BACKGROUNDEndothelial dysfunction is not only an early stage of atherosclerosis, but also involved in the pathogenesis of cerebral small-vessel diseases. Patients with cerebral microbleeds (CMBs) may have arteriolosclerosis as well as systemic atherosclerosis. However, little is known about the associations among CMBs, atherosclerosis of cerebral large arteries, and endothelial function. Our study aimed to investigate the relationships among them.

METHODSThis was a cross-sectional study. Ninety patients hospitalized in Peking University First Hospital with acute ischemic stroke were enrolled consecutively between November 1, 2007 and January 31, 2008. All subjects underwent transcranial Doppler and carotid color duplex ultrasonography to record the intima-media thickness (IMT) of common carotid artery, carotid plaque, and cerebral artery stenosis. Brain magnetic resonance imaging (MRI) routine sequences and gradient recall-echo T2(*)-weighted imaging were performed to count CMBs with clinical data blindness. Endothelial function was evaluated using flow-mediated dilation (FMD) and nitroglycerin-mediated dilation (NMD) of the brachial artery. FMD and NMD were examined by an experienced vascular sonographer using a high-resolution ultrasound.

RESULTSThirty cases (33.3%) had CMBs with counts ranging from 1 to 30. Both FMD ((9.9 ± 4.8)% vs. (15. 2 ± 7.4)%, P = 0.001) and NMD ((13.7 ± 6.1)% vs. (19.0 ± 7.4)%, P = 0.001) were significantly decreased in CMB-positive patients than in CMB-negative patients. No significant relationships were demonstrated between CMBs and intracranial and/or extracranial artery stenosis. The frequencies of CMBs in patients with IMT≥1.0 mm, carotid plaque, and extracranial artery stenosis were 37.5%, 39.4%, and 47.6% respectively, with no significant difference, but much higher than in patients with IMT <1.0 mm (5%, P < 0.05). In Logistic regression analysis, impaired FMD (OR = 5.783, 95% CI 1.652-6.718, P = 0.007) and high pulse pressure (OR = 6.228, 95% CI 1.594-3.891, P = 0.009) were independently associated with the presence of CMBs, as well as previous ischemic stroke. In contrast, NMD was not correlated with CMBs.

CONCLUSIONSCMBs may coexist with cerebral atherosclerosis in ischemic stroke. Endothelial dysfunction may play a role in the pathogenesis of CMBs, but may not simply reflect functional alterations of large arteries.