Case-control studies of gene-environment interactions. When a case might not be the case

Area of Science:

• Genetics
• Epidemiology
• Biostatistics

Background:

• Genome-Wide Association Studies (GWAS) are crucial for understanding complex disease genetics.
• Gene-environment interactions (GxE) are key to elucidating disease etiology.
• Clinical diagnoses can mask underlying distinct pathophysiologic mechanisms with unique genetic bases.

Purpose of the Study:

• To identify bias in GxE estimates when using clinical diagnoses in GWAS.
• To propose a novel statistical method to correct for this bias.
• To validate the method using simulations and a real-world Alzheimer's disease GWAS.

Main Methods:

• Developed a pseudo-likelihood approach to adjust for diagnostic heterogeneity.
• Conducted extensive simulations to assess method performance under various scenarios.
• Applied the corrected method to a Genome-Wide Association Study of Alzheimer's disease.

Main Results:

• Demonstrated that standard GxE analysis is severely biased when disease mechanisms differ and are environment-dependent.
• The proposed pseudo-likelihood method effectively corrects for this bias, yielding more accurate GxE estimates.
• The application to Alzheimer's disease GWAS identified potentially novel genetic associations influenced by environmental factors.

Conclusions:

• Clinical diagnosis alone can lead to misleading conclusions in GxE research.
• The developed statistical method offers a robust solution for analyzing GxE in genetically heterogeneous complex diseases.
• Accurate GxE analysis is essential for advancing our understanding of diseases like Alzheimer's and developing targeted interventions.