PT  - JOURNAL ARTICLE
AU  - Fernando Pires Hartwig
AU  - George Davey Smith
AU  - Jack Bowden
TI  - Robust inference in summary data Mendelian randomisation via the zero modal pleiotropy assumption
AID  - 10.1101/126102
DP  - 2017 Jan 01
TA  - bioRxiv
PG  - 126102
4099  - http://biorxiv.org/content/early/2017/05/10/126102.short
4100  - http://biorxiv.org/content/early/2017/05/10/126102.full
AB  - Background Mendelian randomisation (MR) is being increasingly used to strengthen causal inference in observational studies. Availability of summary data of genetic associations for a variety of phenotypes from large genome-wide association studies (GWAS) allows straightforward application of MR using summary data methods, typically in a two-sample design. In addition to the conventional inverse variance weighting (IVW) method, recently developed summary data MR methods, such as the MR-Egger and weighted median approaches, allow a relaxation of the instrumental variable assumptions.Methods Here, a new method –the mode-based estimate (MBE) – is proposed to obtain a single causal effect estimate from multiple genetic instruments. The MBE is consistent when the largest number of similar (identical in infinite samples) individual-instrument causal effect estimates comes from valid instruments, even if the majority of instruments are invalid. We evaluate the performance of the method in simulations designed to mimic the two-sample summary data setting, and demonstrate its use by investigating the causal effect of plasma lipid fractions and urate levels on coronary heart disease risk.Results The MBE presented less bias and type-I error rates than other methods under the null in many situations. Its power to detect a causal effect was smaller compared to the IVW and weighted median methods, but was larger than that of MR-Egger regression, with sample size requirements typically smaller than those available from GWAS consortia.Conclusions The MBE relaxes the instrumental variable assumptions, and should be used in combination with other approaches in a sensitivity analysis.Key MessagesSummary data Mendelian randomisation, typically in a two-sample setting, is being increasingly used due to the availability of summary association results from large genome- wide association studies.Mendelian randomisation analyses using multiple genetic instruments are prone to bias due to horizontal pleiotropy, especially when genetic instruments are selected based solely on statistical criteria.A causal effect estimate robust to horizontal pleiotropy can be obtained using the mode- based estimate (MBE).The MBE requires that the most common causal effect estimate is a consistent estimate of the true causal effect, even if the majority of instruments are invalid (i.e., the ZEro Modal Pleiotropy Assumption, or ZEMPA).Plotting the smoothed empirical density function is useful to explore the distribution of causal effect estimates, and to understand how the MBE is determined.