The Association between Endometriosis and Immunological diseases

Phen ot ypic as soc iat ion bet we e n end omet riosis and i m mune condit ions The phenotypic association between endometriosis and immunological conditions was investigated in the UK Biobank (UKBB) using both cross-sectional and retrospective cohort study designs, with the latter assuming endometriosis as a diagnostic risk factor preceding an immunological disease diagnosis (see Methods). For the cross-sectional analysis, 8,223 endometriosis cases vs. 265,181 female controls without known endometriosis were included; and 64,620 immunological disease cases vs. 208,784 female controls without known immunological diseases. Supplementary Table 1 shows factors that were determined as potential confounders or mediators in the association analyses between endometriosis and immunological diseases. Adding factors significantly associated with both endometriosis and immunological diseases in a logistic regression model with endometriosis as exposure and any immunological disease as the outcome (see Methods), none were found to be confounders or mediators that significantly affected the effect size of association (>5% change). However, genetically determined ancestry and age at recruitment were included a-priori as potential confounders. In both the cross-sectional and retrospective cohort analyses (Table 1), females with endometriosis vs. those without had a significantly increased risk for all immunological diseases combined (OR: 1.32 (1.25-1.39); HR: 1.32 (1.20-1.45)), classic autoimmune diseases (OR: 1.24 (1.13-1.36); HR: 1.41 (1.15-1.74)), aut oinflammatory diseases (OR: 1.33 (1.26-1.41); HR: 1.29 (1.17- 1.43)), and mixed-pattern diseases (OR: 1.23 (1.10-1.52); HR: 1.88 (1.25-2.81)). Immunological diseases significantly associated with endometriosis in both cross-sectional and cohort analyses were: rheumatoid arthritis (OR:1.22 (1.04-1.41), P = 0.011; HR: 1.57 (1.18-2.10), P = 0.002); coeliac disease (OR: 1.35 (1.06-1.70), P = 0.011; HR: 1.99 (1.30-3.07), P = 0.002); and osteoarthritis (OR: 1.35 (1.27-1.43), P < 0.001; HR: 1.31 (1.19-1.44), P < 0.001). In addition, in the ‘gold standard’ cohort analyses, p soriasis (HR: 1.67 (1.05-2.65), P = 0.030) was significantly . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint associated with endometriosis. One immunological condition significantly associated with endometriosis in cross-sectional analysis, systemic lupus erythematosus (OR:1.62 (1.14-2.24), P=0.005), myasthenia gravis ( OR: 2.55 (1.30-4.49), P=0.003) and gout (OR:1.66 (1.18-2.26), P=0.002), could not be tested in a cohort study design due to insufficient case numbers. Overall, females with endometriosis compared to females without known endometriosis exhibited a 14% increased risk for at least having one immunological disease (OR = 1.14 (1.08-1.21)), a 21% increased risk for at least having two immunological diseases (OR = 1.21 (1.05-1.39)), and a 30% increased risk for having at least three immunological diseases (OR = 1.30 (0.92-1.78)) at any point in their lifetime (P < 0.001) (Supplementary Table 2). When stratifying by menopausal status, gynaecological surgery (hysterectomy/oophorectomy), or hormone replacement therapy (HRT) use, effect sizes for the association between endometriosis and overall immunological disease risk remained largely unchanged (Supplementary Table 3). G en etic c or relatio n between e n dom etr iosis and imm unological dis ease s For a total of eight immunological diseases associated with endometriosis either in cross-sectional or cohort analyses (ankylosing spondylitis, coeliac disease, inflammatory bowel disease, multiple sclerosis, osteoarthritis, psoriasis, rheumatoid arthritis, and systemic lupus erythematosus), we conducted female-only and sex-combined European ancestry GWAS analyses in UKBB (Table 2, See Methods). To achieve the greatest power to detect variants associated with each disease, sex- combined UKBB GWAS results were meta-analysed with existing GWAS summary statistics based on the largest sample sizes where available (Table 2), using the inverse vari ance weighted fixed- effects method as implemented in METAL (Supplementary Figures 1-8). Utilising the GWAS meta-analysis summary results for the immunological diseases and the largest published endometriosis GWAS meta-analysis result s (excluding UKBB GWAS results to achieve sample independence) 16 , we applied linkage disequilibrium (LD)-score regression (LDSC) analysis 17,18 to estimate the genetic correlation (rg) between endometriosis and the eight immunological diseases. Osteoarthritis (rg=0.29, p=3.25x10 -15 ), rheumatoid arthritis (rg=0.26, p=1.54x10 -5 ) and multiple sclerosis (r g=0.09, p=4.00x10 -3 ) showed significant (p<6.25x10 -3 , see Methods) positive genetic correlations (rg) with endometriosis, suggesting a shared genetic component (Table 3). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint M en delian rando mis at ion (MR) anal yse s MR analyses using genetic instrumental variables (IVs) were conducted to further investigate a potential causal relationship between endometriosis (exposure) and the increased risk of osteoarthritis, rheumatoid arthritis, and multiple sclerosis (outcomes) observed in the phenotypic association and genetic correlation analyses. Table 4 shows the results from the main MR-IVW model, with weighted median MR and MR-Egger regression provided as sensitivity analyses to test the robustness of the results. Analyses were conducted utilising 39 genome-wide significant (p<5x10 -8 ) endometriosis associated LD-independent autosomal variants as IVs. Summary statistics for the IVs were extracted from UKBB female-only and meta-analysed sex-combined GWAS results to represent the outcomes of the three immunological diseases (see Methods). For endometriosis vs. rheumatoid arthritis, utilising the 39 IVs illustrated a suggestive causal relationship between endometriosis vs. rheumatoid arthritis in females-only (OR [95%CI] = 1.16 [1.02-1.33], p-value=0.028). For endometriosis vs. osteoarthritis, and vs. multiple sclerosis, the MR analysis did not identify a significant causal relationship in either sex-combined or female-only analyses (Table 4). M u lti-tr ait analysis of endom et riosis and im munological diseases: oste oar thrit is , r heumat oid art hr it i s and mu ltiple sc ler os is To leverage the genetic sharing of association signals between endometriosis and osteoarthritis, rheumatoid arthritis, and multiple sclerosis for the discovery of additional endometriosis risk variants, we conducted a multi-trait analysis of GWAS (MTAG). MTAG capit alises on the genetic correlation between diseases to boost statistical power for detecting associations in genome-wide analyses 19 . The MTAG analysis was conducted for all four diseases simultaneously and identified 42 genome-wide significant lead SNPs for endometriosis (Supplementary Table 4), 6 of which were not reported previously 16 (Supplementary Figure 9a-f). These 6 genetic variants are eQTLs for various genes across multiple tissues including MS RA and PO N 2 protecting and repairing cells from oxidative stress in blood 20,21 , BLK and ZAP70 encoding enzymes belong to Tyrosine kinase family with roles in cell proliferation and differentiation in particular B-cell and T-cell development and adhesion 22,23 , ATRAID , S LC35F6, TMEM214 and XKR6 involved in apoptosis-related pathways 24,25 and, TRPS1 encoding a transcription factor that represses GATA-regulated genes involved in progesterone resistance and endometriosis progression in the pelvis 26 (Supplementary Table 5). The MTAG analysis for osteoarthritis yielded 27 significant variants (Supplementary Table 6), for . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint rheumatoid arthritis yielded 28 significant variants (Supplementary Table 7) and for multiple sclerosis, it identified 64 genome-wide significant variants (Supplementary Table 8). Functional ann otatio n of ide nt ified g enome- wi d e s ignificant var iants and path way a nalysis We mapped genome-wide significant asso ciations from each MTAG analysis for endometriosis, osteoarthritis, rheumatoid arthritis, and multiple sclerosis to the genes whose expression they are associated with, using GTEx V8 (54 tissues) and eQTLGen (31,684 blood datasets). We identified 439 genes regulated by 42 genome-wide significant endometriosis associated variants; 379 genes by 27 genome-wide significant osteoarthritis associated variants (Supplementary Table 6); 490 genes by 28 genome-wide significant rheumatoid arthritis variants (Supplementary Table 7); and 1,113 genes by 64 genome-wide significant multiple sclerosis associated variants (Supplementary Table 8). Of the 439 genes regulated by endometriosis risk variants, 192 were also regulated by a genome-wide significant risk variant of one or more of the other immune diseases (Figure 1). Pathway analysis (see Methods) based on the identified genes per disease identified numerous canonical pathways enriched with these genes (Supplementary Table 9-12). Among the top enriched pathways for endometriosis was ‘ s ignalli ng by receptor t y r os ine kinases’, a major class of cell surface proteins involved in signal transduction which triggers many downstream signalling pathways including NF k B , MA PK and AKT. These pathways are activated in endometriosis and have been suggested to harbour potential targets for non-hormonal therapeutics 27 . Investigating the overlap of enriched genetically driven pathways between endometriosis, osteoarthritis, multiple sclerosis, and rheumatoid arthritis, we discovered that 45 out of the 79 enriched pathways for endometriosis were also enriched in the other immune conditions (Figure 2). In total 7 enriched pathways were shared across all four conditions, including ‘ signalli ng by receptor t yrosi ne kin as e s ’, ‘innate im mune s y stem’ , ‘adaptive immune syst em’ , ‘ extr ac ellular matr ix or ganis at ion ’ , ‘leuk ocyte tr ans- endot helia l migration’ , ‘l ipi d me t abo lis m’, and ‘a rachidonic aci d metabo lis m’ (Supplementary Figure 10). Within these overlapping enriched pathways, there are genes shared between conditions and also genes specific to each condition contributing to the pathway. For example, of the 21 genes enriched from endometriosis in ‘ signa l l ing by reception tyrosine kinas e ’, 8 are shared with osteoarthritis, including N CF4 , LAMB2, RHOA, MS T1, MST1R, MA PKAPK3, D O CK3 , and P TK2 B, and 3 are shared with multiple sclerosis including IT GB3 , PRK CA , and MM P9 (Supplementary Figure 10a). Another enriched pathway across the 4 conditions is . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint ‘ arachidonic ac i d met abo lis m’ ; of the 5 endometriosis genes enriched in this pathway, 4 are shared with the other 3 immune conditions, namely, D P EP3 , G P X1, D PE P2 , and PON2 ( Figure 2i) . Ident ification of shar ed ge netic v a r iants betw een e ndom et riosis and im mune dis eas e s A total of 12 osteoarthritis, rheumatoid arthritis and multiple sclerosis gen ome-wide significant lead SNPs were mapped within 1Mb of endometriosis genome-wide significant lead SNPs, with four of them tagging the same signal (r 2 >0.5) (Table 5, Supplementary Figure 9a-f). Three of these were shared with osteoarthritis ( BMPR2/2q33.1, BSN /3p21.31, and ML L T10 /10p12.31), and one with both osteoarthritis and rheumatoid arthritis ( XKR6/8p23.1). MTAG association results of the 12 genome-wide significant endometriosis SNPs for osteoarthritis, rheumatoid arthritis and multiple sclerosis are provided in Supplementary Table 13. At the BMPR 2/ 2q33.1 locus, the lead SNPs rs72928925 for endometriosis and rs72928605 for osteoarthritis are both eQTLs for BMP R2 in blood and oesophagus muscularis (Supplementary Table 14). BMP R2 encodes a member of the BMP receptor family of transmembrane serine/threonine kinases. The ligands of this receptor are members of the TGF-beta superfamily. The TGF-beta signalling pathway, involved in diverse cellular processes including cell proliferation, differentiation, apoptosis, and migration invasion, was also one of the pathways enriched with 10 eQTL genes regulated by endometriosis, osteoarthritis and multiple sclerosis associated variants (Figure 2, Supplementary Tables 9-12). At the BSN /3p21.31 locus, the lead SNP rs6774202 associated with endometriosis and rs6809879 with osteoarthritis are both eQTLs for a diverse set of overlapping genes (Table 5) that are part of pathways enriched between endometriosis and the other three immune conditions (Supplementary Tables 9-12). In particular, RH O A is part of the ‘ leukocy t e trans-endot hel i al migrat io n’ pathway that was enriched across all four conditions. A thir d sh ar ed locus i s XKR6 / 8p 23.1 , where the lead endometriosis SNP rs12542037 is in strong LD with the lead genome-wide significant osteoarthritis and rheum atoid arthritis SNPs (Table 5). This locus is involved in the regulation of multiple genes, namely BLK, CT S B , and MTM P9 , which play roles in innate and adaptive immune system pathways 28 . In addition, FD F T1 , regulated by the correlated genetic risk variants, encodes for squalene synthase that is involved in cholesterol biosynthesis. The ‘ lipid metabol ism pathway’ is also enriched with genes r egulated by genetic . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint variants in each of the investigated 4 conditions (Supplementary Tables 9-12, Figure 2 and Supplementary Figure 10f). The fourth locus previously implicated and described for endometriosis and osteoarthritis is MLLT10 /10p12.31, which harbours genes such as MLLT10 associated with pain perception and maintenance in multiple tissues 16 . D iscu s sion Our study of UKBB data reveals a significant increase in the risk of auto-immune and auto- inflammatory diseases among endometriosis patients, particularly in rheumatoid arthritis (HR: 1.57 (1.18-2.10), P = 0.002), coeliac disease (HR: 1.99 (1.30-3.07), P = 0.002), osteoarthritis (HR: 1.31 (1.19-1.44), P < 0.001), and psoriasis (HR: 1.67 (1.05-2.65), P = 0.030). Given the age at recruitment into UKBB (40-69 years, in 2006-10) and changes in awareness of endometriosis over time, the proportion of diagnosed females in UKBB (8,223 cases in 273,404 females = 0.03%) is relatively low compared to population prevalence estimates (up to 10% 1 ) . This would have resulted in the undiagnosed females driving effect sizes for associations towards the null 29,30 . Nevertheless, our results are consistent with evidence from previous case/control and cohort studies which showed significant association between endometri osis rheumatoid arthritis (RR:1.46 (0.70-3.03), coeliac disease (RR:1.39( 1.14-1.70)), multiple sclerosis (OR:7.1 (4.4-11.3)) 14 and psoriasis (RR:1.75 (1.10-2.78)) 31 . For systemic lupus erythematosus, cross-sectional evidence showed significant evidence (OR: 1.62 (1.14-2.24)) but we didn’t have enough cases to carry out the cohort study analysis. However, there is previous longitudinal evidence showing increased risk of systemic lupus erythematosus (HR:2.03 (1.17-3.51)) 32 . Moreov er, endometriosis patients compared to females without endometriosis were at increased risk of suffering from multiple immunological diseases which was most pronounced for autoinflammatory diseases (Supplementary table 2). This trend was observed in an early adulthood cohort study which we now expand to a broader age range. Our genetic correlation analysis suggests that genetic factors contribute to the association between endometriosis and the increased risk of rheumatoid arthritis (rg=0.27, P=1.54x10 -5 ), osteoarthritis (rg=0.28, P=3.25x10 -15 ), and to a lesser extent, multiple sclerosis (rg=0.09, P=4.00x10 -3 ). The significant genetic correlation between endometriosis and an immunological . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint condition can be attributed to multiple mechanism s as investigated through MR analyses: (1) endometriosis causally leads to the subsequent development of the immunological condition; (2) endometriosis and the immunological condition share a common genetically driven cause; or (3) endometriosis and the immunological condition share multiple common causes, and the direction of effect between them can be complex 33 . Genetic correlation analyses are more powerful when the genetic architecture between conditions is polygenic involving many causal SNPs of small effect to l everage their aggregated effects, which is the case for endometriosis and the immunological conditions we studied here. The MR analyses yielded no robust evidence of causal relationships between endometriosis and genetically correlated immunological conditions, except for a suggestive causal effect of endometriosis on rheumatoid arthritis in females-only (OR=1.16, 95% CI=1.02-1.33, p=0.028). MR analysis provides insights into whether the association between two complex conditions is causal by utilising genetic variants associated with the exposure (endometriosis). However, it is assumed that the genetic variants utilised as IVs have strong predictive power of the exposure, and the recommendation is to limit these to genome-wide significant (GWS) associated variants. However, even GWS variants as instruments often modestly predict the exposure, which can limit the power of MR an alysis 34 . The 39 endometriosis associated variants included as IVs in our MR analyses explain <2% of heritable variation (5% of stage III/IV disease 16 ) which has implications for the interpretation of non-significant MR results. Our MR instruments would have been weighted towards risk for stage III/IV disease, in particular ovarian endometrioma 16 . Previous studies associating risk of auto-immune and inflammatory conditions with endometriosis included predominantly stage I/II cases 32,35 , although some of this evidence was based on adolescents who may have been genetically predisposed to develop stage III/IV disease later in li fe 35 . Another

we encounter is a lack of available female-specific immunological disease GWAS meta- analysis results, which is surprising given that many immunological conditions exhibit higher prevalence in females. We conducted female-specific GWAS analyses in the UK Biobank, however, sample sizes were limited compared to sex-combined GWAS meta-analysis for these conditions in the literature. The suggestive causal effect of endometriosis on rheumatoid arthritis is intriguing and warrants further exploration in the future, with IVs that explain a greater proportion of the genetic variability for endometriosis. This will require larger endometriosis GWAS to uncover m ore genetic . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint variants contributing to the polygenic component of endometriosis. A large-scale female-specific rheumatoid arthritis GWAS meta-analysis should also result in a more relevant dataset to explore its potential causal basis with endometriosis. While power limitations hamper the interpretation of causal relationships between endometriosis and osteoarthritis, rheumatoid arthritis, or multiple sclerosis, results of the genetic correlation analyses highlight a shared genetic basis. Understanding the basis of genetic sharing regardless of causality is important, as shared biological mechanisms of pathogenesis and pathophysiology could open new avenues for treatment development. The MTAG analyses, leveraging genet ic correlations between endometriosis, osteoarthritis, rheumatoid arthritis and multiple sclerosis, identified 42 genome-wide significant loci for endometriosis, 6 of which were not identified before: AB H D 1 / 2p23.3, TMEM131/ 2q11.2, XRCC4/ 5q14.2, PP P1R9 A/ 7q21.3, XK R 6 / 8p23.1 and TR PS1/8p23.3. These variants were eQTLs for various genes involved in protecting and repairing cells from oxidative stress, in B-cell and T-cell development, apoptosis-related pathways and regulation of progesterone resistance. The MTAG- derived 42 GWS lead endometriosis-associated variants were mapped to 439 genes, 27 GWS lead osteoarthritis SNPs to 379 genes, 28 GWS lead rheumatoid arthritis SNPs to 490 genes and 64 GWS lead multiple sclerosis variants t o 1113 genes. When we considered the overlap between these genes, 43.7% of endometriosis eQTL genes were shared with at least one of the three immunological conditions; the majori ty with osteoarthritis (33%) followed by multiple sclerosis (11.6%) and rheumatoid arthritis (7%). Pathway analysis revealed that 50.6% of pathways enriched in gene lists for endometriosis were also enriched in multiple sclerosis, 26.6% in osteoarthritis and 16.5% in rheumatoid arthritis. Seven pathways were enriched across endometriosis and all three immune conditions, including large, general immune regulatory pathways ‘ i nn ate immune system’ (1100 genes) and ‘adaptive immune sy stem’ (807 genes). A more specific shared pathway was ‘ si gn al ling by receptor tyrosine k inase s ’. Receptor tyrosine kinases are a large family of cell-surface receptors that involved wide-variety inter and intra-cellular signalling. Previous studies have su ggested the involvement of kinase signalling pathways and potential non-hormonal treatment targets therein for endometriosis 27 , which our analyses support. Another shared pathway, ‘ extracellular matr ix or g anisatio n’, included MMP9, PRKCA and ITGB3 . MMP9 encodes for a metalloproteinase that has a purported role in the progression of invasion in endometriosis as well as angiogenesis and fibrosis 36 , has involvement in a variety of inflammatory autoimmune diseases, and has been suggested to be a therapeutic target for autoimmune conditions 37,38 . . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint PRKCA is involved in immune cell trafficking and ITG B 3 is coding for integrin beta3 expression of which is associated with autoimmune conditions including multiple sclerosis 39 . Intriguingly,‘ ar ac hidonic aci d metabo l ism’ was another shared pathway, including four genes shared between endometriosis and osteoarthritis, rheumatoid arthritis and multiple sclerosis (D PEP 3, G P X1, D PE P2 and PON2). Arachidonic acid is an essential fatty acid ingested through diet. Arachidonic acid derived prostaglandins contribute to inflammation through their role as intercellular pro-inflammatory mediators, and promote excitability of the peripheral somatosensory system contributing to pain exacerbation 40 . Among the GWAS loci and effectors genes we identified through MTAG and eQTL analyses, three were shared between endometriosis and osteoarthritis ( BMPR2/2q33.1, BS N /3p21.31, and MLLT10 /10p12.31), and one, XKR6 /8p23.1, between endometriosis and rheumatoid arthritis. MLLT10 has been associated with pain perception and maintenance across multiple tissues and has been previously described 16 . BMPR2 encodes a member of BMP receptor family of transmembrane serine/threonine kinases, acting as receptor for the TGF-beta superfamily. SNP s at the BSN/ 3p21.3 1 locus are eQTLs for various genes in cluding RH O A , part of the ‘ leukoc yte t rans- endothel ial mi grat io n’ pathway that is enriched across all four conditions. Leukocytes migrate from blood into tissues as part of inflammation and immune surveillance. During this process leukocytes bind to cell adhesion molecules and migrate across the vascular endothelium. Another interesting eQTL gene asso ciated with variants at BS N/3p21.31 is HY A L 3 , which is involved in the hyaluronan/hyaluronic acid metabolism and glycosaminoglycan degradation pathways. Hyaluronic acid is a naturally occurring glycosaminoglycan most abundant in the extracellular matrix involved in various physiological processes including wound healing, tissue regeneration and joint lubrication. It is also used for relief of joint pain, wound healing and various other applications, and has been shown to reduce production of proinflammatory mediators, reduce sensitivity associated with osteoarthritis pain 41,42 . Recent in-vivo and in-vitro studies suggest that hyaluronic acid may have the ability to reduce endometriosis lesion size in mice but that it may also promote inflammation when administered acutely 43 , suggesting further research into mechanism of action and therapeutic potential in endometriosis is needed. XKR6/8p23 .1, shared between endometriosis and rheumatoid arthritis, was asso ciated with the regulation of multiple genes including BLK, CTSB, and MTM P9 - which play roles in innate and adaptive immune system pathways 28 – and FDFT1 , involved in cholesterol biosynthesis. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint Cholesterol is a precursor of steroid hormones and essential part of plasma membranes. It is also enriched in lipid rafts which play an important part in many cellular processes including signal transduction pathways, membrane trafficking, cytoskeletal organisation, apoptosis, cell adhesions and migration 44 . In the context of inflammatory conditions, lipid metabolism has been suggested to harbour targets for reducing inflammation without the undesirable side-effects of anti- inflammatory therapies 45 . As mentioned, our analyses were limited by the lack of large-scale female-specific GWAS meta- analyses for immune conditions, particularly those exhibiting higher prevalence in females. It is well established that sex-specific genetic signatures are present for conditions showing variability by sex 46 , and female-specific GWAS r esults for immunological conditions may offer increased genetic correlations with endometriosis and opportunities for discovery and shared genetic signals. Future genetic comorbidity analyses should also explore results for different endometriosis subtypes. Recent GWAS analyses have suggested that ovarian endometriosis has a different genetic basis to peritoneal disease 16 , but the sample sizes for which summary statistics were generated did not allow for sufficiently powered inclusion in the present analyses. Similarly, future analyses should explore signals for different subtypes of immunological diseases, such as osteoarthritis 47 , once larger GWAS datasets become available. Lastly, genetic analyses were limited to European ancestry individuals and larger GWAS across more diverse ancestry groups are needed. In conclusion, our results show that females with endometriosis are at a modestly (30-40%) increased risk of both auto-immune and auto-inflammatory conditions, and that comorbidity with osteoarthritis, rheumatoid arthritis, and to a more limited ex tent multiple sclerosis, is biologically underpinned. In terms of clinical relevance, we suggest awareness among treating physicians of this increased risk of comorbidity, in order to spot early symptoms of immunological conditions among females with endometriosis, and vice versa. While current clinical action is limited to increased vigilance, the results offer a wide range of novel avenues and targets for exploring mechanism s and potential cross-condition treatment development or repurposing. Me t h o d s . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint Phen ot ypic an alysis s t ud y pop ulatio n and disea s e a sc er tainme nt The UK Biobank (UKBB) is comprised of 500K individuals aged 40-69 at time of recruitment (2006- 2010) from across the UK. The biobank was approved by the North West M ulti-Centre Research Ethics Committee (MREC). In the UKBB, information was collected from participants during recruitment using questionnaires on socioeconomic status, behavioural, family history and medical history. Participants were also followed up for cause-specific morbidity and mortality through linkage to disease registries, death registries, hospital admission records and primary care data. In addition, a range of biological samples including blood, urine and saliva were collected from the participants. A more detailed description of the UKBB can be found in the UK Biobank protocol 48 . Given that endometriosis is a gynaecological condition affecting those assigned female at birth, only the individual’s assigned female at birth (N=273,404) were included in the phenotypic association analysis with the immunological conditions. From this point onwards we will refer to those assigned female at birth as females in this manuscript. Endometriosis was identified based on self-reported data from questionnaires and/or hospital records (ICD10/9: N801-809 and 617.1- 9). A total 31 immunological conditions were identified from self-reported data and/or hospital records (ICD 10/9) that were classified into three groups (94) as following: (1) autoinflammatory conditions: Acne (ICD10/9: L70* and 7060, 7061), acute respiratory distress syndrome (ICD10/9: J80, P220, 769, 769.9), erythema nodosum (ICD10/9: L52, 6952) , giant cell/Takayasu arteritis (ICD10/9: M314, M315, M 316, 4465, 4467), gout/pseudogout (ICD10/9: M10*, M11*, 274, 2740, 2741, 2748, 2749, 712, 7120, 7121, 7122, 7123, 7128, 71280, 71281, 71282, 71283, 71284, 71285, 71286, 71287, 71288, 71289, 7129, 71290, 71291, 71292, 71293, 71294, 71295, 71296, 71297, 71298, 71299), total inflammatory bowel disease (ICD10/9: K50, K500, K501, K508, K509, K51, K510, K511, K512, K513, K514, K515, K518, K519, 555, 5550, 5551, 5552, 5559, 556, 5560, 5569), Crohn’s disease (ICD10/9: K50, K500, K501, K508, K509, 555, 5550, 5551, 5552, 5559) Ulcerative colitis (ICD10/9: K51, K510, K511, K512, K513, K514, K515, K518, K519, 556, 5560, 5569), osteoarthritis (ICD10/9: M15, M150, M1500, M151, M152, M153, M154, M158, M159, M1599, M16, M160, M161, M162, M163, M164, M165, M166, M167, M169, M17, M170, M171, M172, M173, M174, M175, M179, M18, M180, M181, M182, M183, M184, M185, M189, M19, M190, M1900, M1911, M1912, M1913, M1914, M1915, M1916, M1917, M1918, M1919, M192, M1920, M1921, M1922, M1923, M1924, M1925, M1926, M1927, M1928, M1929, M198, M1980, M1981, M1982, M1983, M1984, M1985, M1986, M1987, M1988, M1989, M199, M19990, M 1991, M1992, M1993, M1994, M1995, M1996, M1997, M1998, M1999, 715, 7150, 7151, 71510, 71511, 71512, . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint 71513, 71514, 71515, 71516, 71517, 71518, 71519, 7152, 71520, 71521, 71522, 71523, 71524, 71525, 71526, 71527, 71528, 71529, 7153, 71530, 71531, 71532, 71533, 71534, 71535, 71536, 71537, 71538, 71539, 7158, 7159), sarcoidosis (ICD10/9: D860, D861, D862, D863, D868, D869, 135, 1359), (2) classical autoimmune conditions: Addison’s disease (ICD 10/9: E271, E272, 25540, 25542), autoimmune gastritis(ICD10/9: D510, 2810), autoimmune thyroid disease (ICD10/9: E050, E063, 2420, 2452), Graves’ disease (ICD10/9: E050, 2420), Hashimoto’s disease (ICD10/9: E063, 2452), coeliac disease (ICD10/9: K900, 5790), dermatomyositis/polymyositis (ICD10/9: M33*, M360, 7103, 7104), multiple sclerosis (ICD10/9: G35, 340), myasthenia gravis (ICD10/9: G70, G700, G701, G702, G708, G709, 358, 3580, 35800, 35801, 35809, 3581, 3582, 3588, 3589), pemphigus/pemphigoid (ICD10/9: H133, L10, L100, L101, L102, L103, L104, L105, L108, L109, L12, L120, L121, L122, L123, L128, L129, 6944, 6945, 6946), primary biliary cirrhosis (ICD10/9: K743, 5716), rheumatoid arthritis (ICD10/9: M05*, M06*, 714, 7140, 71400-71409, 7141, 71410-71419, 7142, 71420-71429), Sjögren's syndrome (ICD10/9: M350, 7102), systemic lupus erythematosus (ICD10/9: M32*, L90*, 6954, 7100), systemic sclerosis (ICD10/9: M34*, 5172, 7101), type 1 diabetes (ICD10/9: E10*, 25001, 25011, 25021, 25091), vitiligo (ICD10/9: L80, 70901), (3) Combination of autoinflammatory and autoimmune condition categories: Ankylosing spondylitis (ICD10/9: M081*, M45*, 7200), Behcet’s syndrome (ICD10/9: M352, 1361, 7112), reactive arthritis (ICD10/9: M023*, 0993, 7111), psoriasis/psoriatic arthritis/psoriatic arthropathies (ICD10/9: L40*, M070*, M073*, 6961, 6960). A common control set was defined as females without endometriosis diagnosis excluding anyone with diagnoses of any of the 31 immunological conditions. Potential confounding or mediating factors were determined including age of recruitment, genetically determined ancestry, menopause status, age at menarche, parity, body size, BMI and fat distribution 49 , alcohol consumption, smoking, infertility and disease such as ovarian cancer 50 and cardiovascular disease 51 , which h ave been illustrated to be associated with endometriosis and some immunological conditions. Age at recruitment (which represents potential age-related cohort effects) and ancestry were considered a-priori variables to be included in the models. Many of the other factors were assessed only at baseline recruitment into UKBB, which for most females would have followed rather than coincided with, or preceded, an endometriosis diagnosis, and therefore the potential for confounding vs. mediation effects could not be accurately assessed. Nevertheless, to assess their potential impact on the associations, factors associated both with endometriosis and immunological disease were include in a logistic regression model with endometriosis as exposure and immunological disease as outcome. None of these factors either . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint showed >5% change in effect (potential confounders) or removal of effects (mediators), and therefore only a-priori variables age at recruitment and genetically determined ancestry were included in the models. Phen ot ypic as soc iat ion analysis Phenotypic association analysis between endometriosis and immune conditions was conducted utilising two different analysis methods: (1) a cross- sectional analysis to test for a simple association between risk of an immunological disease diagnosis with a diagnosis of endometriosis at any point in time, including all fem ales in the UKBB; (2) a ‘gold standard’ cohort study design to incorporate temporality between diagnoses, where entry time was defined as the recruitment date into UKBB. Cross-sectional analysis was conducted for 26 immunological disease that had at least 100 female cases in UK Biobank. A total of 5 immune conditions were excluded from analysis due to number of cases <100: Reactive arthritis (N=4), Behcet’s syndrome (N=27), acute respiratory distress syndrome (N=79), erythema nodosum associated disease (N=79), pemphigus/pemphigoid (N=95). Cohort analysis was conducted for 9 immunological diseases with a minimum of 1,500 female cases to allow enough number of immunological disease cases after excluding prevalent immunological diseases diagnosed before cohort entry time or before the endometriosis diagnosis. The majority of females had immunological diseases diagnosed after endometriosis (66.8%, 1,275 out of 1,909 females with both diagnoses, had an immunological disease diagnosis after their endometriosis diagnosis). Therefore, endometriosis was treated as the exposure and immunological disease as the outcome in the cohort analyses. This also fits with the observation that many individuals with endometriosis have symptom onset in their teens or twenties, often many years before their ultimate diagnosis 52 . Females with an endometriosis diagnosis at the time of recruitment were classified as exposed, whereas those who had not had an endometriosis diagnosis at the time of recruitment were classified as unexposed. Those individuals who received an endometriosis diagnosis during follow-up, prior to any immunological disease diagnosis, contributed person-time to the unexposed group until the occurrence of endometriosis diagnosis, if any, and subsequently to the exposed group after diagnosis. For each immunological disease, females who had the respective immunological disease diagnosed before endometriosis or those who had the respective immunological disease diagnosed before cohort entry time or had immunological disease diagnosis time missing were excluded from cohort analysis (Table 1). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint In the cross-sectional study analysis, the prevalence of each specific and categorized immunological diseases in females with and without a history of endometriosis diagnosis was investigated using logistic regression models with odds ratios (ORs) as risk measure. Cross- sectional study analysis for each specific and categorized immunological diseases was conducted with adjustment of age and genetically determined ancestry. In the cohort study, the risk of incident immunological diseases in females with and without endometriosis history was investigated using Cox proportional hazards regression models with calculated hazard ratios (HRs). The proportional hazards assumption was tested by function of “cox.zph” in the “survival” R library. In the cohort analysis, time to event was formulated from entry to the cohort until the end of follow-up time. The follow-up time (rather than age) is used as the underlying time variable, since the date of assessment is described in more detail with information on the exact date and months participants attended the assessment centre (to be used as the index date) in the UK Biobank. The end of follow-up time is the date of incident immunological diseases, death, loss to follow-up or end of follow-up (end date of follow-up is the date of last download of the dataset, which is 8 th Jan. 2019), whichever occurred first. Cohort analysis for each specific and categorized immunological diseases was conduct ed with adjustment of age (categorical: =60) and genetically determined ancestry (categorical: white, non-white). All risk estimates were reported with 95% confidence intervals (CIs) and two-sided P-values. Person- years and mean follow-up time for each cohort analysis were calculated. All analyses were carried out using R software. G en om e-wide as s o c iat ion s t ud y (G WA S) and met a-analys is for immune c ondit ions Only genetically determined European ancestry individuals were included in the analysis. GWAS was conducted using UKBB data for females-only and sex-combined study population for 8 immune conditions: inflammatory bowel disease (N = 2,869 females; N = 5,751 sex-combined), osteoarthritis (N = 39,866 females; N = 68,878 sex-combined), ankylosing spondylitis (N = 547 females, N = 1,493 sex-combined), psoriasis (N = 3,036 females, N = 6,591 sex-combined), coeliac disease (N = 1,706 females, N = 2,640 sex-combined), multiple sclerosis ( N = 1,314 females, N = 1,883 sex-combined), rheumatoid arthritis (N = 4,662 females, N = 7,153 sex-combined) and systemic lupus erythematosus (N = 545 females, N = 673 sex-combined). Controls were defined as a common control set without any diagnosis of immunological diseases or endometriosis within UKBB. The linear mixed model (LMM) implemented in BOLT 53 was utilised for GWAS analysis to take into account relatedness in the data and to increase power of analysis by a linear mixed effects model. GWAS results were adjusted for a binary variable denoting the genotyping chip (the . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint UKBB Axiom array or the UK BiLEVE array), for SNPs with minimum MAF filter of < 1% and included SNPs with ≤ 60% missingness. Furthermore, published largest European ancestry GWAS results on these 8 immunological conditions were identified through literature 47,54-60 and downloaded for meta-analysis with UKBB GWAS results. Before meta-analysis, GWAS study-level QC was performed and markers absent in the 1000G reference panel, large missing values (≥ 60%) or lack beta/odds ratio estimates were excluded. GWAS meta-analysis for each immunological disease was carried out using METAL software using an inverse variance weighted fixed effect meta-analysis method 61 . GWAS meta- analysis results were filtered and excluded those with MAF 90), present < 50% effective sample size (Neff; Neff = 4NCases*NControls/( NCases + NControls). All genetic analyses were done on genome reference of homo sapiens (human) genome assembly GRCh37 (hg19). Genetic information on chromosome X was excluded. MHC region of Chr6:24000000-35000000 was excluded as it has a dense clustering of imm une-relevant genes with extreme polymorphism and very strong long-range linkage disequilibrium, which complicates the determination of the exact genes and alleles that are responsible for signals of disease association in the region 62 . G en etic c or relat i o n analys is In genetic correlation analysis, for endometriosis, the GWAS meta-analysis results from the International Endometriosis Genome Consortium (IEGC) were used including 52,350 cases and 504,157 controls from 20 GWAS studies excluding UKBB to prevent overlapping study population 16 . Then genetic correlation analysis was conducted between endometriosis and immunological diseases GWAS meta-analysis results via linkage disequilibrium score regression (LDSC) analysis 17,18 . The LD-score was calculated using software available at (http://github.com/bulik/ldsc), which was based on the 1000 Genomes Eur opean population and estimated within 1-cM windows, the significan ce threshold was set as p-value=0.00625 to account for multiple testing of eight immunological diseases. M en delian rando mis at ion analysi s The potential causal relationship between endometriosis, as exposure, and those genetically correlated immunological disease, as outcome, were investigated by two-sample Mendelian randomisation (MR) using the TwoSampleMR software 63 . As instrumental variables (IVs) we utilised the 39 established genome-wide significant LD-independent lead autosomal SNP s . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint associated with endometriosis to assess whether endometriosis causally affects those genetic correlation immune conditions namely, osteoarthritis, rheumatoid arthritis or multiple sclerosis. Inverse variance weighted MR (MR-IVW) was applied as the initial method to detect causal effect 64 . As sensitivity analysis, other two-sample MR methods including weighted median MR 65 and MR-Egger regression 66 were impl emented in case the assumption of valid IVs was violated. Weighted median MR was shown to have lower Type 1 error rates than the inverse-variance weighted method in a simulation analysis 65 . MR Egger provides a sensitivity analysis to detect evidence of heterogeneity and pleiotropy of IVs 66 . To detect IVs with effect of heterogeneity and pleiotropy, MR PRESSO was applied to identify outliers 67 . Also, scatter plots 68 were generated to present the SNP-outcome association estimates versus the SNP-exposure associations in investigating the causal relationship using the MR models, including IVW, weighted median MR and MR-Egger regression 67 . Furthermore, the strength of the 39 IVs used in this analysis was evaluated by calculating R- squared statistics using the “add_rsq()” function in the TwoSampleMR software and the total R- squared statistics for all 39 IVs is 0.298%. F statistics were calculated for all 39 IVs (a sum of Z statistics for each SNP squared) as 1656.30. Although the F statistics is relatively large for the 39 IVs, given a low R 2 statistics for the 39 IVs used in the MR analysis, the set of IVs used for the MR analysis in this study is limited in power to assess if endometriosis is causal to certain immunological diseases. M u lti-tr ait analysis of G WA S (MTA G) The input files for MTAG are the GWAS meta-analysis summary results files which were pre- processed by filtering out: 1) SNPs with MAF ==20% among datasets; 2) restricting all analyses to a common set of SNPs present among dataset s; 3) multiple SNPs that were mapped to an identical chromosomal position among datasets; 4) SNPs with conflicting alleles among datasets. Z scores (log(OR/SE)) were computed for all SNPs. After variant filtering, a total of 3,873,419 common SNPs between endometriosis, osteoarthritis, rheumatoid arthritis, and multiple sclerosis were included in the MTAG analysis. MTAG is a generalization of the standard inverse variance weighted meta-analysis framework. Here endometriosis, osteoarthritis, rheumatoid arthritis and multiple sclerosis pre-processed GWAS summary statistics were included in a single MTAG analysis. Within MTAG, bi-variate LD score regression is employed to account for possibly unknown sample overlap between GWAS results of different traits. In the . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint results, MTAG outputs trait-specific effects estimated for each SNP, and the resulting p-value can be interpreted and used like those in single-trait GWAS 19 . Ident ification of shar ed ge netic v a r iants and th eir functio nal annotat ion and pat hwa y enrichme nts For each disease, genome-wide significant lead SNPs were identified based on (1) achieving a genome-wide significant P-value (P<5×10 −8 ), (2) being 500kb distant from each other and (3) being independent (r 2 <0.1). Then, the genome-wide significant lead SNPs associated with respective diseases that sit within 1Mb were identified and LD between them was checked. If the LD between lead SNPs of respective disease was r 2 >=0.5, they were considered shared loci between those diseases. The identified shared lead S NPs were looked-up in (1) Genotype-Tissue Expression (GTEx) portal to identify whether they are eQTLs for genes across 49 human tissues from 838 donors with 15,201 samples 69 , and (2) eQTLGen to identify blood eQTLs from 31,684 individuals 70 . Pathway enrichment analysis was conducted in FUMA based on MTAG results for endometriosis, osteoarthritis, rheumatoid arthritis, and multiple sclerosis where pathways were limited to canonical pathways 71 . D at a a va ilab il it y The GWAS meta-analyses for immunological conditions made use of data from the UK Biobank (Application Number 9637) and publicly available GWAS summary statistics for immunological conditions [expand on sources]. GWAS data for endometriosis was based on the latest analyses of International Endogene Consortium: summary statistics excluding 23andMe data is available from EBI GWAS Catalog Stud y Accession GCST90205183; endometriosis GWAS summary statistics from 23andMe, Inc. were made available under a data use agreement that protects participant privacy. Please contact dataset-request@23andme. com or visit research.23andMe.com/collaborate for more information and to apply to access the data. Aut hor Co ntribution s Car r ied out d ata analys is: N.S., N.R. D ra ft e d t h e m a nu s c r i p t: N.R., K.T.Z. Super v is ed th e p hen otyp ic dat a a n al y sis: K.T.Z., C.B., N.R., S.A.M., H.R.H., J.K., H.F. Super v is ed th e gen etic d ata analy s is: K.T.Z., C.B., N.R., A.P.M., C.C., J.K., H.F. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint All authors contributed and discussed the results and commented on the final version of the manuscript. Acknowledg eme nts We thank all the UK Biobank and 23andMe participants. Part of this research was conducted using the UK Biobank Resource under Application Number 9637. N.R. was supported by a grant from Wellbeing of Women UK (RG2031) and the EU Horizon 2020 funded project FEMaLe (101017562). A.P.M. was supported in part by Versus Arthritis (grant 21754). H.F. was supported by National Natural Science Foundation of China (grant 32170663). N.R., S.A.M and K.T.Z. were supported in part by a grant from CDMRP DoD PRMRP (W81XWH-20-PRMRP-IIRA). S.A.M. and K.T.Z. gratefully acknowledge funding provided by the Nezhat Family Foundation on behalf of Worldwide EndoMarch to their research programmes. Comp eting Inte rest s K.T.Z. and C.M.B. reported grants in three years prior, outside the submitted work, from Bayer AG, AbbVie Inc, Volition Rx, MDNA Life Sciences, PrecisionLife Ltd, Roche Diagnostics Inc. S.A.M. reports grants in the three years prior, outside this submitted work, from AbbVie Inc. N.R. is a consultant for Endogene.bio. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint Figure 1. Overlap o f g en es a s soc iat ed w it h GW AS le ad S NPs in e Q TL ana lys e s fo r e ndometri o s i s (42 genom e- wid e sig ni fican t l ead S N Ps ar e e QTL s for 43 9 gen e s), o s teo arth r i t i s (27 g enome- w ide signi fic ant lead SN Ps a r e e Q T Ls for 379 g ene s), mult iple scl er o s i s (64 g enom e- wi d e signi fic an t lea d SN Ps a re e QTLs for 1,113 genes ) and rhe uma toid ar thri ti s (2 8 genome - w ide signi f i cant l e ad SN Ps ar e e QTL s f or 490 gene s) in various ti s s ue s in G T Ex. Figure 2. Overlap o f pathways enriched with eQTL genes that are regulated by GWAS lead SNPs associated with endometriosis, osteoarthritis, rheumatoid arthritis and multiple sclerosis. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint Table 1. Immunological diseas e ri sks among female s with vs. wi thout endomet ri osi s in U K Biobank utili sing two dif ferent study designs : C ross - sectional st udy design (N=26 immunological dis ease with >100 femal e cas es) and cohort s tudy design (N=9 immunol ogical di s ease with >1500 female c ases) Imm u nol ogic al D is ease C ross-s ection a l stud y des ign a C oho rt stud y d esign b T o tal f ema le s End o cas e s (8,2 2 3 ) Fe m al e c o n t r ol s ( 2 65 , 1 8 1) OR ( 95 % C I) P-va l ue Tot a l Endo ca se s Female contro ls Follow- up tim e, Ye a r s H R ( 95%C I ) P -v alue P er son ye a rs Overa l l 646 2 0 ( 23.6 4 % ) 2 0 64 (2 5.1 0 %) 62 556 ( 23. 59%) 1.3 2 (1 .25-1. 39) <0.001 * 14 , 248 4 7 2 13 , 776 9 . 43 1 . 31 (1. 1 9- 1 . 4 4 ) < 0 .0 0 1* 5 , 75 0 . 9 0 Cl a ssic autoimmune di s eas e 147 6 4 ( 5 .40%) 505 ( 6. 14%) 142 5 9 (5 .38%) 1.2 4 (1 .13-1. 36) <0.001 * 2 ,6 27 9 8 2 ,5 29 9 . 78 1 . 42 (1. 1 6- 1 . 7 5 ) < 0 .0 0 1* 6 ,98 2 . 9 9 S yst e mic l u p u s e r ythem atosus 7 8 0 (0.2 9 %) 36 ( 0 . 4 4%) 7 4 4 (0 . 2 8 % ) 1.62 ( 1.14-2. 2 4) 0 . 005* -- -- -- -- -- - - -- Sjo g ren’ s sy n drome 8 0 3 (0.2 9 %) 23 ( 0 . 2 8%) 7 8 0 (0 . 2 9 % ) 1.07 ( 0.68-1. 5 9) 0. 7 6 -- -- -- -- -- - - -- Sys t emic sc l e ro s i s 2 5 5 (0.0 9 %) 10 ( 0 . 1 2%) 2 4 5 (0 . 0 9 % ) 1.60 ( 0.79-2. 8 6) 0. 1 5 -- -- -- -- -- - - -- M u ltip le sclerosis 1 5 71 ( 0 . 57%) 61 ( 0 . 7 4%) 151 0 ( 0 . 57%) 1.23 ( 0.93-1. 5 8) 0. 1 3 1 4 9 9 1 4 0 9 . 8 3 1 . 84 ( 0 . 9 3- 3 . 61 ) 0 . 08 7 , 309.1 6 R h e u m ato id art hri t is 5 8 18 ( 2 . 13%) 187 (2 .27%) 563 1 ( 2 . 12%) 1.22 ( 1.04-1. 4 1) 0 . 011* 1 , 2 6 2 4 8 1 , 214 9 . 8 1 1 . 57 ( 1 . 1 8- 2 . 10 ) 0. 0 0 2 * 7 , 207.0 5 Coeli a c d ise as e 2 0 23 ( 0 . 74%) 76 ( 0 . 9 2%) 194 7 ( 0 . 73%) 1.35 ( 1.06-1. 7 0) 0 . 011* 3 9 1 2 2 3 6 9 9 . 8 2 1 . 99 ( 1 . 3 0- 3 . 07 ) 0. 0 0 2 * 7 , 299.3 7 V i ti li g o 1 74 ( 0 . 06 % ) 6 (0 .07 % ) 1 68 ( 0. 06 %) 1 . 2 9 ( 0. 5 1 - 2 . 6 8 ) 0 .5 4 - - - - - - - - - - - - - - Primary bi l ia ry cirrhosis 2 6 3 (0.1 0 %) 8 ( 0 . 1 0 %) 2 5 5 (0 . 1 0 % ) 1.17 ( 0 .53-2.21) 0. 6 7 -- -- -- -- -- - - -- Addi s on' s disea s e 1 8 0 (0.0 7 %) 6 ( 0 . 0 7 %) 1 7 4 (0 . 0 7 % ) 1.21 ( 0.47-2. 5 0) 0. 6 5 -- -- -- -- -- - - -- Ty p e 1 d ia be tes 1 6 03 ( 0 . 59%) 57 ( 0 . 6 9%) 154 6 ( 0 . 58%) 1.26 ( 0.95-1. 6 4) 0. 0 9 5 11 17 4 9 4 9. 8 2 1. 3 8 ( 0. 8 5-2. 2 5) 0 . 1 9 7 0 74.59 M ya stheni a grav i s 1 7 3 (0.0 6 %) 11 ( 0 . 1 3%) 1 6 2 (0 . 0 6 % ) 2.55 ( 1.30-4. 4 9) 0 . 003* -- -- -- -- -- - - -- Aut o imm une ga stritis 1 4 66 ( 0 . 54%) 48 ( 0 . 5 8%) 141 8 ( 0 . 53%) 1. 2 0 ( 0.88-1. 5 8) 0. 2 2 Derma tomyositi s , P olym yo s i t is 1 1 5 (0.0 4 %) 3 ( 0 . 0 4 %) 1 1 2 (0 . 0 4 % ) 0.97 ( 0 .24-2.59) 0. 9 6 -- -- -- -- -- - - -- Gra ve s' d ise as e 5 4 5 (0.2 0 %) 23 ( 0 . 2 8%) 5 2 2 (0 . 2 0 % ) 1.28 ( 0.80-1. 9 3) 0. 2 7 -- -- -- -- -- - - -- Has h imo to' s d i se as e 2 4 4 (0.0 9 %) 6 ( 0 . 0 7 %) 2 3 8 (0 . 0 9 % ) 0.89 ( 0.35-1. 8 4) 0. 7 9 -- -- -- -- -- - - -- Au t oi m m u n e thyroid di se ase 7 8 8 (0.2 9 %) 29 ( 0 . 3 5%) 7 5 9 (0 . 2 9 % ) 1.17 ( 0 .79-1.69) 0. 4 2 -- -- -- -- -- - - -- Au toin fla mmatory di s eas e 520 2 7 ( 19.0 3 % ) 1 6 38 (1 9.9 2 %) 50 389 ( 19. 00%) 1.3 3 (1 .26-1. 41) <0.001 * 13 , 710 4 4 3 13 , 267 9 . 46 1 . 28 (1. 1 6- 1 . 4 1 ) < 0 .0 0 1* 6 ,08 0 . 8 1 Cro h n' s di se ase 1 3 82 ( 0 . 51%) 48 ( 0 . 5 8%) 133 4 ( 0 . 50%) 1.18 ( 0.87-1. 5 8) 0. 2 7 -- -- -- -- -- - - -- Ulce rativ e co l itis 2 3 65 ( 0 . 87%) 82 ( 1 . 0 0%) 228 3 ( 0 . 86%) 1.24 ( 0.98-1. 5 5) 0. 0 6 4 3 0 13 4 1 7 9 . 8 2 1. 0 4 ( 0. 5 8-1. 8 4) 0. 9 0 7,0 52. 21 Inf la mma t o ry b owel di se ase 3 5 27 ( 1 . 29%) 129 (1 .57%) 339 8 ( 1 . 2 8 %) 1.28 ( 1.06-1. 5 3) 0 . 009* 3 1 6 1 2 3 0 4 9 . 8 2 1 . 40 ( 0 . 7 9- 2 . 50 ) 0 . 25 4 , 247.0 5 Oste o a rth ritis 476 4 2 ( 1 7 . 4 3 % ) 1 4 93 (1 8 . 16 %) 4 6 149 ( 1 7 .40%) 1.35 ( 1.27-1. 4 3) <0.0 0 1 * 13, 3 7 1 4 1 6 12, 9 5 5 9 . 4 8 1 . 31 ( 1 . 1 9- 1 . 44 ) <0. 001* 6 , 196.1 3 Gia nt c ell, Ta ka yas u ce l l 3 6 6 (0.1 3 %) 10 ( 0 . 1 2%) 3 5 6 (0 . 1 3 % ) 1.35 ( 0.67-2. 4 0) 0. 3 5 -- -- -- -- -- - - -- . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint arte ri tis Sa rco id o sis 7 4 1 (0.2 7 %) 26 ( 0 . 3 2%) 7 1 5 (0 . 2 7 % ) 1.39 ( 0.91-2. 0 1) 0. 1 0 -- -- -- -- -- - - -- Ac ne , a c ne fo r m as s o c i a t ed d i s ea s e 2 8 7 (0.1 1 %) 8 ( 0 . 1 0 %) 2 7 9 (0 . 1 1 % ) 0.78 ( 0.35-1. 4 8) 0. 5 0 -- -- -- -- -- - - -- G o ut , ps e ud o g o ut , crystal arthr o pa thy 1 1 17 ( 0 . 41%) 41 ( 0 . 5 0%) 107 6 ( 0 . 41%) 1.66 ( 1.18-2. 2 6) 0 . 002* -- -- -- -- -- - - -- M ix ed -Pattern d is e a se 4 3 79 (1. 60%) 156 ( 1. 90%) 422 3 (1. 59 %) 1.2 3 (1 .10-1. 52) 0 . 0 0 2* 5 2 7 2 7 5 0 0 9 . 82 1 . 81 (1. 2 1- 2 . 7 1 ) 0. 004 * 723 7 .37 3 An k ylo s i ng s pon d y l it is 6 7 6 (0.2 5 %) 28 ( 0 . 3 4%) 6 4 8 (0 . 2 4 % ) 1.59 ( 1.06-2. 3 0) 0 . 019* 9 6 4 9 2 9 . 8 3 1 . 66 ( 0 . 6 1- 4 . 52 ) 0 . 32 7 , 332.2 2 Psori a sis 3 6 97 ( 1 . 36%) 121 (1 .47%) 357 6 ( 1 . 35%) 1.18 ( 0.97-1. 4 1) 0. 0 8 4 2 7 1 9 4 0 8 9 . 8 2 1 . 67 ( 1 . 0 5- 2 . 65 ) 0. 0 3 0 * 7 , 254.4 0 C o nfou n ders inc luded i n t he ana ly se s are age at rec ru itme n t a n d g enetica ll y dete rmin e d a n c estry. -- Insuf fic ient num ber of c ase s to ge n e rat e m ea n ingfu l risk estima tes. a C ross-s e ctio n al d e sign i n c lu ded 8, 223 en d ome tri o sis ca ses a n d fe ma l e co ntr ol s. b C o hort d e sign: F or ea ch imm un o l ogica l co ndi tio n , fem ales who ha d the re sp e ctiv e imm un ologica l di se ase d iag n ose d befo r e end ome trio sis o r th o se who h a d t he resp ec ti v e immun olo g ica l disea s e d ia g no se d before cohort entry t i m e or ha d i m m u nologica l disea s e di a g no sis time m i ss i ng we re e x clu ded f rom c oh o rt a na lysis. Her e a re the num ber of exclud e d i ndivi duals fo r ea ch d isea s e and d ise as e cate gory: Ov era ll imm unol ogical dise as e N= 5 0,67 6 7 , Cl a ssica l a ut oi m m u n e d isea s e N=12, 5 1 1 , m ult iple sc l e ro sis N=1,8 1 5, r h e u m ato i d art hrit i s N=4,94 6 , coel iac di se ase N =2, 0 2 6 , ty p e 1 d ia b e tes N=1,4 8 6 , Aut oinfl a mm a tory d ise as e N= 3 8, 661, ulce rative c o lit i s N= 2 ,32 6 , I n fla m mat ory b owe l d isea s e N=11 5, 4 0 4, o ste oart hri t is N=34,5 68, mixe d -patt ern i mm un ologica l di se as e N=42 4 0 , An k ylo sing s p o n d y l it is N = 9 7 6 , p s o rias i s N= 3 , 662. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint Table 2. N u mber of case s and contr ol s of UKBB based female only and s ex- combined GWAS for immunological c o nditions a n d s umm ary of their me ta-analysi s with the lar g est pub l icl y available Europ e an ancestr y GW A S r e s ult s . Im m un o l o g ic al di s e as es UK B B GWAS (Cases : C ontr o ls) P u blis hed s e x -co m b in e d GWA S (C a s e s : Co ntrols ) Fina l sex-c om bine d G W A S m et a -an a l y s i s (C ase s : Controls ) Femal e -o nly S ex - c ombi ned Anky l osin g spon dy l i tis 5 47 : 162,4 03 1,4 93 : 3 1 9, 532 N /A 1 ,49 3 : 319,5 32 Co e lia c di seas e 1 ,70 6 : 162,4 03 2,6 40 : 3 1 9, 532 4 ,53 3 : 10, 7 50 72 7 ,17 3 : 330,2 82 Infl a m mato ry bo w e l d is eas e 2 ,86 9 : 162,4 03 5,7 51 : 3 1 9, 532 25 ,04 2 : 3 4 ,91 5 54 3 0,7 93 : 3 5 4,4 47 M ultiple s c l e r o si s 1 ,31 4 : 162,4 03 1,8 83 : 3 1 9, 532 14 ,49 8 : 2 4 ,09 1 56 1 6,3 81 : 3 4 3,6 23 Os teoa rt hritis 39 ,86 6 : 1 6 2,4 03 68, 878 : 319 ,53 2 77 ,05 2 : 3 78 ,16 9 73 77 ,05 2 : 3 7 8,1 69 * P sorias i s 3 ,03 6 : 162,4 03 6,5 91 : 3 1 9, 532 15 ,96 7 : 2 8 ,19 4 59 2 2,5 58 : 3 4 7,7 26 Rheuma toi d a r th ritis 4 ,66 2 : 162,4 03 7,1 53 : 3 1 9, 532 14 ,36 1 : 4 3 ,92 3 57 2 1,5 14 : 3 6 3,4 55 S y s t e m ic L up u s eryt he m a tos u s 5 45 : 162,4 03 673 : 3 1 9 ,53 2 5, 874 : 3 2 8 ,59 8 74 6 ,54 7 : 648,1 30 * Pub l i she d G W AS m et a - a na l y s is in c lu ded UK BB h en c e , s u mm a ry st at i st i c s f ro m T a c hm z ido u l e t a l . 73 w e re use d i n th e an a l ys e s . Table 3. G en e t ic cor relations from LDSC a n al y si s bet w een en dometr i o s is and immunol ogical dis ease. Mult iple-testing correction f or number o f diseas e included in the analy s is i s applied (0.05/8=6.2 5x10 -3 ) to det e rmine signific ant c or relations . UKBB end o met r io si s G W AS re sult s we re ex clude d from e ndome trio si s met a- anal y s i s to a void ov e rlap i n LDS C a naly si s wi t h i mmunolog ical c ondition s f or whic h we have analy s e d t his da ta s e t . H 2: He r it abi lity, S E: S tand a r d err or , RG: Gen etic co r r e lat i o n. UK BB fem ale -o n ly GW AS U KBB se x - com b i ned G WAS Se x - com bi n e d GW AS me t a - an a l ys is I m muno l o g ic al dise a s e s H 2 ( S E ) RG ( S E) P -v al u e H2 ( SE) RG ( S E) P-v al u e H2 ( SE) RG ( SE ) P-v a lu e A nky l o si ng s pon d y l itis 0 .006 (0 .003) 0 .308 (0 .124) 0.0 1 3 0.0 0 3 (0.001) 0 .300 (0 .124) 0. 01 6 N/ A N /A N/A C o el i a c d is e a s e 0 .013 (0 .003) 0 .215 (0 .083) 0.0 1 0 0.0 0 9 (0.002) 0 .154 (0 .068) 0. 02 4 0 . 07 7 (0.0 0 5) 0.0 7 6 ( 0.054) 0. 16 1 9 Inflam m a tor y b owe l di s e as e 0 .014 (0 .003) 0 .054 (0 .085) 0.5 2 1 0.0 1 3 (0.002) 0 .007 (0 .063) 0. 91 0 0 . 27 7 (0.0 2 8) -0. 04 1 ( 0.035) 0. 23 7 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint M u l t ipl e s cle r o s is 0 .004 (0 .003) 0 .247 (0 .139) 0.0 7 5 0.0 0 3 (0.001) 0 .273 (0 .116) 0. 01 8 0 . 07 9 (0.0 0 5) 0.0 8 8 ( 0.031) 4. 0 0 x 10 -3 O ste oa rt hrit is 0 .051 (0 .004) 0 .322 (0 .042) 1.7 6 x10 -14 0.0 4 2 (0.002) 0 .278 (0 .038) 2.4 6 x10 -13 0 . 04 6 (0.0 0 2) 0.2 7 8 ( 0.035) 3.2 5 x10 -15 Psori a s i s 0 .012 (0 .003) 0 .043 (0 .084) 0.6 0 5 0.0 1 2 (0.002) 0 .049 (0 .061) 0. 42 0 0 . 23 9 (0.0 2 7) 0.0 6 6 ( 0.037) 0. 07 6 Rh eu m a to id a r th r i t is 0 .013 (0 .003) 0 .277 (0 .079) 0.0 0 1 0.0 1 1 (0.002) 0 .284 (0 .068) 2. 6 5 x 10 -5 0 . 06 4 (0.0 0 9) 0.2 6 6 ( 0.062) 1. 5 4 x 10 -5 S ys t em i c l upu s e r y th em ato su s 0 .003 (0 .003) 0 .154 (0 .154) 0.3 1 6 0.0 0 2 (0.001) 0 .190 (0 .165) 0. 24 8 0 . 33 2 (0.0 4 8) 0.1 2 7 ( 0.071) 0. 07 4 Table 4. Mend e lian Rando mis at i o n ( MR) r es u l t s fo r endome t riosis vs. o s t eoar thr i tis , r heumatoid arth rit is and multipl e sc ler os i s. I mm u nol o gic al dis e ase s SN P s I n v e rs e va r i a n c e w e igh te d We i g ht ed m ed i an M R Egg e r MR E gg e r ( P -va l u es ) O R ( 95% C I) P - va lu e O R ( 95% C I) P - v a lu e O R ( 95 % C I ) P - va l u e H e t a P l e i o t r o p y b Os te oa rt h r i t i s (3 9 IVs ) F ema l e-o nly U KB B 39 1.0 4 ( 0 . 9 7-1. 11 ) 0 . 282 1 . 05 ( 0 . 9 7-1. 13) 0.2 6 7 0.8 9 ( 0. 72 -1. 07 ) 0.2 03 0.0 02 0 . 084 Fe m ale - o n l y UK BB ( ou t lie rs re m o ve d ) 35 1.0 3 ( 0 . 9 8-1. 09 ) 0 . 287 1 . 05 ( 0 . 9 7-1. 13) 0.2 5 0 0.9 5 ( 0. 79 -1. 15 ) 0.5 95 0.5 72 0 . 387 S ex-c o mb in ed meta-a n al y s i s 39 1.0 2 ( 0 . 9 6-1. 07 ) 0 . 533 1 . 02 ( 0 . 9 7-1. 08) 0.4 0 0 1.0 8 ( 0. 92 -1. 26 ) 0.3 53 1.5 6x 1 0 -5 0 . 229 S ex-c o mb in ed meta-a n al y s i s ( o utl i e r s r e mo ve d ) 36 1.0 1 ( 0 . 9 7-1. 06 ) 0 . 495 1 . 03 ( 0 . 9 8-1. 09) 0.2 8 1 1.0 3 ( 0. 90 -1. 18 ) 0.6 55 0.1 59 0 . 800 Rh e uma t o id Ar th r it is ( 3 9 IVs ) F ema l e-o nly U KB B 39 1.1 6 ( 1 . 0 2-1. 33 ) 0 . 028 1 . 15 ( 0 . 9 4-1. 40) 0.1 6 9 1.1 5 ( 0. 78 -1. 70 ) 0.4 94 0.7 56 0 . 959 S ex-c o mb in ed meta-a n al y s i s 31 1.0 6 ( 0 . 9 6-1. 17 ) 0 . 220 1 . 09 ( 0 . 9 4-1. 26) 0.2 6 1 1.0 8 ( 0. 76 -1. 53 ) 0.6 82 0.6 11 0 . 435 M ulti ple S cler osi s ( 39 I V s) F ema l e-o nly U KB B 39 1.1 2 ( 0 . 8 7-1. 43 ) 0 . 376 1 . 23 ( 0 . 8 7-1. 73) 0.2 4 3 1.7 5 ( 0. 84 -3. 64 ) 0.1 43 0.8 15 0 . 211 S ex-c o mb in ed meta-a n al y s i s 39 1.0 6 ( 0 . 9 4-1. 21 ) 0 . 343 1 . 03 ( 0 . 8 9-1. 20) 0.6 9 2 1.2 4 ( 0. 84 -1. 84 ) 0.2 80 0.0 12* 0 . 145 S ex-c o mb in ed Met a-an al ysi s ( o utl i e r s r e mo ve d ) 38 1.0 4 ( 0 . 9 3-1. 18 ) 0 . 484 1 . 02 ( 0 . 8 8-1. 2) 0.7 5 9 1.2 0 ( 0. 84 -1. 73 ) 0.3 25 0.0 63 0 . 202 a Tes t for H ete rogeneit y , b Tes t f or Pleiotr opy, c O u tlier IVs were ident ified by M R PRE SS O so f t w a r e 67 (See m etho ds and Supplemen tary Table 15). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint Table 5. G en ome-wide s ignificant lead SNP s a ssociated with endom etr iosis ( EN D O ) and rheum atoid a rt hr itis (RA) , o s teo a rthr i tis ( O A ) or multiple sclerosis (M S) tha t ar e located wit h in 1Mb, with LD between them. Chr : Chro mos o me, AE: Effective alle le, Fr q: Effective al lel e freq uenc y, LD: Linka ge dis equilibr i u m, eQ TL: Expre s s ion quant itative tr ai t l oc i. Locus Tra its S N P Chr P ositi on E A ( Frq ) OR ( 95 % CI) P-v al u e LD (r 2) eQTLs r e s ul t s* GR E B 1 / 2 p25.1 E ND O r s 12 0 305 76 1 115 8 172 21 G ( 0 . 65 ) 1 . 02 ( 1 .01 - 1 .03 ) 1 .94 x10 -1 4 0 .0 02 TS P A N 2 RA r s 66 7 967 7 1 114 3 038 08 C (0 .1 0) 0 . 85 ( 0 .84 - 0 .87 ) 1 .61 x10 -92 A P4B 1, A PBA2, CD 247, C D 5, C D 6 , CT LA4, D C LR E 1B, FB L N2, F O XP3, GZ MB, I L 10RA, I L2 R A , M A F , ME D1 5 , P H TF1, R N F2 1 4, RT K N2, SLA MF 1, S T8S IA1, STA P1, WLS, ZN F 8 3 1 DNM 3/ 1 q24.3 E ND O r s 24 2 198 5 1 172 0 991 36 T ( 0 .52 ) 0 . 98 ( 0 .97 - 0 .98 ) 1 .94 x10 -1 4 0 .0 001 M E TT L 13 OA r s 55 6 981 00 1 174 2 274 33 T ( 0 .73 ) 1 . 01 ( 1 .01 - 1 .02 ) 7 .97 x10 -1 0 C A CY B P , DA R S 2, G PR 52 , K IAA 0 0 40 , KLHL 20, M R P S 1 4 , P R D X6, R A B G A P1L, RC 3H1, S E RPI N C1 ET A A1/ 2p 1 4 E ND O r s 11 1 261 43 2 678 6 540 8 C (0 .3 2) 1 . 02 ( 1 .01 - 1 .02 ) 6 .08 x10 -1 0 0 .0 1 None M S r s 12 6 226 70 2 686 4 653 6 C (0 .4 6) 0 . 96 ( 0 .98 - 0 .99 ) 1 .56 x10 -9 C NR IP1 , P LEK , PP P3 R 1 BM P R 2/ 2 q33.1 E ND O r s 72 9 289 25 2 203 5 543 24 A (0 .82 ) 1 . 02 ( 1 .02 - 1 .03 ) 9 .61 x10 -1 0 BM P R2, CA R F, R A M11 7B, I CA 1 L, NB E AL 1 OA r s 72 9 286 05 2 203 8 321 20 A (0 .87 ) 1 . 02 ( 1 .01 - 1 .02 ) 4 .33 x10 -8 0 .5 3 BM P R2, CA R F, C Y P 2 0A 1, FA M117 B, ICA 1L, N BEAL 1, NO P58 RA r s 30 8 724 3 2 204 7 389 19 G ( 0 . 45 ) 1 . 03 ( 1 .02 - 1 .04 ) 2 .39 x10 -9 0 .0 01 CT LA4, I CO S BS N / 3 p21.31 E ND O r s 67 7 420 2 3 496 8 777 9 T ( 0 .17 ) 1 . 03 ( 1 .01 - 1 .03 ) 9 .70 x10 -12 0 .9 7 A MT, A RI H2, A P EH, C 3orf62 , C A C NA 2 D2 , CCDC3 6 , C CDC7 1 , CD HR4, DA G 1, D A L RD3 , FA M 2 12 A , GM PP B , G P X 1 , HE M K1, HYA L3, I M PD H 2 , I P 6 K 1, K LH D C 88 , M A P K A PK3 , M ON 1 A, M ST 1R , NAT 6 , N CKI PS D, N D U FAF3, N I CN 1 , P4 H TM, PR K A R2A , Q RI CH 1, RB M6, R H OA, R N F1 23 , TC T A , TR AIP , U B A7 , WDR6 OA r s 68 0 987 9 3 499 3 610 2 G ( 0 . 18 ) 1 . 02 ( 1 .01 - 1 .02 ) 7 .71 x10 -9 A MT, A PEH, A RIH 2 , BSN , CA CNA 2D 2, C A M KV , CD HR 4 , CTD - 2 33 0K9 .3 , DO C K 3, FAM 212A, G M P PB, G PX 1, HEMK 1, H Y A L3 , I P6 K1, KLH D C8B , M A PKAP K 3 , MO N1A , MS T1, MS T 1 R , NAT 6, N I CN 1, P4H T M , RASSF1, RB M5, RB M 6 , RHOA , RN F1 2 3, S E MA3F , T RAI P, U BA7, WD R 6 EBF 1/ 5 q33.3 E ND O r s 29 6 448 5 5 157 9 048 39 G ( 0 . 22 ) 0 . 98 ( 0 .97 - 0 .99 ) 1 .11 x10 -8 0 .0 01 None M S r s 25 4 689 0 5 158 7 599 00 A (0 .52 ) 1 . 05 ( 1 .04 - 1 .07 ) 7 .71 x10 -13 UB LCP1 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint PP P 1 R9A / 7 q21.3 E ND O r s 34 7 510 86 7 945 9 269 1 C (0 .8 7) 1 . 02 ( 1 .02 - 1 .03 ) 3 .23 x10 -8 0 .0 3 A SB 4 , C A SD 1, PEG 10 , P ON 1, P O N 2 , PP P1R 9 A OA r s 69 6 654 0 7 957 2 796 7 T ( 0 .63 ) 0 . 99 ( 0 .98 - 0 .99 ) 1 .46 x10 -9 DY N C1I 1 , S L C25A1 3 XK R6 / 8 p23.1 E ND O r s 12 5 420 37 8 107 5 849 6 A (0 .45 ) 1 . 02 ( 1 .01 - 1 .02 ) 7 .02 x10 -11 BL K, C TSB, FAM 167A, F DF T 1, MSRA, MT M R 9, N EI L2, R P11- 297N 6, R P 1 L 1, S LC3 5 G5 , X KR6 OA r s 42 4 067 3 8 107 8 761 2 T ( 0 .45 ) 1 . 01 ( 1 .01 - 1 .02 ) 2 .24 x10 -12 0 .9 BL K, C TS B , FAM 167A, F DF T1 , MSRA, MT M R 9, N EI L2, R P11- 297N 6. 4, RP 1L1, S LC3 5 G 5 , SO X7, X KR6 RA r s 27 3 634 0 8 113 4 397 3 C (0 .2 6) 0 . 97 ( 0 .96 - 0 .98 ) 2 .78 x10 -9 0 .3 A RG2, BL K, CTS B, F A M16 7 A , F D FT1, GGA 2 , M T MR 9, N E I L2, RP1 1 -2 97N 6 , SL C 3 5G 5, X K R 6 MLLT 10/ 10 p 12 .31 E ND O r s 11 0 127 32 10 218 3 010 4 A (0 .66 ) 0 . 98 ( 0 .97 - 0 .98 ) 3 .84 x10 -1 6 0 .7 C A SC 10 , M L LT10 , N EB L OA r s 12 3 573 21 10 217 9 047 6 G ( 0 . 68 ) 0 . 99 ( 0 .98 - 0 .99 ) 3 .28 x10 -9 C A SC 10 , M L LY10 , N EB L VEZ T/ 12 q 22 E ND O r s 12 3 201 96 12 956 4 538 5 A (0 .52 ) 0 . 98 ( 0 .98 - 0 .99 ) 5 .37 x10 -12 0 .0 03 F G D6, N D UFA 1 2, N R2 C1 , V EZ T OA r s 21 7 112 6 12 941 6 722 0 C (0 .4 9) 0 . 99 ( 0 .98 - 0 .99 ) 2 .88 x10 -9 SO CS 2 DLE U 1/ 13 q 14 .2 E ND O r s 10 2 886 2 13 510 5 513 4 G ( 0 . 86 ) 0 . 98 ( 0 .97 - 0 .98 ) 9 .63 x10 -1 0 0 .0 005 None M S r s 95 9 132 5 13 508 1 122 0 T ( 0 .93 ) 1 . 09 ( 1 .06 - 1 .12 ) 1 .59 x10 -1 0 C OR O1C , DLE U1 , EB PL , KPN A 3 , PH F 1 1 SKA P 1 / 17 q 21 .32 E ND O r s 67 4 467 70 17 462 3 574 5 A (0 .41 ) 0 . 98 ( 0 .98 - 0 .99 ) 8 .37 x10 -1 0 0 .0 02 CB X1, CD K 5 RA P 3 , HO XB2 , H O XB 3, HOX B4, L RR C4 6, M R PL1 0 , NF E2 L 1, S C R N 2, S KA P1 , S N X 1 1 M S r s 11 0 797 84 17 457 0 228 0 T ( 0 .49 ) 0 . 96 ( 0 .94 - 0 .97 ) 1 .97 x10 -1 0 C OPZ2 , EFC A B 13 , HOXB 2 , IT GB 3 , N PEP PS , S CR N2, SKA P 1, T B KB P 1 * F ull e Q T L re s u l ts fr o m 54 G T E x t i s s u e s a nd eQ TLGe n bl o o d t iss ue a re pr ovi d e d i n Su p pl e ment ary T able 14. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint

1. Zondervan, K.T ., Bec ker, C .M . & M is s mer, S.A. Endom etriosis . N Engl J Med 382, 1244- 12 56 (2020). 2. Sampson, J.A. M etas t atic o r Embolic Endometr ios is, due to t he Menstrual D issemination of Endomet rial Ti ssue into the V enous C i r culation. Am J Pa t hol 3 , 93-110 43 (1927) . 3. Halme, J., H am mond , M .G ., H ulka, J. F ., Raj, S. G. & Talber t, L.M. Retr ograde menstr uation in healthy women and in patien t s with endo metri o s is. O b stet G ynec ol 64, 15 1-4 (1 984). 4. Cramer , D.W. & M i ssmer, S.A. The ep idem i o l ogy of en dometr i os is. Ann N Y A c ad Sci 955, 11-22; dis cu ssion 34-6, 396-40 6 (20 0 2). 5. Bulun, S.E. et al. Role of estr ogen receptor -be ta in end ometriosis . Semin Reprod M ed 30, 39-45 (2012) . 6. Han, S.J. et a l . Estr ogen Receptor bet a Modulates Apopt osi s Complexes an d the Inflammasome to D r ive the Pat hogenes is of Endomet riosis . Ce l l 163, 96 0-74 (2015) . 7. Symon s, L. K. et al. The Immu nopatho ph ysiology of Endomet riosi s. Trend s Mol Med 24, 748-762 (20 18). 8. Zondervan, K.T . et al. Endom etriosis . Nat Rev Dis Primer s 4 , 9 (201 8) . 9. Mat a r es e, G ., De Pla cido, G ., Nikas , Y. & A lviggi, C. Pathogenesi s o f endometr ios is: natur al immunit y dysfunction or autoimmu n e dis ease? Tr ends M ol Med 9 , 2 23-8 (2003) . 10. G leicher, N., el-Roeiy, A., Conf ino, E. & Friber g, J. I s endo metriosis an aut oimmune diseas e? O b s tet G ynecol 70 , 1 15- 22 (1987). 11. Eis enb erg, V.H., Zolt i, M . & Sor iano, D. Is there an as s o ciation bet w een autoimmuni t y and endomet r iosis ? Aut oimmun Rev 11, 806-14 (2012) . 12. La u danski, P. e t a l . Aut oan tibody sc r e ening of plas m a and per itonea l flu id of p a t ien ts wi t h endomet r iosis . H um Reprod 38, 629- 6 43 ( 2023). 13. htt ps ://www.auto i m muneinstitut e.or g/dis ea se s _list/ en dometr ios is/ . 14. Shiges i , N . e t al. The as so c iat i o n betw een endome t riosis and auto i m mune disea se s: a sy stemat i c review and meta - anal ysi s . H um Reprod Up date 25, 486- 503 (20 19). 15. Mc Gonagl e, D . & M c D ermo tt, M .F . A proposed clas sific at ion of the immu n ologic al dis ease s. PL o S M e d 3 , e297 (20 06) . 16. Rahmioglu, N. et al. The genetic b asis of endome tr ios is and comorbidit y wit h other pa in and inflamma t ory conditions. Na t Ge n e t 55, 423-4 36 ( 202 3). 17. Bulik-Sullivan, B . e t a l . An a tlas o f genetic cor relations ac r os s human disea ses and tr ai t s . Na t Ge n e t 47 , 123 6- 41 (2015). 18. Bulik-Sullivan, B .K. et al. L D S cor e regress ion d istinguishe s confo unding fro m poly genicity in genome-wide ass o ciation studies. Na t G e n e t 47, 29 1-5 ( 2015). 19. Turley, P . et al. Mult i- tr a i t anal ysi s of g en ome-wide as so cia t ion s u mmary statistics u sin g MTA G . Nat Genet 50, 229-23 7 (2018) . 20. Manco, G. , P orzio, E . & Caru s one, T. M . Human Par a o x on a se-2 (P ON 2): Pr ot ein Func t ions and Modu l at ion. Anti oxid ant s (B as el) 10(2021 ). 21. Shin, S.H. et al. Ar rest def ective 1 regulates the oxidativ e s t ress r e s pon se in human cells and mice by a cetylating met hionine s ulf ox ide reductas e A . Cell D eath Dis 5 , e1490 (201 4). 22. Ichik a wa-Tom ik aw a, N., Sugimot o, K., Kas h i wa gi, K . & Chiba , H. The Src-Fa mily Kinas e s S R C and BLK Contr ibute to t he CLDN6-Ad h e sion Signaling. Cel l s 12(20 23). 23. W ang , H . et al. ZAP-70: an es s ent ial kinase i n T- c ell si g n a ling. Cold Spring Harb Per s pe ct Biol 2 , a0022 79 (2010). 24. Li , C . e t a l . Tr ans m embrane Pr otein 214 (TM EM2 14) mediat es endoplasmic reticulum s t res s - i n duc ed ca spa s e 4 enz yme a ct ivation and apo pt osis . J B i ol Chem 288, 179 08-17 (2013). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint 25. Stelzer, G . et a l . The GeneCard s Suite: From Gene Dat a Mining to D i sease G enome Sequence Analy s es. Curr Protoc Bioinf or matics 54, 1 30 1-1 30 3 3 (20 16). 26. Dy son, M. T. et al. Genome-wide D N A m e t hylation analys is predict s an epigenet ic switch for GATA f ac t or express ion in end om etr i o s is. PL oS Genet 10, e100 4158 (20 14). 27. McKinnon, B. D., Kocbe k, V ., N ir gi anak is, K. , Bersinger, N.A. & M ueller, M.D . Kinas e s ignalling pathwa y s in end ometriosis: potent i a l t a r gets for non-hormo na l therapeut ic s. Hum R eprod U pdate 22, 382-403 (20 16) . 28. Milacic , M . et al. The Reac t ome Path way Knowledgeba s e 2024. Nucle ic Ac ids Res 52, D6 72- D678 ( 2024). 29. Shafrir , A.L . e t al . R is k for an d c onseq uences of endometr ios is: A critical epidemiol ogic review. Bes t Pract Res Clin Obstet Gy naec ol 51, 1-15 (2 018). 30. Zondervan, K.T ., Car don, L.R . & Kenn ed y , S.H . What m a ke s a g o od c a se-co ntr ol s tu dy? Design i ssues f or complex traits su c h as endometr ios is. Hum R eprod 17, 14 15-23 (2 002). 31. Har ris, H .R. et al. Endometriosis , P sor iasis, and P soriatic Ar thritis : A P rospe c t ive Co hort Study. Am J Epi d e m iol 191, 105 0-106 0 (2022). 32. Har ris, H .R. et al. Endometriosis and t he ris k s of s ystemic lupus erythem a t os u s and rheum atoid a rt hr itis in the Nurses ' H ealt h S t udy II. Ann Rheum D is 75, 127 9-84 ( 2016). 33. Kraft , P., Chen, H. & Linds t rom, S. The Us e Of G enet ic Co rrelation A nd Men delian Randomization Stud i es To Increase Our U nder s t anding of Relati onships B etween Complex Traits. Curr Epidemiol Rep 7 , 104-112 (2020). 34. Burges s, S . Sample si z e and power cal culations in M endelian randomi zatio n with a single ins t rumen tal variable and a bina r y o utcome. Int J Epidemio l 43, 922- 9 (201 4). 35. Shafrir , A.L . e t al . C o-occurrence of imm une-m ediated conditions and endo metri o s is among adolesc ent s and adult wom en. Am J Repr od Immuno l 86, e13404 (2 021) . 36. Ke, J., Ye, J. , Li, M. & Z h u, Z. The Role of Matr i x M e tallop rot e inas es in Endo metr ios is: A Potent ial Tar get. Biomolecu les 11(20 21) . 37. Li u , Y. et a l . Incr ea sed Serum M atrix Met allopr ot ei n as e-9 Leve ls are A s so ciat ed wi t h An t i - Jo1 but not Ant i -MD A5 in Myositis P a tients. Agi ng Dis 10, 746-755 (2 019) . 38. Ram, M., Sher er, Y. & Sho enfeld, Y. Matr ix met a ll o pr ot ei n a se-9 and aut o im mune diseas e s. J Clin Immunol 26, 299-3 07 ( 2006). 39. Du, F. et al. Inflamm atory Th17 Cells Ex pr ess Integr i n alphavbet a3 f or Patho ge n i c Fun ction. Cell Rep 16, 1339-1351 (2 016) . 40. Jang, Y. , Kim, M. & Hwan g, S.W. M olec u l ar mechanis m s under l ying th e a ctions of arachidonic a c id-d erived p rostaglandins on peripher al no c iception. J Neuroinf lamma tion 17, 30 (20 20). 41. Migliore, A . & Pr oc o pio, S . Ef fec t iven ess and utility of hyaluronic acid i n o steo a r thritis . Cl i n Cases Miner Bone Metab 12, 31- 3 ( 20 15). 42. Kobaya s h i , T., Chanm ee, T. & Itano, N . Hyaluro nan: Metabo l ism and Function. Biomolecules 10(2020 ). 43. Yu, P .H., Ch ou, P.Y. , L i, W . N ., T sai , S.J . & Wu, M .H . Th e p ro-infla mm atory a nd ant i - inflamma t ory role o f hyalur onic acid in endomet r iosi s. Taiwan J Obs t et G yn ecol 60 , 71 1- 717 (2021). 44. Simons, K. & Toom re, D. Lipid raft s and signal t ransdu ction. Nat Rev Mol Ce ll Bio l 1 , 31 -9 (2000). 45. Robinson, G., P i n e d a-Torr a , I. , Ciur tin, C. & Jur y, E. C. Lipid metabolism in au toimm une rheum atic disease : im pli cation s for modern and convent i o nal t herapi es. J Clin Inve st 132(2022) . 46. Vos kuhl, R. Sex diff erences in aut oim mune diseas e s. Biol S ex Diff er 2 , 1 (2011). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint 47. Boer, C .G . et al. D ec ipher ing os t eoarthr itis genet ic s a c r os s 826, 690 individuals f rom 9 populations. Cell 184, 6003-60 05 (20 21) . 48. Sudlow, C. et al. U K biob a n k : an op en ac cess re source for i d entifying the caus e s of a wide range of complex dise ase s o f middle and old age. PL oS M e d 12, e1001 7 79 ( 2015). 49. Rahmioglu, N. et al. G enom e-wi d e en richment analys i s between endom etr iosis and obesity-related t ra its reveals no v el suscept i b i lit y loci. Hum M ol G enet 24 , 1 18 5-99 (2015). 50. Kvaskoff , M. et al. Endomet riosis : a high-r i sk population f or major chr onic diseases? Hum Reprod Upda te 21, 500- 16 (2015). 51. Atsma, F., Bart eli n k , M .L ., Grobb ee, D .E. & van der Schouw, Y.T . Po stmenop a u s al s t atus and early menopause as independ ent risk fac t or s for cardiovas c u l ar dis ea s e: a m eta- analysis. Menopause 13 , 26 5-79 (200 6). 52. Surr ey , E. , Soliman, A.M ., Trenz , H., Blauer -Peterson, C. & Slui s , A. Im pac t o f Endometr i osi s Diagnostic Dela y s on H ealthcare Resource U t iliz ation and C os t s . Adv Ther 37 , 1087- 10 99 (2020). 53. Loh, P.R. et al. Efficient Bayesia n m ix ed- model ana ly si s incr ease s a sso c iation p ower in large cohort s . Nat Ge n e t 47 , 284-90 (2 015) . 54. de La n g e, K.M . et al. G en ome-wide asso cia t ion s t udy implic at es immun e ac tivation of multip le integ r in genes in inflamm a to ry bowel disease. Nat G enet 49, 25 6- 261 (2017). 55. Inter nationa l G en etic s of An k ylo sing S p ondylitis , C . et a l . Iden tification of mult ip le ris k variants for ankylosin g spond ylitis thr ough high-density genotyping of immune- rela ted loci. Na t Ge n e t 45 , 730 -8 (20 13). 56. Inter nationa l Multiple Sclerosis Genetics, C . Multiple sc ler os i s genomic map implic at es peripher al i m mune cell s and microglia in s us ceptibility. Sc ience 365(201 9). 57. Okada, Y. et al. Genetic s of r heumato id ar thri t is contr ibutes to biol ogy and drug dis cover y. Na t u r e 506, 376-81 ( 201 4). 58. Ricano-Ponce, I . et al. Immunochip m et a-analys is in European and Ar gentinian populati o ns identif ies two novel genetic loc i ass o c iated with celiac diseas e. Eur J Hum Genet 28, 313- 323 (2020). 59. Stuart , P.E. et al. Tran sethnic analysi s of p s or ia sis su s c ept ibility i n Sou th Asians and Europ e an s enhance s f ine-mapping in the M HC and genomewide. HG G A d v 3 (2022). 60. W ang , Y. F. et al. Identif i cation of 38 novel loci fo r s y stemic lupus er ythemato s us and genetic hetero g en ei t y between ance s tr al gr oups. N at Commun 12, 772 (20 21). 61. W iller , C .J., Li , Y. & Abeca s is , G .R. M E TAL: fas t and efficient me ta- a n a lysis of geno mewi d e asso ciation s can s. Bioinform atic s 26, 2190-1 (2 010). 62. Trowsdale, J. & K nig h t, J. C. Major his t oc o mpatibili ty complex g en omics and human diseas e. Annu Rev Ge n omics Hum Genet 14, 3 01-23 (2 013). 63. Hemani, G. et al. The MR- Bas e platf o r m suppor ts sy stematic c au sal inferen ce acro s s t he human pheno me. Eli f e 7 ( 2018). 64. Burges s, S ., But terwor th, A. & Thomp s on, S . G. Mendelian random i zation a naly sis with multip le genetic variants using summar iz ed dat a. Genet Epidemiol 37, 658- 65 (201 3). 65. Bowden, J., D avey Smith , G., Hay co c k, P.C . & Bur ge ss, S . C on s istent E stimat ion in Mendelian Randomiz at i o n w it h S o m e Invali d In s tr ument s Us ing a W eight e d Median Es t imator . G enet Epidemiol 40, 304-1 4 (201 6). 66. Bowden, J., D avey Smith , G. & Burge s s , S. Mendelian r a n domiz at ion wi th in v ali d ins t rumen ts : ef fect es t imation and b i as detection thr ough Eg ger regr ession. Int J Epidemiol 44, 512-25 (2 015) . 67. Verb anc k , M. , C hen, C. Y., N eale, B. & Do, R. D etection of widespread horiz ont al p l ei o tropy in c au s al relationships inferr ed from M endelian rando m i zation be t ween c om plex t ra its and dis ease s. Nat Genet 50, 69 3-698 (20 18 ). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint 68. Bowden, J. & Holm es , M .V. Meta- anal ysi s and M endelian randomization: A r evi ew. Res Sy nt h Methods 10, 48 6-496 ( 2019). 69. Consort ium, G . T. et al. Genetic effects on gene ex pr ession a cros s hum a n t issue s . Natur e 550, 20 4-213 (2017 ). 70. Vosa, U . et al. Large-s ca le ci s - a n d trans-eQTL anal yse s identif y t housa n d s of genet i c loci and polygenic s cores that r eg u l at e blood gene ex pr es sion. Nat Ge net 53 , 1 300-1310 (2021). 71. W at anabe, K. , Tas ke s en, E. , van Boch oven, A. & Posthuma, D . F un c t i o nal m apping a n d annotat ion o f genetic associa t ions wi th FUMA . N at C ommun 8 , 1826 (2 017 ). 72. Dub ois , P. C. et al. M ultiple c o mmon variants for c eliac dis ea se i n fluenci n g imm une gene express ion. Nat G enet 42, 295-30 2 (2 010) . 73. Tac h ma z idou, I. et a l . Iden tification of new ther apeutic target s for os t eoart h ritis throu g h genome-wide analys e s of UK Biobank data. Na t Ge n e t 51 , 230- 236 (2 019). 74. Benth a m , J. et a l. G enet ic as s o ciation analy ses implicate aberr ant regulatio n o f innate and adaptive im munit y genes in the path o ge n es i s of sy stemic lupus er ythemat os u s. Na t Ge n e t 47, 145 7-1464 (201 5 ). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 9, 2024. ; https://doi.org/10.1101/2024.07.08.24310092doi: medRxiv preprint