Serum Uric Acid and Male Erectile Dysfunction: a Mendelian Randomization Study

doi:10.21203/rs.3.rs-4130257/v1

Download PDF

Research Article

Serum Uric Acid and Male Erectile Dysfunction: a Mendelian Randomization Study

https://doi.org/10.21203/rs.3.rs-4130257/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose

To evaluate the potential causal relationship between serum uric acid levels and erectile dysfunction in men, this study performed a Mendelian randomization (MR) analysis.

Methods

The exposure factor was serum uric acid, and the outcome factor was erectile dysfunction. All data come from genome-wide association studies (GWAS), and two-sample MR studies are conducted based on R Studio 4.3.2 and the "TwoSampleMR" R software package. The inverse-variance weighted (IVW) and MR-Egger methods are mainly used for analysis.

Results

Neither the IVW (Beta = 0.068; SE = 0.062; p = 0.274) nor the MR Egger (Beta = 0.079; SE = 0.101; p = 0.436) approaches demonstrated a causal association between the two variables.

Conclusions

According to this MR study, there is no causal relationship between serum uric acid levels and male erectile dysfunction, so medication to lower uric acid levels alone may not be effective in treating erectile dysfunction in patients with high serum uric acid.

Causation

correlation

serum uric acid

erectile dysfunction

Mendelian randomization

Erectile dysfunction (ED) in males refers to the inability to achieve or sustain an erection of the penis that is enough for effective vaginal intercourse. This condition is prevalent among middle-aged and elderly men[1]. Recent studies indicate that various causes contribute to the development of erectile dysfunction (ED), including psychological, neurological, hormonal, vascular, drug-related, and a combination of multiple factors[2]. Vascular factors are the most common among these, including atherosclerosis, hypertension, hyperlipidemia, smoking, diabetes, and pelvic radiotherapy. Endothelial dysfunction is a common mechanism underlying many vascular risk factors that contribute to arterial ED[3]. Endothelial cells create nitric oxide (NO), which is crucial for the erection process. Nitric oxide (NO) can activate soluble guanylyl cyclase (sGC), causing an increase in guanosine 3',5'-cyclic monophosphate (cGMP). This, in turn, causes cytoplasmic calcium depletion and smooth muscle relaxation in the cavernous region[4].

Uric acid (UA) is a weak acid produced by purine metabolism that is mostly present in the form of urate at physiological pH[5]. Serum UA can stimulate vascular smooth muscle cells to proliferate and produce a large amount of angiotensin II. This causes oxidative stress and glycocalyx loss, which reduces nitric oxide activity in blood vessels, boosts inflammatory cytokine levels, and ultimately impairs the function of vascular endothelial cells[6]. Serum UA levels were shown to be positively associated with vasculogenic ED in a previous case-control study [7]. In addition, an observational systematic review and meta-analysis concluded that serum UA is an independent risk factor for male erectile dysfunction[8]. However, current research on UA and ED are retrospective in nature, making them subject to confounding factors such as a small sample size, a short follow-up period, and the combined effect of other factors on erectile dysfunction.

Mendelian randomization (MR) uses an instrumental variable to imitate an individual's random allocation of exposure factors. The instrumental variable is a single nucleotide polymorphism (SNP). During the creation of human gametes, the allele of a specific SNP is randomly assigned to egg or sperm cells. This also accounts for the influence of confounding factors on the exposure-outcome causal link, reducing concerns about reverse causation[9]. This work employs MR to clarify the causal link between UA and ED, resulting in novel clinical approaches for ED treatment and prevention.

2.1 Study Design and Data Sources

The instrumental variable chosen using Mendelian randomization must meet the following three conditions (Fig. 1) [10]. First, the genetic variation of the instrumental variable is inextricably linked to the exposure factors. Second, genetic variation has no relationship with other confounding factors. Third, genetic variation can only be influenced by risk factors. This is the sole way to influence the outcome and cannot be achieved in any other way.

The data in this study are derived from genome-wide association studies (GWAS). The data set related with SNPs and serum uric acid has a sample size of 129,405, containing 12,499,459 SNPs. The data collection related with SNPs and male erectile dysfunction has a sample size of 223,805 and contains 9,310,196 SNPs. Table 1 summarizes the basic information from both databases. We investigated SNPs linked to serum UA. The first phase involved retaining SNPs with genome-wide significance (P < 5×10 − 8) and removing SNPs in linkage disequilibrium (r2 threshold = 0.001 and clumping distance = 10,000 kb). The second stage is to remove SNPs related with confounding factors. We examined the human genotype-phenotype association database (PhenoScanner V2). The third step involves excluding weak instrumental factors, such as SNPs with an F-statistic < 10. We did data harmonization to ensure that the SNP's effect on exposure and outcome corresponded to the same allele. Finally, 59 SNPs were used in the Mendelian randomization analysis. There is no need for extra ethical evaluation for this work because the GWAS database is publically available and the initial study was approved by the ethics review committee.

a: Genetic variation directly affects exposure factors. b: Genetic variation is not associated with any confounders. c: Genetic variation affects outcome only through exposure factors .Dashed lines indicate potential causal relationships that violate Mendelian randomization principles.

Table 1

Basic information on serum UA and ED groups in the study
Exposure	Population	Year	Sample size	Data sources	Web source
Serum UA	East Asian	2021	129,405	Sakaue S et al[11].	https://gwas.mrcieu.ac.uk/datasets/ebi-a-GCST90018757/
ED	European	2018	223,805	Bovijn J et al[12].	https://gwas.mrcieu.ac.uk/datasets/ebi-a-GCST006956/

2.2 Statistical Analysis

This study used five two-sample MR analysis methods: inverse-variance weighted (IVW), MR-Egger, weighted median, simple mode, and weighted mode. The major analytical methods are IVW and MR Egger. In sensitivity analysis, Cochran's Q test was utilized to examine heterogeneity of MR analysis. Significant heterogeneity was recognized when p < 0.05. If there is heterogeneity, a random effects model is employed, with outlier analysis; otherwise, a fixed effects model is utilized. To evaluate if genetic pleiotropy influences the causal relationship, the MR-Egger regression intercept is used. The leave-one-out analysis was done to determine the stability of the results.

This study utilized R Studio version 4.3.2 and the "TwoSampleMR" R software package to conduct MR analysis. A p-value of less than 0.05 was deemed statistically significant.

3.1 Two-sample Mendelian randomization results

The results of Mendelian randomization between serum UA levels and ED are shown in Table 2. All five MR methods showed no causal association between serum UA and ED (all beta values are greater than 0, and all p values are greater than 0.05). The forest plot of individual SNPs associated with serum UA and ED is shown in Fig. 2. Among these 59 SNPs, the 95% confidence intervals of most SNPs span the middle invalid line (odds ratio, OR = 0). After combining the effect sizes, the 95% confidence intervals for the IVW and MR Egger Mendelian randomization methods straddle the middle null line (OR = 0). The lower limit of the 95% confidence interval for the IVW method is 0.947, and the upper limit is 1.210. The lower limit of the 95% confidence interval for the MR Egger method is 0.888, and the upper limit is 1.320.

The scatter plot of the correlation between serum UA and ED is shown in Fig. 3. The x-axis is the impact of SNP on serum UA, and the y-axis is the impact of SNP on ED. Each point in the scatter plot represents each SNP, and the line passing through each point represents the 95% confidence interval. The vertical coordinate of each point is the effect of SNP on ED, and the horizontal coordinate is the effect of SNP on serum UA. The ratio of the two effects is the slope. The straight-line slopes of these five MR methods in the scatter plot are close, and the slope is slightly greater than 0. Combined with the results of the MR analysis, the correlation between serum UA and ED is weak (p > 0.05).

Table 2

MR estimates for each method of assessing the causal effect of serum UA on ED
MR method	Number of SNPs	Beta	SE	P value	OR	Lower limit of 95% CI	Upper limit of 95% CI
MR Egger	59	0.079	0.101	0.436	1.083	0.888	1.320
Weighted median	59	0.072	0.086	0.407	1.074	0.907	1.273
IVW	59	0.068	0.062	0.274	1.071	0.947	1.210
Simple mode	59	0.195	0.149	0.196	1.215	0.907	1.628
Weighted mode	59	0.053	0.072	0.468	1.054	0.915	1.214

3.2 sensitivity analysis

Cochran’s Q test was used to conduct heterogeneity analysis on the results after MR analysis by the MR Egger method and the IVW method. As shown in Table 3, there is no heterogeneity in the MR analysis of the two main methods (p < 0.05). In the funnel plot in Fig. 3, SNPs are distributed roughly symmetrically on both sides of the IVW and Mr. Egger straight lines. The intercept of MR-Egger regression was used to evaluate horizontal pleiotropy in MR analysis. The results showed that the intercept was − 0.000809, the standard error was 0.00582, and the p value was 0.890, indicating that there is no horizontal pleiotropy. The leave-one-out method is used to evaluate the reliability and stability of the results. As shown in Fig. 5, after excluding each SNP, the overall error does not change much, indicating that the results are reliable.

Table 3

Heterogeneity analysis results of Cochran’s Q test
Method		Q-statistics	Degree of freedom	P value
MR Egger		62.401	57	0.291
IVW		62.422	58	0.322

This is the first study to use a two-sample Mendelian randomization strategy to investigate the causal link between serum UA and ED. The data indicate that there is no causal relationship between serum UA and ED. In the MR sensitivity analysis, there was no heterogeneity or horizontal pleiotropy. MR analysis was done after one SNP was removed at a time, and the results changed less, indicating that the results were steady and reliable. Of course, the two are just genetically related and have no causative relationship, but it is possible that serum UA and ED are linked at other levels outside genetics.

Multiple prior systematic reviews and meta-analyses have found that greater serum uric acid levels are related with the development of ED[8, 13, 14]. Furthermore, clinical prediction model studies have demonstrated that serum UA can be used as a risk factor for ED[15].Furthermore, hyperuricemia may be regarded an independent risk factor in addition to known risk factors. This is completely different from the findings of this study, and we feel there are primarily the following reasons: First, all previous research were observational, with no randomization, prospective, or blinding procedures. Differences in study results could be attributed to the limits of observational studies, which are prone to confounding factors and reverse causality. Second, the populations included in these studies may be complicated by other vascular illnesses, such as atherosclerosis, hypertension, and other confounding variables, reducing the trustworthiness of the findings. Interestingly, however, in a multicenter cross-sectional investigation, UA was identified as an independent protective factor for ED, which contradicts the findings of this study[16].Because of the limitations of cross-sectional studies, causality cannot be established; however, there is a correlation between serum UA levels and ED. A cross-sectional investigation in Finland found no apparent risk relation between serum UA levels and ED, which supports the findings of this study[17]. As a result, the association between UA levels and ED remains hotly debated. This study simply confirmed at the genetic level that there is no causal link between serum UA and ED.

Although a lot of case-control studies and meta-analyses have been conducted, the underlying biological mechanism between serum UA and ED remains unknown and requires additional investigation. Serum uric acid contains both pro-oxidative and antioxidant properties in the body. Roumeliotis S. et al. believe that hyperuricemia-induced oxidative stress may contribute to endothelial cell senescence and death[18]. Oxidative stress is caused by an imbalance between pro-oxidant and antioxidant components, which stimulates the production of pro-oxidants. It is assumed that xanthine oxidase produces a large amount of superoxide, causing smooth muscle cells to produce lectin-type oxidized LDL receptor 1 (LOX-1), which harms endothelial cells and reduces NO production. Less NO causes cells to produce less cyclic guanosine monophosphate (cGMP), which is detrimental to smooth muscle and blood flow in the penis' corpus cavernosum[4, 19]. This study demonstrates that there is no genetic relationship between serum UA and ED, implying that several prior observational studies may have been influenced by other confounding factors, resulting in false-positive results. De Becker B et al. investigated serum uric acid levels and related oxidative stress markers and discovered that serum UA had no significant relationship with oxidative stress [20]. This suggests that serum UA affects endothelial cells not just through oxidative stress, but also via additional routes. Masi S et al. observed a U-shaped association between uric acid levels in the blood and microvascular remodeling characteristics such as the media-to-lumen ratio (M/L ratio). They also discovered that both low and high amounts of uric acid in the blood were associated with hypertrophic remodeling of tiny resistance arteries[21]. Serum uric acid is a strong antioxidant at physiological levels and may protect vascular endothelial cells. However, low levels of serum UA, particularly in the extracellular fluid, may drastically reduce the antioxidant ability. High amounts, on the other hand, promote its mobility within cells, triggering pathways for cell proliferation and pro-oxidative inflammation[22].This study's Mendelian randomization analysis implies a linear relationship between serum UA and ED, which may not be the case. It has previously been proposed that damage to the vascular endothelium is a possible biological explanation for ED caused by high serum UA levels. Masi S et al. hypothesized that low serum UA levels could cause endothelial problems, which contradicted the findings of several earlier research. Because observational epidemiological research make it difficult to eliminate biases such as confounding factors and reverse causality, etiological interpretation is limited. As a result, the mechanism by which blood uric acid levels affect vascular endothelial cells has to be investigated further.

This study offers the following benefits: First, there are currently no MR studies that have investigated the causative relationship between serum UA and ED, and the majority of them are observational. MR analysis is a superior statistical method that eliminates the effects of confounding variables and reverse causation. Second, we employ the IVW and MR Egger methods as primary assessment methods, with weighted median, simple mode, and weighted mode as secondary evaluation methods, to assure the results' stability and reliability. Third, the summary statistics generated from GWAS in this study have bigger sample sizes than observational studies, implying that valid causal inferences have more statistical strength.

Of course, this study does have some limitations. First, our study sample was limited to those of European and East Asian heritage; future research is needed to look into other populations. Second, SNPs related to exposure variables and SNPs related to outcomes are from distinct ethnic groups, resulting in a population stratification bias due to disparities in allele frequencies and disease prevalence between ethnic groups. Third, given the F statistics were all greater than 10, this study should be free of weak instrumental variable bias; however, the poor power could be attributed to the small number of SNPs utilized as instrumental variables. Fourth, while our analysis used a large population-based cohort, we acknowledge that statistical power may differ between genetic variations and outcomes. Fifth, we have a poor understanding of the molecular mechanisms that relate genetic variation to exposure, which may influence how results are interpreted.

This MR study shows that there is no causal relationship between serum uric acid levels and male erectile dysfunction. Therefore, in patients with high serum uric acid, simply using medication to reduce uric acid levels may not be effective in treating erectile dysfunction.

Data availability: All data for this study come from the IEU Open GWAS project. The article provides the methods taken to get the original data. You can reach out to the individual authors with any additional questions.

Ethics statement: There is no need for extra ethical evaluation for this work because the GWAS database is publically available and the initial study was approved by the ethics review committee.

Authors contribution: W.C. and W.J. wrote the main manuscript; C.Z. was responsible for the statistical analysis and provided all figures and tables in the article; and G.S. revised and proofread the final manuscript.

Acknowledgments

We are grateful to the IEU Open GWAS project for their cooperation with data.

Conflict of interest: The authors have no competing interests to declare.

Yafi FA, Jenkins L, Albersen M, Corona G, Isidori AM, Goldfarb S, Maggi M, Nelson CJ, Parish S, Salonia A, Tan R, Mulhall JP, Hellstrom WJ. Erectile dysfunction. Nat Rev Dis Primers. 2016;2:16003. 10.1038/nrdp.2016.3. PMID: 27188339; PMCID: PMC5027992.
Shamloul R, Ghanem H. Erectile dysfunction. Lancet. 2013;381(9861):153 – 65. doi: 10.1016/S0140-6736(12)60520-0. Epub 2012 Oct 5. PMID: 23040455.
De Leonardis F, Colalillo G, Finazzi Agrò E, Miano R, Fuschi A, Asimakopoulos AD. Endothelial Dysfunction, Erectile Deficit and Cardiovascular Disease: An Overview of the Pathogenetic Links. Biomedicines. 2022;10(8):1848. 10.3390/biomedicines10081848. PMID: 36009395; PMCID: PMC9405076.
Ghalayini IF. Nitric oxide-cyclic GMP pathway with some emphasis on cavernosal contractility. Int J Impot Res. 2004;16(6):459 – 69. 10.1038/sj.ijir.3901256. PMID: 15229623.
Mandal AK, Mount DB. The molecular physiology of uric acid homeostasis. Annu Rev Physiol. 2015;77:323–45. 10.1146/annurev-physiol-021113-170343. Epub 2014 Nov 12. PMID: 25422986.
Tang J, Liu K, Eshak ES, Cui R, Sakaniwa R, Imano H, Dong JY, Iso H. Association between Serum Uric Acid and Impaired Endothelial Function: The Circulatory Risk in Communities Study. J Atheroscler Thromb. 2022;29(10):1534–46. 10.5551/jat.63199. Epub 2021 Dec 1. PMID: 34853212; PMCID: PMC9529376.
Barassi A, Corsi Romanelli MM, Pezzilli R, Dozio E, Damele CAL, Vaccalluzzo L, Di Dario M, Goi G, Papini N, Massaccesi L, Colpi GM. Melzi d'Eril GV. Levels of uric acid in erectile dysfunction of different aetiology. Aging Male. 2018;21(3):200–205. doi: 10.1080/13685538.2017.1420158. Epub 2018 Jan 12. PMID: 29327639.
Wu X, Liu G, Zhang Y, Zhang W, Dai Y, Jiang H, Zhang X. The association between uric acid and erectile dysfunction: A systematic review and Meta-analysis. Andrologia. 2022;54(3):e14319. 10.1111/and.14319. Epub 2021 Nov 18. PMID: 34796510.
Sekula P, Del Greco MF, Pattaro C, Köttgen A. Mendelian Randomization as an Approach to Assess Causality Using Observational Data. J Am Soc Nephrol. 2016;27(11):3253–65. 10.1681/ASN.2016010098. Epub 2016 Aug 2. PMID: 27486138; PMCID: PMC5084898.
Burgess S, Scott RA, Timpson NJ, Davey Smith G, Thompson SG, EPIC- InterAct Consortium. Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors. Eur J Epidemiol. 2015;30(7):543–52. 10.1007/s10654-015-0011-z. Epub 2015 Mar 15. PMID: 25773750; PMCID: PMC4516908.
Sakaue S, Kanai M, Tanigawa Y, Karjalainen J, Kurki M, Koshiba S, Narita A, Konuma T, Yamamoto K, Akiyama M, Ishigaki K, Suzuki A, Suzuki K, Obara W, Yamaji K, Takahashi K, Asai S, Takahashi Y, Suzuki T, Shinozaki N, Yamaguchi H, Minami S, Murayama S, Yoshimori K, Nagayama S, Obata D, Higashiyama M, Masumoto A, Koretsune Y, FinnGen; Ito K, Terao C, Yamauchi T, Komuro I, Kadowaki T, Tamiya G, Yamamoto M, Nakamura Y, Kubo M, Murakami Y, Yamamoto K, Kamatani Y, Palotie A, Rivas MA, Daly MJ, Matsuda K, Okada Y. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet. 2021;53(10):1415–24. 10.1038/s41588-021-00931-x. Epub 2021 Sep 30. PMID: 34594039.
Bovijn J, Jackson L, Censin J, Chen CY, Laisk T, Laber S, Ferreira T, Pulit SL, Glastonbury CA, Smoller JW, Harrison JW, Ruth KS, Beaumont RN, Jones SE, Tyrrell J, Wood AR, Weedon MN, Mägi R, Neale B, Lindgren CM, Murray A, Holmes MV. GWAS Identifies Risk Locus for Erectile Dysfunction and Implicates Hypothalamic Neurobiology and Diabetes in Etiology. Am J Hum Genet. 2019;104(1):157–63. Epub 2018 Dec 21. PMID: 30583798; PMCID: PMC6323625.
Totaro M, Dimarakis S, Castellini C, D'Andrea S, Parisi A, D'Amato F, Tienforti D, Palazzi S, Baroni MG, Francavilla S, Barbonetti A. Erectile dysfunction in hyperuricemia: A prevalence meta-analysis and meta-regression study. Andrology. 2022;10(1):72–81. 10.1111/andr.13088. Epub 2021 Aug 16. PMID: 34347943.
Wang W, Jing Z, Liu W, Zhu L, Ren H, Hou X. Hyperuricaemia is an important risk factor of the erectile dysfunction: A systematic review and meta-analysis. Andrologia. 2022;54(5):e14384. 10.1111/and.14384. Epub 2022 Feb 7. PMID: 35130578.
Salem S, Mehrsai A, Heydari R, Pourmand G. Serum uric acid as a risk predictor for erectile dysfunction. J Sex Med. 2014;11(5):1118–24. 10.1111/jsm.12495. Epub 2014 Mar 13. PMID: 24621054.
Gao F, Jiang B, Cang Z, Wang N, Han B, Li Q, Chen Y, Chen Y, Xia F, Zhai H, Chen C, Lu M, Meng Y, Lu Y, Shen Z. Serum Uric Acid Is Associated with Erectile Dysfunction: A Population-Based Cross-Sectional Study in Chinese Men. Sci Rep. 2017;7(1):2087. 10.1038/s41598-017-02392-x. PMID: 28522875; PMCID: PMC5437090.
Tuokko AT, Murtola T, Korhonen P, Kaipia A. Hyperuricemia Is Not an Independent Predictor of Erectile Dysfunction. Sex Med. 2021;9(2):100319. 10.1016/j.esxm.2020.100319. Epub 2021 Feb 20. PMID: 33621708; PMCID: PMC8072167.
Roumeliotis S, Roumeliotis A, Dounousi E, Eleftheriadis T, Liakopoulos V. Dietary Antioxidant Supplements and Uric Acid in Chronic Kidney Disease: A Review. Nutrients. 2019;11(8):1911. 10.3390/nu11081911. PMID: 31443225; PMCID: PMC6723425.
Gherghina ME, Peride I, Tiglis M, Neagu TP, Niculae A, Checherita IA. Uric Acid and Oxidative Stress-Relationship with Cardiovascular, Metabolic, and Renal Impairment. Int J Mol Sci. 2022;23(6):3188. 10.3390/ijms23063188. PMID: 35328614; PMCID: PMC8949471.
De Becker B, Van De Borne P. Serum uric acid: a futile bystander in endothelial function? Blood Press. 2023;32(1):2237123. 10.1080/08037051.2023.2237123. PMID: 37470459.
Masi S, Georgiopoulos G, Alexopoulos G, Pateras K, Rosada J, Seravalle G, De Ciuceis C, Taddei S, Borghi C, Grassi G, Rizzoni D, Virdis A, The Study Groups On The Uric Acid Right For heArt Health Urrah, Micro-And Macro-Circulation Of The Italian Society Of Hypertension Siia. The Complex Relationship Between Serum Uric Acid, Endothelial Function and Small Vessel Remodeling in Humans. J Clin Med. 2020;9(7):2027. 10.3390/jcm9072027. PMID: 32605201; PMCID: PMC7409082.
Waring WS, McKnight JA, Webb DJ, Maxwell SR. Uric acid restores endothelial function in patients with type 1 diabetes and regular smokers. Diabetes. 2006;55(11):3127-32. 10.2337/db06-0283. PMID: 17065352.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Serum Uric Acid and Male Erectile Dysfunction: a Mendelian Randomization Study

Status:

Version 1

Abstract

Purpose

Methods

Results

Conclusions

Figures

1. Introduction

2. Methods

2.1 Study Design and Data Sources

2.2 Statistical Analysis

3. Results

3.1 Two-sample Mendelian randomization results

3.2 sensitivity analysis

4. Discussion

5. Conclusion

Declarations

References

Additional Declarations

Status:

Version 1