Dear Sir, Homozygous expansions greater than 66 repeat lengths of a GAA repeat in the first intron of the X25/frataxin gene have been linked with the development of Friedreich's ataxia (FRDA)[1]. Abnormal glucose tolerance is present in 30% of FRDA patients, with about 10% having overt diabetes [2]. Intermediate GAA repeat expansions (10±36 repeats) do not lead to the FRDA phenotype but a recent case control study reported an increased prevalence in Type II (non-insulin-dependent) diabetes mellitus [3]. This has not been replicated in subsequent studies [4±6], although non-diabetic carriers of the intermediate expansion were found to have abnormal insulin secretion [6].

These studies exploring the importance of intermediate expansions in Type II diabetes have all been case control studies, which are susceptible to the generation of spurious results due to population stratification. A family-based association method using parent-offspring trios can overcome this problem as it detects only those associations due to linkage [7]. This approach also allows parent-of-origin effects to be examined, which case control studies cannot detect [7]. This is particularly relevant to FRDA as intergeneration expansion of the GAA repeat has only been reported with maternally transmitted alleles [8], and larger expansions are associated with a more severe clinical phenotype [2].