Likelihood ratio test for detecting gene (G)-environment (E) interactions under an additive risk model exploiting G-E...

Area of Science:

• Epidemiology
• Biostatistics
• Genetic Epidemiology

Background:

• A controversy exists regarding the appropriate risk scale for gene-environment (GxE) interaction testing.
• Logistic models approximate multiplicative risk and are widely used, but additive models better reflect biologic interactions and public health interventions.
• The GxE independence assumption enhances power for multiplicative interactions, but its effect on additive interaction tests was unknown.

Purpose of the Study:

• To develop a statistical test for additive gene-environment interactions in case-control studies.
• To investigate the impact of incorporating the gene-environment independence assumption on the power of additive interaction tests.

Main Methods:

• Developed a likelihood ratio test for additive interactions in case-control studies.
• Incorporated the gene-environment independence assumption into the test.
• Conducted numerical power investigations.
• Applied the method to a bladder cancer study dataset.

Main Results:

• The proposed likelihood ratio test effectively detects additive gene-environment interactions.
• Incorporating the gene-environment independence assumption enhanced the test's efficiency by 2- to 2.5-fold.
• The method was successfully applied to real-world epidemiological data.

Conclusions:

• The developed test provides a more powerful approach for detecting additive gene-environment interactions in case-control studies.
• Leveraging the gene-environment independence assumption is crucial for improving statistical power in this context.
• This method has implications for public health decision-making and understanding disease etiology.