Abstract
Background: Cortical atrophy is a key neuroimaging feature of dementia.
However, the role of subcortical gray matter reduction in cognitive impairment has not
been explored extensively.
Objectives: We examined the risk factors of subcortical structures on
neuroimaging and their association with cognitive impairment and dementia.
Methods: Data from two studies were used: a subsample from the Epidemiology
of Dementia in Singapore (EDIS) study of non-demented community-dwelling subjects
(n = 550) and a case-control study. Subjects underwent similar
neuropsychological tests and brain MRI. Subcortical volumes of accumbens, amygdala,
caudate, pallidum, putamen, thalamus, hippocampus, and brainstem were measured. Cognitive
impairment no dementia (CIND), dementia and its subtypes, vascular cognitive impairment
(VCI), were defined using accepted criteria. Cognitive function was also expressed as both
composite and domain-specific Z-scores.
Results: In the EDIS study, age, female gender, Malay ethnicity, diabetes,
lacunar-infarcts, and white matter lesions were the most important risk factors for
subcortical atrophy. Moreover, smaller volumes of accumbens, amygdala, caudate, thalamus,
and brainstem were significantly associated with lower cognitive composite Z-scores. With
respect to clinical outcomes in the case-control study, structures such as the accumbens,
caudate, putamen, and hippocampus were associated with both CIND and dementia. Smaller
caudate and pallidum volumes were related to VCI whereas amygdalar atrophy was only
associated with non-VCI. Furthermore, subcortical atrophy was related to both VCI and
non-VCI.
Conclusion: Subcortical gray matter atrophy is not only observed in
dementia, but also in the preclinical stages of cognitive impairment. Furthermore, besides
VCI, subcortical structures were also related to non-VCI.