Association between inflammatory bowel disease and osteoporosis in European and East Asian populations: exploring causality, mediation by nutritional status, and shared genetic architecture

Abstract

Background: While previous research has established an association between inflammatory bowel disease (IBD) and osteoporosis (OP), the nature of this association in different populations remains unclear.

Objective: Our study used linkage disequilibrium scores(LDSC) regression analysis and Mendelian randomization(MR) to assess the genetic correlation and causal relationship between IBD and OP in European and East Asian populations.

Methods: We performed separate genetic correlation and causal analyses for IBD and OP in European and East Asian populations, used the product of coefficients method to estimate the mediating effect of nutritional status on the causal relationship, and used multi-trait analysis to explore the biological mechanisms underlying the IBD-nutrition-OP causal pathway.

Results: Our analysis revealed a significant genetic correlation and causal relationship between IBD and OP in the European population. Conversely, no such correlation or causal relationship was observed in the East Asian population. Mediation analysis revealed a significant mediating effect of nutritional status on the causal pathway between IBD and OP in the European population. Multi-trait analysis of the IBD-nutrition-OP causal pathway identified MFAP2, ATP13A2, SERPINA1, FTO and VCAN as deleterious variants.

Conclusion: Our findings establish a genetic correlation and causal relationship between IBD and OP in the European population, with nutritional status playing a crucial mediating role.

Keywords: causality; genetic correlation; harmful variant; inflammatory bowel disease; mediating effect; multi-trait analysis; osteoporosis.

Figure 1

Research flowchart. (A) Flowchart for research on East Asian population; (B) Flowchart for research on European population.

Figure 2

Forest plots of causality. Forest plots including causality between IBD and OP of East Asian population and European population in the analyses and causality between disease subtypes CD, UC and OP. The forest plot includes nSNP, Beta95%CI, and correlation pvalues for all studies in the analysis. ^# represents the number of outliers filtered by MR-PRESSO.

Figure 3

Two-step MR analysis of forest plots. Forest diagrams include analyzing the causal relationship between IBD and IOA. The forest plot includes nSNP, Beta95%CI, and correlation pvalues for all studies in the analysis. ^# represents the number of outliers filtered by MR-PRESSO.

Figure 4

Two-step MR analysis of forest plots. Forest diagrams include analyzing the causal relationship between IOA and OP. The forest plot includes nSNP, Beta95%CI, and correlation pvalues for all studies in the analysis. ^# represents the number of outliers filtered by MR-PRESSO.

Figure 5

GWAS Manhattan chart. (A) GWAS − OP_original Manhattan map, (B) MTAG − OP_{IBD−IOA ( L )} Manhattan map, (C) MTAG − OP_{IBD−IOA ( R )} Manhattan map. Horizontal coordinates are chromosomes and vertical coordinates are -log ₁₀(p). The dashed line indicates p threshold of 1× 10⁻⁶, and the solid line indicates p threshold of 5× 10⁻⁸. Orange dots indicate SNPs 1× 10⁻⁶ <p<5× 10⁻⁸, red dots indicate p<5× 10⁻⁸, and green dots indicate lead SNPs.