Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator

Genetic Epidemiology - Tập 40 Số 4 - Trang 304-314 - 2016
Jack Bowden1, George Davey Smith1, Philip Haycock1, Stephen Burgess2
1Integrative Epidemiology Unit University of Bristol Bristol United Kingdom
2Department of Public Health and Primary Care, University of Cambridge, Cambridge, United Kingdom

Tóm tắt

ABSTRACT

Developments in genome‐wide association studies and the increasing availability of summary genetic association data have made application of Mendelian randomization relatively straightforward. However, obtaining reliable results from a Mendelian randomization investigation remains problematic, as the conventional inverse‐variance weighted method only gives consistent estimates if all of the genetic variants in the analysis are valid instrumental variables. We present a novel weighted median estimator for combining data on multiple genetic variants into a single causal estimate. This estimator is consistent even when up to 50% of the information comes from invalid instrumental variables. In a simulation analysis, it is shown to have better finite‐sample Type 1 error rates than the inverse‐variance weighted method, and is complementary to the recently proposed MR‐Egger (Mendelian randomization‐Egger) regression method. In analyses of the causal effects of low‐density lipoprotein cholesterol and high‐density lipoprotein cholesterol on coronary artery disease risk, the inverse‐variance weighted method suggests a causal effect of both lipid fractions, whereas the weighted median and MR‐Egger regression methods suggest a null effect of high‐density lipoprotein cholesterol that corresponds with the experimental evidence. Both median‐based and MR‐Egger regression methods should be considered as sensitivity analyses for Mendelian randomization investigations with multiple genetic variants.

Từ khóa


Tài liệu tham khảo

10.1093/ije/dym186

10.1002/sim.2487

10.1093/ije/dyv080

10.1201/b18084

10.1093/ije/dyr036

10.1002/gepi.21758

10.1093/aje/kwv017

10.1007/s10654-015-0011-z

CARDIoGRAMplusC4D Consortium, 2013, Large‐scale association analysis identifies new risk loci for coronary artery disease, Nat Genet, 45, 25, 10.1038/ng.2480

10.1111/j.1365-2125.2003.02060.x

10.1093/ije/dyg070

10.1093/ije/dyh132

10.1136/bmj.330.7499.1076

10.1093/hmg/ddu328

10.1001/jama.2009.1619

10.1177/0962280206077743

10.1038/ng.2795

10.1111/j.0006-341X.2000.00455.x

10.1136/bmj.315.7109.629

10.1093/ije/dyt110

10.1038/ng.2797

10.1093/aje/kwr323

10.1002/sim.6522

10.1093/ije/29.4.722

10.1016/j.econlet.2008.09.004

10.1097/01.ede.0000135174.63482.43

10.1093/eurheartj/eht571

JohnsonT.2013.Efficient calculation for multi‐SNP genetic risk scores. Technical report The Comprehensive R Archive Network. Available athttp://cran.r‐project.org/web/packages/gtx/vignettes/ashg2012.pdf[last accessed 2014/11/19].

10.1080/01621459.2014.994705

Kolesár M, 2014, Identification and inference with many invalid instruments, J Bus Econ Stat

10.1002/sim.3034

10.1371/journal.pone.0002986

10.1097/01.ede.0000215160.88317.cb

10.1093/aje/kwj062

Pedersen TR, 1994, Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S), Lancet, 344, 1383

PickrellJ.2015.Detection and interpretation of shared genetic influences on 40 human traits. Technical report bioRxiv. Available athttp://biorxiv.org/content/early/2015/04/16/018150.

10.1093/aje/kwt084

10.1056/NEJMoa1206797

10.1016/j.annepidem.2006.12.005

10.1016/S0140-6736(12)60312-2

WindmeijerF FarbmacherH DaviesN Davey SmithG WhiteI.2015.Selecting (in)valid instruments for instrumental variables estimation. Available athttp://www.hec.unil.ch/documents/seminars/iems/1849.pdf.