Similar results using a large mass of markers have been obtained for other complex traits including multiple sclerosis [6], height [7], cardiovascular risk [8], rheumatoid arthritis [9] and body mass index [10]. In addition, several studies have demonstrated association of a score based on a limited number of top ranking markers [11]–[13]. In some cases, however, the polygenic association is less clear: studies of breast and prostate cancers have been inconclusive, owing in part to technical aspects in analysis but also, potentially, to their sample sizes [14], [15]. An aim of the present work is to determine whether negative results from those studies could be explained by their sample size, or whether a true lack of polygenic effect is the more likely explanation.
