Supplementary file 1_A descriptive multilevel analysis associating COVID-19 with polymyositis: from genetic markers and candidate mediators to clinical hematological profiles.xlsx

ObjectivesObservational studies suggest an association between Coronavirus Disease 2019 (COVID-19) and polymyositis (PM), but causal inference s limited by confounding. This study adopted a multilevel exploratory framework to investigate potential relationships. We used two-sample Mendelian randomization (MR) to assess the causal effect of severe COVID-19 on PM, multi-omic analyses to screen for potential mediators, and retrospectively compared hematological profiles between severe and non-sever COVID-19 cases.
DesignA two-sample MR with mediation analysis and a single-center retrospective cohort analysis.
Setting and participantsGenetic instruments for severe COVID-19 (exposure), PM (outcome) and candidate multi-omic mediators (91 inflammatory proteins, 4,907 circulating plasma proteins, 731 immune-cell traits, and 1,400 plasma metabolites) were obtained from genome-wide association studies (GWAS). The clinical study included 108 hospitalized patients with PCR-confirmed COVID-19, classified into severe and non-severe subgroups.
Main outcomesThe main outcome in the MR analysis was the causal odds ratio (OR) of PM per genetically predicted increase in the risk of severe COVID-19. Secondary outcomes included the identification of mediating biomarkers and differences in hematological indices between severe and non-severe COVID-19 patients.
ResultsMR analysis suggested a potential causal effect of severe COVID-19 on PM (IVW OR = 1.65, 95% CI: 1.36–2.01, p < 0.01) with no reverse causality. Mediation analysis highlighted several candidate mediators (KIAA1024, RNASE1, EGFLAM, CAPZA1, NRG3, IL31). Functional enrichment implicated ErbB signaling and “Overview of proinflammatory and profibrotic mediators pathway.” The clinical analysis showed that severe COVID-19 patients had altered neutrophil, lymphocyte, and platelet counts, and elevated levels of muscle enzymes and IgE. After age and sex adjustment, lymphocyte count, CK-MB, LDH, and HBDH remained independently associated with severe COVID-19 (all adjusted p < 0.05).
ConclusionThis multilevel exploratory study provides preliminary genetic evidence suggesting a potential association between severe COVID-19 and polymyositis, along with candidate molecular mediators. The clinical analysis revealed distinct hematological profiles in patients with severe COVID-19. These findings should be interpreted as preliminary clues rather than definitive mechanistic evidence. Future studies with independent validation cohorts, larger sample sizes, and longitudinal follow-up for incident myositis are needed to confirm these observations.