The causal relationship between hyperthyroidism, hypothyroidism, and osteoporosis: a Mendelian randomization study

doi:10.21203/rs.3.rs-4265217/v1

To investigate the association between hyperthyroidism、hypothyroidism and osteoporosis by bidirectional two-sample Mendelian randomization (MR) to clarify the casual relationship. Independent genetic variants for hyperthyroidism、hypothyroidism and osteoporosis were selected as instruments from published genome-wide association studies (GWAS), mainly in European ancestry. Instrumental variables (IVs) associated with hyperthyroidism、hypothyroidism and osteoporosis were extracted separately from the largest GWAS meta-analysis. MR analyses included inverse variance weighting, weighted median estimator, MR Egger regression, and sensitivity analyses with Steiger fltering and MR PRESSO. In the data samples for hyperthyroidism、hypothyroidism and osteoporosis, there was a negative causal relationship between hyperthyroidism and osteoporosis [IVW, OR = 1.001%, 95% CI: 0.999–1.002, ]; while hypothyroidism and osteoporosis [IVW, OR = 2.648%, 95% CI: 2.582–2.715, p < 0.001]. The results of the bidirectional MR Study suggest that hyperthyroidism and hypothyroidism have negative causal effect on Osteoporosis, which provides implications for clinical treatment decisions in patients with Osteoporosis.

Health sciences/Diseases

Health sciences/Diseases/Endocrine system and metabolic diseases

Health sciences/Medical research

Health sciences/Medical research/Epidemiology

Osteoporosis

Hyperthyroidism

Hypothyroidism

Osteoporosis (OP) is a systemic bone disease characterized by an imbalance between bone resorption and bone formation, leading to skeletal deterioration where bone destruction exceeds bone formation, resulting in decreased bone strength and increased bone fragility due to microstructural damage. (1, 2)。As a disease characterized by decreased bone strength, osteoporosis can significantly increase the risk of fractures and disability, imposing a heavy burden on daily life for patients, particularly with the current trend of an aging society. It is estimated to impact approximately 2 billion people worldwide, with the majority of cases occurring in individuals aged 60 and older. (3-5)。Thyroid hormones play a critical role in development, linear growth, and adult bone turnover and maintenance. Imbalance in thyroid hormones, leading to thyroid dysfunction, can disrupt bone homeostasis. (6, 7)。In a state of hyperthyroidism, bone turnover increases, and the remodeling period shortens. Consequently, hyperthyroidism may lead to decreased bone density and an increased risk of osteoporosis and fractures. Conversely, hypothyroidism can cause decreased bone turnover and prolonged remodeling cycles, creating an imbalance between bone formation and resorption, resulting in negative bone balance. (8)。 Most research studies are currently traditional clinical studies, which cannot avoid the interference of reverse causality and confounding factors. Therefore, it remains unclear whether hyperthyroidism, hypothyroidism, and osteoporosis have a positive causal relationship. Mendelian Randomization (MR) study design follows the Mendelian genetic law of “random allocation of alleles from parent to offspring.” If genotype determines phenotype, and genotype is associated with the disease through the phenotype, genotype can be used as an instrumental variable to infer the association between phenotype and disease. Compared to randomized controlled trials (RCT), MR has advantages such as saving time, saving costs, and greater feasibility. Additionally, as the process between genes and diseases typically flows in one direction, MR can avoid the bias of reverse causality that occurs in traditional observational studies.(9)。In recent years, two-sample Mendelian Randomization (MR) studies have gradually gained recognition among researchers. These studies allow the data on gene exposure, phenotype, and phenotype-disease association to come from two distinct independent populations, thereby increasing research efficiency and statistical power. This study aims to utilize the statistical methods of two-sample MR and leverage the genome-wide association analysis (GWAS) database to select single nucleotide polymorphism (SNP) loci as instrumental variables to investigate the causal relationships between hyperthyroidism, hypothyroidism, and osteoporosis.

Data availability

According to the provided context, the data in this study are all from publicly published research, and all studies have obtained ethical approval. The research results are sourced from publicly available databases. The GWAS data on exposure and outcomes mainly come from the IEU GWAS database, and this study is a secondary data review of previously published research, so no additional ethical approval is required. The study followed Burgess’s guidelines and reported according to the STROBE-MR Statement format requirements.

Exposure data

The acquisition of single-nucleotide polymorphisms (SNPs) data in this article is derived from available and publicly accessible genomic research on hyperthyroidism, hypothyroidism data through Genome-Wide Association Studies (GWAS). Instrumental variables (IVs) need to satisfy the three assumptions of Mendelian randomization: (1) The relevance assumption: The selected independent instrumental variables are directly related to the exposure factor of interest; (2) The independence assumption: The selected instrumental variables are unrelated to any confounding variables between exposure and outcome; (3) The exclusion restriction assumption: The selected instrumental variables do not influence the outcome unless through their association with the exposure(10, 11)。In this study, SNPs meeting the genome-wide significance threshold (P<5×10^-9) and in linkage equilibrium (r2<0.001, 10000 kb) were used as potential instruments to ensure the independence and relevance of the SNPs. (12)。SNPs that meet the GWAS standard (P<5×10-8) and are associated with osteoporosis-related traits were used to construct genetic instruments for each phenotype. Each genetic instrument encompasses one or more genetic variants, and these variants become the characteristics of the instrumental variable. The strength of the instrumental variable is measured by the F-statistic, and a value below 10 indicates that the instrumental variable has weak potential, so it is excluded.The exposure factors hyperthyroidism and hypothyroidism were studied using the UK Biobank dataset (ukb-a-76) and the Finn Gen database (finn-b-HYPOTHY_REIMB) respectively, with the study population primarily comprising European individuals. Furthermore, to avoid the influence of previously published SNPs on the results, a search was conducted in PubMed to identify and exclude relevant SNPs （Table 1）.

Outcome data

The outcome indicator data in this study were summarized from the IEU database ebi-a-GCST90038636) and the UK Biobank dataset ukb-b-17796 (13). (https://gwas.mrcieu.ac.uk/datasets/ebi-a-GCST90038636//) The exposure and outcome data in this article each were sourced from Finn Gen, IEU GWAS database, and the UK Biobank database, and the original data for exposure and outcome came from different research institutions with no overlap. Therefore, there are no duplicate samples in the exposure and outcome data of this study（ Table 1）.

Table 1. Genome-wide Association Study (GWAS) Data Information on Exposure and Outcome

characteristics	Variable type	Panels	Sample size		SNP size	population
characteristics	Variable type	Panels	Cases	Control	SNP size	population
hyperthyroidism	Exposures	UK Biobank	2,547	334,162	10,894,596	European
hypothyroidism	Exposures	Finn Gen	7,183	59,893	16,375,729	European
osteoporosis	Outcomes	UK Biobank	1,976	461,034	9,851,867	European
osteoporosis	Outcomes	IEU GWAS	7,751	476,847	9,587, 836	European

Statistical analysi

We conducted data analysis using the “TwoSampleMR” package in R Studio software (version 4.3.3) and MR Base (https://www.mrbase.org/). We extracted IV-related data from the hyperthyroidism and hypothyroidism outcome data and harmonized the exposure and outcome data. To determine the relationship between hyperthyroidism and osteoporosis, we primarily employed the random-effects inverse-variance weighting (IVW) method as the main outcome measure. MR-Egger, weighted median, and weighted mode were used as supplementary tests in this analysis. MR-Egger intercept detection was used as the main method to detect horizontal pleiotropy, where an intercept not equal to 0 indicates the presence of horizontal pleiotropy between studies. Additionally, the symmetry of the funnel plot was examined to detect horizontal pleiotropy, and leave-one-out analysis was used to assess the impact of each SNP on the outcome. Differences were considered statistically significant at a threshold of P<0.05, and the results were presented as odds ratios (ORs) with corresponding 95% confidence intervals (CIs).

MR analysis

Through screening, 16 strongly correlated SNPs were ultimately included as instruments in the study on the association between hyperthyroidism, hypothyroidism, and osteoporosis (F value >10), indicating the strong correlation of the included SNPs and removing weak instrumental bias (Table 2). Analysis using the IVW method revealed significant statistical significance in the relationship between hyperthyroidism, hypothyroidism, and osteoporosis. However, results from MR-Egger method showed no statistical significance between the two (Table 3).

Table 2. Single Nucleotide Polymorphism (SNP) Information Table

SNPs	EA	UNEA	Effect allele β gene frequencies	β SE	P
rs1794279	T	G	0.198250569	0.055222719	0.000330655
rs2160215	C	T	0.040350298	0.104224887	0.698647769
rs28360997	A	G	0.284210679	0.205670391	0.167010223
rs28383364	G	A	0.086232606	0.096905955	0.37354177
rs3087243	A	G	0.085509995	0.127773636	0.50334855
rs6679677	A	C	-0.303969936	0.163632024	0.063219682
rs10116520	G	A	0.1199	0.0218	3.81E-08
rs10174238	A	G	-0.1477	0.0251	3.81E-09
rs10929817	C	T	-0.1225	0.0219	2.17E-08
rs11571297	C	T	-0.18	0.0222	4.54E-16
rs2858483	C	A	0.1204	0.0217	2.75E-08
rs597808	G	A	-0.2207	0.0216	1.92E-24
rs7030256	G	C	0.1887	0.0225	5.10E-17
rs73192661	T	C	-0.1344	0.022	9.98E-10

Table 3. Results of the Two-Sample Mendelian Randomization Analysis

Exposures

Outcomes

IVW

(P）

Weighted median (P)

MR Egger

(P)

Weighted mode

(p)

hyperthyroidism

osteoporosis

0.03039

0.01059

0.12

0.02233

hypothyroidism

osteoporosis

0.0081

0.0064

0.7137

0.0597

Sensitivity analysis

According to the Single SNP analysis results of SNP by IVR and MR-Egger, it was found that the selected SNPs were strongly associated with hyperthyroidism, and genetic pleiotropy did not bias the results (Figure 1). Furthermore, using IVW, MR-Egger, weighted median, and weighted mode to detect the linear slope showed that the causal associations were consistent among the four detection methods (Figure 2). The funnel plot results indicated a very low risk of horizontal pleiotropy between hyperthyroidism, hypothyroidism, and osteoporosis (Figure 3). Through the Leave-one-out method to sequentially remove one SNP, with the remaining SNPs as IVs, it was found that no single SNP had a significant impact on the results, and all analyses were statistically significant (Figure 4)

With the arrival of global aging population, osteoporosis, as a degenerative disease, its incidence rate is continuously increasing, has become a major health issue of widespread concern worldwide (14, 15). Thyroid hormones, as an essential factor in maintaining bone structure and strength, when thyroid dysfunction occurs (such as hyperthyroidism or hypothyroidism), can become a risk factor for fractures. (16, 17)。However, the role of excessive or insufficient thyroid hormones in the pathogenesis of osteoporosis has long been underestimated. In this study, through two large-scale GWAS research databases, the causal relationship between hyperthyroidism, hypothyroidism, and osteoporosis was investigated using a two-sample MR approach. The results of hypothyroidism showed that for every increase of 1 standard deviation, the risk of developing osteoporosis would increase by about 1%. These MR results suggest that hyperthyroidism and hypothyroidism are important risk factors for osteoporosis, emphasizing the importance of controlling the progression of hyperthyroidism and hypothyroidism in the prevention of osteoporosis. For a long time, hyperthyroidism and hypothyroidism have been considered important risk factors for the development of osteoporosis.(18, 19)。The signaling pathways involved in the interaction between hyperthyroidism, hypothyroidism, and osteoporosis mainly include Human T-cell Leukemia Virus Type I infection, the AGE-RAGE signaling pathway in diabetic complications, Type I diabetes, and graft-versus-host disease. (20–23);TNF, AKT1, ALB, CD4, IL2 may be the main potential targets for treating osteoporosis induced by hyperthyroidism; while ALB, AKT1, TP53, TNF, CTNNB1 may be the main potential targets for treating age-related osteoporosis induced by hypothyroidism. The reason is that TNF (tumor necrosis factor) can directly act on osteoclast precursors, promoting osteoclastogenesis. (24); AKT1, as a signaling mediator in osteoblasts, can control the differentiation of osteoblasts and osteoclasts. (25, 26); the central hydrophobic site of ALB is the binding site for thyroid hormones. (27)༛TP53 and CTNNB1 are the most up-regulated DEGs in the differential expression genes of osteoporosis, indicating that they may play important roles in osteoporosis. (28); Research has also shown that elevated levels of IL2 (Interleukin-2) and high levels of CD4 can lead to the development of osteoporosis. (29)。Xu et al. conducted a meta-analysis of 12 cohort studies with a total of 275,086 participants, following up for 3 months to 13 years based on pre-determined inclusion and exclusion criteria. The results showed that patients with hyperthyroidism had a significantly increased risk of developing osteoporosis. (30)。Some studies suggest that hypothyroidism leads to osteoporosis because 90% of hypothyroid subjects undergoing dual-energy X-ray absorptiometry scans were found to have osteoporosis. (31)。Indeed, this suggests that both hyperthyroidism and hypothyroidism are risk factors for the development of osteoporosis.

This study used two-sample Mendelian Randomization (MR) studies and employed three statistical methods in the analysis, including IVW, weighted median method, and MR-Egger regression, making the statistical analysis results more reliable. Compared to traditional studies, this study effectively avoided the interference of reverse causality and confounding factors, but it also has certain limitations. Due to the database used in this study being from European countries, the applicability of the study conclusions to diverse populations worldwide still requires universal verification. Additionally, due to limitations of the MR-Base platform, direct access to specific data of the study population is not possible, therefore impeding further research on this data. Mendelian Randomization studies can only make preliminary conclusions about causal relationships, and detailed mechanisms of action between hyperthyroidism and osteoporosis still require further investigation.

In summary, the results of the two-sample Mendelian Randomization studies indicate a positive causal relationship between hyperthyroidism, hypothyroidism, and the risk of osteoporosis. A one standard deviation increase in hyperthyroidism is associated with a 2.6% increase in the risk of developing osteoporosis, while a one standard deviation increase in hypothyroidism is associated with a 1% increase in the risk of developing osteoporosis.

Author Contributions: D.WY. and L.H. wrote the main manuscript text .W.S. prepared figures 1-4. F.XL. and W.YL.prepared tabies 1-3.All authors reviewed the manuscript. Especially, D.WY and LH contribute equally to this manuscript.

Funding: This research received no external funding.

Informed Consent Statement: All authors consented for publication.

Data availability

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

Conflicts of Interest: All authors declare no competing financial interest.

Aspray TJ, Hill TR. Osteoporosis and the Ageing Skeleton. Subcell Biochem. 2019;91:453-76.
Vandenbroucke A, Luyten FP, Flamaing J, Gielen E. Pharmacological treatment of osteoporosis in the oldest old. Clin Interv Aging. 2017;12:1065-77.
Inose H, Kato T, Ichimura S, Nakamura H, Hoshino M, Togawa D, et al. Risk factors for subsequent vertebral fracture after acute osteoporotic vertebral fractures. Eur Spine J. 2021;30(9):2698-707.
Brown JP. Long-Term Treatment of Postmenopausal Osteoporosis. Endocrinol Metab (Seoul). 2021;36(3):544-52.
Wright NC, Looker AC, Saag KG, Curtis JR, Delzell ES, Randall S, Dawson-Hughes B. The recent prevalence of osteoporosis and low bone mass in the United States based on bone mineral density at the femoral neck or lumbar spine. J Bone Miner Res. 2014;29(11):2520-6.
Bassett JH, Williams GR. Role of Thyroid Hormones in Skeletal Development and Bone Maintenance. Endocr Rev. 2016;37(2):135-87.
Zhu S, Pang Y, Xu J, Chen X, Zhang C, Wu B, Gao J. Endocrine Regulation on Bone by Thyroid. Front Endocrinol (Lausanne). 2022;13:873820.
Delitala AP, Scuteri A, Doria C. Thyroid Hormone Diseases and Osteoporosis. J Clin Med. 2020;9(4).
Katan MB. Apolipoprotein E isoforms, serum cholesterol, and cancer. Lancet. 1986;1(8479):507-8.
Burgess S, Davey Smith G, Davies NM, Dudbridge F, Gill D, Glymour MM, et al. Guidelines for performing Mendelian randomization investigations: update for summer 2023. Wellcome Open Res. 2019;4:186.
Skrivankova VW, Richmond RC, Woolf BAR, Yarmolinsky J, Davies NM, Swanson SA, et al. Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE-MR Statement. Jama. 2021;326(16):1614-21.
Cheng L, Wang S, Tang H. Type 2 diabetes mellitus plays a protective role against osteoporosis --mendelian randomization analysis. BMC Musculoskelet Disord. 2023;24(1):444.
Dönertaş HM, Fabian DK, Valenzuela MF, Partridge L, Thornton JM. Common genetic associations between age-related diseases. Nat Aging. 2021;1(4):400-12.
Armas LA, Recker RR. Pathophysiology of osteoporosis: new mechanistic insights. Endocrinol Metab Clin North Am. 2012;41(3):475-86.
Yong EL, Logan S. Menopausal osteoporosis: screening, prevention and treatment. Singapore Med J. 2021;62(4):159-66.
You Y, Liu J, Zhang L, Li X, Sun Z, Dai Z, et al. WTAP-mediated m(6)A modification modulates bone marrow mesenchymal stem cells differentiation potential and osteoporosis. Cell Death Dis. 2023;14(1):33.
Huang B, Su Y, Shen E, Song M, Liu D, Qi H. Extracellular vesicles from GPNMB-modified bone marrow mesenchymal stem cells attenuate bone loss in an ovariectomized rat model. Life Sci. 2021;272:119208.
Svensson J, Ohlsson C, Karlsson MK, Lorentzon M, Lewerin C, Mellström D. Subclinical hyperthyroidism is associated with increased risk of vertebral fractures in older men. Osteoporos Int. 2021;32(11):2257-65.
Wang Y, Sun Y, Yang B, Wang Q, Kuang H. The management and metabolic characterization: hyperthyroidism and hypothyroidism. Neuropeptides. 2023;97:102308.
Tan BJ, Sugata K, Reda O, Matsuo M, Uchiyama K, Miyazato P, et al. HTLV-1 infection promotes excessive T cell activation and transformation into adult T cell leukemia/lymphoma. J Clin Invest. 2021;131(24).
Su H, Xue H, Gao S, Yan B, Wang R, Tan G, et al. Effect of Rhizoma Drynariae on differential gene expression in ovariectomized rats with osteoporosis based on transcriptome sequencing. Front Endocrinol (Lausanne). 2022;13:930912.
Eom YS, Wilson JR, Bernet VJ. Links between Thyroid Disorders and Glucose Homeostasis. Diabetes Metab J. 2022;46(2):239-56.
Nederstigt C, Corssmit EP, de Koning EJ, Dekkers OM. Incidence and prevalence of thyroid dysfunction in type 1 diabetes. J Diabetes Complications. 2016;30(3):420-5.
Zhao B, Grimes SN, Li S, Hu X, Ivashkiv LB. TNF-induced osteoclastogenesis and inflammatory bone resorption are inhibited by transcription factor RBP-J. J Exp Med. 2012;209(2):319-34.
Xiong Z, Zheng C, Chang Y, Liu K, Shu L, Zhang C. Exploring the Pharmacological Mechanism of Duhuo Jisheng Decoction in Treating Osteoporosis Based on Network Pharmacology. Evid Based Complement Alternat Med. 2021;2021:5510290.
Mukherjee A, Rotwein P. Selective signaling by Akt1 controls osteoblast differentiation and osteoblast-mediated osteoclast development. Mol Cell Biol. 2012;32(2):490-500.
Farrugia A. Albumin usage in clinical medicine: tradition or therapeutic? Transfus Med Rev. 2010;24(1):53-63.
Xie W, Ji L, Zhao T, Gao P. Identification of transcriptional factors and key genes in primary osteoporosis by DNA microarray. Med Sci Monit. 2015;21:1333-44.
Ji J, Wu S, Bao X, Liu S, Ye Y, Liu J, et al. Mediating oxidative stress through the Palbociclib/miR-141-3p/STAT4 axis in osteoporosis: a bioinformatics and experimental validation study. Sci Rep. 2023;13(1):19560.
Xu N, Wang Y, Xu Y, Li L, Chen J, Mai X, et al. Effect of subclinical hyperthyroidism on osteoporosis: A meta-analysis of cohort studies. Endocrine. 2020;69(1):39-48.
Ahmad T, Muhammad ZA, Nadeem S. Is Hypothyroidism Associated With Outcomes in Fracture Patients? Data From a Trauma Registry. J Surg Res. 2021;268:527-31.

No competing interests reported.

The causal relationship between hyperthyroidism, hypothyroidism, and osteoporosis: a Mendelian randomization study

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1