The association of individual variants with hypertension has been tested by construction of a 3 by 2 contingency table and using χ2 test to compare genotypes in the upper and lower quintiles of the blood pressure distribution. The approach of utilising high and low blood pressure groups obviates the confounding variable of white coat hypertension. This approach is taken in recognition of the possibility that there may be protective as well as susceptibility variants at the β2 adrenergic receptor locus and because it has been suggested this may offer greater power to detect a true difference [65]. In the event that these analyses indicate several variants are associated with hypertension (defined by any of the thresholds tested above) the relative risk for raised blood pressure can be estimated by calculating an odds ratio with 95% confidence intervals. Haplotype reconstruction and haplotype frequency estimation has been done by a maximum likelihood procedure (using the programme EH from the LINKAGE package developed by TERWILLIGER and OTT [176]).
© 2001 Alexander Binder