Skip main navigation

Incidence of Cerebral Microbleeds in the General Population

The Rotterdam Scan Study
Originally publishedhttps://doi.org/10.1161/STROKEAHA.110.607184Stroke. 2011;42:656–661

Abstract

Background and Purpose—

Cerebral microbleeds are frequently seen in the general elderly population, but it is unknown at what rate they occur with aging and whether once present can disappear over time.

Methods—

As part of the Rotterdam Scan Study, 831 persons (mean age, 68.5 years) underwent repeated brain MRI with a mean interval of 3.4 years. We assessed determinants of incident microbleeds in relation to their location with multiple logistic regressions.

Results—

Overall prevalence of microbleeds increased from 24.4% at baseline to 28.0% at follow-up. Eighty-five persons (10.2%) developed new microbleeds. Microbleeds at baseline predicted development of new microbleeds (OR, 5.38; 95% CI, 3.34 to 8.67). In only 6 persons with microbleeds at baseline, fewer microbleeds were present at the follow-up examination. Cardiovascular risk factors, presence of lacunar infarcts, and larger white matter lesion volume at baseline were all associated with incident deep or infratentorial microbleeds, whereas people with the apolipoprotein E ϵ4/ϵ4 genotype or larger white matter lesion volume had a higher risk of incident strictly lobar microbleeds.

Conclusions—

Incidence of microbleeds in the general population over a 3-year interval was substantial and microbleeds rarely disappeared. Risk factors for incident microbleeds were similar to those for prevalent microbleeds and differed according to microbleed location. These results support the assessment of microbleeds on T2*-weighted MRI as a possible marker of both cerebral amyloid angiopathy and hypertensive vasculopathy progression.

Cerebral microbleeds (CMBs) are hypointense lesions seen on T2*-weighted gradient echo MRI that may be indicative of past microhemorrhages.1 Although their prognosis is yet not completely understood, several clinical studies suggested that CMBs might predict future risk of (recurrent) stroke.2,3 Cross-sectional studies have reported on the prevalence and determinants of microbleeds in patients with stroke as well as in the general elderly population.47 Although numbers varied widely across studies due to differences in MRI technique used,8 overall, the prevalence of CMBs both in patients with stroke and in the elderly was high and increased with age. Most consistent risk factors were age and hypertension, but markers of cerebral small vessel disease were found to be related as well.47 In addition, accumulating evidence suggests that the spatial distribution of CMBs may reflect specific underlying vascular pathological changes, in particular cerebral amyloid angiopathy or hypertensive vasculopathy.8

Progression of these underlying vascular pathological changes may be reflected by development of new CMBs. It is unknown, however, at what rate microbleeds occur with aging in community-dwelling elderly and whether once present can disappear over time. Few longitudinal data on CMBs development exist, and existing studies were all small and performed in specific subgroups such as memory clinic patients or patients with cerebral amyloid angiopathy.914 To date, there has been no longitudinal study exploring the incidence of microbleeds and its determinants in the general population.

In the population-based Rotterdam Scan Study, we sought to investigate the incidence of CMBs and the location of these new microbleeds. Furthermore, we studied the determinants of incident CMBs in relation to their location.

Methods

Participants

From 2005 to 2006, 1062 nondemented people (at that time all ≥60 years) underwent a baseline examination that included a brain MRI scan as part of the Rotterdam Scan Study.5,15 From October 2008 to January 2010, these participants were reinvited for follow-up MRI. The Institutional Review Board approved the study.

Of the 1062 participants at baseline, 80 people were not eligible to participate in the second MRI examination (dead, n=54; new MRI contraindications [eg, pacemaker], n=10; institutionalized, n=9; untraceable, n=7). Of 982 eligibles, 848 participated and gave written informed consent (response rate 85%). Due to physical problems (eg, backache), imaging could not be completed in 14 individuals. Of 834 complete MRI examinations, 3 scans were excluded because of severe artifacts, leaving 831 persons with complete and reliable baseline and follow-up MRI examinations.

Brain MRI

We performed an identical MRI protocol on the same 1.5-T scanner (GE Healthcare, Milwaukee, WI) at both time points. A 3-dimensional T2*-weighted gradient-recalled echo sequence was used for microbleed detection.16 The other sequences in the protocol consisted of a T1-weighted sequence, a proton density-weighted sequence, and a fluid-attenuated inversion recovery sequence.5 No scanner software upgrades or hardware alterations were applied during the study period to ensure comparability of scan data over time.

Rating of CMBs

At both time points, all 3-dimensional T2*-weighted gradient-recalled echo scans were reviewed by 1 of 5 trained raters who recorded the presence, number, and location of microbleeds. All raters were blinded to the other MRI sequences, clinical data, and apolipoprotein E (APOE) genotyping; and the 3-dimensional T2*-weighted gradient-recalled echo scan did not reveal the presence of infarcts or white matter lesions. Microbleeds were defined as focal areas of very low signal intensity.8 Signal voids caused by sulcal vessels, symmetrical calcifications in the basal ganglia, choroid plexus and pineal calcifications, and signal averaging from bone were excluded.8 Intraobserver (n=500, 1 rater) and interobserver (n=300) reliabilities were κ=0.87 and κ=0.85, which corresponds to very good agreement. CMBs were categorized into 1 of 3 locations: lobar, deep, or infratentorial.5 Scans of subjects rated positive for CMBs at at least 1 of 2 time points were included in a side-by-side comparison (M.M.P. and M.W.V.) blinded to the time point of the scans to assess the final number and location of microbleeds in each scan.

Cerebrovascular Disease on MRI

Infarcts were rated on fluid-attenuated inversion recovery, proton density-weighted, and T1-weighted sequences at baseline by the same raters who had scored CMBs according to criteria described previously.5 White matter lesion volume was quantified with a validated tissue classification technique.17

Cardiovascular Risk Factors and APOE Genotyping

Cardiovascular risk factors at baseline were examined by interview and laboratory and physical examination as previously described.5 APOE genotyping was performed on coded genomic DNA samples.18 The distributions of APOE genotype and allele frequencies in this population were in Hardy-Weinberg equilibrium.

Data Analysis

We tested differences in baseline characteristics between persons who participated in both examinations and persons who refused or were ineligible to participate in the second MRI using analysis of covariance adjusted for age and sex.

Incidence of CMBs was calculated in 10-year age strata and separately in strata of presence of microbleeds at baseline MRI. Subsequently, we made categories for “strictly lobar incident microbleeds” (persons with ≥1 new microbleeds restricted to a lobar location) and “deep or infratentorial incident microbleeds” (persons with ≥1 new microbleeds in a deep or infratentorial location with or without concomitant lobar microbleeds)5 and assessed whether the incidence differed according to microbleed location at baseline using multiple logistic regressions.

Next, we assessed the relation of vascular risk factors, APOE allele status, and cerebrovascular disease at baseline to CMB incidence with multiple logistic regressions. These analyses were also performed according to microbleed location.

All regression analyses were adjusted for age, sex, and scan interval. To examine independency of risk factors, we used multivariable modeling, including lacunar infarcts, white matter lesion volume, and vascular risk factors.

Analyses were performed using the statistical package SPSS 15.0 (SPSS Inc, Chicago, IL).

Results

Table 1 shows the characteristics of all 1062 participants at baseline. Persons who participated in both MRI examinations were younger compared with persons who participated in the first examination only. Furthermore, participants refusing second MRI were more often APOE ϵ2 carriers, whereas ineligible persons had higher cholesterol, more lacunar infarcts, and a higher white matter lesion volume at baseline, even when differences in age and sex were taken into account.

Table 1. Baseline Characteristics of Participants Who Had a Second MRI Assessment and for Those Who Refused or Were Ineligible

Participants With a Second MRI (N = 831)Participants Who Refused a Second MRI (N=148)Participants Ineligible for a Second MRI (N = 83)
Age, years, mean±SD68.5±6.372.5±8.5*76.1 (8.4)*
Women, no. (%)418 (50.3)83 (56.1)42 (50.6)
Systolic blood pressure, mean±SD143.8±18.1146.1±19.8147.0±22.0
Diastolic blood pressure, mean±SD80.7±10.478.7±10.177.6±9.3
Pulse pressure, mean±SD63.2±15.667.5±17.569.4±21.1
Hypertension
    Mild, no. (%)426 (51.6)76 (52.1)39 (49.4)
    Severe, no. (%)159 (19.2)32 (21.9)23 (29.1)
Smoking
    Never, no. (%)241 (29.3)34 (23.8)15 (19.0)
    Past, no. (%)343 (41.7)63 (44.1)39 (49.4)
    Current, no. (%)239 (29.0)46 (32.1)25 (31.6)
Diabetes mellitus, no. (%)68 (8.3)19 (13.4)8 (10.4)
Serum total cholesterol, mmol/L, mean±SD5.69±0.975.70±0.945.43±0.87*
APOE ϵ2 allele carrier, no. (%)115 (14.6)29 (20.7)*11 (15.3)
APOE ϵ4 allele carrier, no. (%)206 (26.1)44 (31.4)23 (31.9)
Cortical infarct on baseline MRI, no. (%)21 (2.5)9 (6.1)7 (8.4)
Lacunar infarct on baseline MRI, no. (%)64 (7.7)12 (8.1)17 (20.5)*
Subcortical infarct on baseline MRI, no. (%)1 (0.1)1 (0.7)0 (0.0)
White matter lesions on baseline MRI, mL, median (interquartile range)3.8 (2.2–7.4)5.2 (2.5–13.3)8.8 (4.6–18.3)*

*Age- and sex-adjusted mean, median, or percentage is significantly different (P<0.05) from participants with a second MRI.

Participants refusing a second MRI (N=134) and participants with no (complete) examination (N=14).

Ineligible participants (N=80) and participants with ungradeable MRI (N=3).

Data are missing for blood pressure/hypertension (N=11), smoking (N=17), diabetes mellitus (N=24), serum cholesterol (N=17), APOE genotype (N=62), and white matter lesions (N=18).

Mean interval between the 2 MRI assessments was 3.4 years (range, 2.3 to 4.5 years). During this period, overall prevalence of microbleeds increased from 24.4% to 28.0%. Eighty-five of the 831 participants (10.2%) developed new microbleeds on MRI, of whom 38 (4.6%) had multiple new microbleeds (eg, Figure). CMB incidence increased with age from 7.6% in persons aged 60 to 69 years to 18.6% in participants >80 years and no significant differences were observed between sexes. Among persons with new microbleeds, 60% had incident strictly lobar CMBs, whereas 40% had incident deep or infratentorial CMBs (Table 2). Significantly more participants with microbleeds at baseline developed new CMBs during the time interval compared with participants without CMBs at baseline (25.1% versus 5.1%, Table 2; OR, 5.38; 95% CI, 3.34 to 8.67; Table 3). This risk was even higher for persons with multiple CMBs at baseline (OR, 7.15; 95% CI, 4.11 to 12.44). Moreover, CMB location at baseline strongly predicted the location of new CMBs (Table 3).

Table 2. Incidence of CMBs in 10-Year Age Groups in Strata of the Presence of CMBs on Baseline MRI

Age Range, YearsOverall
No Prevalent CMBs on Baseline MRI
Prevalent CMBs on Baseline MRI
No. of PersonsIncident CMBs, No. (%)Multiple Incident CMBs, No. (%)No. of PersonsIncident CMBs, No. (%)Multiple Incident CMBs, No. (%)No. of PersonsIncident CMBs, No. (%)Multiple Incident CMBs, No. (%)
60–6958044 (7.6)21 (3.6)46623 (4.9)11 (2.4)11421 (18.4)10 (8.8)
70–7919230 (15.6)11 (5.7)1248 (6.5)2 (1.6)6822 (32.4)9 (13.2)
80–975911 (18.6)6 (10.2)383 (7.9)0 (0.0)218 (38.1)6 (28.6)
Total83185 (10.2)38 (4.6)62834 (5.4)13 (2.1)20351 (25.1)25 (12.3)

Table 3. Presence of CMBs at Baseline and Incidence of CMBs

Risk of Incident CMBs*
All Incident CMBs (N=85)Strictly Lobar Incident CMBs (N=51)Deep or Infratentorial Incident CMBs (N=34)
CMBs at baseline5.38 (3.34–8.67)4.99 (2.75–9.06)5.97 (2.89–12.28)
Multiple CMBs at baseline7.15 (4.11–12.44)5.69 (2.85–11.36)9.69 (4.48–20.97)
One strictly lobar CMB at baseline1.55 (0.83–2.91)1.66 (0.77–3.57)1.37 (0.51–3.67)
Multiple strictly lobar CMBs at baseline4.32 (2.05–9.11)6.53 (2.86–14.90)1.48 (0.32–6.74)
One deep or infratentorial CMB at baseline2.76 (1.13–6.76)3.24 (1.15–9.12)1.96 (0.44–8.77)
Multiple deep or infratentorial CMBs at baseline7.25 (3.62–14.51)3.02 (1.06–8.56)16.53 (7.16–38.19)

*All values are ORs with 95% CIs, adjusted for age, sex, and scan interval.

With or without concomitant (new) lobar CMBs.

Figure.

Figure. Example of incident microbleeds on 3-dimensional T2*-weighted gradient-recalled echo MRI during a scan interval of 3 years. Arrows indicate new microbleeds on the follow-up scan. A, Baseline scan 2006. B, Follow-up scan 2009.

Only in 6 persons (3% of participants with CMBs at baseline) we found less CMBs at follow-up compared with the baseline MRI scan; these persons were added to the group of no incident microbleeds. In 4 of these 6 persons, 1 CMB was scored at baseline, whereas we assessed no CMBs at follow-up; in 1 person CMB count decreased from 2 to 1; in 1 person 11 CMBs were scored on the baseline scan and only 6 at follow-up. Furthermore, there were another 6 participants in whom not all CMBs that were seen at baseline were recognized at the follow-up scan, yet the total number of CMBs did not decrease over time. Except for 1, all of these had >5 CMBs at baseline and showed an increase in total CMB count at follow-up. In the overall population, we scored 258 new microbleeds, whereas only 18 microbleeds seemed to disappear over time.

Older age, high systolic blood pressure, high pulse pressure, and hypertension at baseline were all associated with development of new microbleeds (Table 4). When stratified according to location, we found an association of high systolic blood pressure, high pulse pressure, and hypertension with incident deep or infratentorial microbleeds but not with new lobar CMBs. With increasing serum total cholesterol, the incidence of microbleeds in a deep or infratentorial location decreased (Table 4).

Table 4. Cardiovascular Risk Factors, APOE Allele Status, and Incidence of CMBs

Risk of Incident CMBs*
All Incident CMBs (N=85)Strictly Lobar Incident CMBs (N=51)Deep or Infratentorial Incident CMBs§ (N=34)
Age per year1.06 (1.03–1.10)1.07 (1.03–1.11)1.05 (1.00–1.10)
Women vs men0.69 (0.43–1.09)0.69 (0.39–1.24)0.68 (0.34–1.37)
Systolic BP per SD increase1.29 (1.03–1.61)1.18 (0.89–1.56)1.43 (1.03–1.98)
Diastolic BP per SD increase1.04 (0.83–1.32)0.95 (0.70–1.27)1.20 (0.84–1.71)
Pulse pressure per SD increase1.33 (1.06–1.67)1.27 (0.95–1.70)1.38 (0.99–1.93)
Hypertension
    Mild vs none1.82 (0.97–3.40)1.68 (0.80–3.53)2.16 (0.71–6.60)
    Severe vs none2.57 (1.27–5.17)1.44 (0.58–3.60)5.39 (1.73–16.83)
Smoking
    Past vs never0.98 (0.56–1.72)0.97 (0.47–2.02)0.96 (0.42–2.19)
    Current vs never0.78 (0.41–1.51)0.92 (0.40–2.10)0.59 (0.21–1.63)
Diabetes mellitus, yes vs no1.22 (0.56–2.69)0.76 (0.23–2.53)1.94 (0.71–5.27)
Serum total cholesterol per SD increase0.89 (0.69–1.15)1.08 (0.79–1.48)0.66 (0.44–0.98)
APOE ϵ4 vs ϵ3/ϵ31.19 (0.69–2.05)1.58 (0.82–3.07)0.70 (0.28–1.78)
APOE ϵ4/ϵ4 vs ϵ3/ϵ34.43 (1.44–13.64)6.60 (1.90–22.89)2.07 (0.25–17.13)
APOE ϵ2 vs ϵ3/ϵ31.20 (0.60–2.37)1.11 (0.44–2.80)1.29 (0.50–3.30)

*All values are ORs with 95% CIs adjusted for age, sex, and scan interval (when applicable).

Additionally adjusted for the use of blood pressure-lowering medication.

Additionally adjusted for the use of lipid-lowering drugs.

§With or without concomitant new lobar CMBs.

BP indicates blood pressure.

There were no significant differences in microbleed incidence in either location for carriers of either the APOE ϵ2 allele or the APOE ϵ4 allele when compared with persons with the ϵ3/ϵ3 genotype (Table 4). However, when we restricted our analyses to APOE ϵ4/ϵ4, we did find an association with development of new microbleeds (OR, 4.43; 95% CI, 1.44 to 13.64). This was especially true for new strictly lobar incident CMBs (OR, 6.60; 95% CI, 1.90 to 22.89), whereas there was no significant association between APOE ϵ4/ϵ4 and deep or infratentorial CMBs. Only 5 participants carried the APOE ϵ2/ϵ2 genotype; none of them developed new microbleeds during the study period.

Cortical infarcts at baseline were not associated with microbleed incidence (Table 5). Lacunar infarcts were strongly related to incident CMBs in participants with new deep or infratentorial microbleeds (OR, 4.46; 95% CI, 1.79 to 11.10), whereas this association in persons with strictly lobar CMBs was less strong and not significant. White matter lesion volume at baseline MRI increased the risk of incident microbleeds in either location (Table 5).

Table 5. Cerebrovascular Disease and Incidence of CMBs

Risk of Incident CMBs*
Incident CMBs (N=85)Strictly Lobar Incident CMBs (N=51)Deep or Infratentorial Incident CMBs (N=34)
Cortical infarcts vs no infarct1.88 (0.60–5.90)1.46 (0.32–6.70)2.73 (0.57–13.04)
Lacunar infarcts vs no infarct3.04 (1.58–5.87)2.26 (0.96–5.34)4.46 (1.79–11.10)
White matter lesion volume per SD increase1.89 (1.47–2.44)1.86 (1.36–2.55)1.96 (1.34–2.86)

*All values are ORs with 95% CIs adjusted for age, sex, and scan interval.

Ln-transformed.

With or without concomitant new lobar CMBs.

Additional multivariable modeling to examine independency of risk factors did not change the results (data not shown).

Discussion

Incidence of microbleeds in this population-based study over a 3-year interval was approximately 10% and microbleeds rarely disappeared. Participants with CMBs at baseline had an almost 5-fold increased risk of developing new microbleeds during the follow-up period compared with persons without CMBs at baseline. This was especially true for persons with multiple microbleeds at baseline. Risk factors for incident microbleeds were similar to those for prevalent microbleeds and differed according to microbleed location. Strengths of this study are its population-based setting and its large number of participants with repeated MRI using a 3-dimensional T2*-weighted gradient-recalled echo sequence with proven high sensitivity for microbleed detection.16 Moreover, we performed an identical MRI protocol on the same 1.5-T scanner at both time points without software or hardware alterations to optimize comparability between scans over time. Another strength is that the raters were blinded to the time point of the scans. This approach prevented overestimation of incident microbleeds and allowed us to assess potential vanishing of CMBs over time.

A possible limitation of the study is that selective dropout may have influenced our results. People who participated were younger and healthier compared with those who refused a second MRI scan or, in particular, were ineligible. If at all, this may have led us to underestimate the true incidence of microbleeds in the population at large. Due to this selection, associations between risk factors and CMB incidence may also have been underestimated. However, when we repeated the cross-sectional analyses we previously reported on at baseline5 in the 831 persons with a second MRI assessment, we found similar associations (data not shown), indicating limited selection bias.

In the previously mentioned cross-sectional analyses,5 we found an association between cardiovascular risk factors, presence of lacunar infarcts, and white matter lesions and prevalent microbleeds in a deep or infratentorial region but not in a lobar location. Moreover, APOE ϵ4 carriers had significantly more often strictly lobar CMBs than noncarriers. In our current study, we observed similar associations between these risk factors and development of new CMBs in the specified locations, indicating that microbleed incidence, in line with microbleed prevalence, may result from different underlying vascular pathology, that is, cerebral amyloid angiopathy and hypertensive vasculopathy.

Few studies in selected subgroups, mainly among patients with stroke and patients with cerebral amyloid angiopathy, have reported on CMB incidence.914 These studies report a wide range of incidence rates (12% to 50%), which may largely be explained by differences in study populations, scan interval time, and MRI protocols. Of note is that the use of an optimized high-resolution sequence like in our study likely results in a higher incidence of microbleeds compared with less sensitive sequences.16 Overall, studies have consistently shown that microbleeds at baseline predict development of new microbleeds.913 Moreover, vascular risk factors have been associated with incident CMBs in previous studies.9,10 However, only 1 study made separate categories according to CMB location and they did not find a relation of vascular risk factors and incidence of deep or infratentorial CMBs.9 In our study, the association between vascular risk factors and incident deep or infratentorial microbleeds remained significant after adjustment for other markers of small vessel disease. This suggests that deep or infratentorial microbleeds are an independent indicator of hypertensive vasculopathy.

Previous studies in memory clinic patients and patients with cerebral amyloid angiopathy found an association between APOE ϵ4 or ϵ2 carriership and incident CMBs.9,13 Although we could not reproduce this in the general population, we did find that the APOE ϵ4/ϵ4 genotype was related to incident CMBs, in particular incident strictly lobar microbleeds. An explanation for the lack of an association with APOE ϵ4 allele in our study compared with others may be the higher prevalence of both determinant and outcome in previous clinical studies compared with our community-dwelling elderly population. In contrast to the baseline cross-sectional study,5 we now also found an association between white matter lesion volume at baseline and incidence of strictly lobar microbleeds, which is in line with other studies.9,11 It may be hypothesized that vascular amyloid deposition alters white matter perfusion and thus causes white matter lesions through vessel stenosis, vasoactive effects of β-amyloid, and smooth muscle cell necrosis that results in loss of vasoreactivity.19,20

In our study, in 12 persons (1.4% of overall study population; 5.9% of persons with CMBs at baseline), some CMBs that were present at baseline seemed to disappear over time. In only 6 of these, this led to a decrease in overall microbleed count at the follow-up examination. This low percentage is comparable with other studies that also found a very small percentage of microbleeds to disappear over time.9,10,12 In contrast, some studies suggested that there may be more widespread dynamism and resolution of CMBs, but these studies were based on very few cases.21 Furthermore, lower quality of imaging will not only influence the incidence rate of microbleeds, but will also result in microbleeds not being detected on either baseline or follow-up MRI scan, which may falsely suggest resolution of microbleeds.8 We cannot, however, rule out the possibility that some CMBs both occurred and disappeared in the time interval between both scans, although this number is likely low based on our available data.

Conclusions

Our results support the assessment of microbleeds on T2*-weighted MRI as a possible marker of both cerebral amyloid angiopathy and hypertensive vasculopathy progression. Strict control of vascular risk factors, especially in persons who already have microbleeds, may potentially slow down progression of pathology and may perhaps prevent symptomatic intracerebral hemorrhage in the long run. Further studies are warranted to investigate this hypothesis.

Sources of Funding

The Rotterdam Study is supported by the Erasmus Medical Center and Erasmus University Rotterdam, the Netherlands Organization for Scientific Research (NWO), the Netherlands Organization for Health Research and Development (ZonMW), the Research Institute for Diseases in the Elderly (RIDE), the Netherlands Genomics Initiative, the Ministry of Education, Culture and Science, the Ministry of Health, Welfare and Sports, the European Commission (DG XII), and the Municipality of Rotterdam. M.W.V. was supported by a grant from the Alzheimer's Association (NIRG-09-13168). This study was further financially supported by the Netherlands Organization for Scientific Research (NWO) grants 948-00-010 and 918-46-615, and an ErasmusMC grant for translational research.

Disclosures

None.

Footnotes

Correspondence to Meike W. Vernooij, MD, PhD,
Department of Radiology, Erasmus MC, PO Box 2040, 3000 CA Rotterdam, The Netherlands
. E-mail

References

  • 1. Schrag M, McAuley G, Pomakian J, Jiffry A, Tung S, Mueller C, Vinters HV, Haacke EM, Holshouser B, Kido D, Kirsch WM. Correlation of hypointensities in susceptibility-weighted images to tissue histology in dementia patients with cerebral amyloid angiopathy: a postmortem MRI study. Acta Neuropathol. 2009; 119:291–302.CrossrefGoogle Scholar
  • 2. Fan YH, Zhang L, Lam WW, Mok VC, Wong KS. Cerebral microbleeds as a risk factor for subsequent intracerebral hemorrhages among patients with acute ischemic stroke. Stroke. 2003; 34:2459–2462.LinkGoogle Scholar
  • 3. Boulanger JM, Coutts SB, Eliasziw M, Gagnon AJ, Simon JE, Subramaniam S, Sohn CH, Scott J, Demchuk AM, Group VS. Cerebral microhemorrhages predict new disabling or fatal strokes in patients with acute ischemic stroke or transient ischemic attack. Stroke. 2006; 37:911–914.LinkGoogle Scholar
  • 4. Sveinbjornsdottir S, Sigurdsson S, Aspelund T, Kjartansson O, Eiriksdottir G, Valtysdottir B, Lopez OL, van Buchem MA, Jonsson PV, Gudnason V, Launer LJ. Cerebral microbleeds in the population based AGES-Reykjavik study: prevalence and location. J Neurol Neurosurg Psychiatry. 2008; 79:1002–1006.CrossrefMedlineGoogle Scholar
  • 5. Vernooij MW, van der Lugt A, Ikram MA, Wielopolski PA, Niessen WJ, Hofman A, Krestin GP, Breteler MM. Prevalence and risk factors of cerebral microbleeds: the Rotterdam Scan Study. Neurology. 2008; 70:1208–1214.CrossrefMedlineGoogle Scholar
  • 6. Roob G, Schmidt R, Kapeller P, Lechner A, Hartung HP, Fazekas F. MRI evidence of past cerebral microbleeds in a healthy elderly population. Neurology. 1999; 52:991–994.CrossrefMedlineGoogle Scholar
  • 7. Jeerakathil T, Wolf PA, Beiser A, Hald JK, Au R, Kase CS, Massaro JM, DeCarli C. Cerebral microbleeds: prevalence and associations with cardiovascular risk factors in the Framingham study. Stroke. 2004; 35:1831–1835.LinkGoogle Scholar
  • 8. Greenberg SM, Vernooij MW, Cordonnier C, Viswanathan A, Al-Shahi Salman R, Warach S, Launer LJ, Van Buchem MA, Breteler MM. Cerebral microbleeds: a guide to detection and interpretation. Lancet Neurol. 2009; 8:165–174.CrossrefMedlineGoogle Scholar
  • 9. Goos JD, Henneman WJ, Sluimer JD, Vrenken H, Sluimer IC, Barkhof F, Blankenstein MA, Scheltens PH, van der Flier WM. Incidence of cerebral microbleeds: a longitudinal study in a memory clinic population. Neurology. 2010; 74:1954–1960.CrossrefMedlineGoogle Scholar
  • 10. Gregoire SM, Brown MM, Kallis C, Jager HR, Yousry TA, Werring DJ. MRI detection of new microbleeds in patients with ischemic stroke: five-year cohort follow-up study. Stroke. 2010; 41:184–186.LinkGoogle Scholar
  • 11. Chen YW, Gurol ME, Rosand J, Viswanathan A, Rakich SM, Groover TR, Greenberg SM, Smith EE. Progression of white matter lesions and hemorrhages in cerebral amyloid angiopathy. Neurology. 2006; 67:83–87.CrossrefMedlineGoogle Scholar
  • 12. Greenberg SM, O'Donnell HC, Schaefer PW, Kraft E. MRI detection of new hemorrhages: potential marker of progression in cerebral amyloid angiopathy. Neurology. 1999; 53:1135–1138.CrossrefMedlineGoogle Scholar
  • 13. Greenberg SM, Eng JA, Ning M, Smith EE, Rosand J. Hemorrhage burden predicts recurrent intracerebral hemorrhage after lobar hemorrhage. Stroke. 2004; 35:1415–1420.LinkGoogle Scholar
  • 14. Liem MK, Lesnik Oberstein SA, Haan J, van der Neut IL, van den Boom R, Ferrari MD, van Buchem MA, van der Grond J. Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy: progression of MR abnormalities in prospective 7-year follow-up study. Radiology. 2008; 249:964–971.CrossrefMedlineGoogle Scholar
  • 15. Hofman A, Breteler MM, van Duijn CM, Janssen HL, Krestin GP, Kuipers EJ, Stricker BH, Tiemeier H, Uitterlinden AG, Vingerling JR, Witteman JC. The Rotterdam Study: 2010 objectives and design update. Eur J Epidemiol. 2009; 24:553–572.CrossrefMedlineGoogle Scholar
  • 16. Vernooij MW, Ikram MA, Wielopolski PA, Krestin GP, Breteler MM, van der Lugt A. Cerebral microbleeds: accelerated 3D T2*-weighted GRE MR imaging versus conventional 2D T2*-weighted GRE MR imaging for detection. Radiology. 2008; 248:272–277.CrossrefMedlineGoogle Scholar
  • 17. de Boer R, Vrooman HA, van der Lijn F, Vernooij MW, Ikram MA, van der Lugt A, Breteler MM, Niessen WJ. White matter lesion extension to automatic brain tissue segmentation on MRI. Neuroimage. 2009; 45:1151–1161.CrossrefMedlineGoogle Scholar
  • 18. Wenham PR, Price WH, Blandell G. Apolipoprotein E genotyping by one-stage PCR. Lancet. 1991; 337:1158–1159.CrossrefMedlineGoogle Scholar
  • 19. Thomas T, Thomas G, McLendon C, Sutton T, Mullan M. Beta-amyloid-mediated vasoactivity and vascular endothelial damage. Nature. 1996; 380:168–171.CrossrefMedlineGoogle Scholar
  • 20. Niwa K, Carlson GA, Iadecola C. Exogenous a beta1–40 reproduces cerebrovascular alterations resulting from amyloid precursor protein overexpression in mice. J Cereb Blood Flow Metab. 2000; 20:1659–1668.CrossrefMedlineGoogle Scholar
  • 21. Kirsch W, McAuley G, Holshouser B, Petersen F, Ayaz M, Vinters HV, Dickson C, Haacke EM, Britt W, Larseng J, Kim I, Mueller C, Schrag M, Kido D. Serial susceptibility weighted MRI measures brain iron and microbleeds in dementia. J Alzheimers Dis. 2009; 17:599–609.CrossrefMedlineGoogle Scholar

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.