Genetic Study Reveals One-Way Causal Link Between Liver Autoimmune Disease and Inflammatory Bowel Disease

A large genetic analysis has uncovered a strikingly one-directional relationship between two chronic autoimmune conditions: primary biliary cirrhosis, a disease in which the immune system slowly destroys the small bile ducts inside the liver, and inflammatory bowel disease, the umbrella term for ulcerative colitis and Crohn’s disease. Using a statistical technique called two-sample bidirectional Mendelian randomization, researchers found that genetic variants that predispose people to primary biliary cirrhosis also raise the risk of developing inflammatory bowel disease, while the reverse influence runs only from ulcerative colitis back to the liver condition—and in the opposite, protective direction. The work, published in Experimental Biology and Medicine, offers some of the strongest evidence to date that the long-observed clinical overlap between these disorders reflects genuine causal biology rather than coincidence or shared diagnostic patterns.

The puzzle the researchers set out to solve is a familiar one in autoimmune medicine. Clinicians have repeatedly noticed that patients with primary biliary cirrhosis, now often called primary biliary cholangitis, are far more likely than the general population to develop ulcerative colitis, and case reports describe Crohn’s disease co-occurring with the liver condition as well. Observational studies, however, cannot untangle whether one disease truly drives the other, whether both are downstream of shared environmental triggers such as infections or smoking, or whether the apparent link simply arises because doctors who find one disease look harder for the second. Disentangling cause from correlation in human populations normally requires randomized trials, which are impossible for diseases that unfold over decades.

Mendelian randomization sidesteps that obstacle by exploiting a natural experiment that everyone carries from conception. Because alleles are shuffled randomly when sperm and eggs form, the genetic variants a person inherits are distributed independently of the environmental exposures, lifestyle factors, and reverse causal chains that confound ordinary observational research. When variants robustly associated with an exposure are also associated with an outcome, the pattern can be read as evidence of causation, in much the same way that a randomized trial assigns participants to groups by chance. The approach relies on three assumptions: the chosen variants must be strongly linked to the exposure, must not influence the outcome through any other pathway, and must not be confounded by other traits.

The research team, led by Mingyi Yang of Honghui Hospital at Xi’an Jiaotong University together with colleagues from several institutions in Xi’an, China, drew on publicly available genome-wide association study summary statistics. For primary biliary cirrhosis they used a dataset of 11,375 individuals of European descent, including 2,861 cases diagnosed by established criteria such as anti-mitochondrial antibodies at a titer of 1:40 or higher, liver biopsy findings, and biochemical markers. The inflammatory bowel disease data came from the Finnish consortium FinnGen: 214,620 participants for ulcerative colitis with 4,320 cases, 211,107 for Crohn’s disease with 807 cases, and 218,792 for inflammatory bowel disease overall with 5,673 cases. After rigorous filtering for linkage disequilibrium, palindromic variants, and potential confounders, the team identified 22 genetic variants strongly associated with primary biliary cirrhosis as instruments for the forward analysis, and smaller sets of variants for the reverse direction.

The forward results were consistent and statistically robust. Genetic liability to primary biliary cirrhosis increased the risk of ulcerative colitis with an odds ratio of 1.081 and of Crohn’s disease with an odds ratio of 1.136, both highly significant, and raised the risk of inflammatory bowel disease as a whole with an odds ratio of 1.076. In plain terms, inheriting the genetic architecture of the liver disease modestly but reliably raises the probability of developing either form of gut inflammation. The effect sizes are small, as is typical for genetically instrumented estimates of complex disease relationships, but their consistency across both inflammatory bowel disease subtypes and the composite phenotype strengthens the case that the liver condition sits upstream in the causal chain.

The reverse direction told a more nuanced story. Ulcerative colitis showed a significant negative causal effect on primary biliary cirrhosis, with an odds ratio of 0.788, meaning that genetic liability to the bowel disease appears to protect against the liver condition. Crohn’s disease, by contrast, exerted no detectable causal influence on primary biliary cirrhosis, and neither did inflammatory bowel disease considered as a whole. This asymmetry is notable because it suggests that the counteracting influence on the liver disease is specific to ulcerative colitis rather than a generic consequence of intestinal inflammation, hinting at a mechanism particular to that subtype.

Because Mendelian randomization estimates can be distorted by heterogeneity among instruments, horizontal pleiotropy in which variants affect outcomes through pathways unrelated to the exposure, or single outliers that dominate the result, the team subjected every analysis to an extensive battery of sensitivity checks. Cochran’s Q and Rucker’s Q statistics tested for heterogeneity, MR-Egger intercept tests and the MR-PRESSO global test probed for pleiotropy, radial analyses and distortion tests flagged outlier variants, leave-one-out analyses confirmed that no single variant drove the estimates, and MR-RAPS verified that residuals followed a normal distribution. In the final models, after iterative exclusion of problematic variants such as rs522127, rs8067378, and rs72678531 in the ulcerative colitis analysis and rs10931468 in the Crohn’s disease analysis, none of these diagnostics signaled residual bias, lending credibility to the central findings.

The biological plausibility of a shared pathway between the biliary tree and the gut has deep roots in the literature. More than half of known inflammatory bowel disease susceptibility loci overlap with loci for other inflammatory and autoimmune diseases, and ulcerative colitis and primary sclerosing cholangitis, a related biliary disorder, share risk variants at genes including MST1, IL2, CARD9, and REL. Both conditions are also associated with the HLA class II region on the short arm of chromosome 6, which encodes regulators of immune recognition. Epidemiologically, the overlap is dramatic: among 412 cohorts of ulcerative colitis patients reviewed in prior work, the prevalence of primary biliary cirrhosis was estimated at roughly 485 per 100,000, nearly thirty times the rate in the general European population, and in most cases the bowel disease had been diagnosed years before the liver disease emerged.

Several mechanistic hypotheses could explain the newly mapped causal architecture. The polyclonal immunoglobulin M response and autoantibodies characteristic of primary biliary cirrhosis cross-react with microbial antigens in the gut, potentially altering microbiome composition and epithelial integrity in ways that promote inflammation. Reduced bile flow and the accumulation of toxic bile products in the liver disease may further disturb the gut and biliary microbiota. On the Crohn’s disease side, shared susceptibility genes such as ATG16L1, IRGM, and LRRK2 implicate defective mitochondrial autophagy, and the mitochondrial autoantigens targeted in primary biliary cirrhosis, including the DLAT/PDC-E2 and OGDH complexes, sit squarely within that same cellular machinery. A Japanese genetic study identified shared risk alleles at CXCR5, ICOSLG, and TNFSF15 in both diseases, while alleles at IL12B, STAT4, and NFKB1 showed opposite effects, suggesting that the IL12-STAT4-NFKB signaling axis may be tuned differently in the two conditions.

The clinical implications, if the findings are replicated, could reshape surveillance practice. The authors suggest that patients diagnosed with primary biliary cirrhosis may warrant heightened vigilance and potentially routine screening for incident ulcerative colitis and Crohn’s disease, since the genetic data position the liver condition as a predisposing factor across the inflammatory bowel disease spectrum. Conversely, the apparent protective effect of ulcerative colitis on primary biliary cirrhosis raises the intriguing possibility that individuals with the bowel disease are less susceptible to the liver condition, a counterintuitive observation that the authors themselves flag as requiring further investigation. The study does carry limitations: the genome-wide association data are restricted to European ancestries, limiting generalizability to other populations where allele frequencies and gene-environment interactions differ, and the number of instrumental variables, particularly in the reverse analyses, is small, which constrains statistical power. Future work incorporating multi-ancestry data, polygenic risk scores, and multi-omics integration will be needed to confirm and extend the causal map. For now, the study stands as a clear demonstration that the relationship between the liver’s bile ducts and the inflamed gut is not a two-way street but a carefully choreographed, direction-dependent genetic dialogue.

Subject of Research: Genetic causal relationship between primary biliary cirrhosis and inflammatory bowel disease using bidirectional Mendelian randomization

Article Title: Primary biliary cirrhosis and inflammatory bowel disease: a two-sample bidirectional Mendelian randomization study

Article References: Primary biliary cirrhosis and inflammatory bowel disease: a two-sample bidirectional Mendelian randomization study. (n.d.). https://doi.org/10.3389/ebm.2026.10940

Image Credits: AI Generated

DOI: 10.3389/ebm.2026.10940

Keywords: primary biliary cirrhosis, inflammatory bowel disease, ulcerative colitis, Crohn's disease, Mendelian randomization, genome-wide association study, genetic causality, autoimmune disease, gut-liver axis, instrumental variables, sensitivity analysis, HLA region