As Mendelian randomization with multiple IVs can be viewed as a meta-analysis of summarized genetic association estimates, methods and diagnostic tools developed for meta-analysis can also be used for Mendelian randomization. This is particularly relevant as summarized genetic association estimates from large consortia are increasingly becoming publicly available (such as those from the CARDIoGRAM consortium used in this paper).44 It has been shown that Mendelian randomization analyses based on summarized data are as efficient as those based on individual-level data.23 Other tools from the meta-analysis literature include methods for bias adjustment, such as the trim-and-fill method,45 and the use of pseudo-data.46 Another diagnostic tool is a heterogeneity test, which tests whether differences between estimates from different studies are compatible with chance variation.47 This can be performed using Cochran’s Q statistic.44 The null hypothesis is that the underlying association is the same in each study. In the Mendelian randomization context, we can test whether causal estimates from different genetic variants are compatible. Considerable heterogeneity would be evidence that the genetic variants are estimating different quantities, and would cast doubt on the
