Figure 2
From: Trick or treat: The effect of placebo on the power of pharmacogenetic association studies
The effect of increasing the ratio between the drug and placebo group sizes on the power of detecting an association between the causative single nucleotide polymorphism and the response to the drug. Estimated power is plotted against the number of persons who received placebo for the different drug to placebo group size ratios: 1:1 -◆-; 2:1 -■-; 4:1 --▲--. The results are based on 1,000 simulations. Genotype relative risk = 3, f2 = 0.8, f0 = 0.266. The results are presented as a function of: (a) the number of placebo-treated individuals (mimicking a drug-only follow-up); (b) the number of individuals in the entire trial.