端粒长度与2型糖尿病：孟德尔随机化研究与多基因风险评分分析

曹岚，李志强，师咏勇，刘赟

研究报告

端粒长度与2型糖尿病：孟德尔随机化研究与多基因风险评分分析

曹岚1,3，李志强2,3，师咏勇3，刘赟4

1. 上海市妇幼保健中心，上海 200062 2. 青岛大学生物医学研究院(暨上海交通大学Bio-X研究院青岛分院)，青岛 266003 3. 上海交通大学Bio-X研究院，遗传发育与精神神经疾病教育部重点实验室，上海 200030 4. 复旦大学生物医学研究院，上海 200032

多项观察性研究表明，端粒长度缩短与2型糖尿病(type 2 diabetes, T2D)之间存在关联。然而，传统观察性研究结果常受到混杂因素和反向因果关联的影响，端粒长度与T2D是否存在因果关联尚不明确。本研究在中国汉族人群中利用孟德尔随机化(Mendelian randomization, MR)和多基因风险评分(polygenic risk score, PRS)方法探索端粒长度与T2D的因果关系。MR研究选取8个与端粒长度相关的独立遗传变异作为工具变量，利用2632例中国汉族人群T2D全基因组关联研究(genome-wide association study, GWAS)数据，检验遗传预测的端粒长度与T2D的关系。利用中国汉族人群GWAS数据，采用PRS分析评价端粒长度PRS与T2D的关系。MR研究共纳入1318例T2D患者和1314例正常对照，逆方差加权、MR-Egger回归、简单中位数和加权中位数法估计的OR值分别为0.78 (95%: 0.36~1.68,= 0.522)、0.23 (95%: 0.01~7.64,= 0.412)、0.60 (95%: 0.28~ 1.28,= 0.185)和0.64 (95%: 0.31~1.33,= 0.233)，遗传预测的较长端粒长度与T2D之间不存在关联。PRS分析未发现端粒长度PRS与T2D显著关联的一致结果。本研究采用MR和PRS方法未发现端粒长度与T2D具有因果关联，后续研究中增大样本量有助于得出更可靠的结论。

孟德尔随机化；多基因风险评分；端粒长度；2型糖尿病

过去几十年中，糖尿病患病率和病例数在全球范围内持续升高[1]。2017年，全球有约4.51亿成人患有糖尿病[2]，而中国估计有超过1亿成人患糖尿病[3]。2型糖尿病(type 2 diabetes, T2D)是一种由遗传和环境因素相互作用导致的复杂疾病[4~6]。T2D的患病率随年龄增加而上升[7]。糖尿病及其并发症给患者家庭和国家造成了巨大的卫生经济负担。

端粒是真核细胞染色体末端的DNA-蛋白质复合体，其功能是维持染色体的完整性[8]。由于DNA末端不能完全复制，正常体细胞端粒会随着细胞分裂逐渐缩短，导致细胞老化[9]。细胞老化是生物老化的重要方面，而端粒长度是细胞老化的重要标志物。端粒长度经常在白细胞中进行测量。白细胞端粒长度(leukocyte telomere length, LTL)具有遗传性，遗传度在36%~84%之间[10]。

多项观察性研究表明，LTL缩短与T2D之间存在关联[11,12]。最近，关于LTL与T2D的meta分析显示缩短的端粒长度与T2D显著相关[13,14]。然而，端粒长度缩短可能是受到疾病或治疗影响并发生在疾病诊断之后，共同的环境因素也可能既影响端粒长度又影响糖尿病风险，导致偏倚的效应估计。

近年，随着全基因组关联研究(genome-wide association study, GWAS)的大量应用，孟德尔随机化(Mendelian randomization, MR)和多基因风险评分(polygenic risk score, PRS)等方法被日益广泛用于发现疾病病因以及因果推断[15~19]。相比传统的观察性流行病学研究，MR研究和PRS分析不会受到常见混杂因素的影响，且因果时序合理。本研究旨在通过MR和PRS方法在中国汉族人群中检验端粒长度与T2D的因果关系。

1 材料与方法

1.1 研究对象

研究对象来自中国汉族人群T2D GWAS的2632名上海居民，包括1318例T2D患者和1314例正常对照。T2D患者均符合WHO糖尿病诊断标准，选取同一地区空腹血糖(fasting plasma glucose, FPG)< 6.1 mmol/L人群作为正常对照[20]。所有2632名研究对象均应用定量PCR测量外周血LTL并进行中国汉族人群LTL GWAS[21]。以上研究已获中国科学院上海生命科学研究院伦理委员会批准(批准号：ER- SIBS-250701)，研究对象均已签署知情同意书。

1.2 孟德尔随机化研究

采用MR方法评估遗传预测的端粒长度与T2D的关系。MR是将与暴露相关联的遗传变异作为工具变量以推断暴露与结局因果关联的一种方法[22]。本研究采用以下标准筛选与端粒长度相关的遗传变异：(1)在已发表的端粒长度GWAS研究中达到全基因组显著性水平(<5×10−8)；(2)在中国人群中的最小等位基因频率(minor allele frequency, MAF)>1%；(3)被选择的遗传变异间不存在明显的连锁不平衡(2<0.01)。符合标准(1)的遗传变异共16个。同时符合标准(1)和标准(2)的遗传变异共12个。本研究最终筛选到8个遗传变异作为工具变量，并获取相关的信息，包括与较长端粒长度相关的等位基因、MAF、效应估计值()、标准误和值。使用已发表端粒长度GWAS中工具变量与端粒长度的效应估计值()和标准误以及2632名中国汉族人群T2D GWAS中工具变量与T2D的效应估计值()和标准误计算因果效应。本研究采用4种MR方法：逆方差加权(inverse-variance weighted, IVW)、MR-Egger回归、简单中位数(simple median estimator, SME)和加权中位数(weighted median estimator, WME)法。此外，通过MR-Egger的截距项评估工具变量是否存在多效性。所有的分析均采用R (version 3.4.0, R Foundation)的软件包‘MendelianRandomization’进行。

1.3 多基因风险评分分析

采用PRS分析检验遗传预测的端粒长度与T2D的关系。PRS分析利用GWAS汇总数据在人群中构建个体遗传评分[23,24]。本研究将2632名研究对象随机分为两组，1316名T2D患者或者正常对照进行T2D GWAS，1316名研究对象进行LTL GWAS。LTL GWAS的研究对象与T2D GWAS的研究对象没有重叠。本研究中端粒长度PRS的构建基于1316名中国人群LTL GWAS的汇总数据。采用PRSice软件[25](http://prsice.info/)进行数据处理和分析，在T2D GWAS研究的1316个个体中计算多个值阈值(P= 0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5)的端粒长度PRS。PRS分析采用Bonferroni法进行多重检验校正，校正后显著性阈值设为0.05/7 = 0.007。

2 结果与分析

2.1 端粒长度与T2D的孟德尔随机化研究

2.1.1 工具变量信息

根据本研究工具变量筛选标准，最终筛选到8个独立的遗传变异作为工具变量[26~28]。表1列出了8个遗传变异的相关信息，包括所在染色体、临近基因、效应等位基因、MAF、与端粒长度关联的系数、与T2D关联的系数等。其中，6个遗传变异与端粒长度和T2D具有相反的效应方向，1个遗传变异与T2D关联的值小于0.05。

2.1.2 孟德尔随机化研究结果

IVW、MR-Egger回归、SME和WME法的OR值分别为0.78 (95%: 0.36~1.68,= 0.522)、0.23 (95%: 0.01~7.64,= 0.412)、0.60 (95%: 0.28~ 1.28,= 0.185)、0.64 (95%: 0.31~1.33，= 0.233)，表明遗传预测的较长端粒长度与T2D之间不存在关联。此外，MR-Egger回归的截距为0.110 (95%: –0.198~0.417,= 0.485)，表明工具变量不存在多效性(图1)。

进一步根据年龄将研究对象分为≤60岁和>60岁两层。在≤60岁的研究对象中，IVW法的OR值为0.60 (95%: 0.27~1.33,= 0.211)。在>60岁的研究对象中，IVW法的OR值为1.22 (95%: 0.36~ 4.08,= 0.751)。在各层均未发现遗传预测的较长端粒长度与T2D具有关联。

表1 与端粒长度相关的遗传变异

SNP：single-nucleotide polymorphism，单核苷酸多态性；Chr：染色体；效应等位基因：与较长端粒长度相关的等位基因；MAF：最小等位基因频率，来自既往GWAS研究；T2D：2型糖尿病；：效应估计值；“*”表示增加一个效应等位基因时端粒长度的增加量(kb)。

图1 不同孟德尔随机化方法分析结果

T2D：2型糖尿病；IVW：逆方差加权法；SME：简单中位数法；WME：加权中位数法。

2.2 端粒长度与T2D的多基因风险评分分析

在1316名T2D或健康对照人群中构建端粒长度PRS以检验端粒长度PRS与T2D的关系。仅有一个值阈值的端粒长度PRS与T2D存在关联(= 0.015)，但经过Bonferroni校正后，此关联无统计学意义(图2)。

3 讨论

到目前为止，多项观察性研究表明端粒长度缩短与T2D之间存在关联。本课题组前期在4016例中国汉族人群中进行的一项病例对照研究也发现较短的LTL与T2D相关(OR = 1.52, 95%: 1.23~1.88,= 0.0001)[29]。最近，一项关于端粒长度与T2D的meta分析显示缩短的端粒长度与T2D的关联有统计学意义(OR = 1.117, 95%: 1.002~1.246,= 0.045)[13]。D’Mello等[14]进行的meta分析也显示缩短的LTL与T2D有关联关系(OR = 1.37, 95%: 1.10~1.72)。端粒孟德尔随机化合作组织[30]于2017年发表的MR研究未发现遗传预测的较长端粒长度与T2D存在关联，但却发现遗传预测的较长端粒长度降低1型糖尿病的风险(OR = 0.71, 95%: 0.51~0.98,= 0.04)。本研究采用MR和PRS方法，在中国汉族人群中评估端粒长度和T2D的因果关系，没有发现遗传预测的较长端粒长度和T2D存在任何显著关联。

图2 端粒长度PRS与T2D的关联

本研究中MR分析选取的工具变量均为欧洲人群发现的与端粒长度相关的遗传变异。本课题组在前期的研究中验证了欧洲人群发现的附近位点rs12696304和rs16847897在中国汉族人群中与LTL相关(= 4.5×10–3和9.5×10–5)[31]。此外，在中国汉族人群GWAS研究中发现上的位点rs2736100与端粒长度相关(= 1.93×10–5)[21]，该发现与欧洲人群研究结果一致[26]。一项在亚洲人群进行的MR研究也表明欧洲人群发现的端粒长度相关遗传变异可以有效应用于亚洲人群[32]。

在传统的病例对照研究中，端粒长度缩短可能发生在疾病诊断之后并由疾病或治疗导致，故其结果常受反向因果关联的干扰，影响其论证因果关系的能力。本研究中遗传预测的端粒长度与抽血、疾病诊断时间无关，遗传变异先于疾病的发生，符合因果推断中“先因后果”的时序性要求。此外，本研究运用遗传预测的端粒长度，有利于将影响端粒长度的遗传因素与非遗传因素进行区分。常见影响端粒长度的非遗传因素包括衰老、氧化损伤等。

与其他研究相比，本研究具有以下优势：(1)选取与端粒长度相关的8个独立的遗传变异作为工具变量，避免连锁不平衡对因果估计结果的影响；(2)采用了多种MR方法。本研究也存在局限性：LTL GWAS和T2D GWAS的样本量较小，PRS分析的把握度较低。

综上所述，本研究在中国汉族人群中采用MR和PRS方法未发现端粒长度与T2D具有因果关联。后续研究中发现更多新的端粒长度相关遗传变异并增大样本量有助于得出更可靠的结论。

[1] NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in diabetes since 1980: a pooled analysis of 751 population-based studies with 4.4 million participants., 2016, 387(10027): 1513–1530.

[2] Cho NH, Shaw JE, Karuranga S, Huang Y, da Rocha Fernandes JD, Ohlrogge AW, Malanda B. IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045., 2018, 138: 271–281.

[3] Bragg F, Holmes MV, Iona A, Guo Y, Du HD, Chen YP, Bian Z, Yang L, Herrington W, Bennett D, Turnbull I, Liu YM, Feng SX, Chen JS, Clarke R, Collins R, Peto R, Li LM, Chen ZM, China Kadoorie Biobank Collaborative Group. Association between diabetes and cause-specific mortality in rural and urban areas of China., 2017, 317(3): 280–289.

[4] Gao KP, Ren YC, Wang JJ, Liu ZC, Li JN, Li LL, Wang BY, Li H, Wang YX, Cao YK, Ohno K, Zhai RH, Liang Z. Interactions between genetic polymorphisms of glucose metabolizing genes and smoking and alcohol consumption in the risk of type 2 diabetes mellitus., 2017, 42(12): 1316–1321.

[5] Huang X, Chen YQ, Xu GL, Peng SH. DNA methylation in adipose tissue and the development of diabetes and obesity., 2019, 41(2): 98–110.黄鑫, 陈永强, 徐国良, 彭淑红. 脂肪组织DNA甲基化与糖尿病和肥胖的发生发展. 遗传, 2019, 41(2): 98–110.

[6] Wang YZ, Zhang YM, Dong XL, Wang XC, Zhu JF, Wang N, Jiang F, Chen Y, Jiang QW, Fu CW. Modification effects of T2DM-susceptible SNPs on the reduction of blood glucose in response to lifestyle interventions., 2020, 42(5): 483–493.王玉琢, 张一鸣, 董晓莲, 王学才, 朱建福, 王娜, 江峰, 陈跃, 姜庆五, 付朝伟. 2型糖尿病易感基因SNP位点对生活方式干预降低血糖应答效果的修饰效应. 遗传, 2020, 42(5): 483–493.

[7] Han CY, Zhang M, Luo XP, Wang CJ, Yin L, Pang C, Feng TP, Ren YC, Wang BY, Zhang L, Li LL, Yang XY, Zhang HY, Zhao Y, Zhou JM, Xie ZH, Zhao JZ, Hu DS. Secular trends in the prevalence of type 2 diabetes in adults in China from 1995 to 2014: A meta-analysis., 2017, 9(5): 450–461.

[8] Moyzis RK, Buckingham JM, Cram LS, Dani M, Deaven LL, Jones MD, Meyne J, Ratliff RL, Wu JR. A highly conserved repetitive DNA sequence, (TTAGGG)n, present at the telomeres of human chromosomes., 1988, 85(18): 6622–6626.

[9] Wolkowitz OM, Reus VI, Mellon SH. Of sound mind and body: depression, disease, and accelerated aging., 2011, 13(1): 25–39.

[10] Aviv A. Genetics of leukocyte telomere length and its role in atherosclerosis., 2012, 730(1–2): 68–74.

[11] Sampson MJ, Winterbone MS, Hughes JC, Dozio N, Hughes DA. Monocyte telomere shortening and oxidative DNA damage in type 2 diabetes., 2006, 29(2): 283–289.

[12] Zee RY, Castonguay AJ, Barton NS, Germer S, Martin M. Mean leukocyte telomere length shortening and type 2 diabetes mellitus: a case-control study., 2010, 155(4): 166–169.

[13] Zhao JZ, Miao K, Wang HR, Ding H, Wang DW. Association between telomere length and type 2 diabetes mellitus: a meta-analysis., 2013, 8(11): e79993.

[14] D'Mello MJJ, Ross SA, Briel M, Anand SS, Gerstein H, Paré G. Association between shortened leukocyte telomere length and cardiometabolic outcomes: systematic review and meta-analysis., 2015, 8(1): 82–90.

[15] Dudbridge F. Polygenic epidemiology., 2016, 40(4): 268–272.

[16] Meng LX, Que XM, Gao X, Wang T. Childhood obesity and coronary artery disease: a Mendelian randomization study., 2019, 40(7): 839–843.孟玲先, 阙喜妹, 高雪, 王彤. 儿童肥胖与冠状动脉疾病的孟德尔随机化研究. 中华流行病学杂志, 2019, 40(7): 839–843.

[17] Polimanti R, Ratanatharathorn A, Maihofer AX, Choi KW, Stein MB, Morey RA, Logue MW, Nievergelt C M, Stein DJ, Koenen KC, Gelernter J, Psychiatric Genomics Consortium Posttraumatic Stress Disorder Working Group. Association of economic status and educational attainment with posttraumatic stress disorder: a Mendelian randomi­zation study., 2019, 2(5): e193447.

[18] Colodro-Conde L, Couvy-Duchesne B, Whitfield JB, Streit F, Gordon S, Kemper KE, Yengo L, Zheng ZL, Trzaskowski M, de Zeeuw EL, Nivard MG, Das M, Neale RE, MacGregor S, Olsen CM, Whiteman DC, Boomsma DI, Yang J, Rietschel M, McGrath JJ, Medland SE, Martin NG. Association between population density and genetic risk for schizophrenia., 2018, 75(9): 901–910.

[19] Chen X, Yazdani S, Piehl F, Magnusson PKE, Fang F. Polygenic link between blood lipids and amyotrophic lateral sclerosis., 2018, 67: 202.e1–202.e6.

[20] Liu Y, Zhou DZ, Zhang D, Chen Z, Zhao T, Zhang Z, Ning M, Hu X, Yang YF, Zhang ZF, Yu L, He L, Xu H. Variants in KCNQ1 are associated with susceptibility to type 2 diabetes in the population of mainland China., 2009, 52(7): 1315–1321.

[21] Liu Y, Cao L, Li ZQ, Zhou DZ, Liu WQ, Shen Q, Wu YT, Zhang D, Hu X, Wang T, Ye JY, Weng XL, Zhang H, Zhang D, Zhang Z, Liu FT, He L, Shi YY. A genome-wide association study identifies a locus on TERT for mean telomere length in Han Chinese., 2014, 9(1): e85043.

[22] Didelez V, Sheehan N. Mendelian randomization as an instrumental variable approach to causal inference., 2007, 16(4): 309–330.

[23] International Schizophrenia Consortium, Purcell SM, Wray NR, Stone JL, Visscher PM, O'Donovan MC, Sul­livan PF, Sklar P. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder., 2009, 460(7256): 748–752.

[24] So HC, Chau KL, Ao FK, Mo CH, Sham PC. Exploring shared genetic bases and causal relationships of schizo­phrenia and bipolar disorder with 28 cardiovascular and metabolic traits., 2019, 49(8): 1286–1298.

[25] Euesden J, Lewis CM, O'Reilly PF. PRSice: Polygenic Risk Score software., 2015, 31(9): 1466– 1468.

[26] Codd V, Nelson CP, Albrecht E, Mangino M, Deelen J, Buxton JL, Hottenga JJ, Fischer K, Esko T, Surakka I, Broer L, Nyholt DR, Leach IM, Salo P, Hägg S, Matthews MK, Palmen J, Norata GD, O'Reilly PF, Saleheen D, Amin N, Balmforth AJ, Beekman M, de Boer RA, Böhringer S, Braund PS, Burton PR, de Craen AJ, Denniff M, Dong Y, Douroudis K, Dubinina E, Eriksson JG, Garlaschelli K, Guo DH, Hartikainen AL, Henders AK, Houwing- Duistermaat JJ, Kananen L, Karssen LC, Kettunen J, Klopp N, Lagou V, van Leeuwen EM, Madden PA, Mägi R, Magnusson PK, Männistö S, McCarthy MI, Medland SE, Mihailov E, Montgomery GW, Oostra BA, Palotie A, Peters A, Pollard H, Pouta A, Prokopenko I, Ripatti S, Salomaa V, Suchiman HE, Valdes AM, Verweij N, Viñuela A, Wang XL, Wichmann HE, Widen E, Willemsen G, Wright MJ, Xia K, Xiao XJ, van Veldhuisen DJ, Catapano AL, Tobin MD, Hall AS, Blakemore AI, van Gilst WH, Zhu HD, Erdmann J, Reilly MP, Kathiresan S, Schunkert H, Talmud PJ, Pedersen NL, Perola M, Ouwehand W, Kaprio J, Martin NG, van Duijn CM, Hovatta I, Gieger C, Metspalu A, Boomsma DI, Jarvelin MR, Slagboom PE, Thompson JR, Spector TD, van der Harst P, Samani NJ. Identification of seven loci affecting mean telomere length and their association with disease., 2013, 45(4): 422–7, 427e1–2.

[27] Mangino M, Hwang SJ, Spector TD, Hunt SC, Kimura M, Fitzpatrick AL, Christiansen L, Petersen I, Elbers CC, Harris T, Chen W, Srinivasan SR, Kark JD, Benetos A, El Shamieh S, Visvikis-Siest S, Christensen K, Berenson GS, Valdes AM, Viñuela A, Garcia M, Arnett DK, Broeckel U, Province MA, Pankow JS, Kammerer C, Liu Y, Nalls M, Tishkoff S, Thomas F, Ziv E, Psaty BM, Bis JC, Rotter JI, Taylor KD, Smith E, Schork NJ, Levy D, Aviv A. Genome-wide meta-analysis points to CTC1 and ZNF676 as genes regulating telomere homeostasis in humans., 2012, 21(24): 5385–5394.

[28] Levy D, Neuhausen SL, Hunt SC, Kimura M, Hwang SJ, Chen W, Bis JC, Fitzpatrick AL, Smith E, Johnson AD, Gardner JP, Srinivasan SR, Schork N, Rotter JI, Herbig U, Psaty BM, Sastrasinh M, Murray SS, Vasan RS, Province MA, Glazer NL, Lu XB, Cao XJ, Kronmal R, Mangino M, Soranzo N, Spector TD, Berenson GS, Aviv A. Genome- wide association identifies OBFC1 as a locus involved in human leukocyte telomere biology., 2010, 107(20): 9293–9298.

[29] Shen Q, Zhao XZ, Yu L, Zhang Z, Zhou DZ, Kan MY, Zhang D, Cao L, Xing QH, Yang YF, Xu H, He L, Liu Y. Association of leukocyte telomere length with type 2 diabetes in mainland Chinese populations., 2012, 97(4): 1371–1374.

[30] Telomeres Mendelian Randomization Collaboration, Haycock PC, Burgess S, Nounu A, Zheng J, Okoli GN, Bowden J, Wade KH, Timpson NJ, Evans DM, Willeit P, Aviv A, Gaunt TR, Hemani G, Mangino M, Ellis HP, Kurian KM, Pooley KA, Eeles RA, Lee JE, Fang SY, Chen WV, Law MH, Bowdler LM, Iles MM, Yang Q, Worrall BB, Markus HS, Hung RJ, Amos CI, Spurdle AB, Thompson DJ, O'Mara TA, Wolpin B, Amundadottir L, Stolzenberg-Solomon R, Trichopoulou A, Onland-Moret NC, Lund E, Duell EJ, Canzian F, Severi G, Overvad K, Gunter MJ, Tumino R, Svenson U, van Rij A, Baas AF, Bown MJ, Samani NJ, van t'Hof FNG, Tromp G, Jones GT, Kuivaniemi H, Elmore JR, Johansson M, Mckay J, Scelo G, Carreras-Torres R, Gaborieau V, Brennan P, Bracci PM, Neale RE, Olson SH, Gallinger S, Li D, Petersen GM, Risch HA, Klein AP, Han J, Abnet CC, Freedman ND, Taylor PR, Maris JM, Aben KK, Kiemeney LA, Vermeulen SH, Wiencke JK, Walsh KM, Wrensch M, Rice T, Turnbull C, Litchfield K, Paternoster L, Standl M, Abecasis GR, SanGiovanni JP, Li Y, Mijatovic V, Sapkota Y, Low SK, Zondervan KT, Montgomery GW, Nyholt DR, van Heel DA, Hunt K, Arking DE, Ashar FN, Sotoodehnia N, Woo D, Rosand J, Comeau ME, Brown WM, Silverman EK, Hokanson JE, Cho MH, Hui J, Ferreira MA, Thompson PJ, Morrison AC, Felix JF, Smith NL, Christiano AM, Petukhova L, Betz RC, Fan X, Zhang XJ, Zhu CH, Langefeld CD, Thompson SD, Wang FJ, Lin X, Schwartz DA, Fingerlin T, Rotter JI, Cotch MF, Jensen RA, Munz M, Dommisch H, Schaefer AS, Han F, Ollila HM, Hillary RP, Albagha O, Ralston SH, Zeng CJ, Zheng W, Shu XO, Reis A, Uebe S, Hüffmeier U, Kawamura Y, Otowa T, Sasaki T, Hibberd ML, Davila S, Xie G, Siminovitch K, Bei JX, Zeng YX, Försti A, Chen BW, Landi S, Franke A, Fischer A, Ellinghaus D, Flores C, Noth I, Ma SF, Foo JN, Liu J, Kim JW, Cox DG, Delattre O, Mirabeau O, Skibola CF, Tang CS, Garcia-Barcelo M, Chang KP, Su WH, Chang YS, Martin NG, Gordon S, Wade TD, Lee C, Kubo M, Cha PC, Nakamura Y, Levy D, Kimura M, Hwang SJ, Hunt S, Spector T, Soranzo N, Manichaikul AW, Barr RG, Kahali B, Speliotes E, Yerges-Armstrong LM, Cheng CY, Jonas JB, Wong TY, Fogh I, Lin K, Powell JF, Rice K, Relton CL, Martin RM, Davey Smith G. Association between telomere length and risk of cancer and non-neoplastic diseases: a Mendelian randomization study., 2017, 3(5): 636–651.

[31] Shen Q, Zhang Z, Yu L, Cao L, Zhou DZ, Kan MY, Li BJ, Zhang D, He L, Liu Y. Common variants near TERC are associated with leukocyte telomere length in the Chinese Han population., 2011, 19(6): 721–723.

[32] Machiela MJ, Hsiung CA, Shu XO, Seow WJ, Wang ZM, Matsuo K, Hong YC, Seow A, Wu C, Hosgood HD, Chen KX, Wang JC, Wen WQ, Cawthon R, Chatterjee N, Hu W, Caporaso NE, Park JY, Chen CJ, Kim YH, Kim YT, Landi MT, Shen HB, Lawrence C, Burdett L, Yeager M, Chang IS, Mitsudomi T, Kim HN, Chang GC, Bassig BA, Tucker M, Wei FS, Yin ZH, An SJ, Qian BY, Lee VH, Lu DR, Liu JJ, Jeon HS, Hsiao CF, Sung JS, Kim JH, Gao YT, Tsai YH, Jung YJ, Guo H, Hu ZB, Hutchinson A, Wang WC, Klein RJ, Chung CC, Oh IJ, Chen KY, Berndt SI, Wu W, Chang J, Zhang XC, Huang MS, Zheng H, Wang JW, Zhao XY, Li YQ, Choi JE, Su WC, Park KH, Sung SW, Chen YM, Liu L, Kang CH, Hu LM, Chen CH, Pao W, Kim YC, Yang TY, Xu J, Guan P, Tan W, Su J, Wang CL, Li HX, Sihoe AD, Zhao ZH, Chen Y, Choi YY, Hung JY, Kim JS, Yoon HI, Cai QY, Lin CC, Park IK, Xu P, Dong J, Kim C, He Q, Perng RP, Kohno T, Kweon SS, Chen CY, Vermeulen RC, Wu JJ, Lim WY, Chen KC, Chow WH, Ji BT, Chan JK, Chu MJ, Li YJ, Yokota J, Li J, Chen H, Xiang YB, Yu CJ, Kunitoh H, Wu G, Jin L, Lo YL, Shiraishi K, Chen YH, Lin HC, Wu TC, Wong MP, Wu YL, Yang PC, Zhou BS, Shin MH, Fraumeni JF Jr, Zheng W, Lin DX, Chanock SJ, Rothman N, Lan Q. Genetic variants associated with longer telomere length are associated with increased lung cancer risk among never-smoking women in Asia: a report from the female lung cancer consortium in Asia., 2015, 137(2): 311–319.

Telomere length and type 2 diabetes: Mendelian randomization study and polygenic risk score analysis

Lan Cao1,3, Zhiqiang Li2,3, Yongyong Shi3, Yun Liu4

Recent epidemiological studies suggest an association between shorter telomere length and higher risk for type 2 diabetes (T2D). However, results from observational studies are susceptible to confounding and reverse causation, and it is not clear whether there is a causal association between telomere length and T2D. Using Mendelian randomization (MR) and polygenic risk score (PRS) approaches, we had evaluated the causal effect of telomere length on T2D in the Chinese Han population. Using 8 telomere-length associated genetic variants as instrumental variables, an analysis of genetically predicted telomere length and T2D risk was performed in the MR study based on data from a T2D genome-wide association study (GWAS) in 2632 individuals (1318 cases and 1314 controls). We also applied a PRS approach to investigate the causal relationship using Chinese GWAS data. The inverse-variance weighted, MR-Egger regression, simple median, and weighted median methods yielded no evidence of association between genetically predicted longer telomere length and risk of T2D (OR = 0.78, 95%: 0.36 ~ 1.68,= 0.522; OR = 0.23, 95%: 0.01 ~ 7.64,= 0.412; OR = 0.60, 95%: 0.28 ~ 1.28,= 0.185; OR = 0.64, 95%: 0.31 ~ 1.33，= 0.233; respectively). Further, PRS analysis did not produce consistent genetic overlap between telomere length and T2D. Accordingly, this study found no evidence supporting a causal association between telomere length and T2D. Further studies with larger cohorts could yield more reliable results and conclusions.

Mendelian randomization; polygenic risk score; telomere length; type 2 diabetes

2020-03-18;

2020-05-22

上海市卫生和计划生育委员会科研课题项目(编号：20164Y0163)资助[Supported by Foundation of Shanghai Municipal Health Commission (No. 20164Y0163)]

曹岚，博士，研究方向：复杂疾病的遗传学。E-mail: caolan@sjtu.edu.cn

曹岚。

10.16288/j.yczz.20-077

2020/9/2 11:40:03

URI: https://kns.cnki.net/kcms/detail/11.1913.R.20200901.1436.001.html

(责任编委: 陈雁)