odds ratios around 1.5 in size with at least 80% power, but the required size of each individual study will depend on whether 1) the analysis will also include case sub-groups; 2) the analysis focuses on candidate genes with a limited number of independent tests or genome-wide associations with many thousands of tests; and 3) there is an a-priori hypothesis to be tested relating to a polymorphism of known allele frequency (for example, in a replication study - see below). The expected effect size to be detected in a study (and thus power) can sometimes be increased by using family history enrichment schemes for case sampling 18. However, as mentioned before, they are not guaranteed to do because of environmental and genetic heterogeneity. If recruitment of cases is more difficult than controls, power can often be increased more economically by increasing the ratio of controls to cases 22. When many SNPs are tested, and testing all of these in many thousands of cases and controls become prohibitively expensive, a multi-stage design can be more economical 32. In such a design, all SNPs are tested in a random subset of cases and controls, and those exhibiting a nominal pre-determined significance level
