Incidence of cerebral microbleeds: a longitudinal study in a memory clinic population.

BACKGROUND: Cerebral microbleeds (MBs) are commonly observed in memory clinic patients. Little is known about occurrence of and risk factors for developing new MBs in this population.

OBJECTIVE: To investigate incidence of lobar and nonlobar MBs in a memory clinic population. Furthermore, to assess risk factors for the development of new MBs and their associations with other MRI changes.

METHODS: A total of 254 patients visiting our memory clinic, with repeat gradient-recalled echo T2*-weighted MRI, were included (scan interval 1.9 +/- 0.9 years). Baseline and incident MBs were regionally counted. White matter hyperintensities (WMH) and progression of WMH were assessed using visual rating scales. Baseline brain volume and whole-brain atrophy rate were estimated automatically. In a subset, APOE was determined.

RESULTS: Thirty-one (12%) patients developed new MBs (range 1-19). Both multiple strictly lobar and nonlobar MBs at baseline predicted incident MBs (odds ratio [OR] 8.4; 95% confidence interval [CI] 2.2-33.2, and OR 33.8; 95% CI 8.1-140.8). Furthermore, baseline WMH grade (OR 1.2; 1.1-1.3), lacunar infarcts (OR 2.8; 1.3-6.0), and APOE epsilon2 carriership (OR 4.2; 1.4-12.5) predicted MB incidence. Incident MB patients had more progression of WMH (p < 0.01) and incident lacunar infarcts (p < 0.05). These relations were most prominent for incident nonlobar MBs. Incident strictly lobar MBs were associated with smoking.

CONCLUSION: In addition to APOE genotype, presence and progression of small-vessel disease and vascular risk factors were predictors of new MBs. The latter are potentially modifiable, suggesting the possibility of preventing incident MBs, hopefully resulting in slower clinical decline.