We read with great interest the article by McNicoll and colleagues who investigated a potential association between the size of the CAG repeat sequence in the Huntington's disease (HD) gene and age of onset in familial Parkinson's disease (PD). 1 The authors hypothesized that as the ATP/ADP ratio, a measure of mitochondrial function, is influenced by the HD CAG repeat sizes in the normal range, 2 and as mitochondrial dysfunction has been implicated in PD pathogenesis, HD CAG repeat size might be related to age of disease onset in PD. 1 Although the authors conclude that HD CAG repeat size does not modify onset age in familial PD, we feel that they have not satisfactorily excluded the influence on age of onset of a potential interaction between the CAG repeat sizes in the two HD alleles. The authors tested for possible interaction effects by dichotomizing their patient cohort at the median CAG repeat size of the larger allele, and by using the sum of the CAG repeat sizes in both alleles as an explanatory variable. 1 However, these methods have only modest statistical power in detecting any interaction effects. The most common, and powerful, way for testing for a statistical interaction between two variables is to test simultaneously for both variables and their product term. 3,4 In this case, this would entail entering of the larger repeat size, the smaller repeat size, as well as the product of the two repeat sizes as predictor variables in the generalized estimating equation model that was used. Additionally, in this way, the effect of the interaction can be assessed while at the same time accounting for the repeated observations within a family. In HD, interaction between the sizes of the normal and expanded CAG repeats has been shown to affect age of disease onset. 5 As in PD, the sizes of the two repeats diverge much less, it would be, in our opinion, very important to satisfactorily exclude a potential interaction effect between the two repeat sizes before assuming that the HD CAG repeat size does not modify age of onset in PD.