Association between Bone Mineral Density and Serum Uric Acid in the Elderly Chinese: Observational and Mendelian Randomization Analyses

doi:10.21203/rs.3.rs-3235353/v1

Summary

As an endogenous antioxidant, serum uric acid (SUA) can promote the proliferation and differentiation of osteoblasts and prevent osteoporosis. We found the higher SUA levels may associate with lower risk of osteoporosis (OP) in postmenopausal women.

Methods

A community-based cross-sectional study was designed on the basis of 2156 postmenopausal women and elderly men (≥50 years old) from Kunming and Jinshan, China. the bone mineral density (BMD) was measured using dual-energy X-ray absorptiometry (DXA).

Results

A total of 1755 participants were recruited. In observational study, a linear association between SUA and BMD existed only in postmenopausal women, and the association was still significant after other confounding factors were adjusted (P<0.05). The binary logistic regression analysis proved that the odds ratio(OR) for OP in SUA quartile 4 was the lowest (Model 1: OR=0.521<1, 95% confidence intervals (CI ), 0.358~0.757, P = 0.01). Restricted cubic spline (RCS) concluded no nonlinear relationship of SUA with BMD in both women and men (P>0.05). Moreover, the Mendelian randomization (MR)analysis showed no causal effect of SUA on BMD at different skeletal sites (IVW causal effect -0.02, 95% CI -0.08 to 0.04, P = 0.55, and MR- -Egger causal effect -0.03, 95% CI -0.13 to 0.06, P =0.50).

Conclusions

Our study concluded a positive linear relationship between SUA and BMD only in the postmenopausal women, and a higher SUA level could predict a higher BMD and a lower risk of OP. However, we found no casual effect of SUA on BMD.

serum uric acid

bone mineral density

osteoporosis

elderly Chinese

Serum uric acid (SUA) accounts for 50% of antioxidative processes in human body^{[1, 2]}. As a purine metabolite, SUA scavenges carbon-centered and peroxyl radicals^[3], inactivates peroxynitrite^[4] into nitro oxide^[5], blocks nitrosylation reactions^[6], and chelates metal ions such as iron ions^[7], thereby reducing oxidative stress in the body. Excessive residual free radicals cause DNA damage and induce protein oxidation, lipid peroxidation and protein carbonyl formation^[8].

Oxidation contributes to the development of osteoporosis (OP)^[9]. Excess reactive oxygen leads to abnormal proliferation, apoptosis or differentiation of bone marrow mesenchymal stem cells, osteoblasts and osteoclasts. As the metabolic balance between bone reconstruction and formation is disrupted, OP occurs^{[9, 10]}. However, the role SUA in bone metabolism is paradoxical. SUA can enhance intracellular oxidation through stimulating the overproduction of reactive oxygen species (ROS) production and reduced nicotinamide adenine dinucleotide phosphate oxidase ( NADPH) oxidase in the presence of oxidants^[11]. Several studies have shown a negative association between high SUA level and vitamin D deficiency^[12–14], but this association has been rebutted by others ^{[15, 16]}.

Clinical studies on the relationship between SUA and osteoporosis have been widely conducted globally. In a study of 1705 men aged over 70 years, the SUA level increased with BMD at all body sites^[16]. However, inconsistent findings have been reported in many other studies of different designs^[17–20]. Here, we furthered the exploration for the association between SUA and BMD in Chinese elderly people using observational and Mendelian randomization analyses.

Ethical Statement

This study was in accordance with the Declaration of Helsinki (revision 2) and approved by the Institutional Review Board of Longhua Hospital, Affiliated to Shanghai University of Traditional Chinese Medicine (No. 2016LCSY065). Every participant enrolled in the study provided informed consent. And the protocol of this study has been registered into ClinicalTrials.gov (ID: NCT02958020).

Study Population and Data Source

This was a community-based cross-sectional study in which all subjects were randomly selected from two communities (Kunming in Yunnan and Jinshan in Shanghai, China) from 2020.01 to 2021.12. All participants had lived in urban or rural areas for more than 5 years, including postmenopausal women and men aged over 50 years. Participants with the following conditions were excluded: (1) mental diseases that failed the participant to cooperate during the study; (2) severe primary diseases such as heart, liver, kidney, or cancer; (3) metabolic diseases such as thyroid, hyperparathyroidism, hypogonadism, and diabetes; (4) autoimmune diseases such as rheumatoid arthritis, systemic lupus erythematosus, and compulsory spondylitis;(5) recent use of bisphosphonates, glucocorticoids, estrogens, diuretics, allopurinol and other drugs that might affect uric acid and bone metabolism indicators; (6) refusal to sign informed consent. Finally, a total of 1575 people were eligible, including 583 males and 992 postmenopausal women. The selection process is shown in Fig.1.

To obtain genetic evidence with MR approach, we selected the largest GWAS published to date for SUA^[21]and dual energy x-ray absorptiometry (DXA) BMD^[22], which were from BioBank Japan Project (BBJ) and 30 epidemiological studies, respectively. GWAS summary statistics for BMD (unit, g/cm2) were downloaded from the Genetic Factors for Osteoporosis Consortium website (GEFOS, http://www.gefos.org/) and GWAS summary statistics for SUA were obtained from IEU GWAS database ( https://gwas.mrcieu.ac.uk/datasets/ ).

Clinical Evaluation and Measurement

Under the guidance of trained doctors, all the participants completed a questionnaire. Collected their socio-demographic data, including name, age, gender, profession, marital status, smoking and alcohol history, lifestyle (mainly physical exercise), menstrual history, family history, bone fracture history, history of complications and drug or supplement use. Height and weight were measured at baseline when the participants wore no shoes and only light clothing. The body mass index (BMI) was calculated as weight divided by height squared.

Biochemical Investigations

All the participants underwent the biochemical analysis in the morning from 8:00 to 10:00. Collected were their fasting SUA, serum creatinine (sCr), serum calcium (sCa), serum phosphorus, serum magnesium, and bone metabolism-related markers (BTMs), including osteocalcin in the form of osteocalcin (OST), procollagen type I N-peptide (PINP), β-cross-linked C-terminal (β-CTX), 25-hydroxyvitamin D[25(OH)D] and alkaline phosphatase(ALP). All the blood samples were collected and tested by Jinyu Medical Testing Center.

Dual-Energy X-Ray Absorptiometry (DXA) Scan and Osteoporosis Diagnosis

The BMD (g/cm2) values of the lumber spine (L1–4), femoral neck and total hip were measured by the DXA (HOLOGIC, American Hologic Wi). Collected were BMD values in the femur t, femur z, total hip t, total hip z, spine t, spine z. The accuracy of the instrument was within 1%, and the repeated measurement error within 1%. All measurements were performed by an experienced operator on the same machine following standardized procedures to reduce the chance of error. After turning on the machine, a standard quality control program was implemented on the machine daily before the participants were checked.

Single Nucleotide Polymorphisins Selection

Eligible instrumental single nucleotide polymorphisins (SNPs) were selected from the GWAS data of SUA, according to genome-wide significance P<5 * 10⁻⁸, linkage disequilibrium r²<0.001, clumping window size kb>10,000 and minor allele frequency (MAF) >0.01. R² values (variance of exposure explained by the instrumental variables) were calculated with MR Steiger test. F statistics were used to assess weak instrumental variable with the formula below, where n represents sample size and k is the number of instrumental variables.

Statistical Analysis

In the observational study, all statistical analyses were performed with SPSS version 25.0 (SPSS, Inc, Chicago, IL, USA) and R software (R Foundation for Statistical Computing, version 4.1.1), with a P value <0.05 considered statistically significant. For continuous variables, nonparametric test was used for variables with skewed distributions or uneven variances. The mean± standard or the median (interquartile range) was used for general description. The classified variables were expressed by frequency or constituent ratio, and the differences between groups were analyzed by chi-square test. Rank sum test was used to compare the rank data. After adjusting for the confounding factors, a multiple linear regression model was used to evaluate the relationship between SUA and BMD. A multiple logistic regression model was used to evaluate the association of OP with SUA in different quartiles. Restricted cubic spline with five knots at 5th, 35th, 50th, 65th, and 95th centiles was used to model the potential nonlinear association of SUA with BMDs. Linear and nonlinear models were both adjusted with age, BMI, job type ,25(OH)D, ALP, PINP, β-CTX and OST.

In the two-sample MR study, the Mendelian randomization estimates for the associations of SUA concentrations with total body BMD were obtained using the inverse-variance weighted meta-analysis (IVW) complemented with the MR-Egger, weighted median, simple mode and weighted mode methods. To eliminate the heterogeneity in each SNP outcome (P values for Cochran’s Q test < 0.05), IVW method was implemented. To define the direction of causality, we further performed the Steiger test. In the sensitivity analysis, to assess the influence of potential pleiotropy on the causal effect, we adopted the MR-Egger method to detect heterogeneity. To identify influential SNPs, we performed a “leave-one-out” sensitivity analysis. All analyses were conducted using the R package “Two Sample MR”.

Basic Characteristics in the Whole and Different Quartiles of Participants

A total of 1575 participants were recruited, including 992 postmenopausal women and 583 older men. As shown in Table 1, except exercise frequency, the age, BMI, cigarette smoking, alcohol, retired or not, job type, family OP history and distributed area were significantly different between the elderly men and postmenopausal women. The distributions of demographic, biochemical, and bone mass characteristics in both men and women of different SUA quartiles are shown in Table 2 and Table 3.

In the elderly men, there were no significant differences in age, cigarette smoking, alcohol, retired or not, job type, family history of OP between males in quartiles of SUA level. However, the BMD values in the lumber spine L1-L4, total hip and femoral neck showed difference between quartiles of SUA level, and the BTMs of β-CTX, OST, P1NP showed difference between quartiles of SUA level (Table 2).

In the postmenopausal women, there was a strong association of BMI, age, job type and some biochemical variables (β-CTX, OST, PINP, 25(OH)D) in different quartiles of SUA value. The BTMs and the BMD except that in the femoral neck also exhibited difference between SUA quartiles (Table 3).

Association between SUA and BMD at Different Skeletal Sites

Multiple linear regression analyses showed association of BMD at multiple skeletal sites with SUA in different quartiles. In the elderly men, the lumbar spine and hip BMD were associated with quartiles of SUA levels after adjustment for age in all participants (P<0.05), but the association disappeared after all other variables were incorporated, including age, BMI, job type ,25(OH)D, ALP, PINP, β-CTX, OST; moreover, the femoral neck BMD had no association with SUA in different quartiles (Table 4). In postmenopausal women, the lumbar spine BMD and hip BMD were associated with the SUA level, and the association was still significant (P<0.05) after all variables (including age, BMI, job type, 25(OH)D, ALP, PINP, β-CTX, OST) were adjusted. The femoral neck BMD had no direct association between quartiles of SUA, but the association was significant after all-variable adjustment (Table 5).

Furthermore, we used RCS to flexibly model and visualize the relation of SUA with BMD (Figure 2 and 3). We found no nonlinear association between SUA and BMD at all skeletal sites (all P for non-linearity >0.05).

Binary Logistic Regression Analysis with OP as a Dependent Variable in Postmenopausal

Women

In the postmenopausal women, the binary logistic regression analysis was performed using OP as a dependent variable and SUA quartiles as an independent variable (Supplementary Table 1). Compared to those in the lowest quartile of SUA levels, the OR in the highest SUA quartile in Model 1 was 0.521 (95%CI, 0.358~0.757, P =0.01), which is lower than SUA quartile levels of 2 and 3. After adjustment with other variables, including age, BMI, job type, 25(OH)D, ALP, PINP, β-CTX and OST in Model 2 and 3, the OR in SUA quartile 4 (Model 2:OR=0.307,95%CI:0.202~0.466; Model 3 : OR=0.320,95%CI: 0.208~0.492) was still the lowest compare to SUA quartile levels of 2 and 3 (P<0.05).

Putative Causal Relationships between SUA and Total BMD

We firstly selected 43 genetic variants associated with SUA at a genome‐wide significant level (P < 5 *10^‐⁸), and excluded one SNP because of its opposite effective direction (supplementary data 1). Total direction was tested using Steiger test to ensure the impact of SUA on total BMD (Fig. S1). F statistics of 42 genetic variants ranged from 30.86 to 1254.8, indicating no weak instrumental variants (supplementary data 2). Cochran’s Q test indicated heterogeneity (P = 4*10^-11) and MR-Egger intercept indicated no pleiotropy (P = 0.7).

The IVW and MR-Egger methods showed that SUA had no significant causal effects on total BMD (IVW causal effect -0.02, 95% CI -0.08 to 0.04, P = 0.55; MR-Egger causal effect -0.03, 95% CI -0.13 to 0.06, P =0.50) (supplementary Fig.4 and Fig.S2). Other MR methods also indicated no significant causal effects (supplementary Table S2).

The sensitivity analysis suggested no significant horizontal pleiotropy (MR-Egger regression intercept close to zero, P>0.05). We performed a “leave-one-out” analysis, finding no SNPs influencing the causal link (supplementary Fig.S3). The funnel plots provided directional horizontal pleiotropy for each outcome (supplementary Fig.S4).

In this community-based observational study, we observed a strong positive relationship between the level of SUA and BMD at the lumbar spine and total hip in postmenopausal women. Our RCS analysis further discovery no significant dose-response curve between the levels of SUA and BMD. In addition, we found no direct association between SUA and femoral neck BMD. However there was a significant association between SUA and femoral neck BMD after adjusting for age, BMI, job type, 25(OH)D, ALP, PINP, β-CTX and OST. Pirro et al ^[23] also reported the relationship may be significant only when other confounding factors, such as adiposity, were included. This may be explained by that the SUA level is affected by various factors, such as BMI and hormonal level. Our finding is consistent with the study reported by Han^[24]that the BMD is positively correlated with SUA in postmenopausal women. Nevertheless, his study only focused on the BMD of lumbar spine, not at all skeletal sites.

In the elderly man, we found the relationship was significant when no confounding factors, or just age, were adjusted. The RCS analysis yielded a similar result with that in the postmenopausal women, that is, there was no causal relationship between SUA and BMD. Our finding is in line with that reported by Yan DD et al^[25] that a positive relationship exists between the SUA and BMD among postmenopausal women, but not in the elderly men. However, some clinical studies have achieved contrary results. Some^[26] suggested that the BMD in the elderly men is also affected by the SUA. These controversial findings may result from the selection of confounding variables. Therefore, we adopted MR to further analyze the relationship between SUA and BMD during the elderly.

In this study, we did not determine a causal effect of SUA on BMD by two-sample MR analysis, which is consistent with the results in other published observational studies. Xiong et al^[27] identified no causal effect between SUA and bone-related outcomes, but a strong association in postmenopausal women and elderly men. The results from the observational study are frequently inconsistent with those from the MR analysis, suggesting the relationship between SUA and BMD may be just a result of confounding factors, rather than a direct effect.

Studies have shown that oxidative stress is the basic mechanism of age-related bone loss, which is caused by excessive ROS and insufficient antioxidants ^[28]. With aging, the level of glutathione, a main antioxidant in the human body, gradually decreases, and the absorption capacity of antioxidants in the intestinal tract weakens. As the antioxidants in the body are not enough to counteract ROS, oxidative stress occurs. Oxidative stress can enhance the differentiation of osteoclasts^[29], increase bone resorption, and even induce apoptosis of osteoblasts and osteocytes^[30], bone remodeling imbalance, all working to cause bone loss. SUA is the main antioxidant in human plasma, accounting for the clearance of 60% free radicals in the serum^[31]. Therefore, it has been suggested that SUA may play a protective role against oxidative stress in diseases, such as OP. More clinical and basic studies have been carried out to explore the relationship between SUA and OP. For example, a cross-sectional study on 5199 subjects in Shanghai, China^[25], suggests that SUA can protect bone metabolism in postmenopausal women. A cross-sectional study on Chinese people over 60 years ^[26] showed that subjects with higher SUA levels had significantly higher BMD and T values than those with lower SUA levels, and the prevalence of OP decreased by 23% to 26% in the former population.

In addition, hyperuricemia may be involved in the pathogenesis of a variety of diseases, such as cardiovascular disease. Hyperuricemia can inhibit the formation of osteoblasts. Studies have evaluated the effect of SUA on BMD in cases of hyperuricemia or gouty arthritis. Paik ^[32]et al. followed up 103799 women for a long term. The results showed that a history of gout can increase the risk of hip fracture by 38%, but does not change the risk of wrist fracture significantly. This contradicts the previous theory. We speculate that SUA only in a normal range can play an antioxidative role to prevent bone loss and OP. Zhang ^[17] et al. found that SUA is independent of BMD after adjusting confounding variables in the multivariate analysis on a large sample of American adults. Therefore, more studies should be performed to verify whether their relationship varies with the concentration of SUA, the location of deposition (trunk bone, limb bone), or the study population. Therefore, considering the protective and toxic effects of SUA, our results should be interpreted carefully in the clinical environment. A proper SUA level should be maintained to reduce cardiovascular risk, but overtreatment should be avoided. The optimal SUA level needs to be individualized after assessing each person's risk of OP and cardiovascular disease.

However, there were some limitations in our study. First, this study was a cross-sectional study, which could not determine the causal relationship because of the limitations of the data. Second, we only explore the relationship in postmenopausal women and elderly men, but not the general population. Therefore, a cohort study needed to be conducted to explore the casual relationship in all age groups.

In summary, our study concluded a positive linear relationship between SUA and BMD in the postmenopausal women, and the higher SUA level predicts a higher BMD and a lower risk of osteoporosis. However, we found no casual effect of SUA on BMD. Nevertheless, further efforts are required to establish a large-cohort study to deep explore the relationship between SUA and BMD.

SUA serum uric acid

OP osteoporosis

BMD bone mineral density

CI confidence intervals

MR mendelian randomization

ROS reactive oxygen species

sCr serum creatinine

sCa serum calcium ,

BTMs bone metabolism-related markers

OST osteocalcin

PINP procollagen type I N-peptide

β-CTX β-cross-linked C-terminal

25(OH)D 25-hydroxyvitamin D

ALP alkaline phosphatase

OR odds ratio

BMI body mass index

DXA dual energy x-ray absorptiometry

IVW inverse-variance weighted meta-analysis

Author Contributions

MX, CY, CH, XJ, HC, HS, JW, LY, YC, JP, HX and JW contributed to the acquisition, analysis, and interpretation of data. MX and CY drafted the manuscript. MX, CY, CH made the statistical analysis. MX, CY, CH contributed equally to this work as first authors. LJ and JW supervised the study and helped revise the manuscript. JW, LJ, ZT and QL had full access to all of the data in the study and take responsibility for the integrity and accuracy of the data analysis. JW, LJ, ZT and QL contributed equally as corresponding authors. All authors contributed to the article and approved the submitted version.

Funding

This work was partially supported by the National Key R&D Program of China (2018YFC1704300), the National Natural Science Foundation of China (81730107, 81973883), Young Qihuang Scholars of the National Administration of Traditional Chinese Medicine, Shanghai Municipal Science and Technology Major Project (2017SHZDZX01), the Shanghai Scientific Research Project (17401971100), Shanghai "Science and Technology Innovation Action Plan" medical innovation research Project (21Y11921400), Traditional Chinese Medicine Research Project of Shanghai Health Commission (2022CX001), Shanghai TCM Medical Center of Chronic Disease (2022ZZ01009) and Three Years Action to Accelerate the Development of Traditional Chinese Medicine Plan (ZY(2018-2020)-CCCX-3003), the Program for Innovative Research Team of Ministry of Education of China (IRT1270), the Program for Innovative Research Team of Ministry of Science and Technology of China (2015RA4002), Yunnan Wang Yongjun Expert Workstation (2019IC023) and Sanming Project of Medicine in Shenzhen (SZSM201808072), Innovation Team and Talents Cultivation Program of National Administration of Traditional Chinese Medicine(ZYYCXTD-C-202202)

Acknowledgments

We would like to thank all the participants for their cooperation, especially in Yunnan Provincial Hospital of Traditional Chinese Medicine, Kunming, Yunnan, China. Thanks all the fundings supported for this study.

Declarations

Ethics approval and consent to participate

This study was approved by the ethics committee of the Longhua Hospital, Affiliated to Shanghai University of Traditional Chinese Medicine (No. 2016LCSY065).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Ames BN, Cathcart R, Schwiers E, et al. Uric acid provides an antioxidant defense in humans against oxidant- and radical-caused aging and cancer: a hypothesis. Proc Natl Acad Sci U S A. 1981;78:6858–62.
Lane NE, Parimi N, Lui LY, et al. Association of serum uric acid and incident nonspine fractures in elderly men: the Osteoporotic Fractures in Men (MrOS) study. J Bone Miner Res. 2014;29:1701–7.
Muraoka S, Miura T. Inhibition by uric acid of free radicals that damage biological molecules. Pharmacol Toxicol. 2003;93:284–9.
Kuzkaya N, Weissmann N, Harrison DG, et al. Interactions of peroxynitrite with uric acid in the presence of ascorbate and thiols: implications for uncoupling endothelial nitric oxide synthase. Biochem Pharmacol. 2005;70:343–54.
Gersch C, Palii SP, Kim KM, et al. Inactivation of nitric oxide by uric acid. Nucleosides Nucleotides Nucleic Acids. 2008;27:967–78.
de La Pomélie D, Santé-Lhoutellier V, Gatellier P. Mechanisms and kinetics of heme iron nitrosylation in an in vitro gastro-intestinal model. Food Chem. 2018;239:86–93.
Pasalic D, Marinkovic N, Feher-Turkovic L. Uric acid as one of the important factors in multifactorial disorders–facts and controversies. Biochem Med (Zagreb). 2012;22:63–75.
Moldogazieva NT, Mokhosoev IM, Mel'nikova TI et al. (2019) Oxidative Stress and Advanced Lipoxidation and Glycation End Products (ALEs and AGEs) in Aging and Age-Related Diseases. Oxid Med Cell Longev 2019:3085756.
Kimball JS, Johnson JP, Carlson DA. Oxidative Stress and Osteoporosis. J Bone Joint Surg Am. 2021;103:1451–61.
Sun J, Ming L, Shang F, et al. Apocynin suppression of NADPH oxidase reverses the aging process in mesenchymal stem cells to promote osteogenesis and increase bone mass. Sci Rep. 2015;5:18572.
Kurajoh M, Fukumoto S, Yoshida S, et al. Uric acid shown to contribute to increased oxidative stress level independent of xanthine oxidoreductase activity in MedCity21 health examination registry. Sci Rep. 2021;11:7378.
Chien KL, Hsu HC, Chen PC, et al. Total 25-hydroxyvitamin D concentration as a predictor for all-cause death and cardiovascular event risk among ethnic Chinese adults: a cohort study in a Taiwan community. PLoS ONE. 2015;10:e0123097.
Faridi KF, Lupton JR, Martin SS, et al. Vitamin D deficiency and non-lipid biomarkers of cardiovascular risk. Arch Med Sci. 2017;13:732–7.
Han Y, Han K, Zhang Y, et al. Serum 25-hydroxyvitamin D might be negatively associated with hyperuricemia in U.S. adults: an analysis of the National Health and Nutrition Examination Survey 2007–2014. J Endocrinol Invest. 2022;45:719–29.
Chen Y, Cheng J, Chen Y, et al. Association between serum vitamin D and uric acid in the eastern Chinese population: a population-based cross-sectional study. BMC Endocr Disord. 2020;20:79.
Nabipour I, Sambrook PN, Blyth FM, et al. Serum uric acid is associated with bone health in older men: a cross-sectional population-based study. J Bone Miner Res. 2011;26:955–64.
Zhang D, Bobulescu IA, Maalouf NM, et al. Relationship between serum uric Acid and bone mineral density in the general population and in rats with experimental hyperuricemia. J Bone Miner Res. 2015;30:992–9.
Dalbeth N, Topless R, Flynn T, et al. Mendelian randomization analysis to examine for a causal effect of urate on bone mineral density. J Bone Miner Res. 2015;30:985–91.
Lee YH, Song GG. Uric acid level, gout and bone mineral density: A Mendelian randomization study. Eur J Clin Invest. 2019;49:e13156.
Pan K, Yao X, Liu M, et al. Association of Serum Uric Acid Status With Bone Mineral Density in Adolescents Aged 12–19 Years. Front Med (Lausanne). 2020;7:255.
Kanai M, Akiyama M, Takahashi A, et al. Genetic analysis of quantitative traits in the Japanese population links cell types to complex human diseases. Nat Genet. 2018;50:390–400.
Medina-Gomez C, Kemp JP, Trajanoska K, et al. Life-Course Genome-wide Association Study Meta-analysis of Total Body BMD and Assessment of Age-Specific Effects. Am J Hum Genet. 2018;102:88–102.
Pirro M, Mannarino MR, Bianconi V, et al. Uric acid and bone mineral density in postmenopausal osteoporotic women: the link lies within the fat. Osteoporos Int. 2017;28:973–81.
Han W, Bai X, Wang N, et al. Association between lumbar bone mineral density and serum uric acid in postmenopausal women: a cross-sectional study of healthy Chinese population. Arch Osteoporos. 2017;12:50.
Yan DD, Wang J, Hou XH, et al. Association of serum uric acid levels with osteoporosis and bone turnover markers in a Chinese population. Acta Pharmacol Sin. 2018;39:626–32.
Chen F, Wang Y, Guo Y, et al. Specific higher levels of serum uric acid might have a protective effect on bone mineral density within a Chinese population over 60 years old: a cross-sectional study from northeast China. Clin Interv Aging. 2019;14:1065–73.
Xiong A, Yao Q, He J, et al. No causal effect of serum urate on bone-related outcomes among a population of postmenopausal women and elderly men of Chinese Han ethnicity–a Mendelian randomization study. Osteoporos Int. 2016;27:1031–9.
Valko M, Leibfritz D, Moncol J, et al. Free radicals and antioxidants in normal physiological functions and human disease. Int J Biochem Cell Biol. 2007;39:44–84.
Huh YJ, Kim JM, Kim H, et al. Regulation of osteoclast differentiation by the redox-dependent modulation of nuclear import of transcription factors. Cell Death Differ. 2006;13:1138–46.
Almeida M, Han L, Martin-Millan M, et al. Skeletal involution by age-associated oxidative stress and its acceleration by loss of sex steroids. J Biol Chem. 2007;282:27285–97.
Fabbrini E, Serafini M, Colic Baric I, et al. Effect of plasma uric acid on antioxidant capacity, oxidative stress, and insulin sensitivity in obese subjects. Diabetes. 2014;63:976–81.
Paik JM, Kim SC, Feskanich D, et al. Gout and Risk of Fracture in Women: A Prospective Cohort Study. Arthritis Rheumatol. 2017;69:422–8.

Table 1. Clinical characteristics of the study samples

		All	male	female	P
n		1755	583	992
Age(years)		65(58,71)	67(59,73)	64(58,69)	<0.01
BMI(Kg/m²)		23.9(22,25.7)	24.5(22.6,26.2)	23.4(21.7,25.4)	<0.01
smoking
	non-smoker	1230(78.1%)	273(46.8%)	957(96.5%)	<0.01
	current-smoker	247(15.7%)	215(36.9%)	32(3.2%)
	ex-smoker	98(6.2%)	95(16.3%)	3(0.3%)
alcohol
	non	1317(83.9%)	353(60.7%)	964(97.7%)	<0.01
	current	200(12.7%)	182(31.3%)	18(1.8%)
	ex	52(3.3%)	47(8.1%)	5(0.5%)
Retire
	yes	368(23.6%)	182(31.4%)	186(19%)	<0.01
	no	1189(76.4%)	398(68.6%)	791(81%)
job type
	1	1230(78.1%)	273(46.8%)	957(96.5%)	<0.01
	2	247(15.7%)	215(36.9%)	32(3.2%)
	3	98(6.2%)	95(16.3%)	3(0.3%)
Exercise frequency
	1	280(19.5%)	96(18%)	184(20.4%)	0.281
	2	35(2.4%)	10(1.9%)	25(2.8%)
	3	109(7.6%)	36(6.8%)	73(8.1%)
	4	1010(70.4%)	391(73.4%)	619(68.7%)
Family history of osteoporosis
	0	1443(92.7%)	561(97.2%)	882(90.1%)	<0.01
	1	113(7.3%)	16(2.8%)	97(9.9%)
area
	southeast	197(12.5%)	44(7.5%)	153(15.4%)	<0.01
	southwest	1378(87.5%)	539(92.5%)	839(84.6%)

Notes: Job type:1 for brain activities;2 for physical activities;3 for both of all.

Exercise frequency:1for (none or almost none) ;2 for (1-3 times/month) ;3 for (3-5 times/week) ;4 for (1-2 times/month). BMI body mass index.

Table 2.Basic population characteristics, BMC levels, and biochemical variables stratified by quartiles of UA in elderly man

		Q1	Q2	Q3	Q4	P
Age (years)		67(59,73)	67(58,74)	68(59,73)	67(60,72)	0.979
BMI (Kg/m²)		23.5(21.2,25.7)	24.5(22.7,26.4)	24.7(23.1,26.4)	25.1(23.4,26.5)	<0.001
smoking
	non-smoker	63(43.2%)	69(47.3%)	74(50.3%)	67(46.5%)	0.48
	current-smoker	62(42.5%)	56(38.4%)	50(34%)	47(32.6%)
	ex-smoker	21(14.4%)	21(14.4%)	23(15.6%)	30(20.8%)
Alcohol
	non	92(63%)	87(59.6%)	90(61.2%)	84(58.7%)	0.835
	current	43(29.5%)	50(34.2%)	42(28.6%)	47(32.9%)
	ex	11(7.5%)	9(6.2%)	15(10.2%)	12(8.4%)
Retire
	yes	95(65.5%）	98(67.6%)	105(71.4%)	100(69.9%)	0.712
	no	50(34.5%)	47(32.4%)	42(28.6%)	43(30.1%)
Job type
	1	50(34.7%)	32(22.2%)	33(22.6%)	43(30.5%)	0.079
	2	63(43.8%)	81(56.3%)	80(54.8%)	60(42.6%)
	3	31(21.5%)	31(21.5%)	33(22.6%)	38(27%)
Exercise frequency
	1	25(18.9%)	21(15.8%)	22(16.3%)	28(21.1%)	0.456
	2	5(3.8%)	1(0.8%)	1(0.7%)	3(2.3%)
	3	5(3.8%)	12(9%)	11(8.1%)	8(6.0%）
	4	97(73.5%)	99(74.4%)	101(74.8%)	94(74.8%)
Family history of osteoporosis
	0	139(96.5%)	142(99.3%)	143(97.9%)	137(95.1%)	0.135
	1	5(3.5%)	1(0.7%)	3(2.1%)	7(4.9%)
Biochemical variables
	25(OH)D (ng/mL)	22.0(18.1,26.4)	23.2(19.2,28.1)	23.2(19.3,26.4)	23.0(18.9,26.9)	0.305
	β-CTX (ng/L)	0.34(0.26,0.44)	0.31(0.23,0.44)	0.25(0.20,0.35)	0.25(0.19,0.38)	<0.001
	OST (ng/L)	12.6(9.6,15.9)	11.9(9.1,14.8)	11.3(8.8,13.6)	10.3(7.9,13.8)	<0.001
	PINP (ng/L)	46.3(36.0,59.4)	45.0(34.8,55.8)	39.9(30.4,50.1)	39.8(32.6,50.8)	0.002
	ALP (U/L)	80(62,92)	73(62,86)	74(62,86)	75(63,88)	0.157
	Phosphorus (mmol/L)	1.06(0.94,1.2)	1.06(0.93,1.22)	1.06(0.94,1.19)	1.09(0.99,1.26)	0.123
	Magnesium（mmol/L）	0.92(0.87,0.99)	0.93(0.88,0.99)	0.93(0.89,0.99)	0.91(0.85,0.98)	0.069
	Calcium (mmol/L)	2.35(2.29,2.4)	2.36(2.29,2.4)	2.35(2.29,2.4)	2.36(2.3,2.44)	0.11
	Creatinine (μmol/L)	82(74,91)	86(78,96)	92(81,102)	94(83,104)	<0.001
BMC (g/cm²)
	Lumbar1-4 BMC	0.86(0.78,0.96)	0.89(0.82,0.99)	0.92(0.83,1.00)	0.90(0.83,1.03)	0.008
	Total Hip BMC	0.84(0.75,0.93)	0.89(0.79,1.01)	0.93(0.82,1.02)	0.91(0.81,0.99)	<0.001
	Femoral neck BMC	0.71(0.63,0.84)	0.77(0.67,0.90)	0.78(0.66,0.88)	0.76(0.68,0.85)	0.041

Notes: Job type:1 for brain activities;2 for physical activities;3 for both of all.

Exercise frequency:1for (none or almost none) ;2 for (1-3 times/month) ;3 for (3-5 times/week) ;4 for (1-2 times/month).

BMI body mass index, 25(OH)D 25-hydroxyvitamin D, β-CTX β-crosslaps of type I collagen containing cross-linked C-telopeptide，OST osteocalcin，P1NP tropocollagen type 1 N-terminal propeptide ，ALP alkaline phosphatase，BMC bone mineral content.

Table 3.Basic population characteristics, BMC levels, and biochemical variables stratified by quartiles of UA in postmenopausal women

		Q1	Q2	Q3	Q4	P
Age (years)		57(53,58)	58(53,61)	57(55,59)	58(55,60)	<0.01
BMI (Kg/m²)		21.8(20.3,23.4)	22.9(21.3,24.4)	23.7(21.9,26.4)	24.2(22.9,26)	<0.01
Menopausal age(years)		50(49,53)	50(48,52)	50(48,53)	50(48,52)	0.217
smoking
	non-smoker	242(96.8%)	238(94.8%)	243(98%)	237(97.5%)	0.227
	current-smoker	8(3.2%)	13(5.2%)	5(2%)	6(2.0%)
	ex-smoker	0	0	0	0
alchole
	non	239(96.8%)	243(96.8%)	244(98.8%)	238(98.3%)	0.559
	current	6(2.4%)	6(2.4%)	2(2.4%)	4(1.7%)
	ex	2(0.8%)	2(0.8%)	1(0.8%)	0(0.0%)
retire
	yes	186(75.9%)	205(82.7%）	198(81.5%)	202(83.8%)	0.128
	no	59(24.1%)	43(17.3%）	45(18.5%)	39(16.2%)
Job type
	1	78(34.4%)	62(25.5%)	68（29.1%）	62(26.5%)	0.007
	2	89(39.3%)	133(54.7%）	119(50.9%)	133(56.8%)
	3	60(26.4%)	48（19.8%）	47(20.1%)	39(16.7%)
exercise frequency
	1	48(21.4%)	41(18%)	44(19.5%)	51(22.9%)	0.91
	2	4(1.8%)	8(3.5%)	6(2.7%)	7(3.1%)
	3	20(8.9%)	18(7.9%)	16(7.1%)	19(8.5%)
	4	152(67.9%)	161(70.6%)	160(70.6%)	146(65.5%)
Family history of osteoporosis
	0	222(91.4%)	222(89.2%)	223(90.3%)	215(89.6%)	0.858
	1	21(8.6%)	27(10.8%)	24(9.7%)	25(10.4%)
Biochemical variables
	25(OH)D (ng/mL)	24.2(20.0,27.5)	22.3(17.3,27.0)	23.7(19.2,28.1)	23.3(18.2,28.0)	<0.001
	β-CTX (ng/L)	0.46(0.34,0.56)	0.42(0.33,0.56)	0.43(0.3,0.56)	0.38(0.26,0.51)	<0.001
	OST (ng/L)	19.6(15.2,22.8)	17.2(13.7,23.5)	17.5(12.9,21.6)	15.3(11.3,19.3)	<0.001
	P1NP (ng/L)	66.5(52.9,76.9)	55.4(45.6,72.7)	57.9(51.0,77.8)	55.1(38.3,71.8)	<0.001
	ALP (U/L)	83(70,99)	82(67,94)	79(68,100)	82(68,98)	0.324
	Phosphorus (mmol/L)	1.15(1.07,1.26)	1.17(1.1,1.27)	1.15(1.07,1.28)	1.21(1.11,1.31)	0.102
	Magnesium (mmol/L)	0.93(0.9,0.97)	0.92(0.87,0.96)	0.93(0.87,0.98)	0.93(0.89,0.97)	0.265
	Calcium (mmol/L)	2.39(2.33,2.44)	2.37(2.31,2.43)	2.39(2.33,2.46)	2.39(2.33,2.43)	0.063
	Creatinine (μmol/L)	61(54,68)	66(60,76)	67(59,76)	68(61,77)	<0.001
BMC(g/cm²)
	Lumbar1-4 BMC	0.82(0.75,0.92)	0.82(0.74,0.91)	0.9(0.77,0.98)	0.88(0.81,0.97)	0.008
	Total Hip BMC	0.79(0.72,0.86)	0.81(0.7,0.89)	0.88(0.76,0.95)	0.84(0.76,0.92)	0.005
	Femoral neck BMC	0.67(0.58,0.73)	0.67(0.6,0.75)	0.72(0.64,0.81)	0.72(0.64,0.76)	0.056

Notes: Job type:1 for brain activities;2 for physical activities;3 for both of all.

Exercise frequency:1for (none or almost none) ;2 for (1-3 times/month) ;3 for (3-5 times/week) ;4 for (1-2 times/month).

BMI body mass index, 25(OH)D 25-hydroxyvitamin D, β-CTX β-crosslaps of type I collagen containing cross-linked C-telopeptide，OST osteocalcin，PINP tropocollagen type 1 N-terminal propeptide ; ALP alkaline phosphatase，BMC bone mineral content.

Table 4 .Multiple linear regression analysis of the relationship between the serum UA levels and bone mineral content(BMC) for total body in different models in postmenopausal women

BMD		Βst	SE	P	95%CI	AR²
Lumbar1-4 BMC
	Mode1	0.097	0.00006	0.002	0.00006~0.0003	0.008
	Mode2	0.176	0.00006	<0.001	0.0002~0.0004	0.139
	Mode3	0.181	0.00006	<0.001	0.0002~0.0004	0.139
	Mode4	0.156	0.00006	<0.001	0.0002~0.0004	0.186
Total Hip BMC
	Mode1	0.092	0.00006	0.004	0.00005~0.0003	0.007
	Mode2	0.166	0.00006	<0.001	0.0002~0.0004	0.117
	Mode3	0.176	0.00006	<0.001	0.0002~0.0004	0.124
	Mode4	0.153	0.00006	<0.001	0.00017~0.0004	0.163
Femoral neck BMC
	Mode1	0.050	0.00006	0.116	-0.00002~0.0002	0.002
	Mode2	0.121	0.00006	<0.001	0.0001~0.0003	0.102
	Mode3	0.127	0.00006	<0.001	0.0001~0.0003	0.104
	Mode4	0.113	0.00006	<0.001	0.00009~0.0003	0.123

Notes:

Model 1: unadjusted;

Model 2: adjusted for age;

Model 3: adjusted for age, BMI, job type;

Model 4: adjusted for age, BMI, job type ,25OHD, ALP, PINP, β-CTX, OST.

Abbreviation: AR², adjusted R². Bst, Beta standardized coefficients.

Table 5. Multiple linear regression analysis of the relationship between the serum UA levels and bone mineral content (BMC) for total body in different models in elderly man

BMD		Βst	SE	P	95%CI	AR²
Lumbar1-4 BMC
	Mode1	0.110	0.00008	0.008	0.00005~0.00034	0.010
	Mode2	0.110	0.00008	0.008	0.00005~0.00034	0.009
	Mode3	0.069	0.00007	0.094	-0.00002~0.0002	0.063
	Mode4	0.043	0.00007	0.291	-0.00007~0.0002	0.107
Total Hip BMC
	Mode1	0.121	0.00007	0.004	0.00007~0.00034	0.015
	Mode2	0.122	0.00007	0.003	0.00007~0.00035	0.024
	Mode3	0.078	0.00007	0.055	-0.000003~0.0002	0.085
	Mode4	0.051	0.00007	0.206	-0.00005~0.0002	0.125
Femoral neck BMC
	Mode1	0.031	0.00008	0.459	-0.00010~0.0002	-0.01
	Mode2	0.032	0.00008	0.443	-0.00009~0.0002	0.002
	Mode3	-0.002	0.00008	0.969	-0.00016~0.00015	0.037
	Mode4	-0.025	0.00008	0.557	-0.0002~0.0001	0.058

Notes:

Model 1: unadjusted;

Model 2: adjusted for age;

Model 3: adjusted for age, BMI, job type;

Model 4: adjusted for age, BMI, job type ,25(OH)D,ALP,PINP, β-CTX, OST.

Abbreviation: AR², adjusted R². Bst, Beta standardized coefficients.

No competing interests reported.

Association between Bone Mineral Density and Serum Uric Acid in the Elderly Chinese: Observational and Mendelian Randomization Analyses

Status:

Version 1

Abstract

Figures

Introduction

Methods

Ethical Statement

Study Population and Data Source

Clinical Evaluation and Measurement

Biochemical Investigations

Dual-Energy X-Ray Absorptiometry (DXA) Scan and Osteoporosis Diagnosis

Single Nucleotide Polymorphisins Selection

Statistical Analysis

Results

Basic Characteristics in the Whole and Different Quartiles of Participants

Association between SUA and BMD at Different Skeletal Sites

Binary Logistic Regression Analysis with OP as a Dependent Variable in Postmenopausal

Women

Putative Causal Relationships between SUA and Total BMD

Discussion

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1