Bordeaux University, ISPED, Centre INSERM U897-Epidemiologie-Biostatistique & CIC-1401, CHU de Bordeaux, Pole de Sante Publique, Bordeaux, France.
Department of Social and Behavorial Sciences, Harvard T.H. Chan School of Public Health, Boston, MA, USA.
Mol Psychiatry. 2015 Oct;20(10):1173-8. doi: 10.1038/mp.2015.62. Epub 2015 Jun 2.
Several genetic polymorphisms have been associated with Late Onset Alzheimer's Disease (LOAD), but there has been limited evidence on whether these polymorphisms predict intermediary stage outcomes such as cognitive changes in prospective community-based studies. Our aim was to evaluate whether polymorphisms previously established as predictors of LOAD also predict worse cognitive function and accelerated decline across multiple cognitive domains. We analyzed data from the 3C-Dijon study, in which 4931 respondents aged 65+ were examined up to 5 times over 10 years with a neuropsychological assessment. We evaluated the associations of polymorphisms in APOE, CR1, BIN1, CLU, PICALM, ABCA7, MS4A6A, CD33, MS4A4E and CD2AP with level and change in 5 neuropsychological tests, assuming a dominant effect model. To optimize measurement, we used a mixed regression model with a latent process for each cognitive domain: global cognition (Mini Mental State Examination); verbal fluency (Isaac's Set Test); visual memory (Benton Visual Retention Test); information processing (Trail Making Test B) and literacy (National Adult Reading Test). APOE was associated with accelerated decline in global cognition and verbal fluency. Only two non-APOE genetic polymorphisms were associated with cognitive decline: CR1 was associated with rate of change in verbal fluency and BIN1 was associated with rate of change in global cognition. In a large prospective population-based study of dementia-free individuals, only a few cognitive domains were associated with established LOAD risk alleles. The most consistent associations were for global cognition and verbal fluency.
几种遗传多态性与晚发性阿尔茨海默病（LOAD）有关，但在基于社区的前瞻性研究中，这些多态性是否能预测中间阶段的结果（如认知变化），证据有限。我们的目的是评估先前被确定为 LOAD 预测因子的多态性是否也能预测认知功能更差和多个认知领域的加速下降。我们分析了 3C-第戎研究的数据，该研究共纳入了 4931 名年龄在 65 岁以上的受访者，在 10 年内接受了多达 5 次神经心理学评估。我们评估了 APOE、CR1、BIN1、CLU、PICALM、ABCA7、MS4A6A、CD33、MS4A4E 和 CD2AP 基因多态性与 5 项神经心理学测试的水平和变化的相关性，假设了显性效应模型。为了优化测量，我们使用了一个混合回归模型，为每个认知领域建立一个潜在过程：整体认知（简易精神状态检查）；言语流畅性（艾萨克的设定测试）；视觉记忆（本顿视觉保留测试）；信息处理（追踪测试 B）和读写能力（全国成人阅读测试）。APOE 与整体认知和言语流畅性的加速下降有关。只有两种非 APOE 遗传多态性与认知下降有关：CR1 与言语流畅性的变化率有关，BIN1 与整体认知的变化率有关。在一项针对无痴呆个体的大型前瞻性基于人群的研究中，只有少数认知领域与已确立的 LOAD 风险等位基因有关。最一致的关联是整体认知和言语流畅性。
