Estimating genotype relative risks in case-parental control studies: An optimal weighting approach

Abstract

The 'case-parental control study' is a novel study design. It can quantify the relations between candidate genes and disease risk. Previous authors have proposed noniterative methods for estimating 'genotype relative risks' (GRRs) in case-parental control studies. Here the authors propose yet another noniterative method. The new method is simple. It involves taking certain weighted averages with weights chosen according to one's educated guess about the likely values of the true GRRs. Monte Carlo simulation shows that the new estimators are approximately unbiased and that they have smaller variances than the previous estimators.