Common genetic influences on autism spectrum disorders and sleep disturbances: evidence from a nationwide twin sample

Autism spectrum disorders (ASD) are frequently comorbid with other conditions. Sleep disorders, such as insomnia and parasomnia, are common comorbidities with ASD; as many as 80% of individuals with ASD either report disturbed... [ view full abstract ]

Autism spectrum disorders (ASD) are frequently comorbid with other conditions. Sleep disorders, such as insomnia and parasomnia, are common comorbidities with ASD; as many as 80% of individuals with ASD either report disturbed sleep or clinically significant sleep problems. It is now relatively well known that ASD shares genetic causes with many other conditions with which it is frequently comorbid; whether ASD and sleep disorders are also associated with similar genetic factors is largely unknown. We therefore aimed to test whether ASD and sleep disorders are associated with similar genetic influences to one another. From 27350 pairs of twins in the Swedish Twin Register born between 1973-2009, we identified 672 individuals with a diagnosis of ASD recorded in nationwide health registries. We also identified 2400 individuals with a recorded diagnosis of a sleep disorder or prescriptions of sleep medications recorded in the Swedish Prescribed Drug Register. The association between these disorders was tested using bivariate liability threshold twin models. Both ASD and sleep disorders are also strongly associated with anxiety and depression, and we hence repeated the analyses while controlling for the effects of comorbid anxiety disorders and depression. There was a significant association between ASD and sleep disorders (β=2.40, SE=0.12, odds ratio=10.98 [8.70-13.78]). ASD and sleep disorders were, individually, under strong genetic influence, with respective heritability estimates of 92% and 65%. There was a moderate phenotypic correlation of .46 (95% confidence intervals=.38-.50). The genetic correlation was also moderate in magnitude (r_a=.46, 95% CI=.41-.52), as was the nonshared environmental correlation (r_e=.67, 95% CI=.44-.84). Shared genetic influences accounted for 76% of the phenotypic correlation between ASD and sleep disorders. Adjusting for anxiety and depression resulted in a reduction in the magnitude of the phenotypic correlation to .29 (95% CI=.27-.37). Similarly, the genetic correlation decreased to .26 (95% CI=.17-.29) and the nonshared environmental correlation decreased to .55 (95% CI=.31-.71). Of note, however, significant associations remained between ASD and sleep disorders even after adjustments for anxiety and depression. These results thus suggest that the genetic pleiotropy seen in relation to ASD and other outcomes, such as ADHD, extends to sleep disorders. Some of this association is accounted for by the strong overlap seen between these disorders individually and anxiety and depression, stressing the need to account for anxiety and depression when studying the association between ASD and sleep disorders in genetic studies. On the other hand, a certain degree of the association between these phenotypes is independent of anxiety and depression, indicating that at least some of the genetic causes of ASD are the same as those that lead to sleep disorders. Ongoing analyses are now investigating these associations longitudinally, and using polygenic risk scores.