<jats:title>Abstract</jats:title><jats:p>GNE myopathy is a distal myopathy caused by biallelic variants in <jats:italic>GNE</jats:italic>, which encodes a protein involved in sialic acid biosynthesis. Compound heterozygosity of the second most frequent variant among Japanese GNE myopathy patients, <jats:italic>GNE</jats:italic> c.620A>T encoding p.D207V, occurs in the expected number of patients; however, homozygotes for this variant are rare; three patients identified while 238 homozygotes are estimated to exist in Japan. The aim of this study was to elucidate the pathomechanism caused by c.620A>T. Identity-by-descent mapping indicated two distinct c.620A>T haplotypes, which were not correlated with age onset or development of myopathy. Patients homozygous for c.620A>T had mildly decreased sialylation, and no additional pathogenic variants in <jats:italic>GNE</jats:italic> or abnormalities in transcript structure or expression of other genes related to sialic acid biosynthesis in skeletal muscle. Structural modeling of full-length GNE dimers revealed that the variant amino acid localized close to the monomer interface, but far from catalytic sites, suggesting functions in enzymatic product transfer between the epimerase and kinase domains on GNE oligomerization. In conclusion, homozygotes for c.620A>T rarely develop myopathy, while symptoms occur in compound heterozygotes, probably because of mildly decreased sialylation, due to partial defects in oligomerization and product trafficking by the mutated GNE protein.</jats:p>