核磁共振显像三维测量预测老年轻度认知功能损害发展为痴呆的价值

A predictive study on mild cognitive impairment in elderly converting into dementia by 3-dimensional magnetic resonance imaging

目的：探索认知障碍发展为痴呆的结构性脑影像的预测指标。方法：前瞻性随访对照研究。研究对象为患轻度认知功能损害的老年人(MCI组)21例,男性5例,女性16例；和认知功能正常的老年人(对照组,NC组)29例,男性12例,女性17例2组。随访对象转为痴呆的标准是根据DSM-Ⅳ的相关标准。认知评价工具为世界卫生组织老年认知功能评价成套神经心理测验(WHO-BCAI)和韦氏记忆测验。基线进行核磁共振显像三维测量(3D-MRI)并以分析基线检测结果对MCI是否发展为痴呆的预测作用。结果：MCI组的颅内总体积为(1316±s111)cm^3,NC组为(1388±90)cm^3,2组间差异有显著意义(P＜0．05)。MCI组的脑灰质体积为(615±66)cm^3,NC组为(674±46)cm^3,2组间差异有非常显著意义(P＜0．01)。MCI组的灰质百分比为46．7%,而NC组为48．6%,差异有非常显著意义(P＜0．01)。MCI组的颅内总体积和脑灰质体积较NC组显著性缩小(P＜0．05)。3a随访时,MCI组有5例发展为痴呆,NC组没有发展为痴呆的病例。以基线脑核磁共振三维测量指标为自变量Logistic回归分析,结果没有指标入选回归方程。结论：3D-MRI的脑灰质减少和侧脑室体积增大有助于诊断MCI,但预测MCI病人是否发展为痴呆的价值有待进一步研究。

AIM： To explore the predictive factors of structural brain imaging for mild cognitive impairment （MCI） converting into dementia during a 3 year＇s follow-up period. METHODS： A cross-section prospective and longitudinal parallel cohort study design was adopted. Subjects were divided into two groups, the elderly with MCI and normal control （NC）. The Neuropsychological Battery of Cognitive Assessment Instruments developed by World Health Organization （WHO-BCAI） and the Wechsler Memory Scale （WMS） were utilized to assess the cognitive function changes at baseline and endpoint. Diagnostic converting criteria for dementia was based on DSM-IV. 3-Dimentional magnetic resonance imaging （3D-MRI） and quantified volumes of hippocampal formation （HF）, gray matter （GM）, white matter （WM）, ventricular and cerebrospinal fluid （CSF） ob- tained at baseline and their predictive capabilites were analyzed. RESULTS： Twenty-one MCI and 29 NC were assessed at baseline with 3D-MRI. The elderly with MCI showed significant decrease of total intracranial volume （P 〈 0.05）, GM volume （P 〈 0.01）, and GM fraction （P 〈 0.01）, and increase of CSF fraction （P 〈 0.05） . The intracranial and GM volume decreased by 5.2 % and 8.8 % respectively, GM fraction decreased by 3.9 %, and CSF fraction increased by 12.3 %. Discriminant analysis demonstrated that variables of GM and ventricular volumes entered the function （P 〈 0.000） with a discriminative accuracy of 74.5 %. Logistic analysis demon- strated that no variables of brain volume had significant predictive capability for MCI converting into dementia. CONCLUSION： 3D-MRI can reveal the brain structural changes of MCI in elderly with decrease of GM and increase of ventricular volumes but still lack of the definite criteria for MCI converting into dementia.