Integrating Family-Based and Mendelian Randomization Designs

Area of Science:

• Genetics
• Epidemiology
• Biostatistics

Background:

• Most Mendelian randomization (MR) studies analyze unrelated individuals.
• These studies are prone to biases such as dynastic effects, assortative mating, population stratification, and horizontal pleiotropy.
• These biases can distort estimates of causal parameters.

Purpose of the Study:

• To review biases affecting MR studies.
• To describe family-based study designs for controlling these biases.
• To highlight the potential of using related individuals in MR.

Main Methods:

• Discussion of biases in MR studies.
• Description of three family-based study designs.
• Review of existing cohort data (twin, birth, population-based).

Main Results:

• Related individuals can help control for or estimate biases in MR.
• Family-based designs offer a robust approach to causal inference.
• Existing cohorts are suitable for implementing these designs.

Conclusions:

• Incorporating family information into MR studies is feasible.
• Family-based MR can significantly enhance understanding of complex traits and diseases.
• This approach promises rich rewards for future etiological research.