dbSNP reports that this variant has 5% allele frequency in a 120 chromosome low-coverage CEU panel (1000 genomes) and 2.6% in a mixed panel of Caucasian & African American samples. Lorenz-Depiereux et al. point to this as suggesting that the variant is a not causal.
The predicted incidence of homozygotes (1 in 400) vs. the incidence of the disease (quite rare) implies that this variant is a highly penetrant causal variant. It may have a modifier or susceptibility effect though (note that the son of the reported family in Lorenz-Depiereux et al. & Rutsch et al. had a more severe effect compared to his father).