Reply|[mdash]|

Stevanin et al.1 and Worth et al.2 suggest that the reduced penetrance of SCA8 in ataxia families and the large alleles found in controls indicate that the CTG expansion may be a non-pathogenic polymorphism tightly linked to an ataxia locus. Because we isolated the CTG expansion from a single ataxia patient using RAPID cloning3, 4, the likelihood that the expansion would be tightly linked to an ataxia locus by chance can be conservatively estimated as 3×10 −4 to 1.