A robust method for finely stratified familial studies with proband-based sampling.

Area of Science:

• Biostatistics
• Epidemiology
• Statistical Genetics

Background:

• Familial studies often involve complex sampling designs, such as proband-based sampling.
• Analyzing stratified familial data requires methods that can handle both individual-level covariates and familial aggregation.
• Existing methods may struggle with bias when dealing with stratum-specific nuisance parameters.

Purpose of the Study:

• To develop a robust statistical inference method for finely stratified familial studies using proband-based sampling.
• To simultaneously estimate marginal effects of covariates and quantify familial aggregation (intrafamilial pairwise odds ratios).
• To reduce bias introduced by stratum-specific nuisance parameters in the analysis.

Main Methods:

• Utilizes an estimating function adapted from proband-based family studies (Zhao et al., 1998).
• Treats stratification effects as fixed nuisance parameters.
• Requires modeling only the first two joint moments of observations, enhancing efficiency.
• Develops an analytical standard error estimator for the proposed method.

Main Results:

• The proposed method significantly reduces bias (by two orders of magnitude) compared to previous approaches.
• Successfully applied to a matched case-control familial study of sleep apnea.
• Simulation studies confirm the method's effectiveness and usefulness.

Conclusions:

• The developed method provides a robust and efficient approach for statistical inference in stratified familial studies.
• It accurately captures both individual covariate effects and familial aggregation patterns.
• The method is validated through application and simulation, demonstrating its practical utility in epidemiological research.