To investigate the feasibility of newborn screening for glycogen storage disease type II (GSDII; Pompe disease or acid maltase deficiency) in the Japanese population, we assayed the acid alpha-glucosidase activity in dried blood spots from 715 Japanese newborns and 18 previously diagnosed patients using a fluorometric procedure. The enzyme activity of apparently healthy newborns showed a bimodal distribution. The median activity of the minor group (31 individuals, 4.3% of the samples) was 6.5 times lower than that of the major group. Four of the 715 control samples (0.56%) fell in the patient range. We then analyzed genomic DNA, extracted from the same blood spots, for the occurrence of two sequence variants, c.1726G > A and c.2065G > A, known to cause "pseudodeficiency". This analysis revealed that 27 of 28 individuals homozygous for c.[1726A; 2065A] belonged to the minor group. One c.[1726A; 2065A] homozygote had just slightly higher activity. Twelve of the 18 patients with GSDII either had one (9 cases) or two (3 cases) c.[1726A; 2065A] alleles. The frequency of this allele was double in the patient compared to the control group (0.42 vs 0.19) at the expense of a lower frequency of the c.[1726G; 2065G] and c.[1726G; 2065A] alleles (0.58 vs 0.71 and 0 vs 0.1). These findings illustrate that c.[1726A; 2065A] homozygosity among apparently healthy individuals (3.9 per 100) complicates newborn screening for GSDII in Japan, and further that one or more pathogenic mutations are associated with the c.[1726A; 2065A] allele. (C) 2009 Elsevier Inc. All rights reserved.