Abstract
BACKGROUND: Membranous nephropathy (MN) is the leading cause of adult-onset nephrotic syndrome, with primary MN of unclear cause accounting for 80% of cases. Retrospective clinical research reported that MN occurring in rheumatoid arthritis (RA) and ankylosing spondylitis (AS) patients was triggered by nephrotoxic drugs or of unknown cause. However, whether RA or AS itself increases the risk of developing MN is unknown. METHODS: We conducted mendelian randomization (MR) analysis to evaluate the causal effects of RA or AS on MN using genome-wide association study (GWAS) statistics. The inverse variance weighted (IVW) method was the primary analysis, and several supplementary analyses and sensitivity analyses were performed to test the causal estimates. RESULTS: We obtained 30 valid instrumental variables (IVs) of RA and 16 valid IVs of AS from large-scale open-access GWASs. The genetically predicted RA significantly increased the risk of MN [IVW odds ratios (OR) = 1.327, 95% confidence interval (CI) = (1.124, 1.565), P = 8.051 × 10(-4)]. Three supplementary MR analyses provided the consistent positive causal effect of RA on MN (all P < 0.05). No horizontal pleiotropy was detected by MR Egger intercept analysis (P = 0.411). However, the genetically predicted AS had no causal effect on MN by IVW and supplementary analysis (all P > 0.05). CONCLUSIONS: Genetically predicted RA could increase the risk of MN, but genetically predicted AS was not associated with MN. Screening for kidney involvement in RA patients should be noted, and active treatment of RA will reduce the public health burden of MN.