Long Xianxian, Yuan Manqiong, Zhang Zeyun, Fang Ya
Key Laboratory of Health Technology Assessment of Fujian Province, School of Public Health, Xiamen University, Xiamen, China; Center for Aging and Health Research, School of Public Health, Xiamen University, Xiamen, China.
Key Laboratory of Health Technology Assessment of Fujian Province, School of Public Health, Xiamen University, Xiamen, China; Center for Aging and Health Research, School of Public Health, Xiamen University, Xiamen, China.
Arch Gerontol Geriatr. 2025 Feb;129:105659. doi: 10.1016/j.archger.2024.105659. Epub 2024 Oct 13.
To derive data-driven subtypes of mild cognitive impairment (MCI) and characterize the complicated changes of general cognitive and daily functions over time in MCI subtypes.
A total of 813 subjects diagnosed as MCI at baseline from the Alzheimer's Disease Neuroimaging Initiative (ADNI) were included. Data-driven MCI subtypes were derived from group-based multi-trajectory modeling (GBMTM) analyses using longitudinal measurement scores in the cognitive domains of visuospatial function, language, and executive function. General cognitive and daily functions were measured by the Mini-Mental State Examination (MMSE) and the Functional Assessment Questionnaire (FAQ), respectively, whose longitudinal trajectory changes were depicted by Linear mixed models.
Three MCI subtypes were derived, which were defined as "Cognitive decline group", "Mild cognitive decline group" and "No cognitive decline group". The "Mild cognitive decline group" had the highest percentage in the sample (46.2 %), followed by the "No cognitive decline group" (35.2 %). Patients in the "Cognitive decline group" had the highest mean age (74.69 years) at baseline, the highest APOE ε4 carriers (63.2 %), and the greatest dementia conversion rate (77.0 %). The changes in MMSE and FAQ score trajectories were fastest in the "Cognitive decline group" in the first 36 months and most slowly in the "No cognitive decline group".
MCI individuals could be subdivided into more fine-grained cognitive subtypes, and identifying these distinct MCI subtypes and their different trajectories of cognitive decline may have important prognostic value for improving clinical course prediction.
推导数据驱动的轻度认知障碍（MCI）亚型，并描述MCI亚型中一般认知和日常功能随时间的复杂变化。
纳入了阿尔茨海默病神经影像学倡议（ADNI）中813名在基线时被诊断为MCI的受试者。数据驱动的MCI亚型通过基于组的多轨迹建模（GBMTM）分析得出，该分析使用了视觉空间功能、语言和执行功能等认知领域的纵向测量分数。一般认知和日常功能分别通过简易精神状态检查表（MMSE）和功能评估问卷（FAQ）进行测量，其纵向轨迹变化通过线性混合模型进行描述。
得出了三种MCI亚型，分别定义为“认知衰退组”、“轻度认知衰退组”和“无认知衰退组”。“轻度认知衰退组”在样本中的比例最高（46.2%），其次是“无认知衰退组”（35.2%）。“认知衰退组”的患者在基线时平均年龄最高（74.69岁），APOE ε4携带者比例最高（63.2%），痴呆转化率也最高（77.0%）。MMSE和FAQ评分轨迹的变化在“认知衰退组”的前36个月最快，在“无认知衰退组”最慢。
MCI个体可细分为更精细的认知亚型，识别这些不同的MCI亚型及其不同的认知衰退轨迹可能对改善临床病程预测具有重要的预后价值。
