Distributional Impact of Brain Microbleeds on Global Cognitive Function in Adults Without Neurological Disorder

Transcription

1 Distributional Impact of Brain Microbleeds on Global Cognitive Function in Adults Without Neurological Disorder Yusuke Yakushiji, MD, PhD; Tomoyuki Noguchi, MD, PhD; Megumi Hara, MD, PhD; Masashi Nishihara, MD; Makoto Eriguchi, MD, PhD; Yusuke Nanri, MD; Masanori Nishiyama, MD; Tatsumi Hirotsu, MD; Junko Nakajima, MD; Yasuo Kuroda, MD, PhD; Hideo Hara, MD, PhD Background and Purpose Brain microbleeds (MBs) are considered to be associated with cognitive decline and can be pathologically and topographically classified as cerebral amyloid angiopathy-related (located in lobar regions) and hypertensive microangiopathy-related (located in deep regions). We examined whether different effects on global cognitive function might be seen with different distributions of MBs. Methods A total of 1279 adults without neurological disorders were studied prospectively. Subjects were divided into 4 groups: without-mbs group; lobar group; deep group; and with in both areas (diffuse group). The Mini-Mental State Examination was administered to determine global cognitive functions, with scores 27 regarded as subnormal. Results MBs were detected in 98 subjects (8%): 36 subjects (3%) classified as lobar group, 48 subjects (4%) as deep group, and 14 subjects (1%) as diffuse group. Subnormal scores were found in 76 subjects (5.9%), associated with age, education, hypertension, severe white matter hyperintensities, and distribution and number of MBs. In the final model of logistic regression analysis, the deep group (OR, 2.79; 95% CI, ) was associated with subnormal scores, whereas the lobar group (OR, 0.77; 95% CI, ) was not. Trend for the diffuse group did not reach the level of significance (OR, 5.01; 95% CI, ). These trends were also seen in analysis using another cut-off point for subnormal score. Scores for total Mini-Mental State Examination and attention and calculation were significantly lower in the deep group and the diffuse groups compared with the without-mbs group. Conclusions This Japanese cross-sectional study demonstrated that MB-related global cognitive dysfunction seems to occur based on hypertensive pathogenesis rather than on cerebral amyloid angiopathy. (Stroke. 2012;43: ) Key Words: microbleeds cognitive impairment hypertension cerebral amyloid angiopathy Brain microbleeds (MBs) on gradient-echo T2*-weighted magnetic resonance imaging (MRI), which are characterized histologically by the presence of hemosiderin around small vessels, are now accepted as a manifestation of cerebral small vessel pathologies, including hypertensive small vessel disease and cerebral amyloid angiopathy (CAA). 1,2 MBs are often detected in patients with stroke, 2,3 and are also reportedly associated with dementia disorders such as Alzheimer s disease, 4 and vascular dementia, 5 or memory loss, 6 as well as cognitive decline However, their pathological impacts on cognitive function remain uncertain. The pathological differences in MBs according to distribution is now well-known, with MBs in deep regions (deep MBs) considered to be associated with hypertensive microangiopathy, whereas strictly lobar MBs (lobar MBs) share risk factors with CAA. 12,13 According to this pathological and distributional concept, we assessed whether different effects on global cognitive function might be seen with different topographical distributions of MBs in a sample of independently living adults without neurological disorder. Materials and Methods Subjects This was a continuation of a previous cross-sectional study using the same system of health-screening tests for the brain. 9 A total of 1573 consecutive adults, who underwent health-screening tests of the brain in our center at their own expense between December 2005 and December 2010, were considered as potential subjects for this study. To be included, subjects had to meet the following criteria: age 20 years; no disability in instrumental activities of daily living; ability to make visits independently for current health-screening tests of the brain; and voluntary provision of written informed consent. Exclusion criteria were: inability to undergo cerebral MRI; history of neurological disorder or brain injury; and abnormality in neurological examination. All protocols were approved by the Institutional Review Board. Of the initial potential subjects, 203 individuals Received December 3, 2011; accepted April 5, From the Division of Neurology, Department of Internal Medicine (Y.Y., M.E., Y.N., Y.K., H.H.), Department of Radiology (T.N., M.N.), Department of Preventive Medicine (M.H.), Saga University Faculty of Medicine, Saga, Japan; and Yuai-Kai Oda Hospital (M.N., T.H., J.N.), Kashima, Japan. The online-only Data Supplement is available with this article at /-/DC1. Correspondence to Yusuke Yakushiji, Division of Neurology, Department of Internal Medicine, Saga University Faculty of Medicine, Nabeshima, Saga , Japan. yakushij@cc.saga-u.ac.jp 2012 American Heart Association, Inc. Stroke is available at DOI: /STROKEAHA

2 Yakushiji et al Microbleed Distributions and Cognition 1801 refused to enroll, 1 individual could not undergo MRI because of severe claustrophobia, and 12 individuals had a history of neurological disorder or brain injury with abnormalities seen on MRI. Among the remaining 1357 subjects, 7 subjects displaying motion artifacts on MRI, 1 subject showing numerous cavernous angiomas, and 47 subjects with incomplete data for analysis were excluded. Finally, 23 subjects with serum triglyceride levels 400 mg/100 ml were also excluded, because low-density lipoprotein cholesterol cannot be accurately estimated from the formula. 14 As a result, we were able to prospectively study 1279 subjects (596 men; median age, years; range, years). Baseline Assessment We recorded age, sex, duration of education, family history of stroke, smoking status, blood pressure (BP), presence of hypertension and diabetes mellitus, and levels of total fasting serum total cholesterol, low-density lipoprotein cholesterol, and high-density lipoprotein cholesterol as clinical characteristics. Hypertension was defined as systolic BP (SBP) 140 mm Hg and/or diastolic BP 90 mm Hg, or use of antihypertensive medication. Diabetes mellitus was defined as fasting serum glucose level 126 mg/dl, hemoglobin A1c levels 6.5%, or use of antidiabetic medication. Patients who were smokers at the time of analysis were classified as current smokers. Family history of stroke and duration of education were obtained from each subject. All subjects were examined by both a general physician and a certified neurosurgeon. Global cognitive function was assessed using the Mini-Mental State Examination (MMSE) 15 by a skilled nurse. In the primary analysis, total scores 27 were defined as subnormal in accordance with the latest MMSE guideline. 16 In an additional analysis, total scores 1.5 SD below the age-related mean were judged as subnormal. 9 Magnetic Resonance Imaging MRI was performed using a 1.5-T scanner (EXCELART Vantage, version 7.0; Toshiba Medical Systems). Gradient-echo T2*- weighted MRI, axial T1-weighted imaging, fluid-attenuated inversion recovery imaging, and fast spin-echo T2-weighted imaging were performed using the same section thickness, matrix, and parameters as described elsewhere. 9 MBs were defined on gradient-echo MRI as rounded areas of signal loss, 10 mm in diameter. Two investigators who were blinded to subject data reviewed the number and location of MBs. Symmetrical hypointensities in the globus pallidum caused by calcification and flow void artifacts of pial vessels were carefully excluded. Subject with MBs were divided into 3 groups by the location of MBs, 12 as follows. The lobar group showed strictly lobar MBs located in the cerebral cortices, subcortical white matter, or periventricular white matter. The deep group had strictly deep MBs located in the basal ganglia (BG); thalamus; white matter of the corpus callosum; internal, external, or extreme capsule; or infratentorially (brain stem or cerebellum). The diffuse group displayed diffuse MBs located in both lobar and deep regions. 17 White matter hyperintensities (WMH), periventricular hyperintensities, and lacunae were independently reviewed by 2 of the authors, who were blinded to subject data, and were rated as described elsewhere. 9,18 Values of interrater reliability for MRI findings, expressed as Cohen s kappa, were as follows: presence of MBs, 0.73; of lacunae, 0.57; of severe WMH, 0.58; and of severe periventricular hyperintensities, Each value of intrarater reliability for MRI findings was determined from 50 randomly selected scans that were scored twice, again expressed as Cohen s kappa. Each value was within the range of 0.65 to Using a computer-assisted processing system (Image J, version 1.38; National Institutes of Health) according to previous studies, 9,19 we calculated percentage brain value (% brain) as an index of cerebral atrophy. Statistics Statistical analysis was performed using the Statistical Analysis System version 9.1 (SAS Institute). We preliminarily examined the Figure. Comparison of the prevalence and distribution of MBs by age group. The prevalence of any-mbs differs among generations ( 2 test, P 0.001), as does the MBs distribution ( 2 test, P 0.012). Blue represents lobar group, orange represents deep group, and green represents diffuse group. relevance of the variables affecting MB distribution. In the primary tests, associations between MB distributions and subnormal MMSE score were investigated. Analyses were performed with the 2 groups (without-mbs versus with any-mbs) and the 4 groups (without-mbs group, lobar group, deep group, and diffuse group). To compare variables among groups, univariate analyses with the 2 test, Mann-Whitney U test, Kruskal-Wallis test, and 1-way analysis of variance using a Dunnett q or C test were performed, as appropriate. Variables with P 0.20 on univariate analysis, in addition to age and sex, were entered into multiple logistic regression analyses. Values of P 0.05 were considered statistically significant. Results MBs were detected in 98 subjects (7.7%). In all, 221 MBs (mean, 0.2; range, 0 26) were identified. Prevalence was 5.2% for a single MB, and 2.5% for 2 MBs. A total of 107 MBs were located in lobar areas (87 MBs in cerebral cortices, 17 MBs in subcortical white matter, and 3 MBs in periventricular white matter), of which 34 MBs (31.8%) were located in the frontal lobe, 17 MBs (15.9%) in the parietal lobe, 31 MBs (29.0%) in the temporal lobe, and 25 MBs (23.3%) in the occipital lobe. A total of 114 MBs were located in deep areas (50 MBs in BG, 37 MBs in thalamus, 3 MBs in white matter of the corpus callosum, 4 MBs in the internal capsule, 3 MBs in the external/extreme capsule, 10 MBs in the brain stem, and 7 MBs in the cerebellum). With regard to the distributions of MBs, 1181 subjects (92.3%) were classified as without-mbs group, 36 subjects (2.8%) as lobar group, 48 subjects (3.8%) as deep group, and 14 subjects (1.1%) as diffuse group. Figure shows a comparison of the prevalence and distribution of MBs by age group. Prevalence and distributions of MBs differed among age groups (P and P 0.012, respectively). Each distribution type showed a different distribution by age group as follows: deep, from age 39 years; lobar, from age 40 years; and diffuse, from age 50 years. (Table S1). For the total subjects, mean MMSE score was (range, 20 30). Scores 27 were found in 76 subjects (5.9%; score 26, n 40; 25, n 17; 24, n 8; 23, n 9; 22, n 1; and 20, n 1). As for the results of the preliminary tests, differences in clinical characteristics and MRI findings among the groups

3 1802 Stroke July 2012 Table 1. Differences in Clinical and MRI Features Between Normal and Subnormal (<27) MMSE Score Groups Variables Normal n 1203 (94.1%) Subnormal n 76 (5.9%) P Value Age, median (IQR), y 58 (50 65) 63 (58 67) Sex, men, n (%) 562 (47) 34 (45) Smoking, n (%) 194 (16) 14 (18) Family history of stroke, n (%) 365 (30) 20 (26) Education, y (IQR) 12 (12 14) 12 (9 12) Hypertension, n (%) 410 (34) 35 (46) SBP, median (IQR), mm Hg 124 ( ) 130 ( ) DBP, median (IQR), mm Hg 78 (70 84) 78 (70 84) Diabetes mellitus, n (%) 98 (8) 11 (15) T-cho, median (IQR), mg/dl ( ) ( ) LDL-C, median (IQR), mg/dl ( ) ( ) HDL-C, median (IQR), mg/dl 60.0 ( ) 58.5 ( ) MR findings Presence of lacunae, n (%) 56 (5) 2 (3) Number of lacunae, median (IQR) 0 (0 0) 0 (0 0) Mean (SD) 0.08 (0.44) 0.03 (1.61)... Presence of MBs, n (%) 84 (7) 14 (18) Number of MBs, median (IQR) 0 (0 0) 0 (0 0) Mean (SD) 0.14 (0.93) 0.64 (3.08)... Distribution of MBs Lobar 34 (3) 2 (3) Deep 40 (3) 8 (10) Diffuse 10 (1) 4 (5) Severe WMH, n (%) 158 (13) 19 (25) Severe PVH, n (%) 35 (3) 1 (1) % Brain, median (IQR), % 84.8 ( ) 84.9 ( ) MRI indicates magnetic resonance imaging; IQR, interquartile range; SBP, systolic blood pressure; DBP, diastolic blood pressure; T-cho, total cholesterol; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; MR, magnetic resonance; MB, microbleed; WMH, white matter hyperintensities; PVH, periventricular hyperintensities. Continuous variables were compared using the Mann-Whitney U test, and are presented as medians with IQRs. Mean with SD is also presented in number of lacunae and MBs. The Chi-square test was used for frequency comparisons. for MBs distribution are shown in S2. In the comparison between without-mbs and any-mbs group, there were differences in age, BPs, % brain, and prevalences of hypertension, diabetes mellitus, lacunae, and severe WMH and periventricular hyperintensities. In the comparison among 4 groups, there were additional differences in sex, high-density lipoprotein cholesterol, and prevalences of current smoker and family history of stroke. The number of MBs was significantly higher in the diffuse group than in the lobar group or the deep group. No difference in number of MBs was apparent between the lobar group and the deep group. In the diffuse group, median (interquartile range) numbers of MBs were 1.5 ( ) in the lobar regions and 1.5 ( ) in the deep regions (data were not shown). The results of logistic regression analyses for factors associated with each distributional pattern of MBs are shown in Table S3. SBP, lacunae, and severe WMH had relevance to subjects with any-mbs, as well as the deep group, and were more pronounced for the diffuse group. Even if hypertension or BP were entered into the logistic regression analysis instead of SBP, similar results were obtained (data not shown). As for the results of the primary tests, differences in variables between normal and subnormal MMSE score ( 27) groups are shown in Table 1. Subnormal scores were associated with age; duration of education; hypertension; SBP; presence, number, and distribution of MBs; and severe WMH. To assess the primary hypothesis, logistic regression analyses were performed (Table 2). As a strong correlation existed between hypertension and SBP (Spearman correlation coefficient (ñ), 0.66) and SBP displayed a lower probability value than did hypertension, SBP was used for multivariate analysis. Models were adjusted for age, sex, and clinical variables (Model 1); then for the variables entered into Model 1, and severe WMH (Model 2); and then for variables entered into Model 2, and number of MBs (Model 3). Results from Models 1 and 2 showed that, compared with the without-mbs

4 Yakushiji et al Microbleed Distributions and Cognition 1803 Table 2. Logistic Regression Analysis of MBs Distributional Pattern Relevant to Subnormal MMSE Score (<27) Groups Crude Model 1 Model 2 Model 3 OR (95% CI) P Value OR (95% CI) P Value OR (95% CI) P Value OR (95% CI) P Value Without-MBs 1.00 ref ref ref ref... Any-MBs 3.01 ( ) ( ) ( ) ( ) Lobar group 1.06 ( ) ( ) ( ) ( ) Deep group 3.61 ( ) ( ) ( ) ( ) Diffuse group 7.22 ( ) ( ) ( ) ( ) Crude: unadjusted. Model 1: adjusted for age (/10 y), sex, duration of education, SBP (/SD mm Hg), diabetes mellitus, and high-density lipoprotein cholesterol (/SD mg/dl). Model 2: adjusted for the variables entered into model 1, and severe WMH. Model 3: adjusted for the variables entered into model 2, and number of MBs. MB indicates microbleed; SBP, systolic blood pressure; WMH, white matter hyperintensities. group, presence of any-mbs was significantly associated with subnormal score. These associations were more pronounced for the deep group and the diffuse group, but were not seen in the lobar group. In Model 3, the deep group still displayed significant associations with subnormal score. In that model, similar trends were also seen in the any-mbs and diffuse groups, but did not reach the level of statistical significance. Even if the presence of lacunae, which has been considered to be associated with cognitive decline (but did not show relevance to subnormal scores on univariate analysis in the present study), was additionally entered into Models 2 and 3, no changes in these trends were identified (Table S4). The additional analysis, in which MMSE scores 1.5 SD below the mean for each given age were defined as subnormal (n 93, 7.3%), demonstrated the same trends (Table S5 6). Table 3 compares MMSE total and subscores among groups. In comparison between the 2 groups, scores for total MMSE and attention and calculation were significantly lower with any-mbs. Among the 4 groups, such scores were significantly lower with the deep group and the diffuse groups compared with the without-mbs group, and lowest in the diffuse group. All scores were similar between the lobar and without-mbs groups. Table 3. Discussion Recent cross-sectional studies investigated the association between MB location and cognitive dysfunction. 10,11 However, these studies included subjects with dementia disorders 10 and previous stroke, 11 thus including the possibility of substantial bias in terms of coexisting pathogeneses for these diseases. The present study investigated the distributional impact of MBs on global cognitive function in adults without neurological disorder. All subjects were living independently without neurological disorder and displayed normal results on neurological examination. The result of significant differences in the impact of MBs on global cognitive function according to the distributional pattern is consistent with our hypothesis. That is, the effect of deep MBs on global cognitive function appears independently of neurological disorder and coexisting risk factors for stroke or dementia. The significant associations in this study between deep MBs and cerebrovascular risk factors are inconsistent with previous studies, 12,13,17 and suggest that deep MBs might be closely attributable to hypertensive microangiopathy. Conversely, no risk factors other than severe periventricular hyperintensities were associated with lobar MBs, which are considered a consequence of underlying CAA. 20,21 Such reduced relevance of lobar MBs to cerebrovascular risk factors is in line with previous studies. 12,13,17 Thus, our main result that subnormal MMSE scores are significantly correlated with deep MBs, but not with lobar MBs, suggests that MB-related global cognitive dysfunction seems to occur based on hypertensive pathogenesis, rather than CAA; this Comparison of MMSE Total Scores and Subscores by Topographical Distribution of MBs Without-MBs Any MBs* Lobar Group Deep Group Diffuse Group n Variables Total score (1.25) (1.80) (1.42) (1.87) (2.21) Orientation 9.99 (0.12) 9.97 (1.73) 9.97 (0.17) 9.98 (0.14) 9.93 (0.27) Immediate recall 3.00 (0.05) 3.00 (0.00) 3.00 (0.00) 3.00 (0.00) 3.00 (0.00) Attention and calculation 4.75 (0.89) 4.43 (1.34) 4.75 (0.94) 4.31 (1.46) 4.00 (1.66) Delayed recall 2.88 (0.42) 2.84 (0.47) 2.89 (0.40) 2.85 (0.46) 2.64 (0.63) Language 8.92 (0.32) 8.87 (0.40) 8.89 (0.40) 8.88 (0.39) 8.79 (0.43) Variables are presented as mean with SD. *Comparison of 2 groups (without-mbs and any-mbs) with Mann-Whitney U test. Comparison of 4 groups (without-mbs group, lobar group, deep group, and diffuse group) with 1-way analysis of variance with Dunnett s test. P 0.05 vs without-mbs group.

5 1804 Stroke July 2012 implies that control of hypertension can reduce MB-related cognitive dysfunction. A close pathological relationship between lobar MBs and CAA, where deposition of A -protein in the vascular media and adventitia leads to loss of integrity of the vessel wall with resulting brain hemorrhage, 22 has been demonstrated with assessments of ApoE genotype in the general population, 12,13 as well as with pathology in brains with Alzheimer s disease. 23 CAA contributes to cognitive decline in Alzheimer s disease, 24 but lobar MBs showed no relevance to global cognitive dysfunction in our analyses with most middle-aged adults. Such discrepancies might be explained largely by the hypothesis that abnormal processing of A -protein may be 1 of the initiating events in the development of Alzheimer s disease, including neuronal dysfunction and neurodegeneration, becoming key pathological processes later in the disease process. 25,26 Diffuse MBs, which were more frequently seen in older persons in this study, had more pronounced relevance to cerebrovascular risk factors, indicating that diffuse MBs might represent chronic extension of deep MBs with or without CAA. 17 The associations between diffuse MBs and global cognitive dysfunction seem to be pronounced, but did not reach the significant level in final model additionally adjusted for number of MBs; this implies that the effect of diffuse MBs on cognition might result partially from increased number of MBs with a more severe hypertensive pathogenesis. However, diffuse MBs could have a mixed pathogenesis involving vascular risk factors and CAA. 1 Therefore, the tendency to have severe cognitive decline in the diffuse group might serve as an additional argument for the notion that both hypertensive microangiopathy and CAA resulting from amyloid cascade act synergistically to cause dementia. 26 The mechanisms underlying pathological associations between hypertensive MBs and global cognitive dysfunction, particularly attention and calculation, remain unclear. In previous studies, 8,9 as well as in the current study, that deep MBs are frequently located in the BG may offer some hints to explaining such associations. Cognitive dysfunction is also common in subjects with BG disease, as represented by Parkinson s disease. Cognitive deficits in the early stages of Parkinson s disease are characterized by executive dysfunction, including impairments in attention 27 and calculation, 28 reflecting involvement of the BG and frontal-subcortical circuits. The predominance of deep MBs in the BG may thus cause cognitive dysfunction resulting from primary or secondary damage to cholinergic pathways in the frontalsubcortical circuits. Several limitations must be considered in the interpretation of this study. First, selection bias might have been present. Given that our subjects were adults who sought health screening tests of the brain at their own expense, a selection bias may have existed toward individuals who were relatively affluent or had a high degree of concern regarding their own health, and might have been affected by differences in educational background and socioeconomic status, which are considered risk factors for cardiovascular disease. 29 Second, we used the MMSE as the sole method to assess cognitive function. This measure reportedly offers relatively low sensitivity for the assessment of cognitive function in patients with cerebral small vessel disease. 30 However, the MMSE is used worldwide as a cognitive assessment tool, and the relatively large sample in this study would reduce the impact of bias. Third, this study included only a relatively small sample of individuals age 70 years old (7.8%), although this age group would have been expected to show the greatest prevalence of CAA-related MBs. The present analyses with mostly middle-aged adults might thus remain insufficient to conclude whether lobar MBs are associated with cognitive decline among individuals without neurological disorders. Finally, the ethnic background of the study population should be considered. Because Asian populations display more prominent cerebrovascular pathology compared with non- Asian populations, additional investigation may be needed to clarify whether our findings are more generally applicable. Some discussion is warranted regarding the late-breaking results from the Rotterdam Scan Study, 31 which have indicated that presence of numerous MBs, particularly in a strictly lobar location, was associated with cognitive decline in a general population. Some methodological differences exist between that study and our own, such as the parameters of gradient-echo MRI, statistical procedures, and the abovementioned limitations. Of those, differences in ethnic background (Asian versus non-asian) might be a central contributor to the contradictory findings. In fact, among the subjects with MBs in these studies, the proportions of subjects in the (strict) lobar group and the other (deep group and diffuse group, as so-called subjects with deep or infratentorial MBs in the Rotterdam Scan Study) were completely reversed, at 68% and 32% in the Rotterdam Scan Study, and 37% and 63% in this study, respectively. The dominance of distributional patterns for MBs according to the ethnic backgrounds of each study sample might thus have influenced the cognitive results. Conclusions In conclusion, this study offers evidence that the impacts of MBs on global cognitive function differ according to the topographical distribution in adults without neurological disorder, indicating that, in the Japanese population, MB-related global cognitive dysfunction seems to occur based on hypertensive pathogenesis, rather than based on CAA. Acknowledgments We wish to thank Mai Miyamoto, Mitsuko Uematsu, Tomomi Ozaki, Dr Masamichi Oda, and the medical staff of Yuai-Kai Oda Hospital for their help in conducting medical interviews, acquiring imaging data, and inputting data. Sources of Funding This work was partially supported by a Grant-in-Aid for University Reform 2006 from the Ministry of Education, Culture, Sports, Science and Technology of Japan. None. Disclosures

6 Yakushiji et al Microbleed Distributions and Cognition 1805 References 1. Fazekas F, Kleinert R, Roob G, Kleinert G, Kapeller P, Schmidt R, et al. Histopathologic analysis of foci of signal loss on gradient-echo T2*- weighted MR images in patients with spontaneous intracerebral hemorrhage: evidence of microangiopathy-related microbleeds. AJNR Am J Neuroradiol. 1999;20: Tanaka A, Ueno Y, Nakayama Y, Takano K, Takebayashi S. Small chronic hemorrhages and ischemic lesions in association with spontaneous intracerebral hematomas. Stroke. 1999;30: Kwa VI, Franke CL, Verbeeten B Jr, Stam J. Silent intracerebral microhemorrhages in patients with ischemic stroke. Amsterdam Vascular Medicine Group. Ann Neurol. 1998;44: Hanyu H, Tanaka Y, Shimizu S, Takasaki M, Abe K. Cerebral microbleeds in Alzheimer s disease. J Neurol. 2003;250: Won Seo S, Hwa Lee B, Kim EJ, Chin J, Sun Cho Y, Yoon U, et al. Clinical significance of microbleeds in subcortical vascular dementia. Stroke. 2007;38: Cordonnier C, van der Flier WM, Sluimer JD, Leys D, Barkhof F, Scheltens P. Prevalence and severity of microbleeds in a memory clinic setting. Neurology. 2006;66: Greenberg SM, Eng JA, Ning M, Smith EE, Rosand J. Hemorrhage burden predicts recurrent intracerebral hemorrhage after lobar hemorrhage. Stroke. 2004;35: Werring DJ, Frazer DW, Coward LJ, Losseff NA, Watt H, Cipolotti L, et al. Cognitive dysfunction in patients with cerebral microbleeds on T2*- weighted gradient-echo MRI. Brain. 2004;127: Yakushiji Y, Nishiyama M, Yakushiji S, Hirotsu T, Uchino A, Nakajima J, et al. Brain microbleeds and global cognitive function in adults without neurological disorder. Stroke. 2008;39: Qiu C, Cotch MF, Sigurdsson S, Jonsson PV, Jonsdottir MK, Sveinbjrnsdottir S, et al. Cerebral microbleeds, retinopathy, and dementia: the AGES-Reykjavik Study. Neurology. 2010;75: van Es AC, van der Grond J, de Craen AJ, Westendorp RG, Bollen EL, Blauw GJ, et al. Cerebral microbleeds and cognitive functioning in the PROSPER study. Neurology. 2011;77: Vernooij MW, van der Lugt A, Ikram MA, Wielopolski PA, Niessen WJ, Hofman A, et al. Prevalence and risk factors of cerebral microbleeds: the Rotterdam Scan Study. Neurology. 2008;70: Poels MM, Vernooij MW, Ikram MA, Hofman A, Krestin GP, van der Lugt A, et al. Prevalence and risk factors of cerebral microbleeds: an update of the Rotterdam scan study. Stroke. 2010;41:S103 S Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18: 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: Folstein MF, Folstein SE, Fanjiang G. Mini-mental state examination: clinical guide. Lutz: Psychological Assessment Resources, Inc; Yakushiji Y, Yokota C, Yamada N, Kuroda Y, Minematsu K. Clinical characteristics by topographical distribution of brain microbleeds, with a particular emphasis on diffuse microbleeds. J Stroke Cerebrovasc Dis. 2011;20: Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalities at 1.5 T in Alzheimer s dementia and normal aging. AJR Am J Roentgenol. 1987;149: Koga H, Yuzuriha T, Yao H, Endo K, Hiejima S, Takashima Y, et al. Quantitative MRI findings and cognitive impairment among community dwelling elderly subjects. J Neurol Neurosurg Psychiatry. 2002;72: Rosand J, Muzikansky A, Kumar A, Wisco JJ, Smith EE, Betensky RA, et al. Spatial clustering of hemorrhages in probable cerebral amyloid angiopathy. Ann Neurol. 2005;58: Mesker DJ, Poels MM, Ikram MA, Vernooij MW, Hofman A, Vrooman HA, et al. Lobar distribution of cerebral microbleeds: the Rotterdam Scan Study. Arch Neurol. 2011;68: Smith EE, Greenberg SM. Beta-amyloid, blood vessels, and brain function. Stroke. 2009;40: Schrag M, McAuley G, Pomakian J, Jiffry A, Tung S, Mueller C, et al. Correlation of hypointensities in susceptibility-weighted images to tissue histology in dementia patients with cerebral amyloid angiopathy: a postmortem MRI study. Acta Neuropathol. 2010;119: Pfeifer LA, White LR, Ross GW, Petrovitch H, Launer LJ. Cerebral amyloid angiopathy and cognitive function: the HAAS autopsy study. Neurology. 2002;58: Jack CR Jr, Knopman DS, Jagust WJ, Shaw LM, Aisen PS, Weiner MW, et al. Hypothetical model of dynamic biomarkers of the Alzheimer s pathological cascade. Lancet Neurol. 2010;9: Cordonnier C, van der Flier WM. Brain microbleeds and Alzheimer s disease: innocent observation or key player? Brain. 2011;134: Muslimovic D, Post B, Speelman JD, Schmand B. Cognitive profile of patients with newly diagnosed Parkinson disease. Neurology. 2005;65: Bronnick K, Emre M, Lane R, Tekin S, Aarsland D. Profile of cognitive impairment in dementia associated with Parkinson s disease compared with Alzheimer s disease. J Neurol Neurosurg Psychiatry. 2007;78: Winkleby MA, Kraemer HC, Ahn DK, Varady AN. Ethnic and socioeconomic differences in cardiovascular disease risk factors: findings for women from the Third National Health and Nutrition Examination Survey, JAMA. 1998;280: O Sullivan M, Morris RG, Markus HS. Brief cognitive assessment for patients with cerebral small vessel disease. J Neurol Neurosurg Psychiatry. 2005;76: Poels MM, Ikram MA, van der Lugt A, Hofman A, Niessen WJ, Krestin GP, et al. Cerebral microbleeds are associated with worse cognitive function: the Rotterdam Scan Study. Neurology. 2012;78: