Cerebral microbleeds (CMBs) may contribute to cognitive deficits in stroke. Cognitive impairment that does not meet the criteria for dementia (cognitive impairment no dementia [CIND]) is common in stroke, and patients with such impairment can revert to normal cognition.

To investigate the association of CMBs and remission of poststroke CIND.

To understand the evolution of poststroke cognitive impairment no dementia (CIND) is bi-directional.

143 patients with CIND at three months after stroke were recruited and followed up for one year. Remission of CIND was defined as a conversion of cognitive status from CIND to cognitively intact at follow-up. MRI variables in terms of infarction, cerebral microbleeds (CMBs), and white matter hyperintensities (WMHs) and hippocampal volume were analyzed. Logistic regression was performed to find the predictors of the remission of poststroke CIND.

30 (21.0%) out of the 143 patients converted to cognitive intact at follow-up. In univariate comparisons, subjects with remission of CIND had younger age (67.1 ± 9.5 vs.73.6 ± 7.6 years, p < 0.001) and higher education years (5.1 ± 4.0 vs.3.6 ± 4.0, p = 0.039). They also had lower WMHs volume (8.2 ± 8.2 vs. 18.6 ± 19.7 cm3, p < 0.001), lower frequency of CMBs (10.0% vs. 31.0%, p = 0.021) and lower volume of the lateral ventricle (33.3 ± 16.5 vs.42.6 ± 19.4 cm3, p = 0.017). In logistic regression, age (odds ratio [OR] = 0.913, 95%C.I. = 0.866–0.962, p = 0.001) and absence of CMBs (OR = 4.292, 95%C.I. = 1.174–15.625, p = 0.028) were significant predictors of remission of CIND.

Younger age and absence of CMBs predict the remission of poststroke CIND.

Longitudinal studies of predicting dementia conversion of poststroke cognitive impairment no dementia (CIND) are limited.

To investigate the clinical and imaging predictors of dementia conversion in poststroke patients with CIND.

To understand dementia conversion of CIND.

143 patients with CIND (defined as impairment in at least one cognitive domain without meeting the criteria of dementia) at three months after stroke were recruited and followed up for one year. Dementia was diagnosed using the criteria of Diagnostic and Statistical Manual of Mental Disorders (4th edition, DSM-IV). MRI measurements including infarction, microbleeds, white matter hyperintensities (WMHs) and hippocampal volume were conducted. Logistic regression was performed to find the predictors of dementia at follow-up.

16 (11.2%) out of the 143 patients developed dementia 15 months after stroke. In univariate comparisons, subjects with dementia at follow-up had older age (78.0 ± 5.3 vs.71.5 ± 8.5 years, p = 0.003) and higher NIHSS score (7.1 ± 3.5 vs.4.7 ± 3.3, p = 0.005) on admission. They also had higher frequency of old infarcts in the thalamus (31.3% vs. 11.0%, p = 0.025), larger volume of old infarcts (4.2 ± 11.2 vs. 0.7 ± 2.6 cm3, p < 0.001) and WMHs volume (33.2 ± 34.0 vs. 14.2 ± 14.1 cm3, p = 0.016). In logistic regression, age (odds ratio [OR] =1.203, 95%C.I.=1.054-1.373, p = 0.006), NIHSS score on admission (OR = 1.324, 95%C.I.=1.082-1.619, p = 0.006) and WMHs volume (OR = 1.045, 95%C.I.=1.007-1.084, p = 0.019) were significant predictors of dementia at follow-up.

WMHs volume predicts dementia in poststroke patients with CIND, suggesting subcortical ischemic vascular disease was an important origin of poststroke delayed dementia.

Prefrontal cortex and sex difference are involved in verbal fluency network described in normal participants. Stroke patients often have prefrontal cortex atrophy.

To investigate whether atrophy in subdivisions of prefrontal cortex and sex difference contribute to verbal fluency in non-aphasic stroke patients.

To understand the relationship between the atrophy of left dorsolateral prefrontal cortex and verbal performance in elderly poststroke women.

30 elderly (age> = 60 years old) women with non-aphasic ischemic stroke and 30 age-controlled stroke men recruited. Automatic segmentation methods were used to assess the volume of both sides of the whole prefrontal cortex, anterior cingulate cortex, orbital frontal cortex and dorsalateral prefrontal cortex (DLPFC), as well as white matter lesions (WMLs) volume. Mini-mental state examination (MMSE) and semantic verbal fluency test (VFT, category: foods and animals) were administered at 3 and 15 months after the index stroke.

The mean (s.d) age was 73.3 ± 7.2 in women and 72.1 ± 6.9 in men. Men had higher education years, less diabetes and higher MMSE scores (p < 0.05). At 3 months after stroke, volume of the left DLPFC was significantly correlated with VFT score in women rather than men, even after controlled by age, education years, neurological deficit, diabetes, WMLs volume and infarct location (partial r = 0.477, p = 0.018). At 15 months, this correlation remained significant (partial r = 0.548, p = 0.006) in women.

Sex difference may be present in the neuropsychological mechanism of verbal fluency impairment in patients with cerebrovascular disease.