Potential association of rheumatic diseases with bone mineral density and fractures: a bi-directional mendelian randomization study

Abstract

Background: Previous studies have implicated rheumatoid arthritis as an independent risk factor for bone density loss. However, whether there is a causal relationship between rheumatic diseases and bone mineral density (BMD) and fractures is still controversial. We employed a bidirectional Mendelian analysis to explore the causal relationship between rheumatic diseases and BMD or fractures.

Methods: The rheumatic diseases instrumental variables (IVs) were obtained from a large Genome-wide association study (GWAS) meta-analysis dataset of European descent. Analyses were performed for the three rheumatic diseases: ankylosing spondylitis (AS) (n = 22,647 cases, 99,962 single nucleotide polymorphisms [SNPs]), rheumatoid arthritis (RA) (n = 58,284 cases, 13,108,512 SNPs), and systemic lupus erythematosus (SLE) (n = 14,267 cases, 7,071,163 SNPs). Two-sample Mendelian randomization (MR) analyses were carried out by using R language TwoSampleMR version 0.5.7. The inverse-variance weighted (IVW), MR-Egger, and weighted median methods were used to analyze the causal relationship between rheumatic diseases and BMD or fracture.

Results: The MR results revealed that there was absence of evidence for causal effect of AS on BMD or fracture. However, there is a positive causal relationship of RA with fracture of femur (95% CI = 1.0001 to 1.077, p = 0.046), and RA and fracture of forearm (95% CI = 1.015 to 1.064, p = 0.001). SLE had positive causal links for fracture of forearm (95% CI = 1.004 to 1.051, p = 0.020). Additionally, increasing in heel bone mineral density (Heel-BMD) and total bone mineral density (Total-BMD) can lead to a reduced risk of AS without heterogeneity or pleiotropic effects. The results were stable and reliable. There was absence of evidence for causal effect of fracture on RA (95% CI = 0.929 to 1.106, p = 0.759), and fracture on SLE (95% CI = 0.793 to 1.589, p = 0.516).

Conclusions: RA and SLE are risk factors for fractures. On the other hand, BMD increasing can reduce risk of AS. Our results indicate that rheumatic diseases may lead to an increased risk of fractures, while increased BMD may lead to a reduced risk of rheumatic diseases. These findings provide insight into the risk of BMD and AS, identifying a potential predictor of AS risk as a reduction in BMD.

Keywords: Bone mineral density; Fracture; Mendelian randomization; Rheumatic diseases.

Fig. 1

The schematic diagram of Mendelian randomization (MR). Three assumptions should be met, as follows: Assumption 1: The SNPs should be closely related to exposures; Assumption 2: The IVs selected are supposed to be independent of confounders; Assumption 3: SNPs should　influence the outcomes just through the exposure. (IVs, instrumental variables; SNPs, single-nucleotide polymorphisms)

Fig. 2

The scatter plot for MR analyses of causal associations between rheumatic diseases AS and BMD. (A) The scatter plot of the relationship between Heel-BMD and AS. (B) The scatter plot of the relationship between Total-BMD and AS.

Fig. 3

The scatter plot for MR analyses of causal associations between rheumatic diseases RA and fractures. (A) The scatter plot of the relationship between RA and fracture of femur. (B) The scatter plot of the relationship between RA and fracture of forearm

Fig. 4

The scatter plot for MR analyses of causal associations between rheumatic diseases SLE and fracture of forearm