AB0154 MENDELIAN RANDOMIZATION STUDY IMPLIES CAUSAL LINKAGE BETWEEN POLYMYOSITIS AND PHYSICAL ACTIVITY
J. Y. Chen,B. Song,Q. Feng,Y. Feng,Q. Yu,P. F. He,X. Li,S. X. Zhang
DOI: https://doi.org/10.1136/annrheumdis-2023-eular.4215
IF: 27.973
2023-01-01
Annals of the Rheumatic Diseases
Abstract:Background Polymyositis (PM) is autoimmune myopathy characterized clinically by proximal muscle weakness, muscle inflammation, extra-muscular manifestations, and frequently, the presence of autoantibodies[1].The causal relationship between physical activity (PA) and PM remains uncertain. Objectives This study aimed to determine the effect of physical activity on PM. Methods The instrumental variables proposed SNP at the genome-wide significance threshold (p <5×10-8).To mitigate collinearity between the contained SNPs, we excluded SNPs in the linkage disequilibrium (LD) (pairwise r2>0.01). F-Statistics, above the threshold of 10, showed the result's sensitivity. Summary statistics of PM were obtained from a genome-wide association study (GWAS) meta-analysis of the descent, including 16,380,451 cases and 213,145 controls[2].The data of PA, including accelerometer-based physical activity(APA), vigorous physical activity(VPA),moderate to vigorous physical activity (MVPA),moderate physical activity 10+ minutes(MPA10+), vigorous physical activity 10+ minutes(VPA10+) and sedentary behavior (Time spent watching television (TV), Time spent using computer and Time spent driving) were available from UK Biobank dataset. To evaluate the causal association between PA and PM, the inverse variance weighted (IVW) was conducted as the primary method. In sensitivity analyses, we used the MR-Egger method to evaluate the potential pleiotropy effects. Moreover, we also applied weighted median and MR-PRESSO to assess the presence of pleiotropy. Meanwhile, leave-one-out analysis was performed to test the influence of outlying values. Results We selected 313 SNPs (10 SNPs for APA,2 SNPs for VPA, 23 SNPs for MVPA, 19 SNPs for MPA10+, 13 SNPs for VPA10+, and 246 SNPs for sedentary). APA was positively associated with the risk of PM (OR 1.53 95%CI 0.03-0.82 P=0.036). MR-Egger, MR-PRESSO, and leave-one-out analysis did not indicate horizontal pleiotropy. Based on the result of IVW method, there was little evidence of causal associations of the remaining seven exposures with the risk of PM. (p>0.05)(Figure1). Conclusion APA was causally associated with the risk of PM. References [1]Mammen, A.L., Dermatomyositis and polymyositis: Clinical presentation, autoantibodies, and pathogenesis. Ann N Y Acad Sci, 2010. 1184: p. 134-53. [2]Klimentidis, Y.C., et al., Genome-wide association study of habitual physical activity in over 377,000 UK Biobank participants identifies multiple variants including CADM2 and APOE. Int J Obes (Lond), 2018. 42(6): p. 1161-1176. Acknowledgements: NIL. Disclosure of Interests None Declared.