Department of Health Psychology, University of Missouri, Columbia, Missouri, USA.
Department of Neurology, University of Texas - Austin, Austin, Texas, USA.
Dement Geriatr Cogn Disord. 2021;50(3):231-236. doi: 10.1159/000516413. Epub 2021 Jun 29.
Our understanding of Alzheimer's disease may be improved by harmonizing data from large cohort studies of older adults. Differences in the way clinical conditions, like mild cognitive impairment (MCI), are diagnosed may lead to variability among participants that share the same diagnostic label. This variability presents a challenge for cohort harmonization and may lead to inconsistency in research findings. Little research to date has explored the equivalence of the diagnostic label of MCI across 2 of the largest and most influential cohort studies in the USA: the National Alzheimer's Coordinating Center (NACC) and the Alzheimer's Disease Neuroimaging Initiative (ADNI).
Participants with MCI due to presumed Alzheimer's disease from the NACC Uniform Data Set (n = 789) and ADNI (n = 131) were compared on demographic, psychological, and functional variables, as well as on an abbreviated neuropsychological battery common to the 2 data sets.
Though similar in terms of age, education, and functional status, the NACC sample was more diverse (17.4% non-White participants vs. 7.6% in ADNI; χ2 = 7.923, p = 0.005) and tended to perform worse on some cognitive tests. In particular, participants diagnosed with MCI in NACC were more likely to have clinically significant impairments on language measures (26.36-31.18%) than MCI participants in ADNI (16.03-19.85%).
The current findings suggest important differences in cognitive performances between 2 large MCI cohorts, likely reflective of differences in diagnostic criteria used in these 2 studies, as well as differences in sample compositions. Such diagnostic heterogeneity may make harmonizing data across these cohorts challenging. However, application of shared psychometric criteria across studies may lead to closer equivalence of MCI groups. Such approaches could pave the way for cohort harmonization and enable "big data" analytic approaches to understanding Alzhei-mer's to be developed.
通过协调老年人大型队列研究的数据，我们对阿尔茨海默病的认识可能会得到提高。轻度认知障碍（MCI）等临床病症的诊断方式的差异可能导致具有相同诊断标签的参与者之间存在差异。这种变异性给队列协调带来了挑战，并可能导致研究结果不一致。迄今为止，很少有研究探讨美国最大和最有影响力的两项队列研究（国家阿尔茨海默病协调中心[NACC]和阿尔茨海默病神经影像学倡议[ADNI]）中 MCI 诊断标签的等效性。
将 NACC 统一数据集（n = 789）和 ADNI（n = 131）中因疑似阿尔茨海默病而患有 MCI 的参与者在人口统计学、心理学和功能变量方面进行比较，以及在两个数据集共有的简短神经心理学测试中进行比较。
尽管 NACC 样本在年龄、教育和功能状态方面相似，但更具多样性（17.4%的非白人参与者，而 ADNI 为 7.6%；χ2 = 7.923，p = 0.005），并且在某些认知测试中表现更差。特别是，在 NACC 中诊断为 MCI 的参与者在语言测试中（26.36-31.18%）更有可能出现临床显著的损伤，而 ADNI 中的 MCI 参与者（16.03-19.85%）则较少。
目前的研究结果表明，两个大型 MCI 队列之间的认知表现存在重要差异，这可能反映了这两项研究中使用的诊断标准以及样本组成的差异。这种诊断异质性可能使这些队列的数据协调具有挑战性。然而，在研究中应用共同的心理测量标准可能会使 MCI 组更加等效。这些方法可以为队列协调铺平道路，并使理解阿尔茨海默病的“大数据”分析方法得以发展。
