Brant Larry J, Sheng Shan L, Morrell Christopher H, Zonderman Alan B
Gerontology Research Center, National Institute on Aging, 5600 Nathan Shock Drive, Baltimore, MD 21224, USA.
J Clin Epidemiol. 2005 Jul;58(7):701-7. doi: 10.1016/j.jclinepi.2005.01.003.
This article presents a computerized method to help predict individuals at risk for developing Alzheimer's disease (AD). This would be a valuable tool for clinicians in developing treatment plans for potential AD patients. Using the initial level and rates of change in visual memory performance, such a method could predict potential AD patients in a fast and inexpensive manner. A longitudinal case-control study of 52 female and 145 male participants was performed in a gerontology research center using premorbid tests of visual memory and neurologic examinations to identify individuals with and without dementia and AD.
The classification method for each individual starts on the second examination and proceeds to compute that person's risk of AD one examination at a time based on all the follow-up information of the remaining individuals.
By performing a crossvalidation study, the optimal combination of sensitivity and specificity derived from a receiver operating characteristic (ROC) curve showed 65% of the Alzheimer cases and 75% of the noncases were correctly classified for females, while 65 and 60% of cases and noncases, respectively, were correctly classified for males.
Longitudinal measurements of cognition can be useful in detecting the presence of AD.
本文介绍了一种计算机化方法，以帮助预测有患阿尔茨海默病（AD）风险的个体。这对于临床医生为潜在的AD患者制定治疗计划将是一个有价值的工具。利用视觉记忆表现的初始水平和变化率，这样一种方法能够以快速且经济的方式预测潜在的AD患者。在一个老年病学研究中心对52名女性和145名男性参与者进行了一项纵向病例对照研究，使用病前视觉记忆测试和神经学检查来识别患有和未患有痴呆及AD的个体。
每个个体的分类方法从第二次检查开始，基于其余个体的所有随访信息，每次检查计算该个体患AD的风险。
通过进行交叉验证研究，从受试者工作特征（ROC）曲线得出的敏感性和特异性的最佳组合显示，女性中65%的AD病例和75%的非病例被正确分类，而男性中病例和非病例的正确分类率分别为65%和60%。
认知的纵向测量有助于检测AD的存在。
