State Key Laboratory of Cognitive Neuroscience and Learning, Faculty of Psychology, Beijing Normal University, Beijing, 100875, China.
School of Systems Science, Beijing Normal University, Beijing, 100875, China.
Alzheimers Res Ther. 2023 Feb 2;15(1):27. doi: 10.1186/s13195-023-01167-z.
Mild cognitive impairment (MCI) has been thought of as the transitional stage between normal ageing and Alzheimer's disease, involving substantial changes in brain grey matter structures. As most previous studies have focused on single regions (e.g. the hippocampus) and their changes during MCI development and reversion, the relationship between grey matter covariance among distributed brain regions and clinical development and reversion of MCI remains unclear.
With samples from two independent studies (155 from the Beijing Aging Brain Rejuvenation Initiative and 286 from the Alzheimer's Disease Neuroimaging Initiative), grey matter covariance of default, frontoparietal, and hippocampal networks were identified by seed-based partial least square analyses, and random forest models were applied to predict the progression from normal cognition to MCI (N-t-M) and the reversion from MCI to normal cognition (M-t-N).
With varying degrees, the grey matter covariance in the three networks could predict N-t-M progression (AUC = 0.692-0.792) and M-t-N reversion (AUC = 0.701-0.809). Further analyses indicated that the hippocampus has emerged as an important region in reversion prediction within all three brain networks, and even though the hippocampus itself could predict the clinical reversion of M-t-N, the grey matter covariance showed higher prediction accuracy for early progression of N-t-M.
Our findings are the first to report grey matter covariance changes in MCI development and reversion and highlight the necessity of including grey matter covariance changes along with hippocampal degeneration in the early detection of MCI and Alzheimer's disease.
轻度认知障碍 (MCI) 被认为是正常衰老和阿尔茨海默病之间的过渡阶段，涉及大脑灰质结构的重大变化。由于大多数先前的研究都集中在单个区域（例如海马体）及其在 MCI 发展和逆转过程中的变化，因此分布在大脑区域之间的灰质协方差与 MCI 的临床发展和逆转之间的关系尚不清楚。
利用来自两个独立研究的样本（北京老龄化大脑 rejuvenation 倡议的 155 例和阿尔茨海默病神经影像学倡议的 286 例），通过基于种子的偏最小二乘分析确定默认、额顶和海马网络的灰质协方差，并应用随机森林模型预测从正常认知到 MCI 的进展（N-t-M）和从 MCI 到正常认知的恢复（M-t-N）。
三种网络中的灰质协方差在不同程度上可以预测 N-t-M 进展（AUC = 0.692-0.792）和 M-t-N 恢复（AUC = 0.701-0.809）。进一步的分析表明，海马体已成为所有三个大脑网络中恢复预测的重要区域，尽管海马体本身可以预测 M-t-N 的临床恢复，但灰质协方差在预测 N-t-M 的早期进展方面显示出更高的预测准确性。
我们的研究结果首次报告了 MCI 发展和恢复过程中的灰质协方差变化，并强调了在 MCI 和阿尔茨海默病的早期检测中需要包括灰质协方差变化和海马体退化。
