Summary
This methodological paper extends Mendelian randomization to the multivariable setting, clarifying how MVMR analyses can estimate direct causal effects of multiple exposures whilst accounting for confounding, mediation, pleiotropy and collider structures. The authors develop tests for assessing instrument strength and validity in both single-sample and two-sample summary data settings, and demonstrate the approach using UK Biobank data. MVMR provides a generalised framework for causal inference across a range of epidemiological scenarios with either individual- or summary-level genetic data.
UK applicability
As a methodological contribution using UK Biobank data, the paper directly supports UK epidemiological research capacity for causal inference in complex exposure settings. The tests and software tools developed are applicable to ongoing UK health research utilising genetic instrumental variables.
Key measures
Instrument strength and validity assessment; causal effect estimates under various confounding and pleiotropy scenarios; simulation performance; UK Biobank genetic and phenotypic data
Outcomes reported
The study evaluated how multivariable Mendelian randomization (MVMR) estimates direct causal effects of multiple exposures on outcomes whilst accounting for confounding, mediation, pleiotropy and collider bias. The methods were illustrated using UK Biobank data to estimate causal effects of education and cognitive ability on body mass index.
Topic tags
Dig deeper with Pulse AI.
Pulse AI has read the whole catalogue. Ask about this record, its theme, or how the findings apply to UK farming and policy — every answer cites the underlying studies.