Hu Yang, Zhang Yan, Zhang Haihua, Gao Shan, Wang Longcai, Wang Tao, Han Zhifa, Sun Bao-Liang, Liu Guiyou
School of Computer Science and Technology, Harbin Institute of Technology, Harbin, 150080, China.
Department of Pathology, The Affiliated Hospital of Weifang Medical University, Weifang, 261053, China.
Mol Psychiatry. 2022 Oct;27(10):4297-4306. doi: 10.1038/s41380-022-01695-4. Epub 2022 Jul 15.
Mendelian-randomization (MR) studies using large-scale genome-wide association studies (GWAS) have identified causal association between educational attainment and Alzheimer's disease (AD). However, the underlying mechanisms are still required to be explored. Here, we conduct univariable and multivariable MR analyses using large-scale educational attainment, cognitive performance, intelligence and AD GWAS datasets. In stage 1, we found significant causal effects of educational attainment on cognitive performance (beta = 0.907, 95% confidence interval (CI): 0.884-0.930, P < 1.145E-299), and vice versa (beta = 0.571, 95% CI: 0.557-0.585, P < 1.145E-299). In stage 2, we found that both increase in educational attainment (odds ratio (OR) = 0.72, 95% CI: 0.66-0.78, P = 1.39E-14) and cognitive performance (OR = 0.69, 95% CI: 0.64-0.75, P = 1.78E-20) could reduce the risk of AD. In stage 3, we found that educational attainment may protect against AD dependently of cognitive performance (OR = 1.07, 95% CI: 0.90-1.28, P = 4.48E-01), and cognitive performance may protect against AD independently of educational attainment (OR = 0.69, 95% CI: 0.53-0.89, P = 5.00E-03). In stage 4, we found significant causal effects of cognitive performance on intelligence (beta = 0.907, 95% CI: 0.877-0.938, P < 1.145E-299), and vice versa (beta = 0.957, 95% CI: 0.937-0.978, P < 1.145E-299). In stage 5, we identified that cognitive performance may protect against AD independently of intelligence (OR = 0.74, 95% CI: 0.61-0.90, P = 2.00E-03), and intelligence may protect against AD dependently of cognitive performance (OR = 1.17, 95% CI: 0.40-3.43, P = 4.48E-01). Collectively, our univariable and multivariable MR analyses highlight the protective role of cognitive performance in AD independently of educational attainment and intelligence. In addition to the intelligence, we extend the mechanisms underlying the associations of educational attainment with AD.
使用大规模全基因组关联研究（GWAS）的孟德尔随机化（MR）研究已经确定了受教育程度与阿尔茨海默病（AD）之间的因果关系。然而，其潜在机制仍有待探索。在此，我们使用大规模的受教育程度、认知能力、智力和AD的GWAS数据集进行单变量和多变量MR分析。在第一阶段，我们发现受教育程度对认知能力有显著的因果效应（β = 0.907，95%置信区间（CI）：0.884 - 0.930，P < 1.145E - 299），反之亦然（β = 0.571，95% CI：0.557 - 0.585，P < 1.145E - 299）。在第二阶段，我们发现受教育程度的提高（优势比（OR）= 0.72，95% CI：0.66 - 0.78，P = 1.39E - 14）和认知能力（OR = 0.69，95% CI：0.64 - 0.75，P = 1.78E - 20）都可以降低AD的风险。在第三阶段，我们发现受教育程度可能依赖于认知能力来预防AD（OR = 1.07，95% CI：0.90 - 1.28，P = 4.48E - 01），而认知能力可能独立于受教育程度预防AD（OR = 0.69，95% CI：0.53 - 0.89，P = 5.00E - 03）。在第四阶段，我们发现认知能力对智力有显著的因果效应（β = 0.907，95% CI：0.877 - 0.938，P < 1.145E - 299），反之亦然（β = 0.957，95% CI：0.937 - 0.978，P < 1.145E - 299）。在第五阶段，我们确定认知能力可能独立于智力预防AD（OR = 0.74，95% CI：0.61 - 0.90，P = 2.00E - 03），而智力可能依赖于认知能力预防AD（OR = 1.17，95% CI：0.40 - 3.43，P = 4.48E - 01）。总体而言，我们的单变量和多变量MR分析突出了认知能力在AD中独立于受教育程度和智力的保护作用。除了智力，我们扩展了受教育程度与AD关联的潜在机制。
