To investigate the 18 putative associations with obesity reported for rare GSVs [10], [16] (see Table 1), we analysed population and case-control cohorts in a similar manner to that successfully used in our replication of the association with obesity of the 16p11.2 593 kb deletion [12]. Using existing genotyping data from cohorts of severely obese (but with no other reported unrelated health problem) French children and adults, similar numbers of non-overweight controls, and a general population cohort from northern Finland [17], [18], each genomic region was analysed for the presence of GSVs.