OBJECTIVETo assess the possible genotype-phenotype correlation for GJB2.

METHODSRetrospectively analyzed GJB2 gene mutations with non-syndromic hearing impairment (NSHI) patients and their families audiological data. Individuals were grouped, according to non-truncated mutant (non-truncating, NT) and truncating mutations (truncating, T), into T/T group, T/NT group and NT/NT group. And according to whether they carry 235delC, grouped into 235delC/235delC group, 235delC/Non-235del group and Non-235delC/Non-235delC group.

RESULTSGrouped according to whether the truncation mutants:Fisher exact statistical analysis showed that the degree of hearing loss among the three groups did not meet the random distribution (P = 0.003) , T/T group was significantly higher than T/NT group (P = 0.000) and NT/NT group (P = 0.000) on the degree of hearing loss. Grouped according to whether they carry 235delC mutation: degrees of hearing loss among the three groups were statistically significant differences. Respectively pairwise comparisons (Fisher exact test) found 235delC/235delC group was significantly higher than 235delC/Non-235delC on the degree of hearing loss group (P = 0.001) and Non-235delC/Non-235delC group (P = 0.000), 235delC/Non-235delC group higher than Non-235delC/Non-235delC group (P = 0.033). In GJB2 mutations homozygous and compound heterozygous mutation genotype:G109A/G109A, 235delC/512insAACG, 299delAT/G109A and 235delC/G109A degree of hearing loss caused by genotype was significantly lower than 235delC/235delC group.

CONCLUSIONS235delC homozygotes have significantly more hearing impairment, when compared with 235delC/non-235delC compound heterozygotes. People with two non-235delC mutations have even less hearing impairment. Patients with non-truncation mutants (G109A) suffer from lighter hearing loss than truncation mutations(235delC, 299delAT).