One annoying thing about this paper is their use of the term "ethnical".
It's surprising to me that they find this difference between groups that are geographically so close, and I'm not too sure how to interpret their statement that their power is 80%.
Role of ethnicity on the association of MAPT H1 haplotypes and subhaplotypes in Parkinson's disease
Susan Winkler, Inke R König, Katja Lohmann-Hedrich, Peter Vieregge, Vladimir Kostic and Christine Klein
European Journal of Human Genetics (2007) 15, 1163–1168
Abstract: An association of the H1 haplotype and subhaplotypes in the microtubule-associated protein Tau (MAPT) gene with Parkinson's disease (PD) has been reported. To further evaluate their role in PD, we genotyped a sample set of 765 cases and controls consisting of two large European subgroups of German (n=418) and Serbian (n=347) origin for the MAPT haplotypes H1 and H2. The H1/H1 carriers were tested for three additional MAPT polymorphisms. In the Serbian sample, there was significant evidence (P=0.0108) of an association of the H1/H1 genotype and PD. Surprisingly, in the German sample, we did not find significant differences in genotype or haplotype frequencies between patients and controls. These results suggest that the role of H1 haplotypes in the etiology of PD may be ethnically dependent.