Sample size requirements for matched case-control studies of gene-environment interaction

Area of Science:

• Epidemiology
• Genetic Epidemiology
• Biostatistics

Background:

• Gene-environment (GxE) interactions are crucial for understanding complex diseases.
• Designing epidemiologic studies to detect GxE interactions requires robust statistical methods.
• Accurate sample size and power calculations are essential for efficient study design.

Purpose of the Study:

• To present a method for calculating sample size and power for detecting GxE interactions.
• To compare the efficiency of different study designs, including matched case-control, case-sibling, and case-parent designs.
• To provide tools for sample size computation in stratified populations.

Main Methods:

• The study utilizes the expected value of the likelihood ratio test statistic.
• Analysis is based on conditional logistic regression models.
• Sample size and power are computed for three specific designs: matched case-control, case-sibling, and case-parent.

Main Results:

• Family-based designs (case-sibling, case-parent) generally require fewer matched sets than case-control designs for detecting GxE interactions.
• The case-sibling design is optimal for dominant gene effects, while the case-parent design is preferred for recessive gene effects.
• Methods are provided for stratified populations, enhancing applicability across diverse ethnic groups.

Conclusions:

• Family-based designs offer increased power and efficiency for detecting GxE interactions compared to traditional case-control studies.
• The choice between case-sibling and case-parent designs depends on the suspected mode of gene action (dominant vs. recessive).
• A freely available software program facilitates the implementation of these sample size and power calculation methods.