Avoiding bias from weak instruments in Mendelian randomization studies.

By Stephen Burgess, S. Thompson. International Journal of Epidemiology, 2011

3,291 citations recorded by the source corpus.

Subfields: Genetics, Pulmonary and Respiratory Medicine

View Paper: https://www.semanticscholar.org/paper/ecbe09df99442653ec12188341b116dd946e9176

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Genetics

1. J. Bowden
2. George Davey Smith
3. Stephen Burgess
4. E. Sanderson
5. S. Thompson
6. Gibran Hemani
7. Nuala Sheehan
8. F. Del Greco M

In Pulmonary and Respiratory Medicine

1. George Davey Smith
2. J. Bowden
3. Stephen Burgess
4. S. Thompson

Citations in the corpus

• Mendelian Randomization Analysis With Multiple Genetic Variants Using Summarized Data (2013, background): “A weak instrument is a variable that satisfies the IV assumptions, but does not explain a large proportion of variation in the risk factor, so that the statistical association between the risk factor and the IV in the dataset is ‘weak’ [Burgess et al., 2011b].”
• An examination of multivariable Mendelian randomization in the single-sample and two-sample summary data settings (2018, background): “This also avoids any potential bias that could arise due to selecting instruments based on their strength.(11)”
• The MR-Base platform supports systematic causal inference across the human phenome (2018, background): “This indicates that the instrument is strong and therefore unlikely to be susceptible to weak instrument bias or bias from sample overlap ( Burgess et al., 2011 ).”
• Assessing the suitability of summary data for two-sample Mendelian randomization analyses using MR-Egger regression: the role of the I2 statistic (2016, positive): “These ideas naturally complement allele score approaches that have been shown to successfully mitigate weak instrument bias when performing standard two-stage least squares or IVW analyses.(18,28) In conclusion, assessing the strength of NOME violation is an important prerequisite to performing caus”
• Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors (2015, background): “However, the data-driven choice of instrumental variables can also lead to weak instrument bias [34]; hence the analysis using all variants regardless of their association with calcium levels is also performed.”
• Bias due to participant overlap in two‐sample Mendelian randomization (2016, background): “…case of extremely weak instruments (those for which the expected strength of association with the risk factor corresponds to a P-value above 0.05 Burgess & Thompson, 2011), indicating that the practical consequences of bias with a single IV are unlikely to be substantial (Angrist & Pischke,…”
• Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator (2016, background): “…were obtained, then the resulting analysis suffers from bias and inflated type 1 error rates when the included variants are “weak” (i.e., they do not explain a substantial proportion of variation in the exposure in the dataset under analysis) (Burgess et al., 2011; Pierce and Burgess, 2013).”