Mendelian randomization analysis using multiple biomarkers of an underlying common exposure

Area of Science:

• Genetics
• Epidemiology
• Biostatistics

Background:

• Mendelian randomization (MR) is a popular method for inferring causal effects.
• Systematic inflammation, a key exposure, is often not directly measurable.
• Multiple biomarkers can reflect underlying latent exposures.

Purpose of the Study:

• To develop and validate a novel MR method for latent exposures (MRLE).
• To leverage multiple related traits for robust causal inference.
• To assess the causal effect of latent inflammation on disease risk.

Main Methods:

• MR analysis on latent exposures (MRLE) using GWAS summary statistics.
• Structural equation modeling to account for indirect and direct genetic effects.
• Utilizing second-order moments of association statistics from multiple biomarkers.

Main Results:

• MRLE demonstrates well-controlled type I error rates and improved statistical power.
• Simulations show MRLE outperforms single-trait MR under pleiotropy.
• MRLE identified potential causal links between inflammation and coronary artery disease, colorectal cancer, and rheumatoid arthritis.

Conclusions:

• MRLE is an effective method for investigating causal effects of latent exposures.
• Inflammation shows potential causal roles in major diseases, which single-trait MR missed.
• This approach advances causal inference in complex biological systems.