a history of general anesthesia than patients with Alzheimer’s disease, whereas those groups did not differ with regard to a family history of dementia or a history of head injury. Finally, Brun9 found that 28.8% of patients with vascular dementia had neuropathological evidence of cerebral hypoperfusion, with either selective incomplete infarction of the cerebral white matter or borderzone infarction. In addition to those variables that were significantly related to incident dementia in our study, including HI disorders, recurrent stroke, and older age, it is also worthwhile to review selected variables that were not related. First, qualitative brain imaging variables, including the location of the index stroke, were not associated with incident dementia, and other studies have reported similar findings.21,23 Although our previous work suggests that certain of those variables were likely to have been related to deficits in baseline cognitive function, which were represented by the baseline MMSE score in our primary Cox proportional-hazards analysis, the effects of those lesions would typically have been static or slightly remitting rather than progressive. Second, similar to the findings of other studies,19,21 vascular risk factors were not related to incident dementia in our stroke cohort, but it is likely that certain of them may have elevated the risk of recurrent stroke and thus indirectly contributed to the incidence of dementia. To the extent that more effective management of vascular risk factors might reduce the risk of recurrent stroke, such an intervention might also reduce the risk of incident dementia. Third, like virtually all dementia studies, we recognized the importance of older age as a risk factor, but education and other demographic variables were unrelated to incident dementia in our primary Cox proportional-hazards analysis. Our failure to recognize an association between education and incident dementia suggests that the “cognitive reserve” hypothesis that has received

Incidence of Dementia After Ischemic Stroke

2259

a great deal of attention in studies of Alzheimer’s disease34 may be of less importance in studies of dementia after stroke. Our study has certain limitations. First, although most of our control cohort was randomly selected from the surrounding community from a Medicare list, the remaining subjects were spouses of stroke patients also enrolled in our study or neighborhood volunteers who came to our attention through advertisements or referrals by friends. Given that our methods may have been biased toward the recruitment of healthier control subjects, we may have slightly overestimated the magnitude of the risk of incident dementia associated with ischemic stroke. Second, we did not have neuropathological confirmation of the dementia subtype in our patients. Thus, we were unable to characterize the importance of concomitant Alzheimer’s disease as a risk factor for incident dementia after stroke. In our Cox proportional-hazards analyses, however, we found that older age was associated with a significantly elevated risk of incident dementia, and it is likely that that variable can be considered a crude surrogate for Alzheimer’s disease. Third, although we focused on the qualitative features of the index stroke as predictors of incident dementia, certain quantitative brain imaging measures (eg, the volume or number of clinically “silent” cerebral infarctions, severity of diffuse white matter disease, severity of atrophy), standardized imaging of symptomatic and clinically “silent”35,36 recurrent stroke, and the findings of state-of-the-art brain imaging techniques (eg, diffusion tensor imaging to assess the integrity of subcortical pathways) might have been relevant to incident dementia. Fourth, we did not examine the contribution of genetic factors. It is becoming clear that genetic factors are important in vascular dementia, whether as risk markers such as the apolipoprotein E ⑀4 allele37 or as primary independent risk factors such as Notch3 mutations in CADASIL,38,39 and these and other genetic factors warrant further study. Fifth, the results of studies such as ours are influenced by the paradigm selected for use in the diagnosis of dementia. To the extent that our use of an alternative diagnostic method might have caused us to identify a larger or smaller number of prevalent cases of dementia, the group of patients who would have been found to be nondemented at baseline and thus at risk of incident dementia would have differed from those on whom this study is based, potentially affecting our estimate of the incidence rate and the risk factors that we identified. In previous work,4 however, we found that the diagnostic method used in this study had greater predictive validity with regard to the adverse outcomes of recurrent stroke and death than less restrictive paradigms based on neuropsychological testing or a reliance on the MMSE or clinical judgment, suggesting that our approach was reasonable. Sixth, we performed our follow-up examinations annually; we may have obtained more precise information regarding the timing of the onset of dementia if we had examined patients more frequently and over shorter intervals. Such an approach would have reduced our ability to recruit and assess such a large cohort of patients, however, and more frequent visits might have caused reduced compliance with follow-up assessments.

Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015

2260

Stroke

September 2002

Acknowledgments This work was supported by grants R01-NS26179, P01-AG07232, K07-AG00959, and K08-NS02051 from the National Institutes of Health. We would like to acknowledge the staff of the Stroke and Aging Research Project for their assistance with data collection.

20.

21. 22.

References 1. Pohjasvaara T, Erkinjuntti T, Vataja R, Kaste M. Dementia three months after stroke: baseline frequency and effect of different definitions of dementia in the Helsinki Stroke Aging Memory Study (SAM) cohort. Stroke. 1997;28:785–792. 2. Barba R, Martínez-Espinosa S, Rodríguez-García E, Pondal M, Vivancos J, Del Ser T. Poststroke dementia: clinical features and risk factors. Stroke. 2000;31:1494 –1501. 3. Desmond DW, Moroney JT, Paik MC, Sano M, Mohr JP, Aboumatar S, Tseng CL, Chan S, Williams JBW, Remien RH, Hauser WA, Stern Y. Frequency and clinical determinants of dementia after ischemic stroke. Neurology. 2000;54:1124 –1131. 4. Desmond DW, Moroney JT, Bagiella E, Sano M, Stern Y. Dementia as a predictor of adverse outcomes following stroke: an evaluation of diagnostic methods. Stroke. 1998;29:69 –74. 5. Desmond DW, Bagiella E, Moroney JT, Stern Y. The effect of patient attrition on estimates of the frequency of dementia following stroke. Arch Neurol. 1998;55:390 –394. 6. Kelsey JL, Thompson WD, Evans AS. Methods in Observational Epidemiology. New York, NY: Oxford University Press; 1986. 7. Moroney JT, Bagiella E, Desmond DW, Paik MC, Stern Y, Tatemichi TK. Risk factors for incident dementia after stroke: role of hypoxic and ischemic disorders. Stroke. 1996;27:1283–1289. 8. Sulkava R, Erkinjuntti T. Vascular dementia due to cardiac arrhythmias and systemic hypotension. Acta Neurol Scand. 1987;76:123–128. 9. Brun A. Pathology and pathophysiology of cerebrovascular dementia: pure subgroups of obstructive and hypoperfusive etiology. Dementia. 1994;5:145–147. 10. Goodglass H, Kaplan E. The Assessment of Aphasia and Related Disorders. 2nd ed. Philadelphia, Pa: Lea & Febiger; 1983. 11. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 3rd ed, revised. Washington, DC: American Psychiatric Association; 1987. 12. Foulkes MA, Wolf PA, Price TR, Mohr JP, Hier DB. The Stroke Data Bank: design, methods, and baseline characteristics. Stroke. 1988;19: 547–554. 13. Stern Y, Andrews H, Pittman J, Sano M, Tatemichi T, Lantigua R, Mayeux R. Diagnosis of dementia in a heterogeneous population: development of a neuropsychological paradigm-based diagnosis of dementia and quantified correction for the effects of education. Arch Neurol. 1992;49:453– 460. 14. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189 –198. 15. Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md Med J. 1965;14:61– 65. 16. Williams JBW. A structured interview guide for the Hamilton Depression Rating Scale. Arch Gen Psychiatry. 1988;45:742–747. 17. Blessed G, Tomlinson BE, Roth M. The association between quantitative measures of dementia and of senile change in the cerebral grey matter of elderly subjects. Br J Psychiatry. 1968;114:797– 811. 18. Tatemichi TK, Foulkes MA, Mohr JP, Hewitt JR, Hier DB, Price TR, Wolf PA. Dementia in stroke survivors in the Stroke Data Bank cohort: prevalence, incidence, risk factors, and computed tomographic findings. Stroke. 1990;21:858 – 866. 19. Bornstein NM, Gur AY, Treves TA, Reider-Groswasser I, Aronovich BD, Klimovitzky SS, Varssano D, Korczyn AD. Do silent brain infarctions

23.

24.

25.

26.

27.

28.

29. 30. 31.

32.

33.

34.

35.

36.

37.

38.

39.

predict the development of dementia after first ischemic stroke? Stroke. 1996;27:904 –905. Hénon H, Durieu I, Guerouaou D, Lebert F, Pasquier F, Leys D. Poststroke dementia: incidence and relationship to prestroke cognitive decline. Neurology. 2001;57:1216 –1222. Loeb C, Gandolfo C, Croce R, Conti M. Dementia associated with lacunar infarction. Stroke. 1992;23:1225–1229. Samuelsson M, Söderfeldt B, Olsson GB. Functional outcome in patients with lacunar infarction. Stroke. 1996;27:842– 846. Kokmen E, Whisnant JP, O’Fallon WM, Chu CP, Beard CM. Dementia after ischemic stroke: a population-based study in Rochester, Minnesota (1960 –1984). Neurology. 1996;46:154 –159. Zhu L, Fratiglioni L, Guo Z, Basun H, Corder EH, Winblad B, Viitanen M. Incidence of dementia in relation to stroke and the apolipoprotein E ⑀4 allele in the very old: findings from a population-based longitudinal study. Stroke. 2000;31:53– 60. Dichgans M, Mayer M, Uttner I, Brüning R, Müller-Höcker J, Rungger G, Ebke M, Klockgether T, Gasser T. The phenotypic spectrum of CADASIL: clinical findings in 102 cases. Ann Neurol. 1998;44:731–739. Desmond DW, Moroney JT, Lynch T, Chan S, Chin SS, Mohr JP. The natural history of CADASIL: a pooled analysis of previously published cases. Stroke. 1999;30:1230 –1233. Dougherty JH, Rawlinson DG, Levy DE, Plum F. Hypoxic-ischemic brain injury and the vegetative state: clinical and neuropathologic correlation. Neurology. 1981;31:991–997. Levy DE, Caronna JJ, Singer BH, Lapinski RH, Frydman H, Plum F. Predicting outcome from hypoxic-ischemic coma. JAMA. 1985;253: 1420 –1426. Volpe BT, Hirst W. The characterization of an amnesic syndrome following hypoxic ischemic injury. Arch Neurol. 1983;40:436 – 440. Alexander MP. Specific semantic memory loss after hypoxic-ischemic injury. Neurology. 1997;48:165–173. Skoog I, Nilsson L, Palmertz B, Andreasson LA, Svanborg A. A population-based study of dementia in 85-year-olds. N Engl J Med. 1993;328:153–158. Longstreth WT, Manolio TA, Arnold A, Burke GL, Bryan N, Jungreis CA, Enright PL, O’Leary D, Fried L. Clinical correlates of white matter findings on cranial magnetic resonance imaging of 3301 elderly people: the Cardiovascular Health Study. Stroke. 1996;27:1274 –1282. Cooper J, Mungas D. Risk factor and behavioural differences between vascular and Alzheimer’s dementias: the pathway to end-stage disease. J Geriatr Psychiatry Neurol. 1993;6:29 –33. Stern Y, Gurland B, Tatemichi TK, Tang MX, Wilder D, Mayeux R. Influence of education and occupation on the incidence of Alzheimer’s disease. JAMA. 1994;271:1004 –1010. Meyer JS, Muramatsu K, Mortel KF, Obara K, Shirai T. Prospective CT confirms differences between vascular and Alzheimer’s dementia. Stroke. 1995;26:735–742. Yoshitake T, Kiyohara Y, Kato I, Ohmura T, Iwamoto H, Nakayama K, Ohmori S, Nomiyama K, Kawano H, Ueda K, Sueishi K, Tsuneyoshi M, Fujishima M. Incidence and risk factors of vascular dementia and Alzheimer’s disease in a defined elderly Japanese population: the Hisayama Study. Neurology. 1995;45:1161–1168. Slooter AJC, Tang MX, van Duijn CM, Stern Y, Ott A, Bell K, Breteler MMB, Van Broeckhoven C, Tatemichi TK, Tycko B, Hofman A, Mayeux R. Apolipoprotein E ⑀4 and the risk of dementia with stroke: a population-based investigation. JAMA. 1997;277:818 – 821. Joutel A, Corpechot C, Ducros A, Vahedi K, Chabriat H, Mouton P, Alamowitch S, Domenga V, Cécillion M, Maréchal E, Maciazek J, Vayssière C, Cruaud C, Cabanis EA, Ruchoux MM, Weissenbach J, Bach JF, Bousser MG, Tournier-Lasserve E. Notch3 mutations in CADASIL, a hereditary adult-onset condition causing stroke and dementia. Nature. 1996;383:707–710. Desmond DW, Moroney JT, Lynch T, Chan S, Chin SS, Shungu DC, Naini AB, Mohr JP. CADASIL in a North American family: clinical, pathologic, and radiologic findings. Neurology. 1998;51:844 – 849.

Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015

Desmond et al

Incidence of Dementia After Ischemic Stroke

2261

Editorial Comment Dementia After Stroke: High Incidence and Intriguing Associations Around 25% of patients with cerebrovascular disease meet operationalized criteria for dementia 3 months after a stroke,1 and a greater number have cognitive impairment short of dementia.2 Compared with individuals without ischemic brain disease, patients who are cognitively intact 3 months after a stroke have a 6- to 9-fold greater risk of developing dementia in the following year,3,4 and whereas the increased risk is greatest in the first 12 months,4 it is still present several years later. The relationship between acute stroke and prevalent and incident dementia has been studied in several hospitalized cohorts; the first one was assembled between 1988 and 1990 by Thomas Tatemichi and his colleagues.3 In this issue of Stroke, Desmond et al present combined longitudinal follow-up data of that cohort and a second group of patients assembled using identical methodologies between 1994 and 1997 at Columbia University.5 Subjects enrolled into the study include 334 stroke patients who were not demented 3 months after the cerebrovascular event and 241 stroke-free controls who either were selected from Medicare lists or were family or community volunteers. Median follow-up was 21.1 months, but a few patients were still in the study after 9 years. The crude incidence rate of dementia (defined according to Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition criteria using a comprehensive neuropsychological and clinical evaluation) in the stroke group was 8.49 cases per 100 person-years while among the controls it was 1.37 cases per 100 person-years. Although its magnitude may be overestimated because many of the controls were volunteers and not a random sample of the population from which the cases were recruited, the relative risk for incident dementia among patients with stroke was 4.4 (95% CI 2.20 to 8.85). The incidence and relative risk figures are similar to those reported by this and other groups3 and highlight the magnitude of the problem. Why do patients with stroke have such a high risk of developing dementia? Traditional concepts of vascular dementia postulate that cognitive decline in patients with cerebrovascular disease can result from the stroke alone when a large volume of brain is affected by infarcts and hemorrhages overcoming the brain’s reserve or compensatory mechanisms, and that strategic lesions can lead to intellectual decline when specific cortical or subcortical areas important for cognition and their connections are damaged. However, recent epidemiological and neuropathological studies have suggested that many patients with stroke develop dementia through the interaction of neurodegenerative and vascular insults to the brain that by themselves may not produce dementia yet in association hasten the decline of the intellect, blurring the sharp dichotomy between Alzheimer’s and vascular dementia. Some patients with dementia after stroke have a progressive course suggestive of a degenerative disorder,4 and

given the high prevalence of cerebrovascular pathology and Alzheimer’s disease in the elderly, it is likely that most have overlapping pathological processes.6 Patients who at autopsy have coexistent Alzheimer-type changes and cerebral infarcts have more severe cognitive impairment (and a higher prevalence of dementia) during life than patients with isolated senile plaques and neurofibrillary tangles.7 Traditional cerebrovascular risk factors have been linked to Alzheimer’s disease.8 Further studies must address whether this interaction is responsible for the increased risk of dementia in hospitalized stroke cohorts. Among Desmond and colleagues’ stroke patients, the risk of dementia was elevated in those who had a loosely defined group of intercurrent illnesses that can produce hypoxia. This finding is intriguing given that cerebral hypoperfusion can lead to cognitive impairment9 and in animal models it enhances amyloid ␤ precursor protein mRNA expression and cleavage of that protein.10 In addition, some conditions considered hypoxic-ischemic in the present study, such as atrial fibrillation, have been identified as risk factors for Alzheimer’s disease. However, the association described by Desmond et al, while previously reported by that group11 and others,12 does not establish causality. The category of hypoxic-ischemic disorders used in the study is too broad; it includes conditions that produce transient (seizures and syncope) and prolonged (heart failure, myocardial infarction) hypoxia, and it encompasses conditions that do not necessarily produce similar alterations in brain oxygenation. The association, however, warrants further studies with precise measures of cerebral blood flow to understand the pathophysiologic processes that lead to dementia in stroke patients. A greater understanding will result in new prophylactic interventions. If the interaction of vascular and neurodegenerative processes is the cause of dementia in patients with stoke, the therapeutic implications are enormous. In recent years, rigorous clinical trials have demonstrated the value of statin agents and inhibitors of angiotensin-converting enzyme in addition to antithrombotic drugs and surgery for the prevention of stroke and Alzheimer’s disease. We can expect that their use will lead to fewer cases of dementia. Conversely, if vascular and neurodegenerative processes indeed interact to produce dementia in a substantial number of cases, it makes sense to treat patients with stroke with acetylcholinesterase inhibitors in an effort to slow the neurodegenerative process. However, there is scant evidence that vascular preventive strategies lead to preserved cognitive function after a stroke. To test this hypothesis, future clinical stroke trials—acute and preventive—must incorporate cognitive evaluations as primary outcome measures. This, of course, opens up exciting new possibilities.

Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015

2262

Stroke

September 2002

José G. Merino, MD, MPhil, Guest Editor Department of Neurology Comprehensive Stroke Program University of Florida Health Sciences Center Jacksonville, Florida

References 1. Pohjasvaara T, Erkinjuntti T, Vataja R, Kaste M. Dementia three months after stroke: baseline frequency and effect of different definitions of dementia in the Helsinki Stroke Aging Memory Study (SAM) cohort. Stroke. 1997;28:785–792. 2. Tatemichi TK, Desmond DW, Stern Y, Paik M, Sano M, Bagiella E. Cognitive impairment after stroke: frequency, patterns, and relationship to functional activity. J Neurol Neurosurg Psychiatry. 1994;57:202–207. 3. Tatemichi TK, Paik M, Bagiella E, Desmond DW, Stern Y, Sano M, Hauser WA, Mayeux R. Risk of dementia after stroke in a hospitalized cohort: results of a longitudinal study. Neurology. 1994;44:1885–1891. 4. Kokmen E, Whisnant JP, O’Fallon WM, Chu CP, Beard CM. Dementia after ischemic stroke: a population-based study in Rochester, Minnesota (1960 –1984). Neurology. 1996;19:154 –159.

5. Desmond DW, Moroney JT, Sano M, Stern Y. The incidence of dementia after ischemic stroke: results of a longitudinal study. Stroke 2002;33:(this issue). 6. Neuropathology Group of the Medical Research Council Cognitive Function and Ageing Study. Pathological correlates of late-onset dementia in a multicentre, community-based population in England and Wales. Lancet. 2001;357:169 –175. 7. Snowdon DA, Greiner LH, Mortimer JA, Riley KP, Greiner PA, Markesbery WR. Brain infarction and the clinical expression of Alzheimer disease: the Nun Study. JAMA. 1997;277:813– 817. 8. Breteler MMB. Vascular risk factors for Alzheimer’s disease: an epidemiologic perspective. Neurobiol Aging. 2000;21:153–160. 9. Meyer JS, Rauch G, Rauch RA, Haque A. Risk factors for cerebral hypoperfusion, mild cognitive impairment, and dementia. Neurobiol Aging. 2000;21:161–169. 10. Shi J, Yang SH, Stubley L, Day AL, Simpkins JW. Hypoperfusion induces overexpression of ␤-amyloid precursor protein mRNA in a focal ischemic rodent model. Brain Res. 2000;853:1– 4. 11. Moroney JT, Bagiella E, Desmond DW, Paik MC, Stern Y, Tatemichi TK. Risk factors for incident dementia after stroke: role of hypoxic and ischemic disorders. Stroke. 1996;27:1283–1289. 12. Hénon H, Durieu I, Guerouaou D, Lebert F, Pasquier F, Leys D. Poststroke dementia: incidence and relationship to prestroke cognitive decline. Neurology. 2001;57:1216 –1222.

Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015

Incidence of Dementia After Ischemic Stroke: Results of a Longitudinal Study David W. Desmond, Joan T. Moroney, Mary Sano and Yaakov Stern Stroke. 2002;33:2254-2262 doi: 10.1161/01.STR.0000028235.91778.95 Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 2002 American Heart Association, Inc. All rights reserved. Print ISSN: 0039-2499. Online ISSN: 1524-4628

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://stroke.ahajournals.org/content/33/9/2254

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Stroke is online at: http://stroke.ahajournals.org//subscriptions/

Downloaded from http://stroke.ahajournals.org/ by guest on July 3, 2015