Race, Stroke Belt, and Cerebral Microbleeds After Stroke

Key Takeaways

• Among REGARDS participants with incident stroke or TIA and available MRI, Black participants had higher adjusted odds of any CMB.
• Stroke-belt residents had lower adjusted odds of any, non-lobar, and lobar CMBs and a lower modeled CMB count.
• Black stroke-belt residents had higher reported odds of any and mixed CMBs.

Cerebral microbleeds (CMBs) detected on magnetic resonance imaging (MRI) in people with incident stroke or transient ischemic attack (TIA) raise a question about whether their distribution differs by race and stroke-belt residence. CMBs are associated with stroke, cognitive impairment, and dementia, making their presence and anatomical distribution relevant to research on vascular brain injury. Available MRI scans allow investigators to examine microbleed location and count across these groups.

The Reasons for Geographic and Racial Differences in Stroke (REGARDS) cohort analysis included 808 participants with incident stroke or TIA and an available MRI, including Black participants and residents of the stroke belt in the United States. Participants were not assigned an intervention. Investigators classified CMBs as any, non-lobar, lobar, or mixed and recorded their counts. Multivariable logistic regression assessed CMB presence, while negative binomial regression assessed counts. The models adjusted for age, sex, and vascular risk factors; regional estimates compared stroke-belt residents with people living outside the stroke belt.

Black participants had higher adjusted odds of any CMB, with an odds ratio (OR) of 1.72. Stroke-belt residents had lower adjusted odds of any CMB, with an OR of 0.51 compared with residents outside the stroke belt. These are separate adjusted associations, not evidence of a race-by-region interaction.

Stroke-belt residents also had a lower modeled CMB count, with a rate ratio (RR) of 0.66. Their adjusted odds of non-lobar and lobar CMBs were lower as well. Investigators reported higher odds of both any and mixed CMBs among Black stroke-belt residents. The reported OR for mixed CMBs was 2.99.

These associations apply to REGARDS participants with incident stroke or TIA for whom MRI was available, rather than everyone who experienced either event. The observational comparisons do not establish that race causes CMBs or that residence in the stroke belt protects against them.

Race and stroke-belt residence had contrasting associations with MRI-defined CMB outcomes. Investigators suggested that the higher odds of any CMB among Black participants may indicate increased subclinical vascular brain injury, without establishing a causal explanation.

Clinician Questions

When were cerebral microbleeds identified relative to stroke or TIA in the REGARDS analysis?

The analysis included people with incident stroke or TIA and an available MRI. The findings cannot establish when their cerebral microbleeds developed.

Can the REGARDS cerebral microbleed odds ratios show absolute risk after stroke or TIA?

Adjusted associations for CMB presence were expressed as odds ratios, not risk ratios. These associations do not supply absolute risk or an individual's prognosis.