0055 Genome-Wide Association Analysis of Accelerometer-Derived Traits Reveals Novel Genetic Loci Associated with Rest-Activity Patterns in the UK Biobank

Abstract

Despite the abundance of epidemiological evidence linking sleep and circadian disturbances to diverse pathologies, little is known about the common genetic factors that underlie these traits. Previous studies have identified genetic loci associated with self-reported sleep and circadian traits. However, this might be subject to bias that could limit the biological relevance of the findings. Wrist-worn accelerometers are increasingly being used, and provide more accurate estimates of rest-activity patterns when compared to self-report. We aimed to understand the common genetic contributions to activity monitor-determined sleep and circadian traits and how they relate to self-reported measures in the UK Biobank.We used 83,726 unrelated individuals of European ancestry with available accelerometer (average of 7 days) and genotyping data. Raw accelerometer recordings were processed to estimate sleep and circadian phenotypes using a validated method (GGIR). We analyzed 11,977,111 genotyped and imputed variants using the Haplotype Reference Consortium reference panel. Genetic associations and SNP-based heritability (hSNP2) were calculated using linear-mixed models.Among all traits, hSNP2 ranged between 0.028 (95%CI=0.020–0.036) for variation in sleep duration to 0.223 (95%CI=0.215–0.231) for number of sleep episodes. A total of 41 genome-wide significant associations (p<5 × 10–8) were found across seven activity monitor-based phenotypes. We replicated loci previously associated with self-reported sleep duration (i.e. PAX8, MEIS1), chronotype (i.e. ALG10B, RGS16, TOX3) and restless legs syndrome (MEIS1, BTBD9). Remarkably, we found 20 genetic loci associated with estimated number of sleep episodes not previously associated with other self-reported sleep traits or sleep disorders. Genome-wide gene-set enrichment analyses identified overrepresentation of genes related to serotonin and other amino compound metabolic processes and synaptic signaling associated with this phenotype.We found novel genetic loci associated with accelerometer-derived measurements of rest-activity patterns, and replicated other previously reported signals. A better understanding of the genetic architecture sleep and circadian rhythm traits should lead to new biological and mechanistic insights, which in turn will allow for better management of related disorders and comorbidities.Medical Research Council, UK; NIH (R01 HL134015)