Skip to main content
NCBI home page
Animal Models and Experimental Medicine logoLink to Animal Models and Experimental Medicine
. 2026 Apr 3;9(5):932–946. doi: 10.1002/ame2.70187

Genetic prediction of blood cell reactivity and its potential causal influence on bone continuity and density disorders

Zhiqin Deng 1, Zhe Zhao 1, Wenting Jiang 2, Xiaoqiang Chen 1, Jianquan Liu 1, Xu Tao 1, Zhengyang Lin 3,4, Zhenhan Deng 3,4,✉, Wencui Li 1,✉
PMCID: PMC13331561  PMID: 41930994

Abstract

Background

This study aimed to investigate the potential causal relationship between genetically predicted human blood cell (HBC) reactivity and disorders of bone continuity, density, and structure.

Methods

We analyzed summary‐level GWAS data for 91 HBC traits and two bone‐related outcomes from publicly available sources, employing the inverse‐variance weighted (IVW) method as the principal Mendelian randomization (MR) technique. The sensitivity analyses comprised MR‐Egger regression and MR‐PRESSO.

Results

MR analysis identified suggestive associations between red blood cell (RBC) perturbation response (Pam3CSK4), neutrophil perturbation response (side fluorescence coefficient of variation of neutrophil 4), Neutrophil perturbation response (colchicine), Monocyte perturbation response (TMAO perturbation) and bone continuity. The MR results are: [β: −0.13, odds ratio (OR): 0.88; 95% confidence interval (CI): 0.77, 0.99; p = 0.040], [β: 0.11, OR: 1.12, 95% CI: 1.02, 1.23; p = 0.016], [β: −0.11, OR: 0.90, 95% CI: 0.81, 0.99; p = 0.029] and [β: −0.04, OR: 0.96, 95% CI: 0.92, 0.99; p = 0.023]. In addition, Neutrophil perturbation response (forward scatter median of neutrophil 1), Unknown cell population perturbation response (nigericin) and other disorders of bone density and structure are also potential causal factors, with MR Result [β: 0.21, OR: 1.24, 95%CI: 1.01, 1.51; p = 0.034], [β: 0.03, OR: 1.03, 95% CI: 1.00, 1.06; p = 0.042]. Reverse Mendelian randomization sensitivity analysis showed a potential bidirectional association between specific HBC features and bone‐related outcomes.

Conclusions

This exploratory study offer valuable preliminary insights into the blood cell functional reactivity and bone health. The findings, while requiring independent validation, highlight plausible biological pathways for further elucidation.

Keywords: disorders of bone continuity, disorders of bone density and structure, human blood cell, Mendelian randomization, osteoporosis

We applied Mendelian randomization to explore causal links between blood cell traits and skeletal disorders. Using genetic instruments from large‐scale summary statistics, we assessed effects on bone continuity, density, and structural integrity. Sensitivity and reverse analyses confirmed robust associations, highlighting potential shared biological pathways between hematologic profiles and bone health.

graphic file with name AME2-9-932-g003.jpg

1. INTRODUCTION

Bone continuity or bone density and structure disorders, primarily characterized by disruptions in the structural integrity of bone, can arise from various factors including trauma, metabolic diseases, and genetic predispositions. The mechanisms underlying these disorders involve complex biological processes that govern bone healing and remodeling. Fractures, which represent a significant form of bone continuity disorder, disrupt the mechanical stability of bone, leading to pathological mobility and loss of support. 1 , 2 , 3 Disorders of bone continuity have a significant genetic basis. Mutations in genes such as LRP5, which is involved in bone homeostasis, can lead to conditions characterized by abnormal bone mass and strength. 4 , 5 These genetic variations can predispose individuals to fractures and affect their healing capacity.

The pathophysiology of bone continuity disorders involves an intricate network of interactions in the bone marrow, where blood cells (particularly hematopoietic stem cells, HSCs), osteoblasts, and the microenvironment are all key players. Hematopoietic disorders, particularly those affecting the bone marrow, can significantly influence bone health and integrity, leading to various bone continuity disorders such as osteoporosis and osteopenia. Hematopoietic disorders, including myelodysplastic syndromes and leukemias, have been shown to disrupt normal bone homeostasis. Research indicates that dysfunction within bone progenitor cells has been linked to the onset of myelodysplasia and secondary leukemia. This evidence highlights how perturbations in osteolineage cells can serve as a catalyst for multifaceted blood disorders while also impairing bone integrity. 6 , 7 This is further supported by evidence showing that low hemoglobin levels in patients with chronic myeloid leukemia (CML) can have profound secondary effects on bone health, suggesting a direct link between hematopoietic dysfunction and bone disorders. 8 , 9 , 10 Furthermore, the function of erythropoietin in modulating erythropoiesis and osteogenesis is well documented. Erythropoietin not only stimulates the production of red blood cells but also influences osteoblast activity, thereby playing a crucial role in maintaining bone mass. 11 , 12 Bone health relies on the homeostasis between osteoclasts and osteoblasts; disturbances from hematopoietic disorders can rapidly lead to bone loss. 13 , 14 Within the bone marrow, a specialized microenvironment comprised of diverse niche cells provides critical support for hematopoietic stem cell (HSC) maintenance and lineage commitment. Alterations in this niche, as observed in acute myeloid leukemia (AML), can lead to impaired hematopoiesis and contribute to cytopenias, which further exacerbate bone health issues. 15 For example, AML has been shown to impede the differentiation of normal HSCs, leading to a failure in producing adequate blood cells, which can indirectly affect bone health by reducing the mechanical loading that bones require for maintenance. 15 Additionally, the interaction between osteoblasts and hematopoietic cells is crucial for maintaining bone integrity. Shiozawa and Taichman emphasized that osteoblasts provide a supportive niche for HSCs, and their ablation can reduce HSC self‐renewal while accelerating leukemia development. 16 , 17 Recently, large‐scale GWAS has moved beyond baseline hematological indices to profile cellular responses to experimental perturbations, revealing the genetic architecture of functional immune and metabolic pathways. 18 Large‐scale GWAS of these perturbation‐response traits providing genetic instruments that reflect how an individual's cells react to challenge, rather than simply how many cells are present at rest. We hypothesized that leveraging these genetic instruments for blood cell reactivity would provide an insightful approach to investigate their causal influence on bone health. This approach allows us to probe whether genetically programmed differences in specific cellular response pathways play a causal role in the etiology of bone continuity and density disorders.

2. METHODS

2.1. Study design overview

This study employed two‐sample MR to explore the potential causal relationship between 91 types of blood cell disturbance response phenotypes and two bone disorders.

2.2. Exposed and outcome data sources

We used the following two GWAS summary statistic datasets from the FinnGen biobank (Release 9) as outcome data: ‘Disorders of continuity of bone (BONECONTINUITY)’ and ‘Other disorders of bone density and structure (BONEDENSOTH)’. These datasets were publicly accessed via their respective manifest entries in the FinnGen R9 summary statistics repository (manifest file: FinnGen R9_manifest.tsv). The GWAS summary statistics include 2543 cases and 358 014 controls of European ancestry for BONECONTINUITY (FinnGen R9 release, ID: finngen_R9_M13_BONECONTINUITY), and 677 cases and 358 014 controls for BONEDENSOTH (FinnGen R9 release, ID: finngen_R9_M13_BONEDENSOTH). These analyses were based on the GRCh38/hg38 genome build. We sourced them directly from the FinnGen R9 release to ensure consistency and traceability. When searching for SNPs from the results, we did not use proxy SNPs primarily because the FinnGen biobank contains a sufficient number of SNPs.

The exposed data use the latest published 91 types of human blood cells reactivity, derived from a large‐scale perturbational phenotyping study. 18 This dataset is based on the GRCh37/hg19 genome build. The open data were downloaded from the GWAS Catalog (https://www.ebi.ac.uk/gwas/), and the range of the GWAS Catalog database is (GCST90257015‐GCST90257105). The author utilized human peripheral blood cells, physical, chemical, and pharmacological perturbations, as well as flow cytometry‐based functional readings, to reveal potential cellular processes. The genetic instruments for our MR analysis thus represent variants associated with inter‐individual differences in cellular responses to these specific stimuli, rather than baseline cell counts. For more information on the exposure dataset, please refer to Table S1. Given the distinct study designs, populations, and recruitment frameworks of the FinnGen biobank and the blood cell perturbation study, we assume no substantial sample overlap between exposure and outcome datasets.

2.3. Selection of genetic instruments

In the entire Mendelian randomization analysis, the same criteria were used to determine the genetic instrumental variables for each inflammatory factor. A genome‐wide significance threshold was set at p < 1 × 10−5, and SNPs with linkage disequilibrium (LD) γ 2 < 0.001 and a genetic distance within 10 000 kb were excluded. The F‐statistic was calculated for all SNPs, and to ensure the strength of the instrumental variables, SNPs with F values <10 were excluded. The detailed F‐statistic for each SNP‐instrument across all traits is available in Table S2. Following clumping (r 2 < 0.001, distance <10 000 kb), the number of independent SNPs retained for the MR analysis ranged from 6 to 28 across the 91 exposure traits, with a median of 14.5 SNPs per trait (Table S2).

2.4. Mendelian randomization analysis

To investigate the potential causal relationships between genetically predicted blood cell reactivity traits and bone disorders, we performed two‐sample Mendelian randomization (MR) analyses. The primary analysis method was the inverse‐variance weighted (IVW) model under a multiplicative random‐effects framework, which provides a consistent causal estimate when all instrumental variables (IVs) are valid. We additionally employed several other MR methods. These included the MR‐Egger regression, which can provide a valid causal estimate even when all IVs are invalid (subject to the Instrument Strength Independent of Direct Effect assumption), and the weighted median method, which yields consistent estimates if at least 50% of the weight comes from valid IVs. Given the exploratory nature of this study and the high correlation among the 91 perturbation‐response exposure traits, we did not apply traditional multiple testing corrections (e.g., Bonferroni) across all 182 tests. All results are presented as preliminary evidence intended to generate hypotheses for future validation in larger, independent studies.

2.5. Sensitivity and validation analyses

To ensure the validity and reliability of our MR findings, we performed the sensitivity analyses. First, Cochran's Q statistic was calculated to assess heterogeneity among the causal estimates from individual SNPs; significant heterogeneity (p < 0.05) would suggest potential violation of the MR assumptions. Second, we evaluated horizontal pleiotropy using the MR‐Egger intercept test and the MR‐PRESSO global test. A significant intercept (p < 0.05) or a significant MR‐PRESSO global test would indicate the presence of directional pleiotropy that could bias the IVW estimate. Third, to probe the possibility of reverse causality, where genetic predisposition to bone disorders might influence blood cell traits, we conducted reverse MR analyses, treating the bone disorder GWAS as exposure and the blood cell reactivity traits as outcomes. It is acknowledged that this analysis is underpowered for definitive conclusions due to sample size limitations. Finally, a leave‐one‐out sensitivity analysis was performed by iteratively removing each SNP to determine if any single variant was disproportionately driving the observed association. These analyses collectively aimed to scrutinize the robustness and directionality of the identified associations.

2.6. Statistical power calculation

Statistical power for MR analyses was assessed using an online power calculator for binary outcomes (https://www.mranalysis.cn/plugins/PowerCalculator/). For each exposure‐outcome pair, the minimal detectable odds ratio (OR) at 80% power and a type I error rate (α) of 0.05 was calculated based on the outcome sample size (N), the case–control ratio (K), and the combined R 2 (the proportion of variance in the exposure explained by all instrumental variants). For reverse MR analyses, the combined R 2 for bone density was extremely low (<0.001), resulting in implausibly large minimal detectable ORs (>2.0); thus, these analyses were considered underpowered to detect plausible effect sizes. We acknowledge that conducting numerous statistical tests increases the risk of false‐positive findings (Type I error). Therefore, all reported associations are considered preliminary. The primary aim was to identify potential signals for future validation in larger, adequately powered studies, rather than to provide definitive causal evidence. We calculated the statistical power for the primary IVW analysis, based on the proportion of variance explained by the instruments (R 2), sample size of the outcome, and assumed OR. Power was estimated at the conventional α = 0.05 level.

2.7. Statistical analysis

All statistical analyses were performed using R software (version 4.3.2; www.r‐project.org). The primary Mendelian randomization analyses were conducted using the TwoSampleMR package (version 0.5.8). Sensitivity analyses, including MR‐Egger regression and the assessment of horizontal pleiotropy, were performed using the MRPRESSO package (version 1.0) via the run_mr_presso function within TwoSampleMR. The Cochran's Q statistic for heterogeneity was calculated using functions within the TwoSampleMR package. Data processing and iterative analyses were implemented using the foreach package (version 1.5.2). Visualization of results, including scatter plots, forest plots, funnel plots, and leave‐one‐out sensitivity plots, was performed using the ggplot2 package (version 4.0.1). Additional specialized forest plots were generated using the forestploter package (version 1.1.2) with support from the grid package (version 4.3.2). All R packages are publicly available from the Comprehensive R Archive Network (CRAN) or Bioconductor. The main causal estimates were derived from the random‐effects IVW approach. Heterogeneity of the aggregated estimates was evaluated using Cochran's Q statistic, with a p < 0.05 suggesting its presence. Horizontal pleiotropy was tested via the MR‐Egger intercept and MR‐PRESSO global tests, with a significance level of p < 0.05. Bonferroni correction was applied based on the types of inflammatory factors. Finally, a leave‐one‐out method was used to assess the effect of individual SNPs. The significance level was set at p < 0.05.

3. RESULTS

3.1. Results of Mendelian randomization on HBC for BONECONTINUITY, or BONEDENSOTH

The F‐statistic ranges from 19.69 to 79.81, all above 10 (Table S2), indicating a low likelihood of weak instrument bias. The MR estimates for the genetic correlation of 91 types of HBC on bone continuity (BONECONTINUITY), bone density, and bone structural disorders (BONEDENSOTH) using different methods are shown in Table 1 and Figure 1. The results indicate that genetic predictions of blood cells are correlated with bone continuity, bone density, and bone structural disorders. Among these, the Red blood cell perturbation response (Pam3CSK4 perturbation), Neutrophil perturbation response (side fluorescence coefficient of variation of neutrophil 4), Neutrophil perturbation response (colchicine perturbation), and Monocyte perturbation response (TMAO perturbation) influence bone continuity, while the Neutrophil perturbation response (forward scatter median of neutrophil 1) and Unknown cell population perturbation response (nigericin perturbation) regulate bone density and bone structural disorders. We found that using the IVW method, the genetic prediction of RBC perturbation response (Pam3CSK4 perturbation) and the risk of bone continuity are negatively correlated [β: −0.13, odds ratio (OR): 0.88; 95% confidence interval (CI): 0.77, 0.99; p = 0.040]. The neutrophil perturbation response (side fluorescence coefficient of variation of neutrophil 4) is positively correlated with the risk of bone continuity [β: 0.11, OR: 1.12, 95% CI: 1.02, 1.23; p = 0.016]. Neutrophil perturbation response (colchicine perturbation) is negatively associated with the risk of bone continuity [β: −0.11, OR: 0.90, 95% CI: 0.81, 0.99; p = 0.029]. Monocyte perturbation response (TMAO perturbation) is negatively associated with the risk of bone continuity [β: −0.04, OR: 0.96, 95% CI: 0.92, 0.99; p = 0.022]. In addition, the IVW method obtained a positive correlation between the genetic prediction of Neutrophil perturbation response (forward scatter median of neutrophil 1) and bone mineral density and bone structure disturbance [β: 0.21, OR: 1.24, 95% CI: 1.01, 1.51; p = 0.034]. The Unknown cell population perturbation response (nigericin perturbation) also showed a positive correlation with bone mineral density and bone structure disturbance [β: 0.03, OR: 1.03, 95% CI: 1.00, 1.06; p = 0.042]. However, after correction using the adjusted p‐value, no potential causal influence of blood cell perturbation response phenotypes on bone continuity, bone density, and bone structural disorder was found in this part. Heterogeneity tests show RBC (Pam3CSK4 perturbation)—BONECONTINUITY (IVW, Q(df) 4.82 (5), p = 0.438), Neutrophil perturbation response—BONECONTINUITY (IVW, Q(df) 9.21 (10), p = 0.513), Neutrophil perturbation response (colchicine perturbation)—BONECONTINUITY (IVW, Q(df) 4.05 (5), p = 0.543), Monocyte perturbation response (TMAO perturbation)—BONECONTINUITY (IVW, Q(df) 12.41 (12), p = 0.413). The heterogeneity test results for the outcome BONEDENSOTH showed a Neutrophil perturbation response‐BONEDENSOTH (IVW, Q(df) 3.99 (9), p = 0.912) and an Unknown wn cell population perturbation response–BONEDENSOTH (IVW, Q(df) 3.05 (9), p = 0.962). Sensitivity analysis results show that in all MR‐Egger regressions, there is no evidence of directional pleiotropy (all intercepts are close to zero, p‐values are all >0.05, Figure 2A, B, Table 1, and Figure S1).

TABLE 1.

MR estimates from different methods of assessing the causal effect of HBC on bone discontinuity or disorders of bone density and structure.

Exposure Outcome SNP Heterogeneity tests Directional horizontal pleiotropy test MR results Average F MR‐PRESSO
Methods Cochran's Q (df) p MR‐Egger intercept (p) Method Beta p Global test p value
Human blood cell perturbation response phenotypes BONECONTINUITY Red blood cell perturbation response (Pam3CSK4 perturbation) MR Egger 4.80 (4) 0.308 9.55E−03 (0.90) MR Egger −0.17 0.604 24.33 0.479
Inverse variance weighted 4.82 (5) 0.438 Inverse variance weighted −0.13 0.040*
Neutrophil perturbation response (side fluorescence coefficient of variation of neutrophil 4) MR Egger 9.19 (9) 0.420 7.70E−03 (0.89) MR Egger 0.07 0.800 24.61 0.554
Inverse variance weighted 9.21 (10) 0.513 Inverse variance weighted 0.11 0.016*
Neutrophil perturbation response (colchicine perturbation) MR Egger 2.68 (4) 0.613 −6.76E−02 (0.31) MR Egger 0.12 0.589 35.80 0.576
Inverse variance weighted 4.05 (5) 0.543 Inverse variance weighted −0.11 0.029*
Monocyte perturbation response (TMAO perturbation) MR Egger 11.80 (11) 0.379 −2.78E−02 (0.47) MR Egger 0.00 0.982 24.25 0.456
Inverse variance weighted 12.41 (12) 0.413 Inverse variance weighted −0.04 0.022*
BONEDENSOTH Neutrophil perturbation response (forward scatter median of neutrophil 1) MR Egger 3.97 (8) 0.860 −1.31E−02 (0.90) MR Egger 0.28 0.571 26.96 0.938
Inverse variance weighted 3.99 (9) 0.912 Inverse variance weighted 0.21 0.034*
Unknown cell population perturbation response (nigericin perturbation) MR Egger 2.64 (8) 0.955 5.04E−02 (0.54) MR Egger 0.00 0.988 22.19 0.960
Inverse variance weighted 3.05 (9) 0.962 Inverse variance weighted 0.03 0.042*
Open in a new tab
*

p < 0.05 indicates statistical significance.

FIGURE 1.

FIGURE 1

Open in a new tab

Mendelian randomization estimates for the associations between 91 human blood cell (HBC) reactivity traits and bone disorders. Estimates were derived using the inverse‐variance weighted (IVW) method. CI, confidence interval; IVW, inverse‐variance weighted; MR, Mendelian randomization; OR, odds ratio; SNP, single nucleotide polymorphism. The forest plot displays odds ratios (squares) and 95% CIs (horizontal lines) for the potential causal effect of each genetically predicted HBC reactivity trait on disorders of bone continuity (BONECONTINUITY) and other disorders of bone density and structure (BONEDENSOTH). Data sources: Exposure data are from the GWAS catalog (blood cell reactivity, GCST90257015‐GCST90257105). Outcome data for BONECONTINUITY (2543 cases; 358 014 controls) and BONEDENSOTH (677 cases; 358 014 controls) are from the FinnGen R9 release.

FIGURE 2.

FIGURE 2

Open in a new tab

Forest plot showing the contribution of individual genetic instruments to the Mendelian randomization estimates. For a representative blood cell reactivity trait, each black dot represents the causal estimate (log OR) generated using a single SNP as an instrument for the association with BONECONTINUITY (A) or BONEDENSOTH (B). Red diamonds represent the combined causal estimate (with 95% CI) obtained using all SNPs as instruments via the inverse‐variance weighted (IVW) and MR‐Egger methods. Data sources and methods: Exposure data are from the GWAS Catalog (blood cell reactivity). Outcome data are from FinnGen R9 (BONECONTINUITY: 2543 cases, 358 014 controls; BONEDENSOTH: 677 cases, 358 014 controls). Analysis was performed using TwoSampleMR R package.

In addition, the MR leave‐one‐out sensitivity analysis was performed on the genetic instrumental variables with p < 0.05 (Figure S1). The scatter plots (Figure 3A,B) display the estimated effect sizes of the SNPs on the exposure (HBC traits) and the outcome (BONECONTINUITY and BONEDENSOTH).

FIGURE 3.

FIGURE 3

Open in a new tab

Scatter plots of SNP effects on the exposure against SNP effects on the outcome for a representative association. Each point represents a single nucleotide polymorphism (SNP) used as an instrumental variable. The X‐axis shows the SNP's effect on a selected human blood cell (HBC) reactivity trait (exposure, in SD units). The Y‐axis shows the SNP's effect on BONECONTINUITY (A) or BONEDENSOTH (B) (outcome, log OR). The slopes of the fitted lines represent the potential causal estimate (β) from five different Mendelian randomization methods. Data sources and methods: Exposure—HBC reactivity GWAS (GWAS Catalog, GCST90257015‐GCST90257105). Outcome—FinnGen R9 bone disorder GWAS (case/control counts as in Figure 1 legend). MR analysis performed using the TwoSampleMR R package.

3.2. Sensitivity analyses

Finally, we utilized a sensitivity analysis to evaluate the potential inverse associations between BONECONTINUITY and BONEDENSOTH with 91 blood cell traits. As shown in Figure 4 and Table 2, The F‐statistic ranges from 16.52 to 25.63, all above 10 (Table S2), using the IVW method, we found a statistically significant association between Neutrophil perturbation response (neutrophil 4 in response to Pam3CSK4 perturbation) and BONECONTINUITY (OR: 1.12, 95% CI: 1.00, 1.26, p = 0.041; Q(df) 28.05 (27), p = 0.408). Reticulocyte perturbation response (reticulocyte 1 in response to ciprofloxacin perturbation) (OR: 1.11, 95% CI: 1.01, 1.21, p = 0.024; Q(df) 27.49 (17), p = 0.051), Red blood cell perturbation response (DMSO perturbation) (OR: 0.78, 95% CI: 0.62, 0.98, p = 0.030; Q(df) 33.41 (15), p = 0.004), Reticulocyte perturbation response (reticulocyte 1 in response to KCl perturbation) (OR: 1.08, 95% CI: 1.01, 1.15, p = 0.034; Q(df) 16.72 (17), p = 0.474), Reticulocyte perturbation response (reticulocyte 1 in response to rotenone perturbation) (OR: 0.91, 95% CI: 0.82, 0.99, p = 0.033; Q(df) 17.34 (17), p = 0.432), Neutrophil perturbation response (neutrophil 2/neutrophil 4 ratio at baseline) (OR: 0.94, 95% CI: 0.88, 0.99, p = 0.029; Q(df) 10.03 (16), p = 0.865). The sensitivity analysis results listed in Figure 5A,B and Table 2 are stable, with the MR‐Egger intercept (p) >0.05. The estimated effect sizes of the SNP on the exposure (BONECONTINUITY and BONEDENSOTH) and the outcome (91 HBC types) are shown in the scatter plots (Figure 6A,B). The MR‐PRESSO global test indicated no significant horizontal pleiotropy for the vast majority of tests (all p > 0.05). A single test involving the ‘Red blood cell perturbation response to DMSO’ yielded a borderline p‐value of 0.06, which did not reach the conventional threshold for statistical significance (p < 0.05).

FIGURE 4.

FIGURE 4

Open in a new tab

Reverse MR sensitivity analysis evaluating the potential effect of bone disorders on 91 human blood cell (HBC) reactivity traits. Estimates were derived using the inverse‐variance weighted (IVW) method. CI, confidence interval; IVW, inverse variance weighted; MR, Mendelian randomization; OR, odds ratio; SNP, single nucleotide polymorphism. The forest plot displays odds ratios and 95% CIs for the potential causal effect of genetically predicted BONECONTINUITY and BONEDENSOTH on each HBC reactivity trait. Data sources and methods for reverse MR analysis: Exposure data (bone disorders) are from the FinnGen R9 release (BONECONTINUITY: 2543 cases, 358 014 controls; BONEDENSOTH: 677 cases, 358 014 controls). Outcome data are the 91 HBC reactivity traits from the GWAS Catalog (GCST90257015–GCST90257105). Analysis was performed using the TwoSampleMR R package.

TABLE 2.

MR estimates from different methods of assessing the causal effect of bone discontinuity or disorders of bone density and structure on HBC.

Exposure Outcome SNP Heterogeneity tests Directional horizontal pleiotropy test MR results Average F MR‐PRESSO
Methods Cochran's Q (df) p MR‐Egger intercept (p) Method Beta p Global test p value
BONECONTINUITY Human blood cell perturbation response phenotypes Neutrophil perturbation response (neutrophil 4 in response to Pam3CSK4 perturbation) MR Egger 26.50 (26) 0.436 −4.78E−02 (0.23) MR Egger 0.41 0.103 18.35 0.393
Inverse variance weighted 28.05 (27) 0.408 Inverse variance weighted 0.12 0.041*
BONEDENSOTH Reticulocyte perturbation response (reticulocyte 1 in response to ciproflaxin perturbation) MR Egger 23.12 (16) 0.111 −8.09E−02 (0.10) MR Egger 0.33 0.028 18.82 0.054
Inverse variance weighted 27.49 (17) 0.051 Inverse variance weighted 0.10 0.024*
Red blood cell perturbation response (DMSO perturbation) MR Egger 32.20 (14) 0.004 9.30E−02 (0.48) MR Egger −0.52 0.205 18.72 0.006
Inverse variance weighted 33.41 (15) 0.004 Inverse variance weighted −0.25 0.030*
Reticulocyte perturbation response (reticulocyte 1 in response to KCl perturbation) MR Egger 14.55 (16) 0.558 −5.61E−02 (0.16) MR Egger 0.23 0.055 18.82 0.496
Inverse variance weighted 16.72 (17) 0.474 Inverse variance weighted 0.07 0.034*
Reticulocyte perturbation response (reticulocyte 1 in response to rotenone perturbation) MR Egger 17.34 (16) 0.364 −1.4E−04 (1.00) MR Egger −0.10 0.527 18.82 0.483
Inverse variance weighted 17.34 (17) 0.432 Inverse variance weighted −0.10 0.033*
Neutrophil perturbation response (neutrophil 2/neutrophil 4 ratio at baseline) MR Egger 7.96 (15) 0.926 −4.71E−02 (0.17) MR Egger 0.06 0.512 18.87 0.858
Inverse variance weighted 10.03 (16) 0.865 Inverse variance weighted −0.07 0.029*
Open in a new tab
*

p < 0.05 indicates statistical significance.

FIGURE 5.

FIGURE 5

Open in a new tab

Forest plot from the reverse MR sensitivity analysis, showing the contribution of individual SNPs. For a representative analysis where a bone disorder is the exposure, each black dot represents the causal estimate (log OR) generated using a single SNP as an instrument for the association with a selected HBC reactivity trait (outcome). Red diamonds represent the combined IVW estimate (with 95% CI). Data sources and methods: Exposure—FinnGen R9 bone disorder GWAS (case/control counts as in Figure 4 legend). Outcome—HBC reactivity GWAS from the GWAS Catalog. Analysis performed using the TwoSampleMR R package.

FIGURE 6.

FIGURE 6

Open in a new tab

Scatter plots from the reverse Mendelian randomization analysis. Each point represents an instrumental variable SNP. The X‐axis shows the SNP's effect on genetically predicted BONECONTINUITY (A) or BONEDENSOTH (B) (exposure, log OR). The Y‐axis shows the SNP's effect on a selected HBC reactivity trait (outcome, in SD units). The slopes of the fitted lines represent the potential causal estimate (β) from five MR methods, testing the effect of the bone disorder on the blood cell trait. Data sources and methods: Exposure—FinnGen R9 bone disorder GWAS. Outcome—HBC reactivity GWAS from the GWAS Catalog. MR analysis performed using the TwoSampleMR R package.

It is of note that, consistent with the exploratory design of this study, these p‐values are not adjusted for multiple testing. They should therefore be interpreted as highlighting potential signals that warrant further investigation rather than as definitive potential causal evidence.

Power calculations for forward MR analyses are presented in Table S3. The minimal detectable ORs ranged from 1.06 to 1.31, indicating adequate power to detect modest to large effects for most examined associations. In contrast, reverse MR analyses were substantially underpowered due to the weak instrument strength (low R 2) for bone density, and thus their null results should be interpreted with caution.

4. DISCUSSION

Our exploratory Mendelian randomization analysis identifies genetic associations suggesting that genetically predicted variations in blood cell functional reactivity may be linked to bone disorders. It is crucial to note that our exposure instruments reflect genetically may influence differences in how cells respond to specific perturbations, not their resting state. This implies that the potential causal relationships we identify likely operate through the activation of latent biological pathways. This MR study identified that Red blood cell perturbation response (Pam3CSK4 perturbation), Neutrophil perturbation response (side fluorescence coefficient of variation of neutrophil 4), Neutrophil perturbation response (colchicine perturbation), Monocyte perturbation response (TMAO perturbation) affect bone continuity, Neutrophil perturbation response (forward scatter median of neutrophil 1) and Unknown cell population perturbation response (nigericin perturbation) regulate bone density and bone structure disruption. Reverse MR sensitivity analysis found that Neutrophil perturbation response (neutrophil 4 in response to Pam3CSK4 perturbation) has a statistically significant association with BONECONTINUITY. The blood cells with statistically significant associations with BONEDENSOTH are Reticulocyte perturbation response (reticulocyte 1 in response to ciproflaxin perturbation), Red blood cell perturbation response (DMSO perturbation), Reticulocyte perturbation response (reticulocyte 1 in response to KCl perturbation), Reticulocyte perturbation response (reticulocyte 1 in response to rotenone perturbation), and Neutrophil perturbation response (neutrophil 2/neutrophil 4 ratio at baseline).

Notably, our findings implicate several distinct biological pathways. The association of RBC reactivity with Pam3CSK4 suggests a previously underappreciated role for Toll‐like receptor signaling in erythroid cells on bone homeostasis. 19 Pam3CSK4 is a specific TLR1/2 agonist. TLR activation drives the production of pro‐inflammatory cytokines such as TNF‐α and IL‐6, thereby regulating osteogenic and osteoclastogenic functions in the bone marrow microenvironment. 20 Another study found that patients with severe bone trauma showed the expression of genes encoding receptors that have an inhibitory downstream effect on erythropoiesis, such as TLR4. 19 This suggests that inter‐individual variations in TLR1/2 pathway reactivity within erythroid or myeloid cells may influence the inflammatory state of the bone marrow microenvironment, consequently affecting bone remodeling. While the function of TLRs in osteoclasts is known, our finding suggests that immune signaling in red blood cell precursors may indirectly influence the bone marrow microenvironment. Besides, the association of neutrophil reactivity to colchicine with bone continuity disorders strongly supports the central role of the NLRP3 inflammasome pathway in bone health. This provides a genetic correlation perspective supporting the therapeutic use of colchicine in inflammatory conditions with bone involvement (e.g., acute gouty arthritis). 21 In addition, the effect of monocyte reactivity to TMAO aligns with growing evidence of a gut‐bone axis, suggesting that diet and microbiome‐derived metabolites may influence bone health by priming innate immune cells. 22 While these perturbational traits are not conventional stable biomarkers, their genetic instruments capture innate, heritable differences in how an individual's blood cells respond to specific challenges.

Beyond these core pathways, our analysis implicated additional biological processes. The association of nigericin perturbation (a potent NLRP3 inflammasome activator) with bone density disorders further reinforces the critical role of inflammasome signaling. 23 Similarly, responses to ciprofloxacin (an antibiotic) and rotenone (a mitochondrial complex I inhibitor) point to potential roles of the gut microbiome and cellular metabolic stress in bone homeostasis, respectively. 24 , 25 While the precise mechanisms linking these specific reactivities to bone remain to be elucidated, they highlight promising avenues for future research.

Some of our findings, such as those involving DMSO (a common solvent) or KCl (an osmotic stimulus). These perturbations may reflect broader aspects of cellular robustness, membrane integrity, or volume regulation. Furthermore, traits like the neutrophil 2/neutrophil 4 ratio at baseline suggest that even unstimulated cellular composition, as captured by high‐dimensional cytometry, holds potential causal information for bone health.

The interplay between blood cells and bone density is a complex relationship influenced by various hematological and physiological factors. Bone health is significantly compromised in individuals with hematopoietic disorders such as sickle cell anemia and thalassemia, as evidenced by observational data. The high prevalence of diminished bone density and fracture risk in sickle cell anemia patients is primarily attributable to ongoing inflammation and aberrant bone remodeling processes. 26 , 27 Similarly, thalassemia is associated with chronic bone marrow hyperplasia, which leads to decreased bone mineral density (BMD) and increased fracture risk. 26 , 28 These conditions illustrate how blood disorders can disrupt normal bone metabolism, highlighting the critical role of hematopoietic cells in maintaining bone integrity. The connection between blood cell counts and skeletal density is further supported by evidence from research focusing on the osseous microvasculature. The vascular network in bone is essential for nutrient delivery and waste removal, which are vital for bone health. 29 , 30 , 31 Age‐related declines in vascular function can compromise these processes, leading to impaired bone remodeling and increased susceptibility to osteoporosis. 29 , 31

Moreover, research indicates that circulating blood cell levels, particularly white blood cells, are negatively correlated with BMD during osteoporosis progression, suggesting that hematopoietic changes may directly influence bone homeostasis. 32 , 33 This interplay underscores the importance of vascular health and blood cell dynamics in the context of bone density.

However, this study has some limitations. While Mendelian randomization analysis suggests a causal relationship, this inference must be interpreted with caution, acknowledging that confounding elements from sources including age, sex, and lifestyle cannot be fully discounted. In addition, the findings are based exclusively on European GWAS data, and their transferability to other ethnic groups requires further investigation in ethnically diverse populations. We also performed a large number of statistical tests on a novel set of exposure phenotypes. While this exploratory approach is valuable for hypothesis generation, it increases the risk of false positive findings (Type I error). The fact that several associations did not survive stringent Bonferroni correction warrants careful consideration. This may be attributable to these factors: (i) the potentially complex and polygenic architecture of these functional cellular responses, which might not be fully powered by current GWAS sample sizes; and (ii) limited statistical power for the bone disorder outcomes, particularly for BONEDENSOTH. Future confirmatory studies with larger, targeted designs are essential. Fourth, we performed a reverse MR analysis as a sensitivity check for reverse causality. However, the low statistical power of this analysis, due to the limited sample size, precludes definitive conclusions about directionality. Its primary utility is to highlight that our main exploratory findings are not obviously contradicted by a simple reverse analysis. We also note that one reverse MR analysis yielded a borderline MR‐PRESSO p‐value (0.06). Given that this is an isolated finding among numerous tests, this result is most parsimoniously attributed to chance variation or the instability of the test near the significance threshold, rather than to robust evidence of horizontal pleiotropy.

5. CONCLUSION

In this exploratory genetic study, we observed associations between genetically predicted blood cell perturbation response phenotypes and disorders of bone continuity, density, and structure. Our screen identified that several traits, including Red blood cell perturbation response (Pam3CSK4 perturbation), Neutrophil perturbation response (side fluorescence coefficient of variation of neutrophil 4), Neutrophil perturbation response (colchicine perturbation), and Monocyte perturbation response (TMAO perturbation), influence bone continuity, while Neutrophil perturbation response (forward scatter median of neutrophil 1) and Unknown cell population perturbation response (nigericin perturbation) are potential candidates linked to bone density and bone structure abnormalities.

Reverse MR sensitivity analysis suggested potential bidirectional relationships, such as between Neutrophil perturbation response (neutrophil 4 in response to Pam3CSK4 perturbation) and bone continuity, and associations between bone density and bone structure abnormalities with blood cell characteristics such as Reticulocyte perturbation response, Red blood cell perturbation response, and Neutrophil perturbation response.

AUTHOR CONTRIBUTIONS

Zhiqin Deng: Conceptualization; data curation; formal analysis; methodology; resources; software; writing – original draft; writing – review and editing. Zhe Zhao: Data curation; formal analysis; software. Wenting Jiang: Data curation; formal analysis; software. Xiaoqiang Chen: Data curation; formal analysis; software. Jianquan Liu: Data curation; formal analysis; software. Xu Tao: Data curation; formal analysis; software; supervision. Zhengyang Lin: Formal analysis; software; supervision; writing – review and editing. Zhenhan Deng: Data curation; formal analysis; funding acquisition; resources; supervision; writing – review and editing. Wencui Li: Formal analysis; funding acquisition; resources; supervision; writing – review and editing.

FUNDING INFORMATION

This study was supported by the National Natural Science Foundation of China (81800785, 81972085, 82172465), the Natural Science Foundation of Guangdong Province (2023A1515010102, 2024A1515220060), Guangdong Provincial Key Clinical Discipline‐Orthopedics (2000005), Guangdong Province Medical Science and Technology Research Foundation Project (A2024359), the Sanming Project of Shenzhen Health and Family Planning Commission (SZSM202311008), Shenzhen Science and Technology Planning (JCYJ20240813141041053, JCYJ20240813141011015, JCYJ20250604180734044, JCYJ20250604180551064), the Municipal Financial Subsidy of Shenzhen Medical Key Discipline Construction (SZXK025), Team‐based Medical science Research Program (2024YZZ13). Shenzhen Portion of Shenzhen‐Hong Kong Science and Technology Innovation Cooperation Zone (HTHZQSWS‐KCCYB‐2023060). Shenzhen Second People's Hospital Clinical Research Fund of Shenzhen High‐level Hospital Construction Project (20253357005).

CONFLICT OF INTEREST STATEMENT

Zhiqin Deng, Zhe Zhao, Wenting Jiang, Xiaoqiang Chen, Jianquan Liu, Xu Tao, Zhengyang Lin, Zhenhan Deng and Wencui Li declare that they have no other conflict of interest. Zhenhan Deng is an editorial board member of Animal Models and Experimental Medicine (AMEM) and a corresponding author of this article. To minimize bias, he was excluded from all editorial decision making related to the acceptance of this article for publication.

ETHICS STATEMENT

This study utilized exclusively publicly available, de‐identified data from IEU Open GWAS and GWAS Catalog. Therefore, ethical approval for this specific analysis was not required, as confirmed by the institutional review board of Shenzhen Second People's Hospital. The original data collection in IEU Open GWAS and GWAS Catalog was conducted under ethical approval from its respective governing bodies. All participants in the original studies provided informed consent.

Supporting information

Figure S1.

AME2-9-932-s001.jpg (506.2KB, jpg)

Table S1.

AME2-9-932-s002.xlsx (63.2KB, xlsx)

ACKNOWLEDGMENTS

None.

Contributor Information

Zhenhan Deng, Email: dengzhenhan@wmu.edu.cn.

Wencui Li, liwencui@email.szu.edu.cn.

DATA AVAILABILITY STATEMENT

The data used in this paper are derived from publicly available data. The outcome data for “Disorders of continuity of bone (BONECONTINUITY, ID: finngen_R9_M13_BONECONTINUITY)” and “Other disorders of bone density and structure (BONEDENSOTH, ID: finngen_R9_M13_BONEDENSOTH)” were sourced from the FinnGen biobank (Release 9), and can be accessed via the FinnGen public data portal: gs://finngen‐public‐data‐r9/summary_stats/finngen_R9_M13_BONECONTINUITY.gz, or: gs://finngen‐public‐data‐r9/summary_stats/finngen_R9_M13_BONEDENSOTH.gz. The exposed data was downloaded from the GWAS Catalog (https://www.ebi.ac.uk/gwas/), and the range of the GWAS Catalog database is (GCST90257015‐GCST90257105). The custom R scripts used for the Mendelian randomization analyses, sensitivity tests, and generation of figures in this study are not deposited in a public repository, but are available from the corresponding author upon reasonable request.

REFERENCES

  • 1. Walker MD, Shane E. Postmenopausal osteoporosis. N Engl J Med. 2023;389(21):1979‐1991. doi: 10.1056/NEJMcp2307353 [DOI] [PubMed] [Google Scholar]
  • 2. Wang T, Wang L, Zhang L, Long Y, Zhang Y, Hou Z. Single‐cell RNA sequencing in orthopedic research. Bone Res. 2023;11(1):10. doi: 10.1038/s41413-023-00245-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Deng Z, Chen X, Lin Z, et al. The homeostasis of cartilage matrix remodeling and the regulation of volume‐sensitive ion channel. Aging Dis. 2022;13(3):787‐800. doi: 10.14336/ad.2021.1122 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Semenov MV, He X. LRP5 mutations linked to high bone mass diseases cause reduced LRP5 binding and inhibition by SOST. J Biol Chem. 2006;281(50):38276‐38284. doi: 10.1074/jbc.M609509200 [DOI] [PubMed] [Google Scholar]
  • 5. Diegel CR, Kramer I, Moes C, et al. Inhibiting WNT secretion reduces high bone mass caused by Sost loss‐of‐function or gain‐of‐function mutations in Lrp5. Bone Res. 2023;11(1):47. doi: 10.1038/s41413-023-00278-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Raaijmakers MH, Mukherjee S, Guo S, et al. Bone progenitor dysfunction induces myelodysplasia and secondary leukaemia. Nature. 2010;464(7290):852‐857. doi: 10.1038/nature08851 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Colom Díaz PA, Mistry JJ, Trowbridge JJ. Hematopoietic stem cell aging and leukemia transformation. Blood. 2023;142(6):533‐542. doi: 10.1182/blood.2022017933 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Steer K, Stavnichuk M, Morris M, Komarova SV. Bone health in patients with hematopoietic disorders of bone marrow origin: systematic review and meta‐ analysis. J Bone Miner Res. 2017;32(4):731‐742. doi: 10.1002/jbmr.3026 [DOI] [PubMed] [Google Scholar]
  • 9. Yuan Y, Fang Y, Zhu L, et al. Deterioration of hematopoietic autophagy is linked to osteoporosis. Aging Cell. 2020;19(5):e13114. doi: 10.1111/acel.13114 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Ambrosi TH, Marecic O, McArdle A, et al. Aged skeletal stem cells generate an inflammatory degenerative niche. Nature. 2021;597(7875):256‐262. doi: 10.1038/s41586-021-03795-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Singbrant S, Russell MR, Jovic T, et al. Erythropoietin couples erythropoiesis, B‐lymphopoiesis, and bone homeostasis within the bone marrow microenvironment. Blood. 2011;117(21):5631‐5642. doi: 10.1182/blood-2010-11-320564 [DOI] [PubMed] [Google Scholar]
  • 12. Suresh S, de Castro LF, Dey S, Robey PG, Noguchi CT. Erythropoietin modulates bone marrow stromal cell differentiation. Bone Res. 2019;7:21. doi: 10.1038/s41413-019-0060-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Kim KJ, Lee Y, Hwang HG, Sung S, Lee M, Son YJ. Betulin suppresses osteoclast formation via down‐regulating NFATc1. J Clin Med. 2018;7(6):154. doi: 10.3390/jcm7060154 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Deng Z, Zhang Q, Zhao Z, et al. Crosstalk between immune cells and bone cells or chondrocytes. Int Immunopharmacol. 2021;101(Pt A):108179. doi: 10.1016/j.intimp.2021.108179 [DOI] [PubMed] [Google Scholar]
  • 15. Miraki‐Moud F, Anjos‐Afonso F, Hodby KA, et al. Acute myeloid leukemia does not deplete normal hematopoietic stem cells but induces cytopenias by impeding their differentiation. Proc Natl Acad Sci USA. 2013;110(33):13576‐13581. doi: 10.1073/pnas.1301891110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Shiozawa Y, Taichman RS. Getting blood from bone: an emerging understanding of the role that osteoblasts play in regulating hematopoietic stem cells within their niche. Exp Hematol. 2012;40(9):685‐694. doi: 10.1016/j.exphem.2012.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Bowers M, Zhang B, Ho Y, Agarwal P, Chen CC, Bhatia R. Osteoblast ablation reduces normal long‐term hematopoietic stem cell self‐renewal but accelerates leukemia development. Blood. 2015;125(17):2678‐2688. doi: 10.1182/blood-2014-06-582924 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Homilius M, Zhu W, Eddy SS, et al. Perturbational phenotyping of human blood cells reveals genetically determined latent traits associated with subsets of common diseases. Nat Genet. 2024;56(1):37‐50. doi: 10.1038/s41588-023-01600-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Kelly LS, Apple CG, Darden DB, et al. Transcriptomic changes within human bone marrow after severe trauma. Shock. 2022;57(1):24‐30. doi: 10.1097/shk.0000000000001826 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Kim J, Yang J, Park OJ, et al. Lipoproteins are an important bacterial component responsible for bone destruction through the induction of osteoclast differentiation and activation. J Bone Miner Res. 2013;28(11):2381‐2391. doi: 10.1002/jbmr.1973 [DOI] [PubMed] [Google Scholar]
  • 21. Park K, Shin I, Kim Y, et al. A novel NLRP3 inhibitor as a therapeutic agent against monosodium urate‐induced gout. Front Immunol. 2023;14:1307739. doi: 10.3389/fimmu.2023.1307739 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Chittimalli K, Manandhar I, Adkins S, et al. Angiotensin‐(1‐7) modulates gut‐bone marrow axis in diabetes. Biochem Pharmacol. 2025;241:117168. doi: 10.1016/j.bcp.2025.117168 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Yan LJ, Qi S, Wu C, et al. Hypocrellin a from an ethnic medicinal fungus protects against NLRP3‐driven gout in mice by suppressing inflammasome activation. Acta Pharmacol Sin. 2025;46(4):1016‐1029. doi: 10.1038/s41401-024-01434-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Zhao Z, Ning J, Bao XQ, et al. Fecal microbiota transplantation protects rotenone‐induced Parkinson's disease mice via suppressing inflammation mediated by the lipopolysaccharide‐TLR4 signaling pathway through the microbiota‐gut‐brain axis. Microbiome. 2021;9(1):226. doi: 10.1186/s40168-021-01107-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Dethlefsen L, Relman DA. Incomplete recovery and individualized responses of the human distal gut microbiota to repeated antibiotic perturbation. Proc Natl Acad Sci USA. 2011;108(1):4554‐4561. doi: 10.1073/pnas.1000087107 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Valderrábano RJ, Lui LY, Lee J, et al. Bone density loss is associated with blood cell counts. J Bone Miner Res. 2017;32(2):212‐220. doi: 10.1002/jbmr.3000 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Batte A, Kasirye P, Baluku R, et al. Mineral bone disorders and kidney disease in hospitalized children with sickle cell anemia. Front Pediatr. 2022;10:1078853. doi: 10.3389/fped.2022.1078853 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Di Paola A, Marrapodi MM, Di Martino M, et al. Bone health impairment in patients with hemoglobinopathies: from biological bases to new possible therapeutic strategies. Int J Mol Sci. 2024;25(5):2902. doi: 10.3390/ijms25052902 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Guderian S, Lee S, McLane MA, Prisby RD. Progressive ossification of the bone marrow vasculature with advancing age corresponds with reduced red blood cell count and percentage of circulating lymphocytes in male Fischer‐344 rats. Microcirculation. 2019;26(8):e12550. doi: 10.1111/micc.12550 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Tomlinson RE, Silva MJ. Skeletal blood flow in bone repair and maintenance. Bone Res. 2013;1(4):311‐322. doi: 10.4248/br201304002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Grunewald M, Kumar S, Sharife H, et al. Counteracting age‐related VEGF signaling insufficiency promotes healthy aging and extends life span. Science. 2021;373(6554):eabc8479. doi: 10.1126/science.abc8479 [DOI] [PubMed] [Google Scholar]
  • 32. Li Y, Hao W, Guan J, et al. Relationship between indices of circulating blood cells and bone homeostasis in osteoporosis. Front Endocrinol. 2022;13:965290. doi: 10.3389/fendo.2022.965290 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Wang Q, Xu F, Chen J, Xie YQ, Xu SL, He WM. Serum leukocyte cell‐derived chemotaxin 2 (LECT2) level is associated with osteoporosis. Lab Med. 2023;54(1):106‐111. doi: 10.1093/labmed/lmac080 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1.

AME2-9-932-s001.jpg (506.2KB, jpg)

Table S1.

AME2-9-932-s002.xlsx (63.2KB, xlsx)

Data Availability Statement

The data used in this paper are derived from publicly available data. The outcome data for “Disorders of continuity of bone (BONECONTINUITY, ID: finngen_R9_M13_BONECONTINUITY)” and “Other disorders of bone density and structure (BONEDENSOTH, ID: finngen_R9_M13_BONEDENSOTH)” were sourced from the FinnGen biobank (Release 9), and can be accessed via the FinnGen public data portal: gs://finngen‐public‐data‐r9/summary_stats/finngen_R9_M13_BONECONTINUITY.gz, or: gs://finngen‐public‐data‐r9/summary_stats/finngen_R9_M13_BONEDENSOTH.gz. The exposed data was downloaded from the GWAS Catalog (https://www.ebi.ac.uk/gwas/), and the range of the GWAS Catalog database is (GCST90257015‐GCST90257105). The custom R scripts used for the Mendelian randomization analyses, sensitivity tests, and generation of figures in this study are not deposited in a public repository, but are available from the corresponding author upon reasonable request.

Articles from Animal Models and Experimental Medicine are provided here courtesy of Wiley

RESOURCES