An association study in the Taiwan Biobank elicits the GABAA receptor genes GABRB3, GABRA5, and GABRG3 as candidate loci for sleep duration in the Taiwanese population.

Background Gamma-aminobutyric acid type A (GABAA) receptors mainly mediate the effects of gamma-aminobutyric acid, which is the primary inhibitory neurotransmitter in the central nervous system. Abundant evidence suggests that GABAA receptors play a key role in sleep-regulating processes. No genetic association study has explored the relationships between GABAA receptor genes and sleep duration, sleep quality, and sleep timing in humans. Methods We determined the association between single-nucleotide polymorphisms (SNPs) in the GABAA receptor genes GABRA1, GABRA2, GABRB3, GABRA5, and GABRG3 and sleep duration, sleep quality, and sleep timing in the Taiwan Biobank with a sample of 10,127 Taiwanese subjects. There were 10,142 subjects in the original study cohort. We excluded 15 subjects with a medication history of sedative-hypnotics. Results Our data revealed an association of the GABRB3-GABRA5-GABRG3 gene cluster with sleep duration, which has not been previously identified: rs79333046 (beta = - 0.07; P = 1.21 × 10-3) in GABRB3, rs189790076 (beta = 0.92; P = 1.04 × 10-3) in GABRA5, and rs147619342 (beta = - 0.72; P = 3.97 × 10-3) in GABRG3. The association between rs189790076 in GABRA5 and sleep duration remained significant after Bonferroni correction. A variant (rs12438141) in GABRB3 was also found to act as a potential expression quantitative trait locus. Additionally, we discovered interactions between variants in the GABRB3-GABRA5-GABRG3 gene cluster and lifestyle factors, such as tea and coffee consumption, smoking, and physical activity, that influenced sleep duration, although some interactions became nonsignificant after Bonferroni correction. We also found interactions among GABRB3, GABRA5, and GABRG3 that affected sleep duration. Furthermore, we identified an association of rs7165524 (beta = - 0.06; P = 2.20 × 10-3) in GABRA5 with sleep quality and an association of rs79465949 (beta = - 0.12; P = 3.95 × 10-3) in GABRB3 with sleep timing, although these associations became nonsignificant after Bonferroni correction. However, we detected no evidence of an association of individual SNPs in GABRA1 and GABRA2. Conclusions Our results indicate that rs189790076 in GABRA5 and gene-gene interactions among GABRB3, GABRA5, and GABRG3 may contribute to sleep duration in the Taiwanese population.