Genome-wide association analysis of actigraphic sleep phenotypes in the LIFE Adult Study

Abstract

The genetic basis of sleep is still poorly understood. Despite the moderate to high heritability of sleep-related phenotypes, known genetic variants explain only a small proportion of the phenotypical variance. However, most previous studies were based solely upon self-report measures. The present study aimed to conduct the first genome-wide association (GWA) of actigraphic sleep phenotypes. The analyses included 956 middle- to older-aged subjects (40–79 years) from the LIFE Adult Study. The SenseWear Pro 3 Armband was used to collect 11 actigraphic parameters of night- and daytime sleep and three parameters of rest (lying down). The parameters comprised measures of sleep timing, quantity and quality. A total of 7 141 204 single nucleotide polymorphisms (SNPs) were analysed after imputation and quality control. We identified several variants below the significance threshold of P ≤ 5× 10−8 (not corrected for analysis of multiple traits). The most significant was a hit near UFL1 associated with sleep efficiency on weekdays (P = 1.39 × 10−8). Further SNPs were close to significance, including an association between sleep latency and a variant in CSNK2A1 (P = 8.20 × 10−8), a gene known to be involved in the regulation of circadian rhythm. In summary, our GWAS identified novel candidate genes with biological plausibility being promising candidates for replication and further follow-up studies.