Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression.
Bowden J, Davey Smith G, Burgess S. · 2015
Read Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. by Bowden J, Davey Smith G, Burgess S. (2015) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc4469799.txt.
Description
Jack Bowden, George Davey Smith, and Stephen Burgess adapted a bias test for Mendelian randomization in the International Journal of Epidemiology in 2015. The method uses gene variants as instruments to infer cause, and studies now use many variants at once for precision. Some variants, though, affect the outcome through other traits as well, a flaw called pleiotropy, which breaks their validity.
Viewing the variants as a small meta-analysis opens a fix. Bias from pleiotropy resembles the small-study bias familiar in meta-analysis, which Egger regression was built to detect, and a funnel plot can reveal it visually. The adapted Egger test still gives a valid check of no causal effect. Its slope estimates the true effect, even when every variant is invalid. That holds provided each variant's strength is independent of its side effect.
Catalog record
| CREATOR | Bowden J, Davey Smith G, Burgess S. |
|---|---|
| YEAR | 2015 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |