Key PointsQuestion听
To what extent are shared genetic determinants in the comorbidities and associations between gastrointestinal tract diseases and psychiatry disorders involved in the gut-brain axis?
Findings听
In this genome-wide pleiotropic association study using genome-wide association summary statistics from publicly available data sources, pervasive genetic correlations and genetic overlaps between gastrointestinal tract diseases and psychiatric disorders were found. The pleiotropic genetic determinants between them were extensively distributed across the genome.
Meaning听
These findings not only support the shared genetic basis underlying the gut-brain axis but also have important implications for intervention and treatment targets of these 2 types of diseases simultaneously.
Importance听
Comorbidities and genetic correlations between gastrointestinal tract diseases and psychiatric disorders have been widely reported, with the gut-brain axis (GBA) hypothesized as a potential biological basis. However, the degree to which the shared genetic determinants are involved in these associations underlying the GBA is unclear.
Objective听
To investigate the shared genetic etiology between gastrointestinal tract diseases and psychiatric disorders and to identify shared genomic loci, genes, and pathways.
Design, Setting, and Participants听
This genome-wide pleiotropic association study using genome-wide association summary statistics from publicly available data sources was performed with various statistical genetic approaches to sequentially investigate the pleiotropic associations from genome-wide single-nucleotide variation (SNV; formerly single-nucleotide polymorphism [SNP]), and gene levels and biological pathways to disentangle the underlying shared genetic etiology between 4 gastrointestinal tract diseases (inflammatory bowel disease, irritable bowel syndrome, peptic ulcer disease, and gastroesophageal reflux disease) and 6 psychiatric disorders (schizophrenia, bipolar disorder, major depressive disorder, attention-deficit/hyperactivity disorder, posttraumatic stress disorder, and anorexia nervosa). Data were collected from March 10, 2021, to August 25, 2021, and analysis was performed from January 8 through May 30, 2022.
Main Outcomes and Measures听
The primary outcomes consisted of a list of genetic loci, genes, and pathways shared between gastrointestinal tract diseases and psychiatric disorders.
Results听
Extensive genetic correlations and genetic overlaps were found among 22 of 24 trait pairs. Pleiotropic analysis under a composite null hypothesis identified 2910 significant potential pleiotropic SNVs in 19 trait pairs, with 83 pleiotropic loci and 24 colocalized loci detected. Gene-based analysis found 158 unique candidate pleiotropic genes, which were highly enriched in certain GBA-related phenotypes and tissues, whereas pathway enrichment analysis further highlighted biological pathways primarily involving cell adhesion, synaptic structure and function, and immune cell differentiation. Several identified pleiotropic loci also shared causal variants with gut microbiomes. Mendelian randomization analysis further illustrated vertical pleiotropy across 8 pairwise traits. Notably, many pleiotropic loci were identified for multiple pairwise traits, such as 1q32.1 (INAVA), 19q13.33 (FUT2), 11q23.2 (NCAM1), and 1p32.3 (LRP8).
Conclusions and Relevance听
These findings suggest that the pleiotropic genetic determinants between gastrointestinal tract diseases and psychiatric disorders are extensively distributed across the genome. These findings not only support the shared genetic basis underlying the GBA but also have important implications for intervention and treatment targets of these diseases simultaneously.
The comorbidities and associations between gastrointestinal tract diseases and psychiatric disorders have been widely reported,1-3 which was likely to be regulated by the gut-brain axis (GBA). The GBA is characterized by bidirectional interactions between the gastrointestinal tract and the central nervous system (CNS) and would link intestinal dysfunction and inflammation with brain function and psychiatric disorders.4,5 Various biological mechanisms were involved in the GBA, such as inflammatory immune responses, the autonomic nervous system, and enteric nervous system, where the role of the composition of gut microbiota and related metabolites has been particularly highlighted.4-7 Underlying the conceptual framework of the GBA, the shared genetic etiology might be involved in the associations between gastrointestinal tract diseases and psychiatric disorders.
Genome-wide association studies (GWAS) have identified multiple genetic variants (ie, single-nucleotide variations [SNVs]; formerly single-nucleotide polymorphisms [SNPs]) associated with gastrointestinal tract diseases and psychiatric disorders.8-10 Genetic correlations have been suggested between these 2 types of diseases using linkage disequilibrium (LD) score regression (LDSC).10-14 However, it remains unclear whether the overall genetic correlation would be attributed to a few loci or across the genome.15 Indeed, there would be genetic overlap even without any genetic correlation. Although previous studies have investigated genetic overlap,16 shared susceptibility genes,17,18 and causal relationships19-21 between these 2 types of diseases, they mainly focused on inflammatory bowel disease (IBD) and psychiatric disorders with limited sample sizes. Recently, 2 studies10,22 have conducted GWAS of specific gastrointestinal tract diseases as well as systematic post-GWAS analyses and pointed out the necessity to explore the shared genetic risk across traits to improve the understanding of the disordered brain-gut interactions. Therefore, it is of great importance to further seek out the specific genomic variants or loci accounting for genome-wide genetic correlation and to deeply probe into the shared genetic etiology between these 2 types of diseases. Shared genetic etiology also indicates the potential pleiotropy, which often acts as genetic confounding of the associations between trait pairs.23-25 Cross-trait analysis has been proposed to investigate the pleiotropic genetic variants or loci among multiple traits by leveraging the correlation of GWAS signals,23,26-29 where the pleiotropic loci could serve as intervention targets with the potential to simultaneously prevent or treat these diseases.
In this genome-wide pleiotropic association study, using large-scale GWAS summary data, we performed a genome-wide pairwise trait pleiotropic analysis between 4 gastrointestinal tract diseases (IBD, irritable bowel syndrome [IBS], peptic ulcer disease [PUD], and gastroesophageal reflux disease [GERD]) and 6 psychiatric disorders (schizophrenia, bipolar disorder [BIP], major depressive disorder [MDD], attention-deficit/hyperactivity disorder [ADHD], posttraumatic stress disorder [PTSD], and anorexia nervosa [AN]) through various statistical genetic approaches to sequentially investigate the pleiotropic associations from genome-wide, SNV, and gene levels and biological pathways to disentangle the underlying shared genetic etiology. Of note, under the framework of pleiotropic analysis, we first performed the SNV-level analysis to detect pleiotropic variants and loci, followed by pairwise colocalization analysis to determine colocalized loci and gene-level analysis to identify candidate pleiotropic genes, based on which we further performed parallel phenotype and tissue-specific enrichment analysis to characterize the phenotype and tissue specificity as well as additional gene-level analysis to identify the tissue-specific and cell type-specific pleiotropic genes. We also highlighted the role of gut microbiomes in interpreting the shared genetic etiology, followed by mendelian randomization analysis to evaluate pairwise causal associations and partly characterize different types of pleiotropy (vertical pleiotropy or horizontal pleiotropy).
We sought GWAS summary statistics from publicly available data sources with European ancestry, owing to the limited availability of well-powered GWAS with other ancestries, and selected GWAS with sample sizes larger than 50 000 to ensure statistical power. GWAS for GERD, IBD, and PUD were obtained from the same gastrointestinal tract GWAS based on 456 327 individuals from UK Biobank (UKB).10 GWAS for IBS were obtained from a larger meta-analysis with 486 601 individuals (53 400 cases and 433 201 controls).22 GWAS for the 6 psychiatric disorders were from the Psychiatric Genomics Consortium, including schizophrenia,30 BIP,31 MDD,32 ADHD,33 PTSD,34 and AN.35 In addition, GWAS for early age-related macular degeneration (AMD)36 and cataract37 were obtained to serve as a common set of negative controls for both gastrointestinal tract diseases and psychiatric disorders, given these 2 disorders are relatively limited to the pathological lesions of intraocular contents, and previous studies also showed no significant genetic correlations between AMD and MDD38 as well as among cataract, GERD, and psychiatric symptoms.39 All GWAS were approved by relevant ethic committees, and written informed consent was obtained from all participants, with details provided in the eTable 1 in Supplement 1. Data were collected from March 10, 2021, to August 25, 2021, and analyzed from January 8 through May 30, 2022. This genome-wide pleiotropic association study followed the Strengthening the Reporting of Genetic Association Studies () reporting guideline.
All analyses were performed after excluding SNVs in the major histocompatibility complex region (chromosome 6: 25-35鈥塵egabase [Mb]) due to its complex LD structure and restricted to biallelic SNVs with minor allele frequency larger than 0.01. Details of these methods are provided in Figure 1 and eMethods in Supplement 1.
Both LDSC40 and high-definition likelihood41 were used to assess genome-wide genetic correlations for 24 pairwise traits. The intercept from LDSC could also indicate potential sample overlap between 2 GWAS. Given that genetic correlation only reflects the overall correlation across the genome, we further applied genetic analysis incorporating pleiotropy and annotation (GPA)29 to explore the overall genetic overlap between traits. The Bonferroni-corrected significant threshold was set at P鈥�&濒迟;鈥�2.08鈥壝椻赌�10鈭�3 (.05/24). In addition, negative control analysis was performed through LDSC between AMD and cataract and the 4 gastrointestinal tract diseases as well as the 6 psychiatric disorders, with Bonferroni-corrected significant threshold set at P鈥�&濒迟;鈥�2.5鈥壝椻赌�10鈭�3 (.05/20).
For the union set of pairwise traits with significant genetic correlation or genetic overlap, we used pleiotropic analysis under composite null hypothesis (PLACO) to identify potential pleiotropic SNVs.26 Single-nucleotide variations with P鈥�&濒迟;鈥�5鈥壝椻赌�10鈭�8 for PLACO were considered significant pleiotropic variants. Functional mapping and annotation of genetic associations (FUMA)42 was applied to characterize potential pleiotropic loci, based on which a bayesian colocalization analysis43 was performed to further identify shared causal variants in each pleiotropic locus. We declared a colocalized locus with posterior probability of H4 (PP.H4) larger than 0.7.
Based on PLACO results, we further explored the shared biological mechanisms of these pleiotropic loci. We performed gene-level multimarker analysis of GenoMic annotation (MAGMA)44 analysis on the genes located in or overlapped with the pleiotropic loci based on both PLACO outputs and single-trait GWAS to identify candidate pleiotropic genes, with the significance declared at the locus-specific Bonferroni-corrected 2-sided P鈥�<鈥�.05 for MAGMA analysis on PLACO results and 2-sided P鈥�<鈥�.05 for MAGMA analyses based on original single-trait GWAS. Further, we performed phenotype enrichment analysis based on the Mouse Genome Informatics platform45 to characterize the phenotype specificity of these pleiotropic genes against that of nonpleotropic genes by examining the differences in the proportions of genes associated with certain phenotypes in the pleiotropic gene group against that in the nonpleotropic gene group using the Fisher exact test. Then, we performed tissue-specific enrichment analyses to illustrate the tissue specificity of these pleiotropic genes using the deTS tissue-specific enrichment method46 based on 2 different reference panels, Genotype-Tissue Expression project (GTEx)47 and the Encyclopedia of DNA Elements project (ENCODE).48 We declared the significance with a nominal threshold (2-sided P鈥�<鈥�.05) for these 2 parallel enrichment analyses. In addition, we used E-MAGMA (expression quantitative trait loci [eQTL]鈥搃nformed MAGMA)49 and transcriptome-wide association study analysis using joint-tissue imputation50 to further investigate the tissue-specific genes, parallelized with H-MAGMA (Hi-C鈥揷oupled MAGMA)51 to indicate the cell-type specificity, in which we declared the significance with locus-specific Bonferroni correction. Gene set enrichment analysis was also performed to identify potential biological pathways using the clusterProfiler package.52 Significantly enriched pathways were declared with normalized enrichment score greater than 2 and adjusted P鈥�<鈥�.05. Multitrait colocalization analysis with HyPrColoc (hypothesis prioritization for multi-trait colocalization)53 incorporating host-microbiome GWAS was performed to highlight the critical role of gut microbiome.
Last, we conducted bidirectional mendelian randomization analysis for 24 pairwise traits between gastrointestinal tract diseases and psychiatric disorders to investigate the potential causal trait pairs (ie, vertical pleiotropy), along with negative control analysis using AMD and cataract. We used the inverse-variance weighted method54 as main analysis and in total 6 alternative mendelian randomization methods with different model assumptions as additional analyses for further validation. For main analysis, we chose a false discovery rate approach for multiple testing correction with the significance threshold being false discovery rate鈥揳djusted P鈥�<鈥�.05, given that the commonly used Bonferroni correction is often too stringent for multiple nonindependent tests. The nominally significant threshold (P鈥�<鈥�.05) was used for alternative mendelian randomization methods. The Latent Heritable Confounder Mendelian Randomization (LHC-MR) method, which could account for sample overlap, was used for pairwise traits with potential sample overlap to further validate the mendelian randomization results.55
Genomic Loci Characterization and Functional Annotation
For significant pleiotropic SNVs from PLACO, we applied FUMA to identify independent variants, characterize genomic risk loci, and annotate the functions of variants using LD information from the 1000 Genome Project phase 3 reference panel of European population.42 We characterized independent SNVs with r2 less than 0.6000 and lead SNVs with r2 less than 0.1000 within 1 Mb. Genomic risk loci were defined by merging genomic regions if the physical distance between lead SNVs was less than 250 kilobase.42
Functional annotations, including ANNOVAR software tool categories (Bioinformatics), combined annotation-dependent depletion (CADD) scores, and RegulomeDB scores, were also provided by FUMA. Single-nucleotide variations with a CADD score larger than 12.37 were considered a potentially deleterious variant.42 Single-nucleotide variations with P鈥�&濒迟;鈥�5鈥壝椻赌�10鈭�8 in each single-trait GWAS were also annotated by FUMA for comparison.
Genetic Correlations and Genetic Overlaps Between Gastrointestinal Tract Diseases and Psychiatric Disorders
Using genome-wide association summary statistics from publicly available data sources, pervasive significant genome-wide genetic correlations and genetic overlaps were found across 24 pairwise traits, among which 14 were identified from LDSC and 21 were identified from GPA (Table 1). Notably, 8 trait pairs were identified with nonsignificant genetic correlation but with significant genetic overlap, producing a final union set of 22 pairwise traits for subsequent analysis. In addition, the results from the LDSC were highly consistent with those from high-definition likelihood (eTable 2 and eFigure 1 in Supplement 1) and suggested significant sample overlap for 5 trait pairs (Table 1). For negative control analysis, no significant genetic correlations were detected as expected (eTable 3 in Supplement 1).
Shared Loci Between Gastrointestinal Tract Diseases and Psychiatric Disorders
A total of 2910 SNVs (2284 unique) were identified as potential pleiotropic variants by PLACO in 19 trait pairs (eFigure 2 in Supplement 1). FUMA identified 83 independent genomic risk loci as pleiotropic loci, involving 54 unique chromosomal regions (Figure 2 and eTables 4 and 5 in Supplement 1). Some pleiotropic regions were identified in multiple pairwise traits, such as 11q23.2 (mapped gene: NCAM1 [OMIM ]) and 1q32.1 (mapped gene: INAVA [OMIM ]), suggesting the extensive pleiotropic effects of these loci. The top SNVs in 83 FUMA-annotated pleiotropic loci showed mixed directions of allelic associations (eTable 6 in Supplement 1). Overall, 53 of 83 top SNVs (64%) showed concordant associations with a certain pair of traits, indicating these variants could simultaneously decrease or increase the risk of gastrointestinal tract diseases and psychiatric disorders. The remaining 30 of 83 top SNVs (36%) showed discordant associations, suggesting the possibly distinct biological mechanisms.
ANNOVAR category annotation illustrated that 34 of 83 index SNVs (41%) were intronic variants and 27 of 83 (33%) were intergenic variants. Only 7 of 83 index SNVs (8%) (6 unique) were exonic variants, including 5 messenger RNA (mRNA) exonic variants and 2 noncoding RNA exonic variants (eTable 5 in Supplement 1). Of note, 5 mRNA exonic variants (4 unique) were located in 3 loci and mapped to 3 genes. Specifically, the index SNV rs20551 at 22q13.2 locus (PLACO P鈥�=鈥�5.07鈥壝椻赌�10鈭�11 for IBS-schizophrenia) was a significant eQTL for EP300 (OMIM ) encoding p300 protein, which plays an important role in cell proliferation and differentiation. The index SNVs rs681343 (PLACO P鈥�=鈥�1.36鈥壝椻赌�10鈭�12 for PUD-schizophrenia) and rs601338 (PLACO P鈥�=鈥�5.07鈥壝椻赌�10鈭�11 for PUD-ADHD) at 22q13.2 were in high LD (r2鈥�=鈥�0.996) and had almost identical eQTL regulation information (eTable 7 in Supplement 1), regulating the expression of FUT2 (OMIM ). The index SNV rs13107325 at 4q24 locus (PLACO P鈥�=鈥�1.78鈥壝椻赌�10鈭�14 for GERD-schizophrenia; PLACO P鈥�=鈥�4.19鈥壝椻赌�10鈭�8 for GERD-PTSD) was a significant eQTL for SLC39A8 (OMIM ), which encodes ZIP8 metal cation transporter. In addition, 7 index SNVs (6 unique) were identified with CADD scores larger than 12.37, in which 2 mRNA exonic variants had higher CADD scores, including rs601338 (CADD score: 52; mapped gene: FUT2 [OMIM 182100]) and rs13107325 (CADD score: 34; mapped gene: SLC39A8).
Further colocalization analysis identified 24 of 83 potential pleiotropic loci (29%) with PP.H4 larger than 0.7, in which 22 top SNVs of corresponding loci were identified as candidate shared causal variants (Table 2, Figure 2, and eFigure 3 in Supplement 1). Interestingly, the 19q13.33 locus, which was identified as pleiotropic loci for 2 pairs of traits, was only colocalized between PUD and schizophrenia (PP.H4鈥�=鈥�0.8927) rather than PUD and ADHD (PP.H4鈥�=鈥�0.3552), with the same potential shared causal variant rs681343 identified (mapped gene: FUT2). In addition, 7 pleiotropic loci were identified with PP.H3 larger than 0.7000, indicating there might be different causal variants in these loci (eTable 8 in Supplement 1).
Prioritization of Candidate Pleiotropic Genes and Characterization of Phenotype and Tissue Specificity
MAGMA analysis based on 295 potential pleiotropic genes that located in or overlapped with 83 pleiotropic loci identified 196 significant pleiotropic genes (158 unique), in which 38 genes were detected in 2 or more trait pairs (eTable 9 in Supplement 1). For example, NCAM1 was identified as a significant pleiotropic gene in 5 pairs of traits, followed by INAVA, CACNA1S (OMIM ), and KIF21B (OMIM ) in 4 pairs of traits and BANK1 (OMIM ), CAMKV (OMIM ), DPYD (OMIM ), MON1A (OMIM ), MST1R (OMIM ), PSCA (OMIM ), RBM5 (OMIM ), RBM6 (OMIM ), and SLC39A8 in 3 pairs of traits.
Further parallel phenotype and tissue-specific enrichment analysis suggested both higher phenotype and tissue specificity (Figure 3). Specifically, we obtained in total 19 326 genes with available phenotype annotations resorting to Mouse Genome Informatics platform, in which 144 of 158 unique pleiotropic genes were included. These pleiotropic genes were associated with several GBA-related phenotypes with the analyses performed on each phenotype (Figure 3A) and were associated with the 2 phenotypes simultaneously (behavior-neurological phenotype and digestive-alimentary phenotype), contrasting with that of nonpleotropic genes with less association with these phenotypes (7.64% vs 3.61%; P鈥�=鈥�1.80鈥壝椻赌�10鈭�2) (Figure 3B). Details of these genes were provided in eTable 10 in Supplement 1. Tissue-specific enrichment analysis using deTS based on GTEx and ENCODE reference panels suggested higher tissue specificity (Figure 3) in several gastrointestinal tract and brain tissues, such as stomach, transverse colon, terminal ileum of small intestine, Peyer鈥檚 patch, and diencephalon. The hypothalamic-pituitary-adrenal axis is an important component of the GBA and could provide primary biological response to stressful stimuli, with adrenal gland especially highlighted. Details of these tissue-specific genes were also provided (eFigure 4 and eTables 11 and 12 in Supplement 1).
In addition, E-MAGMA and joint-tissue imputation analysis also confirmed the tissue specificity not only in gastrointestinal tract and brain tissues but also in other GBA-related tissues. In total, 52 pleiotropic genes (22 unique) across 8 pleiotropic loci were significantly detected in multiple GBA-related tissues in at least 2 trait pairs, for example, PSCA (8q24.3) for 3 pairwise traits (PUD-schizophrenia, PUD-BIP, and schizophrenia-AN), LPR8 (1p32.3) for 2 pairwise traits (IBS-schizophrenia and IBS-BIP), and FUT2 (19q13.33) for 2 pairwise traits (PUD-schizophrenia and PUD-ADHD). H-MAGMA analysis further suggested the cell-type specificity of these pleiotropic genes. Details of these gene-level analyses were also provided (eTables 13-15 in Supplement 1).
Synaptic and Immune-Related Mechanisms Shared Between Gastrointestinal Tract Diseases and Psychiatric Disorders
Gene set enrichment analysis identified 127 significantly enriched pathways (normalized enrichment score >2 and adjusted P鈥�<鈥�.05), including 99 Gene Ontology (GO) terms and 28 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (eTable 16 in Supplement 1). These enriched pathways are mainly involved in cell adhesion, synaptic structure and function, and immune cell differentiation. For example, homophilic cell adhesion via plasma membrane adhesion molecules (GO:0007156), which plays important roles in neural and immune mechanisms, was identified in 5 pairs of traits. A 纬-aminobutyric acid (GABA)鈥搈odifying synapse (GO:0098982), which typically functions as an inhibitory synapse using GABA as neurotransmitter, was significantly enriched in 3 pairs of traits. Notably, both T helper 1 (TH1) and TH2 cell differentiation (KEGG has04658) and TH17 cell differentiation (KEGG has04659), which involve in immunoinflammatory responses, were highlighted in 2 pairs of traits.
Multitrait Colocalization Analysis to Pinpoint Critical Gut Microbiomes
Multitrait colocalization analysis from HyPrColoc highlighted 15 pleiotropic loci (11 unique) that were colocalized to share a causal variant, supporting the significant role of gut microorganisms (Figure 2 and eTable 17 in Supplement 1). Seven of 11 unique loci were identified to be colocalized among schizophrenia, gastrointestinal tract diseases, and gut microorganisms. For example, 19q13.33 was particularly colocalized between PUD and schizophrenia not only with the prevalence of Ruminococcaceae species but with the abundance of Bacteroides species, where the index SNV rs681343 (mapped gene: FUT2) was also identified as shared causal variant.
Mendelian Randomization and Associations Between Gastrointestinal Tract Diseases and Psychiatric Disorders
Bidirectional mendelian randomization analyses using inverse-variance weighted method showed 11 significant positive associations (eFigure 5 and eTable 18 in Supplement 1). In addition, no significant bidirectional associations were detected in negative control analysis (eTable 19 in Supplement 1). The results from several alternative mendelian randomization methods were largely consistent with that from main analysis (eTables 18 and 19 in Supplement 1). Notably, among 5 pairwise traits with potential sample overlap, LHC-MR further validated the results (eTable 20 in Supplement 1).
In this genome-wide pleiotropic association study, we found extensive genome-wide genetic correlations and genetic overlaps between gastrointestinal tract diseases and psychiatric disorders. Further comprehensive analyses highlighted the pleiotropic genetic variants and loci, potential shared causal variants, pleiotropic genes, biological pathways, and a potential genetic basis associated with gut microbiome. All these findings supported the role of GBA in the shared genetic etiology underlying these 2 types of diseases.
Using multiple methods with different model assumptions could provide complementary evidence and allow deep investigation of the underlying pleiotropic associations from different perspectives. The GPA identified substantial genetic overlap between several trait pairs even without any significant genetic correlation. The mixed directions across the genomic risk loci, which indicate the existence of concordant and discordant pleiotropy, also help explain the polygenic overlap despite nonsignificant genetic correlation. PLACO analysis further determines the pleiotropic variants, followed by colocalization analysis to identify potential shared causal variants in each pleiotropic locus. Mendelian randomization analysis could detect causal trait pairs and partially reflect vertical pleiotropy. Of note, we conducted LDSC to examine potential sample overlap and performed pleiotropic analysis using PLACO as well as mendelian randomization analysis using LHC-MR to alleviate sample overlap issue. The LDSC and mendelian randomization analyses were of comparable consistency with previous studies, with detailed comparisons provided (eTables 21-23 in Supplement 1).
Overall, pleiotropic variants between gastrointestinal tract diseases and psychiatric disorders were extensively distributed, with several loci especially highlighted between certain trait pairs, such as 1q32.1 (INAVA), 19q13.33 (FUT2), 11q23.2 (NCAM1), and 1p32.3 (LRP8). Several loci previously identified to be associated with gastrointestinal tract diseases were illustrated to be potential pleiotropic loci shared with psychiatric disorders and vice versa (eDiscussion in Supplement 1). The pleiotropic genes were more likely to be enriched in GBA-related phenotypes, such as behavioral, neurological, and digestive phenotypes, as well as GBA-related tissues, especially gastrointestinal and brain tissues.
Shared genetic determinants also reflect common biological pathways, among which the TH1, TH2, and TH17 cell differentiation pathways were highlighted between IBD and schizophrenia, BIP, and AN. The TH17 cells play important roles in the development of inflammatory responses and autoimmune diseases.56-58 Previous studies suggested the significant role of dysfunction of TH17 cell differentiation and the accumulation of TH17-related cytokines in the pathological process of IBD.59,60 In addition, TH17 cells and related cytokines would provoke CNS neuroinflammation and neurotoxicity under pathological status and disrupt the blood-brain barrier and thus alter the permeability of the blood-brain barrier.61,62 Bacterially produced bile acid metabolites have been shown to inhibit TH17 cell differentiation, which may be related to the pathophysiology of IBD.63 Therefore, TH17 cell differentiation and function play a vital role in the regulation of brain-gut-microbiome axis.
Our study is not without limitations. First, we were unable to assess the causal effects of gut microbiome on gastrointestinal tract diseases or psychiatric disorders through mendelian randomization analyses since SNVs with associated gut microorganisms were hard to obtain due to the limited sample size of gut microbiome GWAS. Therefore, caution should be used on the interpretation of the role of gut microorganisms. Second, our study was restricted to European ancestry and may not generalize to other ancestries.
Given strong evidence of genetic correlation and genetic overlap between gastrointestinal tract diseases and psychiatric disorders, we found that pleiotropic genetic variants, loci, and genes were extensively distributed across the genome, with higher phenotype and tissue specificity. We highlighted some genetic determinants previously associated with gastrointestinal tract diseases that were shared with psychiatric disorders and vice versa. More importantly, shared biological mechanisms concerning immune response, synaptic structure and function, and potential gut microbiome were identified. Our findings not only support the shared genetic basis underlying GBA but provide novel insight into the intervention and treatment targets of these diseases.
Accepted for Publication: December 4, 2022.
Published Online: February 8, 2023. doi:10.1001/jamapsychiatry.2022.4974
Open Access: This is an open access article distributed under the terms of the CC-BY License. 漏 2023 Gong W et al. JAMA Psychiatry.
Corresponding Author: Zhongshang Yuan, PhD, Department of Biostatistics, School of Public Health, Cheeloo College of Medicine, Shandong University, 44, Wenhua West Road, Jinan, Shandong 250012, China (yuanzhongshang@sdu.edu.cn).
Author Contributions: Dr Yuan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Mr Gong and Ms Guo contributed equally.
Concept and design: Yuan.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Gong, Guo.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Gong, Guo, Li, L. Liu, Yan, S. Liu, Zhou.
Obtained funding: Yuan.
Administrative, technical, or material support: Wang, Xue, Yuan.
Supervision: Yuan.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by grants 82173624 and 81872712 from the National Natural Science Foundation of China (Dr Yuan), grant ZR2019ZD02 from the Natural Science Foundation of Shandong Province (Dr Yuan), the Taishan Scholar Project of Shandong Province (Dr Yuan), and Cheeloo Young Talent Program of Shandong University (Dr Yuan).
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: The Program in Complex Trait Genomics (), GWAS Catalog (), the Psychiatric Genomics Consortium (), Regensburg GEM Platform (), and GWAS Atlas () provided access to genome-wide association study data. We thank all research participants who provided DNA samples for these studies.
Data Sharing Statement: See Supplement 2.
1.Shah
听E锘�, Rezaie
听A锘�, Riddle
听M锘�, Pimentel
听M锘�. 听Psychological disorders in gastrointestinal disease: epiphenomenon, cause or consequence?听锘� 听Ann Gastroenterol. 2014;27(3):224-230.
2.Person
听H锘�, Keefer
听L锘�. 听Psychological comorbidity in gastrointestinal diseases.听锘� 听Prog Neuropsychopharmacol Biol Psychiatry. 2021;107:110209.
3.Barberio
听B锘�, Zamani
听M锘�, Black
听CJ锘�, Savarino
听EV锘�, Ford
听AC锘�. 听Prevalence of symptoms of anxiety and depression in patients with inflammatory bowel disease.听锘� 听Lancet Gastroenterol Hepatol. 2021;6(5):359-370.
4.Cryan
听JF锘�, O鈥橰iordan
听KJ锘�, Cowan
听CSM锘�,
听et al. 听The microbiota-gut-brain axis.听锘� 听Physiol Rev. 2019;99(4):1877-2013.
5.Mayer
听EA锘�, Tillisch
听K锘�, Gupta
听A锘�. 听骋耻迟/brain axis and the microbiota.听锘� 听J Clin Invest. 2015;125(3):926-938.
6.Carabotti
听M锘�, Scirocco
听A锘�, Maselli
听MA锘�, Severi
听C锘�. 听The gut-brain axis.听锘� 听Ann Gastroenterol. 2015;28(2):203-209.
7.Ma
听Q锘�, Xing
听C锘�, Long
听W锘�, Wang
听HY锘�, Liu
听Q锘�, Wang
听RF锘�. 听Impact of microbiota on central nervous system and neurological diseases.听锘� 听J Neuroinflammation. 2019;16(1):53.
8.Sullivan
听PF锘�. 听The psychiatric GWAS consortium.听锘� 听狈别耻谤辞苍. 2010;68(2):182-186.
9.Bycroft
听C锘�, Freeman
听C锘�, Petkova
听D锘�,
听et al. 听The UK Biobank resource with deep phenotyping and genomic data.听锘� 听狈补迟耻谤别. 2018;562(7726):203-209.
10.Wu
听Y锘�, Murray
听GK锘�, Byrne
听EM锘�, Sidorenko
听J锘�, Visscher
听PM锘�, Wray
听NR锘�. 听GWAS of peptic ulcer disease implicates Helicobacter pylori infection, other gastrointestinal disorders and depression.听锘� 听Nat Commun. 2021;12(1):1146.
11.Tylee
听DS锘�, Lee
听YK锘�, Wendt
听FR锘�,
听et al. 听An atlas of genetic correlations and genetically informed associations linking psychiatric and immune-related phenotypes.听锘� 听JAMA Psychiatry. 2022;79(7):667-676. doi:锘�
12.Tylee
听DS锘�, Sun
听J锘�, Hess
听JL锘�,
听et al; 23 and Me Research Team; Inflammation Working Group of the CHARGE Consortium; METASTROKE Consortium of the International Stroke Genetics Consortium; Netherlands Twin Registry; neuroCHARGE Working Group; Obsessive Compulsive and Tourette Syndrome Working Group of the Psychiatric Genomics Consortium. 听Genetic correlations among psychiatric and immune-related phenotypes based on genome-wide association data.听锘� 听Am J Med Genet B Neuropsychiatr Genet. 2018;177(7):641-657.
13.Pouget
听JG锘�, Han
听B锘�, Wu
听Y锘�,
听et al; Schizophrenia Working Group of the Psychiatric Genomics Consortium. 听Cross-disorder analysis of schizophrenia and 19 immune-mediated diseases identifies shared genetic risk.听锘� 听Hum Mol Genet. 2019;28(20):3498-3513.
14.Duncan
听LE锘�, Shen
听H锘�, Ballon
听JS锘�, Hardy
听KV锘�, Noordsy
听DL锘�, Levinson
听DF锘�. 听Genetic correlation profile of schizophrenia mirrors epidemiological results and suggests link between polygenic and rare variant (22q11.2) cases of schizophrenia.听锘� 听Schizophr Bull. 2018;44(6):1350-1361.
15.Zhang
听Y锘�, Lu
听Q锘�, Ye
听Y锘�,
听et al. 听SUPERGNOVA: local genetic correlation analysis reveals heterogeneous etiologic sharing of complex traits.听锘� 听Genome Biol. 2021;22(1):262.
16.Wang
听Q锘�, Yang
听C锘�, Gelernter
听J锘�, Zhao
听H锘�. 听Pervasive pleiotropy between psychiatric disorders and immune disorders revealed by integrative analysis of multiple GWAS.听锘� 听Hum Genet. 2015;134(11-12):1195-1209.
17.Uellendahl-Werth
听F锘�, Maj
听C锘�, Borisov
听O锘�,
听et al. 听Cross-tissue transcriptome-wide association studies identify susceptibility genes shared between schizophrenia and inflammatory bowel disease.听锘� 听Commun Biol. 2022;5(1):80.
18.Zhu
听S锘�, He
听M锘�, Liu
听Z锘�, Qin
听Z锘�, Wang
听Z锘�, Duan
听L锘�. 听Shared genetic susceptibilities for irritable bowel syndrome and depressive disorder in Chinese patients uncovered by pooled whole-exome sequencing.听锘� 听J Adv Res. 2020;23:113-121.
19.Luo
听J锘�, Xu
听Z锘�, Noordam
听R锘�, van Heemst
听D锘�, Li-Gao
听R锘�. 听Depression and inflammatory bowel disease.听锘� 听J Crohns Colitis. 2022;16(4):633-642.
20.Qian
听L锘�, He
听X锘�, Gao
听F锘�,
听et al. 听Estimation of the bidirectional relationship between schizophrenia and inflammatory bowel disease using the mendelian randomization approach.听锘� 听Schizophrenia (Heidelb). 2022;8(1):31.
21.Sadik
听A锘�, Dardani
听C锘�, Pagoni
听P锘�,
听et al; iPSYCH Autism Spectrum Disorder Working Group. 听Parental inflammatory bowel disease and autism in children.听锘� 听Nat Med. 2022;28(7):1406-1411.
22.Eijsbouts
听C锘�, Zheng
听T锘�, Kennedy
听NA锘�,
听et al; 23andMe Research Team; Bellygenes Initiative. 听Genome-wide analysis of 53 400 people with irritable bowel syndrome highlights shared genetic pathways with mood and anxiety disorders.听锘� 听Nat Genet. 2021;53(11):1543-1552.
23.Solovieff
听N锘�, Cotsapas
听C锘�, Lee
听PH锘�, Purcell
听SM锘�, Smoller
听JW锘�. 听Pleiotropy in complex traits: challenges and strategies.听锘� 听Nat Rev Genet. 2013;14(7):483-495.
24.Verbanck
听M锘�, Chen
听CY锘�, Neale
听B锘�, Do
听R锘�. 听Detection of widespread horizontal pleiotropy in causal relationships inferred from mendelian randomization between complex traits and diseases.听锘� 听Nat Genet. 2018;50(5):693-698.
25.Sivakumaran
听S锘�, Agakov
听F锘�, Theodoratou
听E锘�,
听et al. 听Abundant pleiotropy in human complex diseases and traits.听锘� 听Am J Hum Genet. 2011;89(5):607-618.
26.Ray
听D锘�, Chatterjee
听N锘�. 听A powerful method for pleiotropic analysis under composite null hypothesis identifies novel shared loci between type 2 diabetes and prostate cancer.听锘� 听PLoS Genet. 2020;16(12):e1009218.
27.Hackinger
听S锘�, Zeggini
听E锘�. 听Statistical methods to detect pleiotropy in human complex traits.听锘� 听Open Biol. 2017;7(11):170125.
28.Lee
听CH锘�, Shi
听H锘�, Pasaniuc
听B锘�, Eskin
听E锘�, Han
听B. 听PLEIO: a method to map and interpret pleiotropic loci with GWAS summary statistics.听锘� 听Am J Hum Genet. 2021;108(1):36-48.
29.Chung
听D锘�, Yang
听C锘�, Li
听C锘�, Gelernter
听J锘�, Zhao
听H锘�. 听GPA: a statistical approach to prioritizing GWAS results by integrating pleiotropy and annotation.听锘� 听PLoS Genet. 2014;10(11):e1004787.
30.Pardi帽as
听AF锘�, Holmans
听P锘�, Pocklington
听AJ锘�,
听et al; GERAD1 Consortium; CRESTAR Consortium. 听Common schizophrenia alleles are enriched in mutation-intolerant genes and in regions under strong background selection.听锘� 听Nat Genet. 2018;50(3):381-389.
31.Mullins
听N锘�, Forstner
听AJ锘�, O鈥機onnell
听KS锘�,
听et al; HUNT All-In Psychiatry. 听Genome-wide association study of more than 40 000 bipolar disorder cases provides new insights into the underlying biology.听锘� 听Nat Genet. 2021;53(6):817-829.
32.Howard
听DM锘�, Adams
听MJ锘�, Clarke
听TK锘�,
听et al; 23andMe Research Team; Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium. 听Genome-wide meta-analysis of depression identifies 102 independent variants and highlights the importance of the prefrontal brain regions.听锘� 听Nat Neurosci. 2019;22(3):343-352.
33.Demontis
听D锘�, Walters
听RK锘�, Martin
听J锘�,
听et al; ADHD Working Group of the Psychiatric Genomics Consortium (PGC); Early Lifecourse & Genetic Epidemiology (EAGLE) Consortium; 23andMe Research Team. 听Discovery of the first genome-wide significant risk loci for attention deficit/hyperactivity disorder.听锘� 听Nat Genet. 2019;51(1):63-75.
34.Nievergelt
听CM锘�, Maihofer
听AX锘�, Klengel
听T锘�,
听et al. 听International meta-analysis of PTSD genome-wide association studies identifies sex- and ancestry-specific genetic risk loci.听锘� 听Nat Commun. 2019;10(1):4558.
35.Watson
听HJ锘�, Yilmaz
听Z锘�, Thornton
听LM锘�,
听et al; Anorexia Nervosa Genetics Initiative; Eating Disorders Working Group of the Psychiatric Genomics Consortium. 听Genome-wide association study identifies eight risk loci and implicates metabo-psychiatric origins for anorexia nervosa.听锘� 听Nat Genet. 2019;51(8):1207-1214.
36.Winkler
听TW锘�, Grassmann
听F锘�, Brandl
听C锘�,
听et al. 听Genome-wide association meta-analysis for early age-related macular degeneration highlights novel loci and insights for advanced disease.听锘� 听BMC Med Genomics. 2020;13(1):120.
37.Watanabe
听K锘�, Stringer
听S锘�, Frei
听O锘�,
听et al. 听A global overview of pleiotropy and genetic architecture in complex traits.听锘� 听Nat Genet. 2019;51(9):1339-1348.
38.Amare
听AT锘�, Vaez
听A锘�, Hsu
听YH锘�,
听et al. 听Bivariate genome-wide association analyses of the broad depression phenotype combined with major depressive disorder, bipolar disorder or schizophrenia reveal eight novel genetic loci for depression.听锘� 听Mol Psychiatry. 2020;25(7):1420-1429.
39.Choquet
听H锘�, Melles
听RB锘�, Anand
听D锘�,
听et al; 23andMe Research Team. 听A large multiethnic GWAS meta-analysis of cataract identifies new risk loci and sex-specific effects.听锘� 听Nat Commun. 2021;12(1):3595.
40.Bulik-Sullivan
听BK锘�, Loh
听PR锘�, Finucane
听HK锘�,
听et al; Schizophrenia Working Group of the Psychiatric Genomics Consortium. 听LD score regression distinguishes confounding from polygenicity in genome-wide association studies.听锘� 听Nat Genet. 2015;47(3):291-295.
41.Ning
听Z锘�, Pawitan
听Y锘�, Shen
听X锘�. 听High-definition likelihood inference of genetic correlations across human complex traits.听锘� 听Nat Genet. 2020;52(8):859-864.
42.Watanabe
听K锘�, Taskesen
听E锘�, van Bochoven
听A锘�, Posthuma
听D锘�. 听Functional mapping and annotation of genetic associations with FUMA.听锘� 听Nat Commun. 2017;8(1):1826.
43.Giambartolomei
听C锘�, Vukcevic
听D锘�, Schadt
听EE锘�,
听et al. 听Bayesian test for colocalisation between pairs of genetic association studies using summary statistics.听锘� 听PLoS Genet. 2014;10(5):e1004383.
44.de Leeuw
听CA锘�, Mooij
听JM锘�, Heskes
听T锘�, Posthuma
听D锘�. 听MAGMA: generalized gene-set analysis of GWAS data.听锘� 听PLoS Comput Biol. 2015;11(4):e1004219.
45.Blake
听JA锘�, Baldarelli
听R锘�, Kadin
听JA锘�, Richardson
听JE锘�, Smith
听CL锘�, Bult
听CJ锘�; Mouse Genome Database Group. 听Mouse Genome Database (MGD): knowledgebase for mouse-human comparative biology.听锘� 听Nucleic Acids Res. 2021;49(D1):D981-D987.
46.Pei
听G锘�, Dai
听Y锘�, Zhao
听Z锘�, Jia
听P锘�. 听deTS: tissue-specific enrichment analysis to decode tissue specificity.听锘� 听叠颈辞颈苍蹿辞谤尘补迟颈肠蝉. 2019;35(19):3842-3845.
47.Consortium
听GT锘�; GTEx Consortium. 听The Genotype-Tissue Expression (GTEx) project.听锘� 听Nat Genet. 2013;45(6):580-585. doi:锘�
48.Consortium
听EP锘�; ENCODE Project Consortium. 听An integrated encyclopedia of DNA elements in the human genome.听锘� 听狈补迟耻谤别. 2012;489(7414):57-74.
49.Gerring
听ZF锘�, Mina-Vargas
听A锘�, Gamazon
听ER锘�, Derks
听EM锘�. 听E-MAGMA: an eQTL-informed method to identify risk genes using genome-wide association study summary statistics.听锘� 听叠颈辞颈苍蹿辞谤尘补迟颈肠蝉. 2021;37(16):2245-2249.
50.Zhou
听D锘�, Jiang
听Y锘�, Zhong
听X锘�, Cox
听NJ锘�, Liu
听C锘�, Gamazon
听ER锘�. 听A unified framework for joint-tissue transcriptome-wide association and mendelian randomization analysis.听锘� 听Nat Genet. 2020;52(11):1239-1246.
51.Sey
听NYA锘�, Hu
听B锘�, Mah
听W锘�,
听et al. 听A computational tool (H-MAGMA) for improved prediction of brain-disorder risk genes by incorporating brain chromatin interaction profiles.听锘� 听Nat Neurosci. 2020;23(4):583-593.
52.Yu
听G锘�, Wang
听LG锘�, Han
听Y锘�, He
听QY锘�. 听clusterProfiler: an R package for comparing biological themes among gene clusters.听锘� 听翱惭滨颁厂. 2012;16(5):284-287.
53.Foley
听CN锘�, Staley
听JR锘�, Breen
听PG锘�,
听et al. 听A fast and efficient colocalization algorithm for identifying shared genetic risk factors across multiple traits.听锘� 听Nat Commun. 2021;12(1):764.
54.Burgess
听S锘�, Butterworth
听A锘�, Thompson
听SG锘�. 听Mendelian randomization analysis with multiple genetic variants using summarized data.听锘� 听Genet Epidemiol. 2013;37(7):658-665.
55.Darrous
听L锘�, Mounier
听N锘�, Kutalik
听Z锘�. 听Simultaneous estimation of bi-directional causal effects and heritable confounding from GWAS summary statistics.听锘� 听Nat Commun. 2021;12(1):7274.
56.Yang
听J锘�, Sundrud
听MS锘�, Skepner
听J锘�, Yamagata
听T锘�. 听Targeting TH17 cells in autoimmune diseases.听锘� 听Trends Pharmacol Sci. 2014;35(10):493-500.
57.Korn
听T锘�, Bettelli
听E锘�, Oukka
听M锘�, Kuchroo
听VK锘�. 听IL-17 and TH17 cells.听锘� 听Annu Rev Immunol. 2009;27(1):485-517.
58.Zhao
听J锘�, Lu
听Q锘�, Liu
听Y锘�,
听et al. 听TH17 cells in inflammatory bowel disease: cytokines, plasticity, and therapies.听锘� 听J Immunol Res. 2021;2021:8816041.
59.Rovedatti
听L锘�, Kudo
听T锘�, Biancheri
听P锘�,
听et al. 听Differential regulation of interleukin 17 and interferon gamma production in inflammatory bowel disease.听锘� 听骋耻迟. 2009;58(12):1629-1636.
60.Fujino
听S锘�, Andoh
听A锘�, Bamba
听S锘�,
听et al. 听Increased expression of interleukin 17 in inflammatory bowel disease.听锘� 听骋耻迟. 2003;52(1):65-70.
61.Kebir
听H锘�, Kreymborg
听K锘�, Ifergan
听I锘�,
听et al. 听Human TH17 lymphocytes promote blood-brain barrier disruption and central nervous system inflammation.听锘� 听Nat Med. 2007;13(10):1173-1175.
62.Huppert
听J锘�, Closhen
听D锘�, Croxford
听A锘�,
听et al. 听Cellular mechanisms of IL-17-induced blood-brain barrier disruption.听锘� 听FASEB J. 2010;24(4):1023-1034.
63.Paik
听D锘�, Yao
听L锘�, Zhang
听Y锘�,
听et al. 听Human gut bacteria produce TH17-modulating bile acid metabolites.听锘� 听狈补迟耻谤别. 2022;603(7903):907-912.