Предлагаемые причинно-следственные связи между 17 аутоиммунные заболевания и панкреатит: A two- Выбрать менделианское исследование случайности.

1. Introduction

Pancreatitis, caused by the auto-digestive action of pancreatic proteases, is a complex and potentially life-threatening gastrointestinal emergency, particularly in developed countries. [ 1 ] acute pancreatitis (AP) occurs at an incidence of approximately 34 cases per 100,000 people in high-income countries, [ 2 ] and up to 20% of cases progress to moderate or severe forms with substantial morbidity and mortality. Approximately 3% to 35% of patients develop chronic pancreatitis (CP) within several years, [ 3 – 5 ] which can result in persistent abdominal pain, pancreatic insufficiency, diabetes, and pancreatic cancer. [ 6 ] Alcohol consumption is a major risk factor and is more commonly associated with pancreatitis in Western countries and Japan than in other Asian regions, [ 7 ] leading to 2 subtypes: alcohol-induced acute pancreatitis (AAP) and alcohol-induced chronic pancreatitis (ACP).

In addition to well-established risk factors such as gallstones, alcohol, and hyperlipidemia, [ 7 , 8 ] autoimmune diseases (ADs) have drawn increasing attention as potential contributors to pancreatitis. Epidemiological studies have suggested an association between pancreatitis and various ADs, including rheumatoid arthritis (RA), ulcerative colitis (UC), celiac disease (CeD), Hashimoto’s thyroiditis, and primary biliary cirrhosis. [ 9 – 16 ] However, these observational findings are vulnerable to residual confounding and reverse causality, and the causal nature of these associations remains unclear.

Mendelian randomization (MR) leverages genetic variants as instrumental variables to strengthen causal inference. [ 17 – 21 ] While previous MR analyses have explored selected immune-mediated diseases, such as inflammatory bowel disease in relation to pancreatitis, [ 22 ] a comprehensive evaluation across multiple ADs and distinct forms of pancreatitis is lacking. Given the possible heterogeneity in pathophysiology, this study investigated the genetic causal relationships between 17 common ADs and 4 clinically important pancreatitis subtypes: AP, CP, AAP, and ACP, representing both acute and chronic, alcohol-related and non-alcohol-related forms. This approach may help clarify shared biological pathways and inform preventive strategies.

2. Methods study design

MR was used to study the causal relationship between ADs and the different types of pancreatitis. Single nucleotide polymorphisms (SNPs) associated with these risk factors were used as instrumental variables (IVs). The following 3 assumptions served as the foundation for the MR study: the SNPs are closely related to the risk factors, the SNPs are irrelevant to various confounders, and the SNPs only influence the outcomes through the risk factors (Fig. 1 ). Since the data used were based on published research and public databases (genome-wide association study (GWAS), European genome-wide association studies, and Finngen datasets open to IEU), there was no need for additional ethical approval from the Institutional Review Committee.

Flowchart of the MR study between pancreatitis and 17 ADs. AAP = alcohol-induced acute pancreatitis, ADs = autoimmune diseases, ACP = alcohol-induced chronic pancreatitis, CP = chronic pancreatitis, MR = Mendelian randomization, AP = acute pancreatitis.

2.1. Study cohorts and GWAS

Seventeen ADs closely related to pancreatitis were selected for use in this analysis (Literature Retrieval Process ADs Related to Pancreatitis in Table 1 ). In addition, datasets for each AD were obtained from the GWAS and Finngen datasets open to IEU as exposures, including RA, systemic lupus erythematosus (SLE), Eczema, Psoriasis, Vitiligo, multiple sclerosis (MS), myasthenia gravis (MG), narcolepsy, type 1 diabetes (T1D), graves’ disease (GD), hashimoto’s thyroiditis (HT), primary sclerosing cholangitis (PSC), primary biliary cirrhosis (PBC), UC, crohn’s disease (CD), CeD, and autoimmune hepatitis (AH). The diseases were classified into 4 categories: connective tissue diseases, neuromuscular diseases, endocrine diseases, and digestive system diseases. The outcomes were 4 different types of pancreatitis: AP, CP, AAP, and ACP. The latest and largest publicly available GWAS summary data were used; all case and control groups were of European origin, and there was no significant overlap between GWAS populations (Table 2 ). Preliminary studies have provided detailed information on registration procedures and diagnostic criteria.

Table 1 Literature retrieval process autoimmune diseases related to pancreatitis. Database Retrieval strategies Merge PubMed #1“Pancreatitis”[Mesh] OR “Acute pancreatitis” OR “Chronic pancreatitis” OR “Alcohol induced acute pancreatitis” OR “Alcohol induced chronic pancreatitis” #2 “Autoimmune Diseases” [Title/Abstract] OR “Autoimmunity “[Mesh] OR “Rheumatoid Arthritis”[Mesh] OR “Systemic Lupus Erythematosus”[Mesh]]OR “Multiple Sclerosis”[Mesh] OR “Type 1 Diabetes” #1 AND #2 Embase #1 “pancreatitis*”:ab,ti OR “acute pancreatitis *”:ab,ti OR “Chronic pancreatitis”:ab,ti OR “Alcohol induced acute pancreatitis”:ab,ti OR “Alcohol induced chronic pancreatitis”/ exp #2 “autoimmune disease”:ab,ti OR “autoimmunity”/exp.OR “rheumatoid arthritis”:ab,ti OR “systemic lupus erythematosus”:ab,ti OR “multiple sclerosis,”:ab,ti OR “type 1 diabetes”:ab,ti #1 AND #2 Scopus #1 TITLE-ABS-KEY (“Pancreatitis*”) OR TITLE-ABS-KEY (“Acute pancreatitis *”)OR TITLE-ABS-KEY (“Alcohol induced acute pancreatitis*”) OR TITLE-ABS-KEY (“Alcohol induced chronic pancreatitis*”) #2 TITLE-ABS-KEY (“autoimmune diseases”) OR TITLE-ABS-KEY (“autoimmunity “) OR TITLE-ABS-KEY(“rheumatoid arthritis”) OR TITLE-ABS-KEY(“autoimmune hepatitis”) OR TITLE-ABS-KEY(“systemic lupus erythematosus”) OR TITLE-ABS-KEY(“multiple sclerosis”) OR TITLE-ABS-KEY (“systemic lupus erythematosus”) OR TITLE-ABS-KEY (“type 1 diabetes”) #1 AND #2

Literature retrieval process autoimmune diseases related to pancreatitis.

Table 2 Characteristics of the ADs and pancreatitis GWAS cohorts. Disease Study Journal Cases Controls Sample size GWAS ID RA Sakaue S et al.2021 Nat Genet 8255 409,001 417,256 ebi-a-GCST90018910 SLE Sakaue S et al.2021 Nat Genet 647 482,264 482,911 ebi-a-GCST90018917 Eczema Loh PR et al.2018 Nat Genet NA NA 461,199 ebi-a-GCST90029017 Psoriasis Stuart PE al.2022 HGG Adv 15,967 28,194 44,161 ebi-a-GCST90019016 Vitiligo NA NA 131 207,482 NA finn-b-L12_VITILIGO MS Andlauer TF et al.2016 Sci Adv 4888 10,395 15,283 ebi-a-GCST003566 MG Sakaue S et al.2021 Nat Genet 197 354,945 355,142 ebi-a-GCST90018876 Narcolepsy Faraco J et al.2013 PLoS Genet 1886 10,421 12,307 ebi-a-GCST005522 T1D Chiou J et al.2021 Nature 18,942 501,638 520,580 ebi-a-GCST90014023 GD Sakaue S et al.2021 Nat Genet 1678 456,942 458,620 ebi-a-GCST90018847 HT Sakaue S et al.2021 Nat Genet 15,654 379,986 395,640 ebi-a-GCST90018855 PSC Sun-Gou Ji et al.2017 Nat Genet 2871 12,019 14,890 ieu-a-1112 PBC Liu JZ et al.2012 Nat Genet 2861 8514 11,375 ebi-a-GCST005581 UC Saori Sakaue et al.2021 Nat Genet 5371 412,561 417,932 ebi-a-GCST90018933 CD Neale et al.2017 NA 732 336,467 337,199 ukb-a-552 CeD Trynka et al.2017 Nat Genet 12,041 12,228 24,269 ieu-a-1058 AH Sakaue S et al.2021 Nat Genet 821 484,413 485,234 ebi-a-GCST90018785 AP Sakaue S et al.2021 Nat Genet 3798 476,104 479,902 ebi-a-GCST90018789 CP Sakaue S et al.2021 Nat Genet 1424 477,528 477,528 ebi-a-GCST90018821 AAP NA NA 457 218,335 NA finn-b-ALCOPANCACU ACP NA NA 977 217,815 NA finn-b-ALCOPANCCHRON AAP = alcohol-induced acute pancreatitis, ACP = alcohol-induced chronic pancreatitis, AH = autoimmune hepatitis, AP = acute pancreatitis, CD = crohn’s disease, CeD = celiac disease, CP = chronic pancreatitis, GD = graves’ disease, HT = hashimoto’s thyroiditis, MG = myasthenia gravis, MS = multiple sclerosis, PBC = primary biliary cholangitis, PSC = primary sclerosing cholangitis, RA = rheumatoid arthritis, SLE = systemic lupus erythematosus, T1D = type 1 diabetes, UC = ulcerative colitis.

Characteristics of the ADs and pancreatitis GWAS cohorts.

AAP = alcohol-induced acute pancreatitis, ACP = alcohol-induced chronic pancreatitis, AH = autoimmune hepatitis, AP = acute pancreatitis, CD = crohn’s disease, CeD = celiac disease, CP = chronic pancreatitis, GD = graves’ disease, HT = hashimoto’s thyroiditis, MG = myasthenia gravis, MS = multiple sclerosis, PBC = primary biliary cholangitis, PSC = primary sclerosing cholangitis, RA = rheumatoid arthritis, SLE = systemic lupus erythematosus, T1D = type 1 diabetes, UC = ulcerative colitis.

2.2. Instrumental variables selection

SNPs were selected for each AD, and default settings in the “TwoSampleMR” R package were used. Genome-wide significant SNPs were selected based on a P -value < 5 × 10 −6 , and standard clumping parameters were used to select independent SNPs (clumping window of 10,000 kb, LD r 2 cutoff 0.001). The intensity of the instrumental variables (IVs) was estimated by calculating the proportion of variance in the disease explained by SNPs (R 2 ), and the F-statistic was used to satisfy the first MR hypothesis. F statistics ( F  = beta 2 /se 2 ) were calculated for each SNP, and a general F statistic was calculated for all SNPs for the corresponding exposure. [ 23 ] F  > 10 was considered to indicate sufficient strength. Reverse MR imaging was performed using the same approach.

2.3. Mendelian randomization analysis

Two-sample MR analyses were conducted using random-effects inverse-variance weighted (IVW), [ 23 ] weighted median, and MR-Egger. The traditional IVW analysis method may be affected by invalid instrumental bias or multivariate validity; therefore, the present study conducted sensitivity analysis to test IVW. Validity and robustness of results. To minimize the bias caused by horizontal pleiotropy (affecting the results through causal pathways rather than exposure), this study used pleiotropic residual sums and outliers (MR-PRESSO) to test for extensive horizontal pleiotropy. Two other established MR methods, including weighted median and MR-Egger regression methods, were used for the association between ADs and pancreatitis. Further tests were conducted; finally, to assess the robustness of the results, further heterogeneity tests, including the MR-Egger intercept test, sensitivity analysis, and modified Cochrane’s Q statistic, were performed for statistically significant results. [ 23 ] Finally, a leave-one-out analysis was performed to screen for any individual SNPs that were disproportionately accountable for the outcome of each MR. As the outcome was a dichotomous variable, the effect estimates were converted to odds ratios (OR) to visualize the relationship between ADs and pancreatitis [ 23 ] .pancreatitis. [ 23 ] Statistical significance was set at P  < .05.

Multiple hypothesis test corrections were performed using the Bonferroni correction. A Bonferroni-corrected significance level of P  < 2.94 × 10 −3 (0.05/17) was used, and P values ranging from 2.78 × 10 −3 to .05 were classified as suggestive causal associations. All MR analyses were implemented using the 2 Sample MR package version 0.5.8 in R (version 4.3.1).

3. Results

3.1. Causal effects of 17 autoimmune diseases on acute pancreatitis

The IVW analysis revealed that only T1D and UC were causally associated with AP (T1D: OR = 1.032, 95CI% = 1.008–1.056, P  = .0086; UC: OR = 1.065, 95CI% = 1.012–1.121, P  = .0151). Other findings were that all 11 ADs increased the risk of AP (RA: OR = 1.041, P  = .2897; SLE: OR = 1.002, P  = .946; Eczema: OR = 1.064, P  = .7899; Psoriasis: OR = 1.014, P  = .4487; MS: OR = 1.014, P  = .4245; MG: OR = 1.010, P  = .4739; GD: OR = 1.002, P  = .9124; HT: OR = 1.026, P  = .3909; PSC: OR = 1.036, P  = .0767; CD: OR = 30.345, P  = .3909; AH: OR = 1.036, P  = .1616), but there was no causality (Fig. 2 ).The weighted median, MR-Egger test, pleiotropy, heterogeneity, and leave-one-out results are presented in Supplementary Figure 1 , Supplemental Digital Content 1 and Supplementary Table 1 , Supplemental Digital Content 3.

Forest plot to visualize the causal effect of 17 ADs on AP using the inverse variance-weighted method. ADs = autoimmune diseases, AP = acute pancreatitis.

3.2. Causal effects of 17 autoimmune diseases on chronic pancreatitis

The IVW analysis showed that only PSC was causally associated with CP (PSC: OR = 1.080, 95% CI = 1.013–1.153, P  = .0194). Other results showed 7 ADs increased the risk of CP (RA: OR = 1.094, P  = .1371; Psoriasis: OR = 1.048, P  = .134; MS: OR = 1.033, P  = .2312; MG: OR = 1.015, P  = .4967; GD: OR = 1.029, P  = .4743; UC: OR = 1.056, P  = .1654; CD: OR = 1.249, P  = .967), but there was no causal relationship (Figure 3 ).The weighted median, MR-Egger test, pleiotropy, heterogeneity, and leave-one-out results are presented in Supplementary Figure 1 , Supplemental Digital Content 1 and Supplementary Table 2 , Supplemental Digital Content 4.

Forest plot to visualize the causal effect of 17 ADs on CP using the inverse variance-weighted method. ADs = autoimmune diseases, AP = acute pancreatitis.

3.3. Causal effects of 17 autoimmune diseases on alcohol-induced pancreatitis

The IVW analysis showed no causal relationship between ADs and alcohol-induced pancreatitis. Other results were that 7 ADs increased the risk of AAP (Psoriasis: OR = 1.020, P  = .7351; Narcolepsy: OR = 1.321, P  = .1283; PSC: OR = 1.033, P  = .6065; PBC: OR = 1.091, P  = .0535; UC: OR = 1.152, P  = .0867; CD: OR = 5.699, P  = .8850; AH: OR = 1.031, P  = .7069), and the presence of 12 ADs was associated with an increased risk of ACP (RA: OR = 1.085, P  = .3171; Eczema: OR = 1.035, P  = .9437; Psoriasis: OR = 1.047, P  = .2563; Vitiligo: OR = 1.031, P  = .4239; MG: OR = 1.005, P  = .8986; Narcolepsy: OR = 1.079, P  = .4791; GD: OR = 1.014, P  = .781; HT: OR = 1.103, P  = .1378; PBC: OR = 1.057, P  = .1056; UC: OR = 1.020, P  = .7537; CD: OR = 32.643, P = .613; AH: OR = 1.018, P = .7585), but there was no causal relationship (Fig. 4 ) The weighted median, MR-Egger test, pleiotropy, heterogeneity, and leave-one-out results are presented in Supplementary Figure 1 , Supplemental Digital Content 1 and Supplementary Tables 3 , Supplemental Digital Content 5 and 4 , Supplemental Digital Content 6.

Forest plot to visualize the causal effect of 17 ADs on alcohol-induced pancreatitis using the inverse variance-weighted method. ADs = autoimmune diseases, AP = acute pancreatitis.

3.4. Reverse MR analysis: causal effects of 17 autoimmune diseases on pancreatitis

Reverse MR analysis showed that only AAP was potentially associated with a lower risk of HT (OR = 0.962, 95% CI = 0.936–0.989, P  = .0056). AP, CP, AAP, and ACP were not causally associated with 17 ADs (Fig. 5 )The IVW results are shown in Supplementary Figure 2 , Supplemental Digital Content 2 and Supplementary Table 5 , Supplemental Digital Content 7.

Summary data to visualize the causal effect of 4 Pancreatitis on 17 ADs using the inverse variance-weighted method. AAP = alcohol-induced acute pancreatitis, ACP = alcohol-induced chronic pancreatitis, ADs = autoimmune diseases, AP = acute pancreatitis, CP = chronic pancreatitis.

4. Discussion

ADs are caused by the immune system’s inability to recognize foreign invaders and self-antigens, [ 24 ] thus, inflammatory mediators (inflammatory factors and inflammasomes) play an important role in the pathogenesis of ADs. Innate immune cells (macrophages and granulocytes) proliferate abnormally, secrete large amounts of inflammatory factors, and stimulate the abnormal infiltration of T and B cells, causing progressive inflammation and tissue damage. [ 25 , 26 ] Pancreatitis is caused by the self-digestion of pancreatic tissue by trypsin. Under normal circumstances, trypsinogen in the pancreatic juice is inactive. When influenced by certain factors, such as gallstones, alcoholism, high triglycerides, infection, immune diseases, or drug toxicity, inflammation occurs, causing trypsin to be activated, followed by other enzyme reactions such as elastase and phospholipase A, which cause self-digestion of the pancreas and promote its necrosis and dissolution. Pancreatitis has obvious complications and a high mortality rate, with a mortality rate of approximately 2% to 5%. [ 27 ] The occurrence and progression of pancreatitis are multifaceted, resulting from the interactions between genetic and environmental factors. Epidemiological studies have reported that various ADs increase the risk of pancreatitis; however, the causal impact of genetics on ADs and pancreatitis is still unclear.

4.1. Acute pancreatitis (AP)

Our analysis demonstrated that genetically predicted type 1 diabetes mellitus (T1DM) and UC) were associated with an increased risk of AP. This is in line with previous epidemiological studies reporting a higher incidence of AP in patients with ADs such as RA, UC, and CeD. For connective tissue diseases (RA, SLE, psoriasis, eczema, and vitiligo), [ 28 ] however, we did not observe genetic causality with pancreatitis. Notably, 2 large population-based cohort studies from Taiwan showed that RA was associated with an increased AP risk (HR = 1.62; 95% CI: 1.43–1.83), [ 9 ] and psoriasis with an increased CP risk (HR = 1.81; 95% CI: 1.53–2.15), but these risks were reduced with corticosteroids, nonsteroidal anti-inflammatory drugs, or methotrexate treatment. [ 29 ] Conversely, drug withdrawal and the use of certain disease-modifying antirheumatic drugs have also been linked to increased AP risk. [ 30 , 31 ] These contradictory findings suggest that pharmacological factors, rather than genetic predisposition, may play a more important role in the development of pancreatitis in connective tissue diseases.

With regard to T1DM, biological plausibility is supported by shared autoimmune pathways. [ 32 ] T1DM frequently coexists with CeD owing to common HLA haplotypes, while non-hla genetic and environmental factors (viral infections, gut microbiota, and gluten exposure) may also contribute. Pancreatic β -cells aberrantly express tissue transglutaminase under stress, potentially linking T1DM autoimmunity to exocrine pancreatic injury. Although autoimmune pancreatitis (AIP) is exceedingly rare in children and is not considered a relevant comorbidity of CeD, immune-mediated overlap may exist and warrant further investigation. [ 33 ] Our findings confirm that T1DM increases the risk of AP, supporting the hypothesis that immune dysregulation in diabetes may extend to the pancreatic exocrine tissues.

Observational studies have suggested a 2 to 3-fold higher risk of pancreatitis for CeD, potentially due to malnutrition-related acinar atrophy, altered intestinal hormone secretion, and shared Th1-driven immune responses. However, our MR results do not support a causal link. Importantly, a recent pediatric review concluded that pancreatitis, including autoimmune forms, is not a relevant comorbidity in children with CeD, highlighting possible age-related differences.

4.2. Chronic pancreatitis (CP)

We identified a causal genetic association between PSC and CP. This is biologically plausible, as PSC shares immunopathological features with AIP including elevated serum IgG4 levels, which have been reported in 9% to 36% of patients with PSC, and lymphoplasmacytic infiltration of hepatobiliary tissues. [ 34 ] High IgG4 levels have also been observed in the liver, gallbladder, and extrahepatic bile ducts, which are commonly affected by AIP. Whether PSC and AIP develop independently or represent variations within the same IgG4-related disease spectrum remains uncertain, but their immunological overlap provides mechanistic support for our findings. [ 35 ]

In contrast, although psoriasis has been associated with CP in observational studies, our MR analysis did not confirm a genetic relationship, suggesting that environmental or therapeutic factors may account for the observed risk. Similarly, no causal associations were detected between CP and other autoimmune conditions such as HT or primary biliary cholangitis (PBC). [ 36 ] Previous studies have suggested immunological overlaps between primary biliary cholangitis and AIP including HLA-DQB1 susceptibility, dysregulation of cytotoxic T-lymphocyte antigen-4, predominant T-cell–mediated mechanisms, and B-cell infiltration. [ 16 , 37 ] Nevertheless, our MR analysis did not reveal a genetic relationship between primary biliary cholangitis and pancreatitis, and comorbidity has only been described in isolated case reports.

4.3. Alcohol-related pancreatitis (AAP and ACP)

Alcohol is a well-established risk factor for pancreatitis; however, its interaction with ADs remains poorly understood. In the present reverse MR analysis, we observed a suggestive protective causal association between AAP and HT. Specifically, genetically predicted AAP was associated with a lower risk of HT. However, this association did not remain statistically significant after Bonferroni correction and therefore should be interpreted with caution.

The biological mechanisms underlying this observation remain unclear. Given the exploratory nature of this finding, it is premature to infer a direct protective effect of AAP on thyroid autoimmunity. Instead, the observed association may reflect shared genetic backgrounds, immune regulatory pathways, or other unmeasured factors that warrant further investigation. [ 36 ] Future experimental and epidemiological studies are needed to validate this finding and clarify the underlying mechanisms. No significant associations were observed between other ADs and alcohol-related pancreatitis, underscoring the need for further validation in larger and more diverse populations.

4.4. Reverse Mendelian Randomization findings

In addition to the forward MR results, reverse analysis indicated that AAP may increase the risk of HT, suggesting a bidirectional relationship between alcohol-related pancreatic inflammation and thyroid autoimmunity. However, no other reverse causal links have been identified between pancreatitis subtypes and ADs.

4.5. Strengths and limitations

This is the first bidirectional MR study to systematically evaluate the causal relationship between 17 ADs and 4 pancreatitis subtypes. The bidirectional approach strengthens causal inference and reduces residual confounding compared with observational studies. Nevertheless, this study has several limitations that must be acknowledged. First, our datasets were derived exclusively from European populations, which may limit their generalizability to other ethnic groups. Second, sex-specific analyses could not be performed, although many ADs are more prevalent in women. Finally, only 4 pancreatitis subtypes (AP, CP, AAP, and ACP) were analyzed because of the absence of large-scale GWAS data for other clinically relevant forms such as AIP and acute recurrent pancreatitis (ARP). Future studies that include more diverse populations, sex-stratified analyses, and expanded GWAS resources are warranted to validate and extend these findings.

5. Conclusion

This MR study suggests potential genetic associations between certain ADs and pancreatitis in European populations. In particular, T1DM and UC may have causal links with AP, while PSC appears related to CP. No strong genetic evidence was found for other ADs. These results warrant cautious interpretation and further confirmation in other populations and pancreatitis subtypes.

Acknowledgments

The authors acknowledge the UK Biobank, FinnGen Biobank, Sakaue S et al., Loh PR et al., Stuart PE et al., Faraco J et al., Chiou J et al., Andlauer TF et al., and others, for their publicly available summary data.

Author contributions

Conceptualization: Dehua Huang, Zhenyu Li, Yi Miao.

Data curation: Dehua Huang, Zhenyu Li, Chunhui Lu, Chenchen Li.

Investigation: Zhenyu Li, Huijuan Wang.

Methodology: Zhenyu Li, Shangnan Dai, Lingdi Yin.

Formal analysis: Chongfa Chen, Guangfu Wang, Feihu Sun.

Software: Hanxiao You.

Supervision: Jinfan Zhang.

Funding acquisition: Yi Miao.

Writing – original draft: Dehua Huang, Yun Li, Zhenyu Li, Yi Miao.