By Halina White, MD

Assistant Professor of Neurology, Weill Cornell Medical College

Dr. White reports no financial relationships relevant to this field of study.

Synopsis: Cerebral microbleeds in non-demented elderly people are associated with a global reduction in cerebral blood flow and a high prevalence of beta-amyloid deposition in the brain.

Source: Gregg NM, et al. Incidental cerebral microbleeds and cerebral blood flow in elderly individuals. JAMA Neurol doi:10.1001/jamaneurol.2015.1359. Published online July 13, 2015.

Cerebral microbleeds are small, ovoid, hypointense lesions (“black dots”) seen on susceptibility-weighted MRI sequences in many older individuals. They are caused by leakage of red blood cells from small cerebral vessels. Cerebral amyloid angiopathy affects cortical and leptomeningeal vessels, leading to microbleeds in the superficial areas of the cerebral hemispheres. Hypertension and cardiovascular risk factors lead to arteriosclerosis and lipohyalinosis, and are associated with microbleeds in the basal ganglia and deep hemispheric white matter. In some studies, patients with cerebral amyloid angiopathy, as evidenced by cortical lobar microbleeds, have been found to have cognitive decline. This study sought to determine possible mechanisms through which this cognitive decline could occur by quantifying how cerebral microbleeds are related to cerebral blood flow and cognition in the very elderly.

Fifty-five cognitively normal elderly individuals (mean age 86.8 years) underwent detailed neuropsychiatric testing and comprehensive MRI brain evaluation, including gradient echo and susceptibility-weighted imaging to detect microbleeds, arterial spin labelling to evaluate cerebral blood flow, and T1 and T2 sequences to evaluate for microvascular disease burden, hippocampal volume, and total cerebral volume. Subjects also underwent FDG-PET, Pittsburgh compound B-PET, apolipoprotein E (ApoE4) genotyping and evaluation for systemic vascular disease burden with carotid ultrasound, Ankle-Brachial Pressure Index, EKG, and serum cystatin C (a measure of renal microvascular resistance). The subject’s microbleeds were divided into lobar cortical, lobar subcortical, and deep.

In the study, 38% of participants were found to have cortical microbleeds suggestive of cerebral amyloid angiopathy. Interestingly, the presence of cortical microbleeds was not associated with age, ApoE4 status, the presence of cardiovascular risk factors, or systemic vascular disease burden. Participants with any cerebral microbleeds had a trend toward decreased regional cerebral blood flow in the bilateral frontal and occipital lobes, the right parietal and temporal lobes, and the precuneate cortex. However, this trend was not statistically significant when corrected for age, sex, and global Pittsburgh compound B distribution. Compared to the rest of the cohort, participants with cortical cerebral microbleeds had significantly reduced cerebral blood flow to all lobes and to all deep structures. Cortical microbleeds were not significantly associated with brain metabolism as measured by FDG-PET, with amyloid A-beta as measured by Pittsburgh compound B-PET, nor with cerebral atrophy or hippocampal volume on MRI. Cortical microbleeds were significantly associated with the presence of infarcts, but not with total microvascular disease burden. Cortical microbleeds were also significantly associated with a non-zero score on the clinical dementia rating scale, but not with any other neuropsychiatric testing metric.

In summary, in this cohort of apparently cognitively normal elderly individuals, cerebral cortical microbleeds, likely related to cerebral amyloid angiopathy, were significantly related to globally decreased cerebral blood flow. Moreover, cerebral cortical microbleeds were also significantly associated with the presence of infarcts and subtle decreases in neuropsychiatric testing metrics.


These findings suggest that elderly individuals with cortical cerebral microbleeds may be exposed to chronic global cerebral hypoperfusion. As this was a cross-sectional cohort study, no assertions as to the causal relationships between these findings can be made. However, it is interesting to hypothesize that perhaps patients with cerebral amyloid angiopathy have chronically poor cerebral blood flow leading to neuronal injury and a greater burden of infarcts, and that these may contribute to their decreased cognitive performance.

In this study ApoE4 status and A-beta amyloid deposition were not significantly related to the presence of cerebral cortical microbleeds, as one might expect if cerebral cortical microbleeds are simply a biomarker for cerebral amyloid angiopathy. The writers postulate that this may have been caused by the very high prevalence of A-beta amyloid deposition (81%) in this cohort of very elderly patients.

Poor cerebral perfusion associated with the cortical microbleeds also was not correlated with cerebral glucose hypometabolism, suggesting that cerebral hypometabolism is not driving the perfusion deficit. Exactly how cerebral amyloid angiopathy may lead to globally poor cerebral blood flow, however, is unclear.

Study limitations included the advanced age of the subjects, small sample size, and cross-sectional study design. Strengths included the fact that this was a study of cognitively normal elderly subjects in the presymptomatic stage of dementia.

This study suggests that MRI arterial spin labelling measures of cerebral blood flow may be an early marker for cerebral amyloid angiopathy. This early marker could be useful in diagnosis and in treatment. Further work with repeated evaluation of these subjects is needed to elucidate the temporal relationships between cerebral microbleeds, cerebral hypoperfusion, and cognitive decline.