Premium
Investigating neural correlates of mild cognitive impairment using estimated clinical status from neuropsychological test battery: LASI‐DAD
Author(s) -
Keller Brenton James,
Lee Jinkook,
Toga Arthur W,
Jovicich Jorge
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.038440
Subject(s) - neuropsychology , cognition , dementia , cognitive impairment , medicine , audiology , psychology , clinical psychology , psychiatry , disease
Background Having ways to quickly assess cognitive impairment is critical for developing nations. Here we examine how different combinations of neuropsychiatric tests identify cognitive impairment using the Longitudinal Aging Study in India ‐ Diagnostic Assessment of Dementia (LASI‐DAD). We then investigate for brain regions that show structural morphological differences between cognitive impairment and cognitive normal groups. Methods A total of 103 subjects (female = 46, age = 68 ± 6 years) were studied from three clinical 3T MRI centers using the ADNI 3 MRI protocol (NIMHANS, Bangalore, N=52, NM Medical, Mumbai, N=23, and Institute of Kolkata N=28). Cognitive impairment clinical status was missing for 62 of the subjects. Cognitive impairment status was investigated using the LASI‐DAD main dataset (clinical status n = 1486). We examined different combinations of factors from the LASI‐DAD neuropsychological assays to predict clinical impairment status. After estimating cognitive impairment status, brain morphometry data was studied with Freesurfer using volumetric ROI‐analysis and vertex‐based cluster‐corrected structural analysis. Results We found a combination of JORM IQCODE, Cognitive Activity Score, years of education, HMSE, and the interaction between HMSE and general cognition to best predict clinical status (AUC=0.90). Clinical status was calculated for the MRI subjects lacking a clinical diagnosis (CN=38, CI=24). Cognitive impairment was found to be associated with thickness and volumetric decline, especially the right hemisphere transverse temporal gyrus (10‐15%) and medial orbitofrontal cortex (6‐9%), and bilateral volumetric decline of the hippocampus (presubiculum and molecular layer, 5‐8%). Furthermore, age was not found to affect hippocampal volume, yet there was a significant negative effect on hippocampal volume with the interaction of cognitive impairment and age (6‐8%). A significant interaction between cognitive impairment and age was also found to reduce the curvature of the left hemisphere fusiform. Conclusions We investigated combinations of neurophysiological tests to predict cognitive impairment status. Using the predicted cognitive impairment status, we found significant morphological changes in the brain between groups. This demonstrates how quick neurocognitive exams can be used in areas lacking clinicians who can diagnosis dementia.