Estimating the strength of genetic effects: a comparison of maximum likelihood and transmission disequilibrium methods in the study of ankylosing spondylitis.

Here we compare several methods for evaluating associations in the analysis of family data. We applied the transmission disequilibrium test (TDT), and logistic regression, familial aggregation, and segregation analyses for the study of Ankylosing Spondylitis. All methods detected effects from B27 upon susceptibility for Ankylosing Spondylitis. However, the TDT was unable to detect effects from B40 and also could not include sex as a covariate during the analysis. Conditioning or not conditioning the data for the ascertainment event led to little difference in the estimated covariate effects, although the genetic model for residual effects not due to the covariates did change. Results from these studies provide further characterization of the relationship between HLA B27, B40, and sex upon risk for developing Ankylosing Spondylitis. We also identified potential effects from further genetic susceptibility.