Article Text
Abstract
Introduction Insomnia significantly impacts lifetime morbidity and thus has substantial socioeconomic costs. In developed, high-income countries insomnia prevalence is increasing. However, little is known about insomnia in less urbanised, lower-income populations. Baependi is a Brazilian rural town, which has been shown to maintain sleep cycles synchronised to natural daylight, in spite of electrification. We aimed to investigate the components of insomnia in the family-based Baependi Heart Study cohort, using the Insomnia Severity Index (ISI) questionnaire.
Methods and materials Descriptive analysis was performed on data collected from the Baependi population (n=1,202) using R software. Heritability analysis was calculated using polygenic mixed modelling. Genome-wide association analysis (GWAS) was subsequently performed on the Baependi data, in order to interrogate for associations with polymorphisms previously related with insomnia symptoms (n= 811).
Results Descriptive regression analysis categorised 7.6% of the participants as suffering from ‘clinical insomnia’ based on their ISI scores, with an average total score of 6.5±5.0 (SD). Heritability of ISI score, based on the best-fit model adjusted for sex, age, education, and depression, was 19%. GWAS yielded four associations of genome-wide significance with single-nucleotide polymorphisms (SNP) rs869481, rs62037617 and rs3747579, which are located in the CORO7 gene, and rs3789038, located on the neighbouring HMOX2 gene on chromosome 16.
Conclusion This is one of the first studies of ISI score distribution in a general population. The heritability value observed is consistent with previously published literature, which have used different measures of insomnia symptoms. In addition, this is the first reported GWAS analysis for ISI score, identifying the first significant genome-wide genetic associations of ISI score. Thus, this study confirms the reliability and suitability of ISI as a measure for genetic studies in population.
Acknowledgements This study was supported by the Santander Universities Researcher Mobility Award and CNPq (PVE 400791/2014-5).