Rxivist logo

MR-pheWAS with stratification and interaction: Searching for the causal effects of smoking heaviness identified an effect on facial aging

By Louise A C Millard, Marcus R. Munafò, Kate Tilling, Robyn Wootton, George Davey Smith

Posted 19 Oct 2018
bioRxiv DOI: 10.1101/441907 (published DOI: 10.1371/journal.pgen.1008353)

Mendelian randomization (MR) is an established approach for estimating the causal effect of an environmental exposure on a downstream outcome. The gene x environment (GxE) study design can be used within an MR framework to determine whether MR estimates may be biased if the genetic instrument affects the outcome through pathways other than via the exposure of interest (known as horizontal pleiotropy). MR phenome-wide association studies (MR-pheWAS) search for the effects of an exposure, and a recently published tool (PHESANT) means that it is now possible to do this comprehensively, across thousands of traits in UK Biobank. In this study, we introduce the GxE MR-pheWAS approach, and search for the causal effects of smoking heaviness - stratifying on smoking status (ever versus never) - as an exemplar. If a genetic variant is associated with smoking heaviness (but not smoking initiation), and this variant affects an outcome (at least partially) via tobacco intake, we would expect the effect of the variant on the outcome to differ in ever versus never smokers. If this effect is entirely mediated by tobacco intake, we would expect to see an effect in ever smokers but not never smokers. We used PHESANT to search for the causal effects of smoking heaviness, instrumented by genetic variant rs16969968, among never and ever smokers respectively, in UK Biobank. We ranked results by: 1) strength of effect of rs16969968 among ever smokers, and 2) strength of interaction between ever and never smokers. We replicated previously established causal effects of smoking heaviness, including a detrimental effect on lung function and pulse rate. Novel results included a detrimental effect of heavier smoking on facial aging. We have demonstrated how GxE MR-pheWAS can be used to identify causal effects of an exposure, while simultaneously assessing the extent that results may be biased by horizontal pleiotropy.

Download data

  • Downloaded 363 times
  • Download rankings, all-time:
    • Site-wide: 42,344 out of 85,187
    • In epidemiology: 653 out of 1,556
  • Year to date:
    • Site-wide: 70,583 out of 85,187
  • Since beginning of last month:
    • Site-wide: 81,557 out of 85,187

Altmetric data


Downloads over time

Distribution of downloads per paper, site-wide


PanLingua

Sign up for the Rxivist weekly newsletter! (Click here for more details.)


News