Mind the gap: characterizing bias due to population mismatch in two-sample Mendelian randomization

Area of Science:

• Genetics
• Statistical Epidemiology
• Bioinformatics

Background:

• Mendelian randomization (MR) estimates causal effects using genetic variants as instrumental variables.
• Two-sample MR (2SMR) utilizes separate samples for exposure and outcome associations.
• Valid 2SMR inference requires participants from the same population to avoid bias.

Purpose of the Study:

• To investigate the impact of using different populations in 2SMR.
• To quantify the bias introduced by population mismatch in 2SMR estimates.
• To explore strategies for improving 2SMR power in understudied groups.

Main Methods:

• Conducted a broad survey of 546 2SMR trait pairs across 2-7 ancestries.
• Compared 2SMR estimates using matching versus mismatching populations.
• Analyzed the relationship between genetic distance and bias magnitude.

Main Results:

• Population mismatch in 2SMR biases causal estimates towards zero on average.
• Bias increases with greater genetic distance between study populations.
• Bias is observed even between populations of the same continental ancestry.
• Larger, mismatched exposure studies can improve precision and power.

Conclusions:

• Violation of the same-population assumption in 2SMR introduces bias.
• Even intra-continental population differences can impact causal estimates.
• Leveraging larger, mismatched ancestry studies offers potential to enhance 2SMR power for understudied populations.