duration of illness, etc.). By contrast, in ENIGMA1, the rs7294919 genetic variant was associated with a hippocampal volume decreased by 47.6 mm3 or by 1.2 % of the average hippocampal volume per risk allele. Bearing in mind that the cause of the effects is quite different, the effect of ENIGMA’s top hippocampal SNPs on hippocampal volume is approximately 5 % of the mean effect of Alzheimer’s disease, around one-third of the effect of schizophrenia, and around 40 % of the effect of bipolar disorder. If such effect sizes are typical, sample size requirements will generally be larger in genetic association studies than in neuroimaging studies of disease effects on the same phenotypes, but not vastly larger.
