Petersen R C, Smith G E, Ivnik R J, Kokmen E, Tangalos E G
Department of Neurology, Mayo Clinic, Rochester, MN 55905.
Neurology. 1994 May;44(5):867-72. doi: 10.1212/wnl.44.5.867.
The detection of very early Alzheimer's disease (AD) can be important for both theoretical and practical reasons. Typically, a memory impairment is the first sign of incipient disease, but the early clinical diagnosis can be challenging. We investigated several aspects of memory function in AD and normal aging to determine which indices of performance were most sensitive at detecting early impairments. We evaluated 106 pairs of patients with probable AD and matched controls from the Mayo Clinic Alzheimer's Disease Patient Registry using a logistic regression model that included measures of memory function, verbal and nonverbal intelligence, attention, and language. Results indicated that an index of learning, especially with semantic cuing, was most sensitive at separating the two groups. We then matched subsets of individuals from the larger groups of AD and control subjects on the Mini-Mental State Examination (range of scores, 24 to 26). A logistic regression analysis on these matched groups yielded the same results. A measure of learning with facilitation of performance using cues appears to be the best discriminator at detecting very mild AD. These measures can be useful in selecting patients for interventional strategies.
出于理论和实际原因,极早期阿尔茨海默病(AD)的检测都可能很重要。通常,记忆障碍是疾病初期的首个迹象,但早期临床诊断可能具有挑战性。我们研究了AD患者和正常衰老个体记忆功能的几个方面,以确定哪些表现指标在检测早期损伤时最为敏感。我们使用逻辑回归模型评估了106对可能患有AD的患者以及来自梅奥诊所阿尔茨海默病患者登记处的匹配对照组,该模型纳入了记忆功能、语言和非语言智力、注意力及语言的测量指标。结果表明,尤其是在语义提示下的学习指标,在区分两组时最为敏感。然后,我们在简易精神状态检查表(得分范围为24至26)上对AD和对照受试者较大群体中的个体子集进行匹配。对这些匹配组进行的逻辑回归分析得出了相同结果。使用提示促进表现的学习测量指标似乎是检测极轻度AD的最佳鉴别指标。这些测量指标在选择适合干预策略的患者时可能会很有用。