Xue Gao1, Ling-Xian Meng2, Kai-Li Ma3, Jie Liang4, Hui Wang5, Qian Gao6, Tong Wang7. 1. Department of Health Statistics, School of Public Health, Shanxi Medical University, 56 Xinjiannanlu Street, Taiyuan, Shanxi, 030001, China. Electronic address: dr2016038@d.sxmu.edu.cn. 2. Department of Health Statistics, School of Public Health, Shanxi Medical University, 56 Xinjiannanlu Street, Taiyuan, Shanxi, 030001, China. Electronic address: 201700510655@b.sxmu.edu.cn. 3. Department of Health Statistics, School of Public Health, Shanxi Medical University, 56 Xinjiannanlu Street, Taiyuan, Shanxi, 030001, China. Electronic address: 201600510595@b.sxmu.edu.cn. 4. Department of Health Statistics, School of Public Health, Shanxi Medical University, 56 Xinjiannanlu Street, Taiyuan, Shanxi, 030001, China. Electronic address: liangjie_sxmu@sina.com. 5. Department of Health Statistics, School of Public Health, Shanxi Medical University, 56 Xinjiannanlu Street, Taiyuan, Shanxi, 030001, China. Electronic address: annieyuner@163.com. 6. Department of Health Statistics, School of Public Health, Shanxi Medical University, 56 Xinjiannanlu Street, Taiyuan, Shanxi, 030001, China. Electronic address: 051510608@b.sxmu.edu.cn. 7. Department of Health Statistics, School of Public Health, Shanxi Medical University, 56 Xinjiannanlu Street, Taiyuan, Shanxi, 030001, China. Electronic address: tongwang@sxmu.edu.cn.
Abstract
BACKGROUND: Several observational studies have investigated the association of insomnia with psychiatric disorders. Such studies yielded mixed results, and whether these associations are causal remains unclear. Thus, we aimed to identify the causal relationships between insomnia and five major psychiatric disorders. METHODS: The analysis was implemented with six genome-wide association studies; one for insomnia and five for psychiatric disorders (attention-deficit/hyperactivity disorder, autism spectrum disorder, major depressive disorder, schizophrenia, and bipolar disorder). A heterogeneity in dependent instrument (HEIDI) approach was used to remove the pleiotropic instruments, Mendelian randomization (MR)-Egger regression was adopted to test the validity of the screened instruments, and bidirectional generalized summary data-based MR was performed to estimate the causal relationships between insomnia and these major psychiatric disorders. RESULTS: We observed significant causal effects of insomnia on the risk of autism spectrum disorder and bipolar disorder, with odds ratios of 1.739 (95% confidence interval: 1.217-2.486, p = 0.002) and 1.786 (95% confidence interval: 1.396-2.285, p = 4.02 × 10-6), respectively. There was no convincing evidence of reverse causality for insomnia with these two disorders (p = 0.945 and 0.546, respectively). When insomnia was considered as either the exposure or outcome variable, causal estimates for the remaining three psychiatric disorders were not significant. CONCLUSIONS: Our results suggest a causal role of insomnia in autism spectrum disorder and bipolar disorder. Future disease models should include insomnia as a factor for these two disorders to develop effective interventions. More detailed mechanism studies may also be inspired by this causal inference.
BACKGROUND: Several observational studies have investigated the association of insomnia with psychiatric disorders. Such studies yielded mixed results, and whether these associations are causal remains unclear. Thus, we aimed to identify the causal relationships between insomnia and five major psychiatric disorders. METHODS: The analysis was implemented with six genome-wide association studies; one for insomnia and five for psychiatric disorders (attention-deficit/hyperactivity disorder, autism spectrum disorder, major depressive disorder, schizophrenia, and bipolar disorder). A heterogeneity in dependent instrument (HEIDI) approach was used to remove the pleiotropic instruments, Mendelian randomization (MR)-Egger regression was adopted to test the validity of the screened instruments, and bidirectional generalized summary data-based MR was performed to estimate the causal relationships between insomnia and these major psychiatric disorders. RESULTS: We observed significant causal effects of insomnia on the risk of autism spectrum disorder and bipolar disorder, with odds ratios of 1.739 (95% confidence interval: 1.217-2.486, p = 0.002) and 1.786 (95% confidence interval: 1.396-2.285, p = 4.02 × 10-6), respectively. There was no convincing evidence of reverse causality for insomnia with these two disorders (p = 0.945 and 0.546, respectively). When insomnia was considered as either the exposure or outcome variable, causal estimates for the remaining three psychiatric disorders were not significant. CONCLUSIONS: Our results suggest a causal role of insomnia in autism spectrum disorder and bipolar disorder. Future disease models should include insomnia as a factor for these two disorders to develop effective interventions. More detailed mechanism studies may also be inspired by this causal inference.
Authors: Joseph Firth; Marco Solmi; Robyn E Wootton; Davy Vancampfort; Felipe B Schuch; Erin Hoare; Simon Gilbody; John Torous; Scott B Teasdale; Sarah E Jackson; Lee Smith; Melissa Eaton; Felice N Jacka; Nicola Veronese; Wolfgang Marx; Garcia Ashdown-Franks; Dan Siskind; Jerome Sarris; Simon Rosenbaum; André F Carvalho; Brendon Stubbs Journal: World Psychiatry Date: 2020-10 Impact factor: 49.548
Authors: Christina Dardani; Lucy Riglin; Beate Leppert; Eleanor Sanderson; Dheeraj Rai; Laura D Howe; George Davey Smith; Kate Tilling; Anita Thapar; Neil M Davies; Emma Anderson; Evie Stergiakouli Journal: Int J Epidemiol Date: 2021-06-07 Impact factor: 9.685