The role of inflammatory cytokines and blood metabolites in osteoporosis: A Mendelian randomization study on pathways and therapeutic targets

Liu Yuxi, Dongrui Feng, Hai Jiang, Fabing Tang, Liming Wu, Sui‐Lin Mo, Jianhong TaoMedicine 2025 · DOI

TL;DR

This study uses Mendelian randomization to uncover causal links between inflammatory cytokines, blood metabolites, and osteoporosis (OP). It identifies three inflammatory cytokines that contribute to OP and pinpoints specific metabolites that mediate these effects, offering potential new therapeutic targets for OP prevention and treatment.

What Problem Does This Paper Address?

Osteoporosis (OP) is a major global health concern, characterized by reduced bone mineral density and increased fracture risk, particularly in aging populations. While inflammation and metabolic dysregulation are suspected contributors, the precise causal pathways remain unclear. Understanding these mechanisms is crucial for developing effective prevention and treatment strategies. This paper aims to untangle these complex relationships by investigating the causal roles of inflammatory cytokines and blood metabolites in OP development.

Methods

The core methodology employed is Mendelian Randomization (MR), a powerful genetic epidemiology technique that uses genetic variants as instrumental variables (IVs) to infer causal relationships between an exposure and an outcome. This study utilizes a two-sample MR approach, meaning summary statistics from different genome-wide association studies (GWAS) are combined.

Key methodological components include:

  • Bidirectional MR: To assess causal relationships in both directions (cytokines on OP, and OP on cytokines).
  • Two-step MR (TSMR): To identify potential mediating metabolites. This involves two stages: first, assessing the effect of metabolites on OP, and second, assessing the effect of inflammatory cytokines on metabolites.
  • Multivariable MR (MVMR): To validate the mediation findings from TSMR and estimate controlled direct effects, accounting for potential confounding and correlated exposures.
  • Sensitivity Analyses: Rigorous checks for potential violations of MR assumptions, including:
  • MR-Egger regression: To detect and adjust for horizontal pleiotropy (where genetic variants affect the outcome through pathways other than the exposure).
  • Weighted Median method, Weighted Mode, Maximum Likelihood: Alternative MR methods to assess robustness.
  • MR-PRESSO: To identify and remove outlier genetic variants that might bias results.
  • Leave-one-out analysis: To assess the influence of individual genetic variants.
  • Cochran Q statistic: To assess heterogeneity among genetic variants.
  • False Discovery Rate (FDR) correction: Applied to account for multiple testing in large-scale genomic analyses.

Datasets

The study leverages publicly available summary statistics from large-scale GWAS:

  • Inflammatory Cytokines: GWAS data from 8293 healthy Finnish participants.
  • Blood Metabolites: Summary statistics for 1400 blood metabolites from the GWAS Catalog.
  • Osteoporosis (OP): Data from the FinnGen Consortium's R12 release, specifically focusing on the "OP" phenotype.

Key Insights and Ideas

  • Inflammatory Cytokines Drive OP, Not Vice Versa: A significant finding is that while certain inflammatory cytokines causally contribute to osteoporosis, the reverse is not true – OP does not appear to causally influence these cytokines. This suggests a unidirectional causal pathway from inflammation to bone health decline.
  • Metabolites as Key Mediators: The study highlights that the link between inflammatory cytokines and OP is not always direct but is often mediated by specific blood metabolites. This reveals a crucial intermediate step in the disease pathway.
  • Specific Cytokines and Metabolites Identified: The research pinpoints CTACK and TRAIL as key inflammatory cytokines associated with OP. Furthermore, it identifies Oxalate, 2,4-di-tert-butylphenol, and the Phosphate to linoleoyl-arachidonoyl-glycerol ratio as metabolites that mediate the effects of CTACK on OP. TRAIL's effects are linked to Trans 3,4-methyleneheptanoate and X-11478.
  • Therapeutic Potential: By identifying these specific cytokines and mediating metabolites, the study offers concrete targets for future drug development and personalized treatment strategies for osteoporosis.

Key Results

  • Three inflammatory cytokines were found to be causally associated with OP.
  • IL-18 showed the strongest causal effect among the identified cytokines, increasing OP risk (OR = 1.086, 95% CI = 1.024–1.152, P = .006).
  • In two-step MR, several metabolites were significantly associated with OP, including Spermidine to (N(1) + N(8))-acetylspermidine ratio, Oxalate, and 2,4-di-tert-butylphenol.
  • MVMR analysis confirmed that the association between CTACK and OP was mediated by:
  • Spermidine to (N(1) + N(8))-acetylspermidine ratio (mediated proportion: -15.6%, P = .001).
  • Oxalate (mediated proportion: 7.8%, P = .014).
  • 2,4-di-tert-butylphenol (mediated proportion: -15.6%, P = .005).
  • Sensitivity analyses indicated robust findings with no significant evidence of horizontal pleiotropy or substantial heterogeneity.

Limitations and Future Work

The authors acknowledge several limitations:

  • Linearity Assumption: MR assumes linear relationships, which may not fully capture the complex biological interactions between cytokines and metabolites.
  • Generalizability: The study primarily used GWAS data from individuals of European ancestry, necessitating validation in more diverse populations.
  • Sex-Specific Effects: Sex-specific GWAS estimates were not consistently available, limiting the ability to fully explore sex differences in causal effects.
  • FDR Correction: While appropriate for large-scale studies, some marginal q-values warrant further independent validation.

Future work should focus on:

  • Integrating non-linear MR models and interaction analyses to better understand complex biological systems.
  • Replicating findings in diverse cohorts to ensure generalizability.
  • Conducting sex-specific analyses to explore gender differences in OP pathways.
  • Further validation of identified metabolites and cytokines through experimental studies and clinical trials to confirm their therapeutic potential.

Summary updated

Abstract

The precise mechanisms underlying the development of osteoporosis (OP) remain unclear, but evidence suggests that inflammatory cytokines and blood metabolites play a significant role. We aim to investigate, as a strictly genomic study based on genome-wide association studies (GWAS) summary statistics, the causal relationships between inflammatory cytokines, blood metabolites, and OP by using Mendelian randomization (MR) analysis and uncover potential therapeutic targets. We utilized inflammatory cytokines from a GWAS summary containing 8293 healthy participants, 1400 blood metabolites from GWAS Catalog, and OP data from the FinnGen repository, all of which are sourced from the largest GWAS conducted to date. Employing bidirectional MR analyses, we investigated the causal relationships between inflammatory cytokines and OP. Additionally, we conducted 2 mediation analyses, two-step MR and multivariable MR (MVMR), to identify potential mediating metabolites. Three inflammatory cytokines were causally associated with OP, while OP did not have a significant causal effect on them. In the two-step MR analysis, CTACK and TRAIL, along with metabolites Spermidine to (N(1) + N(8))-acetylspermidine ratio, Oxalate, 2,4-di-tert-butylphenol, Phosphate to linoleoyl-arachidonoyl-glycerol (18:2-20:4) ratio, Trans 3,4-methyleneheptanoate and X-11478, were all significantly associated with OP (all P < .05). MVMR analysis revealed that the associations between CTACK and OP were mediated by Spermidine to (N(1) + N(8))-acetylspermidine ratio (-15.6%, P = .001), Oxalate (7.8%, P = .014), and 2, 4-di-tert-butylphenol (-15.6%, P = .005). The present MR study offers evidence supporting the causal relationships between several specific inflammatory cytokines and OP, as well as identifying potential mediating metabolites.
BibTeX
@inproceedings{yuxi2025role,
  author = {Liu Yuxi and Dongrui Feng and Hai Jiang and Fabing Tang and Liming Wu and Sui‐Lin Mo and Jianhong Tao},
  title = {{The role of inflammatory cytokines and blood metabolites in osteoporosis: A Mendelian randomization study on pathways and therapeutic targets}},
  booktitle = {Medicine},
  year = {2025},
  doi = {10.1097/md.0000000000046664},
}
Figure