A genome-wide association study of caffeine-related sleep disturbance: confirmation of a role for a common variant in the adenosine receptor

Abstract

Objectives: To identify common genetic variants that predispose to caffeine-induced insomnia and to test whether genes whose expression changes in the presence of caffeine are enriched for association with caffeine-induced insomnia.

Design: A hypothesis-free, genome-wide association study.

Setting: Community-based sample of Australian twins from the Australian Twin Registry.

Participants: After removal of individuals who said that they do not drink coffee, a total of 2,402 individuals from 1,470 families in the Australian Twin Registry provided both phenotype and genotype information.

Measurements and results: A dichotomized scale based on whether participants reported ever or never experiencing caffeine-induced insomnia. A factor score based on responses to a number of questions regarding normal sleep habits was included as a covariate in the analysis. More than 2 million common single nucleotide polymorphisms (SNPs) were tested for association with caffeine-induced insomnia. No SNPs reached the genome-wide significance threshold. In the analysis that did not include the insomnia factor score as a covariate, the most significant SNP identified was an intronic SNP in the PRIMA1 gene (P = 1.4 × 10⁻⁶, odds ratio = 0.68 [0.53 - 0.89]). An intergenic SNP near the GBP4 gene on chromosome 1 was the most significant upon inclusion of the insomnia factor score into the model (P = 1.9 × 10⁻⁶, odds ratio = 0.70 [0.62 - 0.78]). A previously identified association with a polymorphism in the ADORA2A gene was replicated.

Conclusions: Several genes have been identified in the study as potentially influencing caffeine-induced insomnia. They will require replication in another sample. The results may have implications for understanding the biologic mechanisms underlying insomnia.

Keywords: Caffeine; genetics; insomnia.

Figure 1

(A) Quantile-quantile (Q-Q) plot for analysis with age and sex as covariates only. (B) Q-Q plot for analysis with insomnia factor score included as a covariate.

Figure 2

Manhattan plots for the 2 association analyses.