Deciphering the causal links between gut microbiota and osteoporosis via two-sample Mendelian randomization

Sheng Wang1, #, Yilin Cao 2, #, Jinliang Ni 3, #, Di Du4, Lingtong Kong5, *, Hongyue Zhang5, *

1Department of Traumatic Orthopedics, Shanghai Fourth People's Hospital, School of Medicine, Tongji University, Shanghai, China

2Department of Clinical Medicine, Naval Military Medical University, Shanghai, China

3Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts, USA; Department of Urology, Shanghai Tenth People’s Hospital, School of Medicine, Tongji University, Shanghai, China

5Department of Orthopedics, Shanghai New Start Rehabilitation Hospital, Shanghai, China

6Department of Orthopedics, The First Affiliated Hospital of Naval Medical University, Changhai Hospital, Shanghai, China

#These authors contributed equally to this work and share first authorship.

*Corresponding Author: Lingtong Kong, MD, konglingtong1024@163.com; Hongyue Zhang, MD, zhanghongyue204@163.com.

Abstract

Objectives:

As global aging accelerates, the disease burden associated with age-related osteoporosis continues to grow. Beyond conventional risk factors, the gut microbiota has received considerable attention as a potential modulator; however, it remains unclear if its causal relationship with osteoporosis is age-specific. This study aimed to investigate the causal relationship between gut microbiota and osteoporosis as well as bone mineral density across different age groups using two-sample Mendelian randomization analysis, with a focus on age-specific differences in association patterns.

Methods:

A two-sample Mendelian randomization design was performed in this study. Exposure data were derived from the MiBioGen consortium’s genome-wide association studies (GWAS), covering 211 gut microbial taxonomic units. Outcome data included GWAS-aggregated data for osteoporosis and bone mineral density across all ages, as well as in two age stages: 45–60 years and > 60 years. We selected single nucleotide polymorphisms significantly associated with microbiota as instrumental variables (P < 5 × 10⁻⁵, linkage disequilibrium R² < 0.001, distance > 10,000 kb, F-statistic > 10). Inverse variance weighting was the primary method for causal estimation, supplemented by weighted median estimation and Mendelian randomization-Egger regression. Multicollinearity, heterogeneity, and sensitivity analyses were performed using MR-Egger intercept tests, Cochran’s Q tests, leave-one-out validation, and Mendelian Randomization Pleiotropy RESidual Sum and Outlier.

Results:

The analysis identified 33 gut microbial taxonomic units that have potential causal associations with osteoporosis or bone mineral density, most exhibiting age dependency. Among these, six taxa were directly associated with osteoporosis risk: Peptococcus (β = 0.0015, P = 0.001), Terrisporobacter (β = 0.0014, P = 0.005), and Dorea (β = 0.0013, P = 0.041) as risk factors; Veillonella (β = –0.0011, P = 0.021), Euryarchaeota (β = –0.0008, P = 0.037), and Bilophila (β = –0.0013, P = 0.039) were protective factors. In age-stratified analyses, over 27 bacterial groups showed differential associations with bone mineral density across age cohorts: for example, Eubacterium coprostanoligenes group correlated with higher bone mineral density in both all ages and > 60 years groups, while Oxalobacter exhibited a more pronounced negative association with bone mineral density in the > 60 years cohort; In the 45–60 age group, Romboutsia and other bacterial groups showed specific negative correlations.

Conclusion:

This study provides genetic evidence for causal associations between gut microbiota and osteoporosis/bone mineral density, and reveals age-specific patterns of these associations. The identified age-associated microbiota may influence bone health through diverse mechanisms, including regulation of lipid metabolism and calcium homeostasis, with their effects evolving throughout the lifespan. These findings emphasize the importance of integrating age-related factors into gut microbiota intervention strategies, which could offer valuable targets for designing age-appropriate methods to prevent and treat osteoporosis.

通过双样本孟德尔随机化解析肠道菌群与骨质疏松症的因果关联

摘要

目的：随着全球老龄化进程加速，与年龄相关的骨质疏松症疾病负担持续加重。除传统风险因素外，肠道微生物群作为潜在调节因子备受关注；但其与骨质疏松症的因果关系是否具有年龄特异性尚不明确。本研究旨在通过双样本孟德尔随机化分析，探究不同年龄组肠道微生物群与骨质疏松症及骨矿物质密度的因果关联，重点关注关联模式的年龄特异性差异。

方法：本研究采用双样本孟德尔随机化设计。暴露数据源自MiBioGen联盟的全基因组关联研究（GWAS），涵盖211个肠道微生物分类单元。结局数据包括所有年龄段及45-60岁与60岁以上两个年龄阶段的骨质疏松症和骨矿物质密度GWAS聚合数据。我们选取与微生物群显著相关的单核苷酸多态性作为工具变量（P < 5×10⁻⁵，连锁不平衡R² < 0.001，距离>10,000 kb，F统计量>10）。因果估计主要采用逆方差加权法，辅以加权中位数估计法和孟德尔随机化-Egger回归分析。通过MR-Egger截距检验、Cochran Q检验、逐一剔除验证及孟德尔随机化多效性残差和异常值分析，实施多重共线性、异质性及敏感性分析。

结果：分析识别出33个与骨质疏松症或骨矿物质密度存在潜在因果关联的肠道微生物分类单元，其中多数表现出年龄依赖性。其中六类分类单元直接关联骨质疏松风险：肽球菌（β = 0.0015, P = 0.001）、土孢杆菌（β = 0.0014, P = 0.005）和多雷菌（β = 0.0013, P = 0.041）为风险因子； 而维氏乳杆菌（β = –0.0011, P = 0.021）、广古菌门（β = –0.0008, P = 0.037）和双歧杆菌（β = –0.0013, P = 0.039）则表现为保护性因子。在年龄分层分析中，超过27个细菌群在不同年龄组呈现差异性关联：例如粪杆菌属与所有年龄组及60岁以上组的骨矿物质密度呈正相关，而草酸杆菌在60岁以上组与骨矿物质密度呈更显著的负相关； 在45-60岁组中，Romboutsia属及其他细菌群呈现特异性负相关。

结论：本研究为肠道菌群与骨质疏松症/骨矿物质密度之间的因果关联提供了遗传学证据，并揭示了这些关联的年龄特异性模式。所识别的年龄相关菌群可能通过多种机制影响骨骼健康，包括调节脂质代谢和钙稳态，其效应在生命周期中持续演变。这些发现强调将年龄相关因素纳入肠道微生物干预策略的重要性，这为设计预防和治疗骨质疏松症的适龄方法提供了宝贵靶点。