Developmental regression and GABA receptor genes in multiple racial-ethnic groups

Background: Developmental regression (DR), which occurs in 20-49% of individuals with AUT and may co-exist with seizures, is a potential subsetting variable. The relationship between DR and seizures implicates the GABAergic system, as GABA transmission influences seizure susceptibility. SNPs in GABA receptor genes on chromosome 4 (GABRA4) are associated to AUT in Caucasians (CA) and African Americans (AA). However, the associated SNPs differed in the two racial groups. Objectives: Examine the effect of DR on association to AUT in GABA receptor genes and its significance for understanding observed differences in associated SNPs in AA and CA families. Methods: Using the ADI-R we identified a DR subset (N = 263; 35%) from our overall sample of 606 AUT families (54 AA, 552 CA). DR families were those with positive ADI-R regression scores in an affected individual. Using the pedigree disequilibrium test (PDT) we tested for association in the DR and non-DR subsets in 36 SNPs in GABRA4 and GABRB1. Data were analyzed by race and contrasted with overall findings. Results: In the AA-DR subset (N=17), allelic association (p<.05) was detected for the two SNPs in GABRA4 (rs2280073, rs16859786) previously identified in the overall AA group as well as for two additional SNPs in GABRA4 (rs13151769, rs2351299). In contrast, no SNPs identified in the overall CA group showed association in the CA-DR subset (N=246) although three SNPs (rs17599165, rs1912960, rs17599416) in the overall set remained significant in the CA non-DR complement. Conclusions: Subsetting on DR had an effect on association and may explain the different patterns of associated SNPs in the two racial groups. This is the first study of DR, GABA genes, and AUT. Our results suggest that DR should be explored in other candidate gene studies of AUT and extended to other racial groups.