Dystonia in complex regional pain syndrome type I

We recently used array-based expression profiling to identify differentially regulated genes in laser-captured substantia nigra neurons of 8 Parkinson disease (PD) patients and 9 controls. We subsequently tested for genetic association of candidate genes in idiopathic PD. Association analysis was limited to genes surviving Bonferroni correction for multiple testing, although several known PD-associated genes were also differentially regulated. We found an association of rs2010795 in the pyridoxal kinase (PDXK) gene in German PD cases versus population-based controls, and replicated this association in 2 further case-control samples from the United Kingdom and Italy. In an attempt to replicate the association with the PDXK locus, Vilariño-Güell et al and Guella et al tested rs2010795 in 7 different case control samples from various populations and found no association. In contrast, a recently published genomewide association study (GWAS) shows significant p values for this PDXK single nucleotide polymorphism (SNP). In this study, separate analysis of the US samples (971 PD patients and 3,034 controls) also reveals a signal. A joint analysis of the German studies, the United States, and the dataset provided by Vilariño-Güell et al and Guella et al (a total of 5420 cases, 8,570 controls) generates a p value of 3.6E-05 (Fig). As the effect size of the reported polymorphism is small, the sample sizes of studies performed so far are underpowered. In addition, genome-wide SNP marker sets have to be used to exclude a gross genetic substructure, as was done in the GWAS controls. Meta-analysis of large datasets, as reported for traits such as obesity or diabetes, must be awaited to resolve the issue.