Abstract
Introduction
The aim of this study was to investigate the causal relationship between Parkinson’s disease (PD) and myocardial infarction (MI), atrial fibrillation and flutter (AF), and venous thromboembolism (VTE) by Mendelian randomization (MR) analysis.
Methods
By using data from publicly available genome-wide association studies from databases, single nucleotide polymorphisms were screened as instrumental variables, and the MR analysis was finished by inverse-variance weighted (IVW), MR-egger, weighted median methods.
Results
The primary IVW method showed a negative association between genetically predicted PD and risk of MI (OR = 0.9989; 95% CI: 0.9980–0.9998; p = 0.02). However, PD was not significantly associated with AF or VTE.
Conclusion
This study suggests a negative association between PD with MI, which implies that PD has a protective effect on MI.
Keywords: Mendelian randomization analysis, Parkinson’s disease, Myocardial infarction, Venous thromboembolism, Causality
Introduction
Parkinson’s disease (PD) is a neurodegenerative disorder typically characterized by resting tremor, bradykinesia, increased muscle tone, gait, and balance disorders [1]. The patients’ social functioning is severely affected with great suffering. Meanwhile, cardiovascular diseases associated with aging, whether myocardial infarction (MI), atrial fibrillation and flutter (AF), or venous thromboembolism (VTE), can also lead to serious consequences, threatening the lives of individuals and creating a huge social burden. Some studies have reported that patients with PD are more prone to acute MI compared to normal controls [2, 3]. However, there is also a meta-analysis demonstrating that PD is associated with a reduced risk of MI [4]. Meanwhile, a meta-analysis indicated that MI was not significantly different in PD versus non-PD populations [5]. One study demonstrated that the premotor and early stages of PD were comorbid with atrial fibrillation, whereas the risk of atrial fibrillation was lower in the later stages [6]. Previous studies have indicated that PD is associated with an increased risk of ischemic stroke [7–9], while atrial fibrillation has been found to be associated with stroke in various patient groups [10]. However, some observational findings do not support an increased risk of atrial fibrillation in patients with PD [11, 12]. Whether causal relationship exists between PD and AF warrants further study. In addition, most of the studies of these above are observational, prone to multiple forms of bias, and face challenges such as potential confounders and reverse causality, which affect their causal inference and need to be further discussed. One article mentions pulmonary embolism as the second most common cause of death in 60 autopsies of patients with PD [13]. Moreover, asymptomatic pulmonary embolism often occurs in patients with deep venous thrombosis (DVT) [14]. However, there are few studies on the association between PD and VTE. Although a Mendelian randomization (MR) analysis explored no significant association between PD and MI [15], it was different from the results of most of the observational studies, and it used a European population for the PD dataset but a mixed population for the MI dataset. Unlike their study, the PD and MI datasets we have chosen were all from European populations, and the MI dataset we have chosen has a larger sample size. In addition, we set up this study to explore and validate the causal relationship between PD and a more comprehensive spectrum of cardiovascular diseases which included MI, AF, and VTE.
ME analysis is a type of randomized controlled study based on the availability of large-scale genome-wide association studies (GWASs), which, using genetic variation and following the laws of independent categorization, allows for the assessment of causal relationships between modifiable exposures or risk factors and clinically relevant outcomes [16, 17]. MR analysis has been widely used to assess potential causal relationships between various exposures and clinical outcomes. According to Mendel’s laws of inheritance, alleles are randomly passed from parents to offspring, a process that occurs earlier than disease, effectively avoiding the influence of lifestyle and environmental factors [18]. Using publicly available GWAS data, we set genetic variants associated with PD as instrumental variables (IVs), expressed as single nucleotide polymorphisms (SNPs). Some studies have shown that risk factors common to cardiovascular disease and PD include diabetes, obesity, hypertension [19, 20]. We excluded obesity, hypertension, hypercholesterolemia, and diabetes mellitus as confounders in our study and explored the causal relationship between PD and cardiac and vascular diseases including MI, AF, VTE by a two-sample MR analysis.
Materials and Methods
Research Design
The schematic research design with three key hypotheses for MR is shown in Figure 1: (1) SNPs are associated with PD; (2) SNPs are not associated with known confounders; (3) SNPs can only affect heart and vascular disease through PD and not through other pathways to influence outcomes [17].
Fig. 1.
The three key assumptions of the MR are as follows: (1) SNPs strongly associated with PD; (2) SNPs are independent of confounders; (3) SNPs can only affect heart and vascular disease through Parkinson’s and cannot act on outcomes through other pathways.
Data Sources
The analysis used publicly available summary statistics from the MRC IEU Open GWAS database available online, and the main study population consisted of men and women. The GWAS database for AF is for mixed populations, all others are for European populations. As shown in Table 1, the GWAS summary statistics for PD (n = 482,730) were obtained from the International Parkinson’s Disease Genomics Consortium [21]. GWAS summary statistics for MI (n = 462,933) were obtained from the MRC-IEU Consortium. GWAS summary statistics for AF (n = 337,199) were obtained from the Neale Lab Consortium. GWAS summary statistics for VTE (n = 361,194) were obtained from Neale laboratory. The current analysis requires no ethical approval, as all included GWAS data are publicly available and have been approved by the appropriate ethical review boards.
Table 1.
Basic sources and information about GWASs
| Trait | Gwas ID | Sample size | Year |
|---|---|---|---|
| PD | ieu-b-7 | 482,730 | 2019 |
| MI | ukb-b-15829 | 462,933 | 2018 |
| AF | ukb-a-536 | 337,199 | 2017 |
| VTE | ukb-d-I9_VTE | 361,194 | 2018 |
Screening and Validation of SNPs
MR studies are epidemiologic research methods that use SNPs as IVs to derive causal issues of exposure and outcome. Four criteria were used to select appropriate SNPs in order to satisfy the three key assumptions of MR studies as described above.
First, in the GWAS dataset of PD, the genome-wide significance threshold (p < 5 × 10−8) is used to select SNPs that are highly correlated with PD, then the selected SNPs selected as IVs are strongly associated with exposure. Second, in the analysis, we need to remove the linkage disequilibrium (LD) between the selected IVs, LD is the phenomenon that genetic variants physically close to each other on the same chromosome tend to be inherited together. If there is an LD, the genetic variants will not be distributed independently, increasing the impact of the IVs associated with the causal variants selected in the analysis on the other confounders associated with the possibility that the IVs associated with the causal variants chosen in the analysis may have an effect on other confounders associated with the outcome, biasing the results. To ensure that the IVs are independent of each other, we evaluated the independence of the selected SNPs based on pairwise linkage disequilibrium [22] and controlled the parameters of r2 < 0.001 and clumping window >10,000 kb when selecting SNPs, so as to filter out SNP loci that are independent from others. After these two steps, we selected 23 SNPs. Considering that there are multiple steps that may remove some SNPs before the subsequent MR analysis, resulting in too few samples and causing bias, we relaxed the genome-wide significance threshold (p < 5 × 10−6) to reintroduce the above two criteria for screening. Third, F-statistics were calculated to verify the strength of individual SNP. When the F-statistics were less than 10, SNPs were considered weak IVs with high potential risk of bias and needed to be excluded. Fourth, the selected SNPs were examined one by one in the phenoscanner database (http://phenoscanner.medschl.cam.ac.uk/) to eliminate the influence of potential confounders (high-risk and protective factors that may affect the outcome), and it should be ensured that the selected SNPs are not directly related to the outcome and remove SNPs which are directly and highly correlated with outcome. Finally, we obtained 50 SNPs for MR analysis (online suppl. Table S1; for all online suppl. material, see https://doi.org/10.1159/000536484), and data were performed prior to the MR analysis harmonization step because the effects of SNPs on exposure and outcome must correspond to the same alleles.
Data Analysis Methods
All analyses were performed by the “TwoSampleMR” package in R version 4.3.1. Inverse-variance weighted (IVW) meta-analysis under the random-effects model was used as the main analysis, supplemented by weighted median and MR-Egger (p < 0.05 considered statistically significant). The weighted median method provides valid estimates if more than 50% of the information comes from valid IVs [23]. Heterogeneity was assessed using Cochran’s Q test, and the MR-Egger method was used to assess horizontal pleiotropy of IVs [24]. Horizontal pleiotropy is a phenomenon in which IVs affect outcome through pathways other than exposure and is a potential source of bias. It is therefore important to detect horizontal pleiotropy, which, when detected, indicates that the second hypothesis of MR has not been met. The MR-Egger Intercept test is robustly sensitive to directional bias due to pleiotropy, and its p value >0.05 indicates that there is no horizontal pleiotropy problem. There are often multiple cohort studies in large GWAS database studies, which can result in some variation in the acquisition of SNPs across cohort studies. Heterogeneity, on the other hand, represents the problem of variability among the data of a study. In this study, we assessed the potential heterogeneity between IVs at each analysis using Cochran’s Q test, where a p value >0.05 means that there is no significant heterogeneity problem. The leave-one-out (LOO) method, which reassesses the causal effect after removing 1 SNP from the analysis, is effective in assessing the extent to which the results of MR analyses depend on a particular SNP. A key feature of MR is that the results of the analysis are summarized graphically. Scatter plots show both heterogeneity and the absence of heterogeneity in the causal estimates obtained based on different genetic variants. Among other graphical presentations, forest plot and funnel plot can also assess heterogeneity [25].
Results
IVW analysis was first performed for PD combined with MI, AF, VTE, supplemented by the weighted median and MR-Egger methods, respectively (as shown in Fig. 2). IVW analysis showed that genetically predicted PD was negatively associated with MI. No significant association with PD was observed for AF, VTE. MI (OR = 0.9989; 95% CI: 0.9980–0.9998; p = 0.02), AF (OR = 1.0000; 95% CI: 0.9994–1.0007; p = 0.96), VTE (OR = 1.0000; 95% CI: 0.9994–1.0007; p = 0.95). MR-Egger analysis showed that MI (OR = 0.9982; 95% CI: 0.9957–1.0008; p = 0.18), AF (OR = 1.0002; 95% CI: 0.9987–1.0016; p = 0.83), VTE (OR = 0.9998; 95% CI: 0.9983–1.0012; p = 0.77). In weighted median analysis, MI (OR = 0.9986; 95% CI: 0.9973–0.9999; p = 0.03), AF (OR = 1.0000; 95% CI: 0.9991–1.0010; p = 0.92), VTE (OR = 1.0004; 95% CI: 0.9994–1.0014; p = 0.43). Combining the three methods of analysis, IVW, weighted median, and MR-Egger, the analytical estimates were generally consistent.
Fig. 2.
MR estimates the causal relationship between PD with heart and vascular disease. IVW, inverse-variance weighted; OR, odds ratio; CI, confidence interval.
The results of MR-Egger Intercept and Cochran’s Q test for horizontal pleiotropy and heterogeneity are shown in Table 2, which shows that the values of Intercept are all small, with p value >0.05, which proves that horizontal pleiotropy is not significant. The results of Cochran’s Q test, which are all showing p value >0.05, indicate that heterogeneity is negligible and does not bias the results. The heterogeneity was examined in scatter plots, and no significant heterogeneity was seen in any of them, and the obtained conclusions were in line with the results of Cochran’s Q analyses (online suppl. Fig. S1–3). No single SNPs were seen in leave-one-out analyses that had a disproportionate effect on the overall results (online suppl. Fig. S4–6). The forest plots and funnel plots in online supplementary Figures S7–12 also showed the same performance as the above results.
Table 2.
Results for heterogeneity and pleiotropy
| Exposure | Outcome | Pleiotropy | Heterogeneity | ||
|---|---|---|---|---|---|
| intercept | p value | Q | p value | ||
| PD | MI | 0.00010 | 0.56 | 28.21 | 0.82 |
| AF | −0.00002 | 0.83 | 50.62 | 0.37 | |
| VTE | 0.00004 | 0.72 | 42.97 | 0.68 | |
Finally, our study found the existence of negative correlation between PD and MI. However, no causal relationship was found between PD and AF or VTE; this result can help reduce unnecessary clinical trials and save the cost of scientific research, but on the other hand, because the bias of the dataset is equally likely to be present, further research is needed to justify it.
Discussions
This MR analysis utilized publicly available databases to obtain extensive genetic data to assess the causal association of PD with MI, AF, and deep vein thromboembolism. The above studies suggest that PD is negatively associated with MI, which implies that PD has a protective effect on MI. However, there is no evidence to support a causal relationship between PD and AF, VTE.
PD is a chronic neurodegenerative disease that affects more than 1 million people in the USA and more than 10 million people worldwide [26]. The prevalence of PD increases with age, affecting 1% of the population over 60 years of age [27]. A Singaporean study shows that people with PD place a considerable burden on patients, the health system, and society as a whole [28]. Cardiovascular diseases are also common and become more prevalent with age [29]. VTE, on the other hand, can lead to diseases such as pulmonary embolism, which is associated with significant mortality, especially in critically ill patients [30]. Therefore, our study of the causal relationship between PD with including MI, AF, VTE appears to be of great importance, and is an extremely important guideline for the development of protocols for prevention, treatment, and care.
It is interesting to note that one article mentioned that smoking and cholesterol appear to be negatively associated with PD and cardiovascular disease [31]. One other study noted that elevated levels of total cholesterol and low-density lipoprotein cholesterol were associated with a reduced risk of PD in men, but this association was not significant in women [32]. This conflicts with the traditional understanding, and it needs further validation. Also, there was a study which suggested that the changes in cholesterol levels, smoking and alcohol consumption, and other lifestyle factors might contribute to the risk of MI in PD patients [33]. The potential role of cholesterol in disease protection associated with PD deserves further investigation, and we will be very interested in this area of research in future work.
A systematic review and meta-analysis concluded that PD was associated with a reduced risk of MI [4], which was similar to the findings of our study. An animal experiment found that recombinant PD protein 7 (Park7) was protective against MI-induced injury and reduced oxidative stress in MI model mice [34], which is supportive of our findings. We hope that more basic studies will further be performed to elucidate the exact mechanism of PD and the reduction of MI risk.
However, there are still some limitations to our study. First, most of the GWASs data used in our analysis were from European populations, and the applicability of these findings to other ethnic groups needs to be further determined. Second, in our study, the results of the analysis of PD with risk of AF or VTE were not statistically significant (p > 0.05), and we cannot completely rule out an association between them; what’s more, more findings are needed to support the evidence. Third, it is difficult to completely exclude the effect of potential directional pleiotropy in any MR study, though the presence of horizontal pleiotropy was not seen in the MR-Egger Intercept test in our study, and similar results could be observed in the sensitivity analysis.
Conclusion
In summary, our MR analysis demonstrated a negative correlation between PD and MI, which implies that PD has a protective effect on MI. There is a lack of evidence to support a causal relationship between PD and AF, VTE. More studies are needed to validate our findings and elucidate the underlying mechanisms.
Statement of Ethics
Ethical approval and consent were not required as this study was based on publicly available data.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
This study was no supported by any sponsor or funder.
Author Contributions
Writing-original draft and visualization: Lize Chen; methodology: Qiushi Zhang and Shiduo Li; data curation: Haoran Chen, Jing Guo, and Zongmao Zhao; and writing-review and editing: Jing Tong. All authors have read and agreed to the published version of the manuscript.
Funding Statement
This study was no supported by any sponsor or funder.
Data Availability Statement
Data available in a publicly accessible repository do not issue DOIs. Publicly available datasets were analyzed in this study. These data can be found here (https://gwas.mrcieu.ac.uk/datasets/). Further inquiries can be directed to the corresponding author.
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
Supplementary Material
References
- 1. Tarakad A, Jankovic J. Recent advances in understanding and treatment of Parkinson’s disease. Fac Rev. 2020;9:6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Liang HW, Huang YP, Pan SL. Parkinson disease and risk of acute myocardial infarction: a population-based, propensity score-matched, longitudinal follow-up study. Am Heart J. 2015;169(4):508–14. [DOI] [PubMed] [Google Scholar]
- 3. Sheen SH, Hong JB, Kim H, Kim J, Han IB, Sohn S. The relationship between Parkinson’s disease and acute myocardial infarction in korea : a nationwide longitudinal cohort study. J Korean Neurosurg Soc. 2022;65(4):507–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Nabizadeh F, Valizadeh P, Sharifi P, Zafari R, Mirmosayyeb O. Risk of myocardial infarction in Parkinson’s disease: a systematic review and meta-analysis. Eur J Neurol. 2023;30(8):2557–69. [DOI] [PubMed] [Google Scholar]
- 5. Alves M, Caldeira D, Ferro JM, Ferreira JJ. Does Parkinson’s disease increase the risk of cardiovascular events? A systematic review and meta-analysis. Eur J Neurol. 2020;27(2):288–96. [DOI] [PubMed] [Google Scholar]
- 6. Hong CT, Chan L, Wu D, Chen WT, Chien LN. Association between Parkinson’s disease and atrial fibrillation: a population-based study. Front Neurol. 2019;10:22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Ben-Shlomo Y, Marmot MG. Survival and cause of death in a cohort of patients with parkinsonism: possible clues to aetiology? J Neurol Neurosurg Psychiatry. 1995;58(3):293–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Huang YP, Chen LS, Yen MF, Fann CY, Chiu YH, Chen HH, et al. Parkinson’s disease is related to an increased risk of ischemic stroke-a population-based propensity score-matched follow-up study. PLoS One. 2013;8(9):e68314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Abugroun A, Taha A, Abdel-Rahman M, Patel P, Ali I, Klein LW. Cardiovascular risk among patients ≥65 years of age with Parkinson’s disease (from the national inpatient sample). Am J Cardiol. 2020;136:56–61. [DOI] [PubMed] [Google Scholar]
- 10. Manolio TA, Kronmal RA, Burke GL, O’Leary DH, Price TR. Short-term predictors of incident stroke in older adults. The Cardiovascular Health Study. Stroke. 1996;27(9):1479–86. [DOI] [PubMed] [Google Scholar]
- 11. Alves M, Abrantes AM, Portugal G, Cruz MM, Reimão S, Caldeira D, et al. Does Parkinson’s disease increase the risk of atrial fibrillation? Insights from electrocardiogram and risk scores from a case-control study. Front Neurol. 2021;12:633900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Cereja F, Alves M, Ferreira JJ, Caldeira D. Atrial fibrillation risk on Parkinson’s disease - a systematic review and meta-analysis. J Thromb Thrombolysis. 2023;55(4):747–50. [DOI] [PubMed] [Google Scholar]
- 13. Mosewich RK, Rajput AH, Shuaib A, Rozdilsky B, Ang L. Pulmonary embolism: an under-recognized yet frequent cause of death in parkinsonism. Mov Disord. 1994;9(3):350–2. [DOI] [PubMed] [Google Scholar]
- 14. Li F, Wang X, Huang W, Ren W, Cheng J, Zhang M, et al. Risk factors associated with the occurrence of silent pulmonary embolism in patients with deep venous thrombosis of the lower limb. Phlebology. 2014;29(7):442–6. [DOI] [PubMed] [Google Scholar]
- 15. Wang M, Zhang Z, Liu D, Ji L, Huang S, Cao L, et al. Genetic predisposition to Parkinson’s disease and risk of cardio and cerebrovascular disease: a Mendelian randomization study. Parkinsonism Relat Disord. 2022;94:49–53. [DOI] [PubMed] [Google Scholar]
- 16. Sekula P, Del Greco M F, Pattaro C, Köttgen A. Mendelian randomization as an approach to assess causality using observational data. J Am Soc Nephrol. 2016;27(11):3253–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Emdin CA, Khera AV, Kathiresan S. Mendelian randomization. JAMA. 2017;318(19):1925–6. [DOI] [PubMed] [Google Scholar]
- 18. Ebrahim S, Davey Smith G. Mendelian randomization: can genetic epidemiology help redress the failures of observational epidemiology? Hum Genet. 2008;123(1):15–33. [DOI] [PubMed] [Google Scholar]
- 19. Simon KC, Chen H, Schwarzschild M, Ascherio A. Hypertension, hypercholesterolemia, diabetes, and risk of Parkinson disease. Neurology. 2007;69(17):1688–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Camargo Maluf F, Feder D, Alves de Siqueira Carvalho A. Analysis of the relationship between type II diabetes mellitus and Parkinson’s disease: a systematic review. Parkinsons Dis. 2019;2019:4951379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Nalls MA, Blauwendraat C, Vallerga CL, Heilbron K, Bandres-Ciga S, Chang D, et al. Expanding Parkinson’s disease genetics: novel risk loci, genomic context, causal insights and heritable risk. BioRxiv; 2019. [Google Scholar]
- 22. Machiela MJ, Chanock SJ. LDlink: a web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinforma. 2015;31(21):3555–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Burgess S, Bowden J, Fall T, Ingelsson E, Thompson SG. Sensitivity analyses for robust causal inference from mendelian randomization analyses with multiple genetic variants. Epidemiology. 2017;28(1):30–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Bowden J, Spiller W, Del Greco M F, Sheehan N, Thompson J, Minelli C, et al. Improving the visualization, interpretation and analysis of two-sample summary data Mendelian randomization via the Radial plot and Radial regression. Int J Epidemiol. 2018;47(6):2100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Marras C, Beck JC, Bower JH, Roberts E, Ritz B, Ross GW, et al. Prevalence of Parkinson’s disease across north America. NPJ Parkinsons Dis. 2018;4:21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Tysnes OB, Storstein A. Epidemiology of Parkinson’s disease. J Neural Transm. 2017;124(8):901–5. [DOI] [PubMed] [Google Scholar]
- 28. Zhao YJ, Tan LC, Li SC, Au WL, Seah SH, Lau PN, et al. Economic burden of Parkinson’s disease in Singapore. Eur J Neurol. 2011;18(3):519–26. [DOI] [PubMed] [Google Scholar]
- 29. Benjamin EJ, Muntner P, Alonso A, Bittencourt MS, Callaway CW, Carson AP, et al. Heart disease and stroke statistics-2019 update: a report from the American heart association. Circulation. 2019;139(10):e56–e528. [DOI] [PubMed] [Google Scholar]
- 30. Chan CM, Shorr AF. Economic and outcomes aspects of venous thromboembolic disease. Crit Care Clin. 2012;28(1):113–23. vii. [DOI] [PubMed] [Google Scholar]
- 31. Potashkin J, Huang X, Becker C, Chen H, Foltynie T, Marras C. Understanding the links between cardiovascular disease and Parkinson’s disease. Mov Disord. 2020;35(1):55–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Rozani V, Gurevich T, Giladi N, El-Ad B, Tsamir J, Hemo B, et al. Higher serum cholesterol and decreased Parkinson’s disease risk: a statin-free cohort study. Mov Disord. 2018;33(8):1298–305. [DOI] [PubMed] [Google Scholar]
- 33. Reichmann H, Csoti I, Koschel J, Lorenzl S, Schrader C, Winkler J, et al. Life style and Parkinson’s disease. J Neural Transm. 2022;129(9):1235–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Pan G, Yang S, Han X, Wang X, Kou L, Xie J, et al. Parkinson’s disease protein 7 protected against oxidative stress of myocardial infarction direct through p47phox and nicotinamide adenine dinucleotide phosphate oxidase 4. Hum Exp Toxicol. 2022;41:9603271221124099. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data available in a publicly accessible repository do not issue DOIs. Publicly available datasets were analyzed in this study. These data can be found here (https://gwas.mrcieu.ac.uk/datasets/). Further inquiries can be directed to the corresponding author.