Association between Rheumatoid Arthritis and Renal Function: A Bidirectional Mendelian Randomization Study

doi:10.21203/rs.3.rs-2998964/v1

Download PDF

Research Article

Association between Rheumatoid Arthritis and Renal Function: A Bidirectional Mendelian Randomization Study

https://doi.org/10.21203/rs.3.rs-2998964/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

This article investigates whether there is a causal relationship between Rheumatoid arthritis (RA) and renal function (measured by estimated glomerular filtration rate) using Mendelian randomization method. The appropriate single nucleotide polymorphisms were extracted from the GWAS of RA and eGFR as gene instrumental variables. Bidirectional two-sample MR analyses were conducted using inverse variance weighting, weighted median method, and MR Egger regression. The outcome and exposure populations were both East Asians. After screening, a total of 11 single-nucleotide polymorphisms (SNPs) of genome-wide significance were included as RA instrumental variables. The analysis did not show a significant correlation between RA and eGFR (β = 0.007, P = 0.717 in IVW method). Likewise, using 54 genome-wide significant SNPs as eGFR instrumental variables, no causal relationship between eGFR and RA was found (odds ratio = 0.890, 95% CI: 0.691—1.144, p = 0.363). Methods such as MR-Egger regression and weighted median also support the result that there is no evidence of any causal relationship between RA and eGFR in any direction, except for the IVW method.

bidirectional two-sample Mendelian randomization

Rheumatoid arthritis

renal function

Rheumatoid arthritis (RA) is a chronic, autoimmune, inflammatory joint disease [1]that is characterized by persistent synovitis and systemic inflammation. It can occur at any age, and its etiology is not yet clear. Recent studies have shown that its mechanism is closely related to genetics, environment, and immune response [2]. Genes such as MHC Ⅱ and HLA-DR4 are involved in the occurrence of RA.Bone resorption, cartilage, and synovial destruction are its main three pathogenic pathways, leading to joint disorders and deformities in patients. At the cellular level, dendritic cells, macrophages, and mast cells bind to autoantigens, further triggering joint inflammation and releasing cytokines such as IL-6,IL-1 and TNF-α to promote osteoclast maturation. These osteoclast secretes hydrochloric acid and proteolytic enzyme cathepsin k, degrades osteonectin, resulting in chronic destruction of joints. When the human body is exposed to smoke, it may induce citrullination on the surface of the lungs and other mucosa, activate the immune response and promote the occurrence of RA [3]. The number of patients with RA accounts for about 1% of the world's population [4]. Bone degradation, cartilage and synovial destruction are the three main pathways to cause RA, resulting in joint disorders and deformities, as well as the loss of motor ability. Therefore, it is necessary to understand the potential risk factors that have a causal relationship with RA. RA mainly affects the whole body skeleton, cardiovascular system and nervous system [5], which is also an important reason for the increase of mortality in RA. In terms of treatment, High-dose methotrexate and combination therapy are commonly used for early treatment [6–7]. The glomerular filtration rate (GFR) is usually used for the diagnosis, classification, and prognosis evaluation of chronic kidney disease and is one of the most commonly used indicators in clinical practice [8–9],and it is also the preferred index to evaluate glomerular filtration function.The GFR can be measured or estimated [10]. Although the former is more accurate, the latter is more convenient and widely used, eGFR is often used to measure renal function in clinical studies [11–14].It is calculated indirectly by measuring the clearance rate of a marker in the kidney or plasma. RA rarely involves the kidneys, and patients with impaired renal function are not included as high-risk groups in the 2021 American Rheumatology Association's rheumatoid arthritis treatment guidelines [15]. However, in the course of the disease, its effect on the renal function of patients can not be ignored. This has also been described in many articles over the years. When the disease is not well controlled, amyloidosis may occur in the kidney, resulting in an increase in mortality. It has been reported that when RA occurs, the disease itself or anti-RA therapy causes renal morphological changes, which leads to the occurrence of kidney diseases [16].Teja et al. suggest that renal dysfunction in RA patients is mainly due to the nephrotoxicity of RA treatment, such as nonsteroidal anti-inflammatory drugs (NSAIDs) and disease-modifying antirheumatic drugs (DMARDs). In recent years, with the emergence of biological agents and methotrexate (MTX) [17], and changes in treatment strategies, the possibility of renal function impairment in RA patients has decreased [18]. However, CKD seems to be mainly related to cardiovascular metabolic risk factors such as hypertension and insulin resistance, rather than the severity of RA. But Lee et al. [19] conducted a retrospective cohort study aimed to observe the safety of methotrexate (MTX) in patients with renal function decline, and the results showed that about 30% of RA and renal dysfunction patients receiving MTX treatment had nephrotoxicity, while the use of Hydroxychloroquine showed antagonism against MTX toxicity. The results of a multicenter cross-sectional study reported by Couderc et al. showed that renal function impairment is relatively common in RA, and it is related to cardiovascular risk factors such as age, hypertension, and SCORE equation, but not related to the activity or severity of RA [20]. Märker-Hermann et al. believe that renal involvement is more common in RA. Renal biopsy usually shows mesangial proliferative glomerulonephritis and IgA nephropathy, while clinically, the most common and relevant are the progressive renal failure caused by other sequelae of chronic systemic inflammation [21]. A retrospective cross-sectional study conducted by Yang et al. showed that neutrophils greater than 7.5 × 10⁹/L are associated with a high risk of renal dysfunction in RA, suggesting that neutrophils may be a biomarker of renal dysfunction in RA patients [22]. and this mechanism may be related to its increased oxidative stress response and release of extracellular traps .Zhang et al.'s study showed that IgA nephropathy (IgAN) is the most common in 56 RA patients with renal involvement, and renal dysfunction is also mainly seen in this group of patients [23]. Other studies have shown that renal function is directly related to renal amyloidosis in RA, and there is a significant correlation between the area of amyloid protein deposition and renal function parameters [24].

In summary, traditional observational epidemiological studies have shown an association between rheumatoid arthritis (RA) and kidney function. However, there may be important differences in confounding factors between study groups in observational studies, so differences in results may be caused by differences in exposure itself or measurable and unmeasurable confounding factors between groups, or both, and there may also be a lack of clear causal timing. Two-sample Mendelian randomization (TSMR) studies in epidemiology can address these limitations and provide new insights for exploring the etiology of diseases [25]. This method is based on the principles of Mendelian randomization [26] and uses publicly available summary databases from genome-wide association studies (GWAS) to investigate causal associations between two complex diseases. Figure 1 illustrates the design approach of this study.

The underlying tenet of TSMR resides in the fortuitous allotment of genetic variation during conception, whereby a trait typically exhibits no discernible correlation with other traits. This mechanism closely parallels the haphazard assignment of participants in a randomized controlled trial. Furthermore, the MR design effectively mitigates the perils of reverse causality, as alleles are definitively established at birth and remain impervious to the advent or progression of a malady. By harnessing genetic variation as an instrumental variable to deduce causal associations between exposure and outcome, one can curtail the confounding factors and impediments stemming from reverse causality, while simultaneously reaping the rewards of reduced temporal investment and diminished demands on manpower and resources.

First, data acquisition and preparation. The GWAS databases of RA and OA were obtained from publicly published GWAS databases, and the data were initially organized. Second, instrumentalvariables (IVs) are selected and validated according to three hypothetical premises that need to be satisfied by Mendelian randomization[27]: ①IVs are related to RA. ②IVs are not directly related to eGFR. ③IVs produce effects on eGFR only through RA. While in reverse MR analysis,IVs have other limiting conditions.Third, TSMR analysis. Within this scholarly inquiry, the predominant methodology employed to procure preliminary analytical findings was the inverse variance weighted (IVW) technique. Furthermore, auxiliary methods such as the weighted median approach, MR-egger regression methodology, simple model, and weighted model were enlisted to bolster the TSMR analysis. Given the fulfillment of the IV hypothesis by each genetic variant, the amalgamation of IVW with Wald ratio estimations concerning the causal ramifications of distinct SNPs bestows an unwavering appraisal of the causal influence of exposure on the outcome [28]. Nevertheless, it is imperative to exercise vigilant discretion due to IVW's susceptibility to feeble instrumental biases, thereby necessitating meticulous scrutiny. The weighted median (WM) represents the median derived from the distribution function subsequent to arranging all individual SNP effect values based on their weight. Remarkably, WM persists in offering resilient estimations even when confronted with a considerable proportion of erroneous instrumental variables, extending up to 50%, alongside substantial disparities in precision amidst instrumental variable estimations. [29]. Fourth, quality control was performed to ensure the accuracy of the study results, including ① IVs heterogeneity test: in order to test the differences among IV.② Multiplicity test: mainly to test whether there is horizontal multiplicity among individual IVs, using the intercept representation of the MR Egger method, which uses the inverse of the variance of the outcome as the weight to fit, and it adds the intercept term to the regression [30], rather than forcing zero at the origin of the coordinates of the scatter plot as in the fitted curves like other methods. This method is able to assess whether genetic variation has a multiplicative effect on the outcome that differs from zero on average, as well as providing consistent estimates of causal effects under weaker assumptions [31]. If there is horizontal pleiotropy, it implies that the instrumental variables directly affect the outcome, violating the three assumptions under which the MR analysis holds; Leave-one-outsensitivitytest is to calculate MR results for all remaining IVs after removing a particular IV one by one, and we need to remove IVs that have a large impact on the outcome; we also use, for example, MR PRESSO [32] to test whether there is variation on the disease beyond its effect on the Mendelian randomization exposure: whether there is horizontal pleiotropy makes the results less reliable, and if horizontal pleiotropy is detected and there is a significant difference before and after removal of outliers, SNPs with pleiotropy are removed to calculate the correct results. ③Other TSMR analysis methods for screening and validation of instrumental variables and results. ④bidirectional Mendelian randomization analysis methods.

This study used a fixed-effects model, with RA as the exposure factor, and SNPs significantly associated with RA as instrumental variables to measure the outcome variable of kidney function, as measured by estimated glomerular filtration rate (eGFR). Reverse Mendelian randomization (MR) analysis was also conducted, using eGFR as the exposure and RA as the outcome, to determine if there was any causal relationship between the two. The analysis was based on the observation of the β values, P values, OR values, and 95% confidence intervals.

3.1. Study Design

The study design, as shown in Fig. 1, describes a bidirectional MR design between RA and eGFR, using summary data from GWAS. Forward MR Analysis uses RA as exposure, eGFR as outcome, and reverse MR Switches exposure and outcome.

3.2. Data Source

The data pertaining to Rheumatoid Arthritis (RA) was sourced from the Biobank Japan database, which, since its inception in 2003, has endeavored to establish one of the world's most expansive disease biobanks. Its overarching objective is to lay the groundwork for medical research geared towards the customization of healthcare based on the distinctive characteristics of each individual patient. With the consent of a comprehensive cohort encompassing 260,000 patients and 440,000 instances of primarily multifactorial (common) ailments, the Biobank Japan (BBJ) has conscientiously accumulated DNA, serum samples, medical records containing clinical information, and other pertinent data. A staggering assortment of no fewer than 5,800 meticulously scrutinized data points are readily available for scientific inquiry, including survival information concerning the patients, with a remarkable 95% of the cohort diligently monitored for an average period of ten years.The sample size for RA was 212,453, with 8,885,805 SNPs. The data for kidney function also came from the Japanese biobank database, with eGFR as the measurement, and a sample size of 143,658 and 6,108,953 SNPs. Both samples were of East Asian ancestry. OR values and 95% confidence intervals were used to measure the causal relationship between exposure and outcome in the reverse MR analysis. All data used in this study were publicly available, so ethical approval was not necessary.

3.3. Selection of Instrumental Variables

Based on the three assumptions of MR, we selected SNPs strongly associated with RA from the GWAS database with a P value less than 5×10^− 8, which satisfies the association hypothesis, setting the linkage disequilibrium parameter to LD r2 < 0.001, and sets the genetic distance threshold to 10,000 kb to ensure the independence of the IVs. We used the online website PhenoScanner (http://www.phenoscanner.medschl.cam.ac.uk) to search and exclude SNPs related to other phenotypes, which satisfies the independence hypothesis. We used the MR-Egger [33] regression intercept term test for multiple effects analysis to determine whether there is genetic pleiotropy between the SNP and eGFR, and if there is no pleiotropy, the exclusivity hypothesis is satisfied. In addition, the F statistic [34] was calculated to evaluate the presence of weak IV bias, where F < 10 indicates weak instrument strength [35]. When the F value is very weak, it provides very little information about the exposure, which can lead to imprecise estimates of causal effects and bias that contradicts other core assumptions [3]. In two-sample MR, weak instruments can lead to regression dilution bias and underestimate the exposure-outcome association [37]. The formula for calculating the F value is F=(N-K-1/K)×[R2/(1-R2)], where N represents the sample size of the exposure GWAS study, k represents the number of IVs, R2 is the degree to which the IV explains the exposure (the coefficient of determination of the regression equation), and R value can be calculated using the formula R2 = 2×(1-MAF)×MAF×(β/SD)2, where MAF is the minor allele frequency, β is the effect size of the SNP on the exposure, and SD is the standard deviation. MAF and β can be obtained directly, and in calculating R2, it is equivalent to the effect allele frequency (EAF), and SD = SE×√N, where SE is the standard error of β, which can be obtained directly, and N in the F statistic formula is consistent with the N in the exposure GWAS sample size, which represents the sample size of the exposure.All SNPs are shown in Table3 and Table4.

We initially screened 11 SNPs strongly associated with RA, PhenoSacnner is used to check whether there are SNPs associated with recognized confounding factors that may interfere with MR Results. For example, if SNPS are associated with phenotypes such as hypertension, diabetes and obesity, they are removed.A search of the Pub-med database found no evidence that the included SNPs were strongly correlated with confounding factors. The above formula was used to calculate F value, and the result was that all SNPS were qualified with F value > 10. Then, MRPRESSO method was used to test outliers and polypotency. No outliers were reported, and P value was 0.520 > 0.05, indicating no horizontal polypotency, which means that instrumental variables are qualified.According to the 11 related SNPS, corresponding eGFR results were found in BBJ, and then synergies were performed.However, not all SNPS can correspond to the outcome (eGFR). Rstudio automatically eliminated this part of SNPS and finally got 8 SNPS for our analysis.In the reverse MR, the same steps were taken to obtain 54 SNPs.Removing palindrome sequence as rs10840341, rs2337106, rs28607641, rs35925637, rs549752, rs936873 are two outliers, and then we wiped off confounding factors, when the phenotype is not recognized and outcomes associated with a particular SNPS associated, should retrieve the document verification authority.When the corresponding phenotype could not be retrieved, snp was temporarily retained as not directly related to outcome (RA). In addition to IVW, heterogeneity testing was performed to ensure the rigor of the outcome. MR-Egger and leave-one-out [38] were used to evaluate the results model to further verify the reliability of the results. The latter mainly involves sequentially removing each IV and then calculating the combined effect of all remaining SNPs to assess the effectiveness and stability of the results. If a certain MR result has no significant statistical difference from the overall result, it means that the IV will not have a non-specific effect on the effect estimate results; a forest plot is drawn to observe whether individual SNPs affect the main causal relationship.The assumption of MR-Egger method is that the intensity of IV is independent of the effect of IV. The intercept term of this method is an estimate of multiplicity. The P value of intercept term was not significant (P > 0.05), which indicated that there was no pleiotropy. Statistical analysis was carried out in Two-Sample-MR software package.

4.1 TSMR analysis

In the forward MR analysis, the IVW method showed P = 0.717, β = 0.007. The results were consistent with those obtained by WME and MR Egger methods, indicating that RA has no significant causal effect on kidney function. In the reverse MR analysis, the IVW method showed an odds ratio of 0.890, 95% CI: 0.691–1.144, and a p-value of 0.363. The results were consistent with those obtained by other methods, indicating that eGFR has no significant causal effect on RA. The results of the two-sample MR analysis are shown in Table 1. Figure 2.1–3.4 are visual manifestations of MR analysis.

4.2 Sensitivity Analysis

In the forward MR analysis, after preliminary screening of SNPs, the MR PRESSO test was performed and no outliers were reported. After horizontal pleiotropy testing, the P-value was 0.910 > 0.05, indicating that there was little possibility that the causal evaluation of this study was affected by genetic pleiotropy, and the reliability of the results was relatively high. The leave-one-out method was used to analyze the IVW method, and the graph indicated that there was no significant impact on the results when any sensitive SNP was removed, indicating the reliability of the IVW method.In the reverse MR analysis, the same steps were taken. Five palindromic sequences were removed and the MR PRESSO test was performed again. The result showed P < 0.001, indicating the presence of outliers, which made the instrumental variable have pleiotropy, making it unacceptable. Therefore, we removed the outliers and then used the F-value to assess the strength of the instrumental variables. The F-value of each SNP was calculated, and weak instrumental variables were removed when F < 10. Finally, 54 reliable IVs were obtained. After horizontal pleiotropy testing, the P-value was 0.848 > 0.05, and the IVs met the assumptions, indicating that the results were reliable. The leave-one-out method also confirmed that the results were credible. The funnel plots in the bidirectional analysis were symmetrical with respect to SNPs. The results of the sensitivity analysis are shown in Table 2.

In this bidirectional two-sample MR analysis, there was no evidence to support a causal relationship between RA and kidney function in any direction. In previous studies on independence, most tended to believe that RA would lead to a decline in kidney function, which could be due to confounding factors or reverse causality [39]. However, we completed a reverse MR analysis in our study, which was able to eliminate the latter. LaTonya J Hickson et al. suggested that over time, RA patients are more likely to experience a decline in kidney function than healthy individuals, which may be related to cardiovascular disease [40]. SangHeonSuh et al. found that patients with RA had a higher incidence of end-stage renal disease than subjects without RA, and that the risk of end-stage renal disease(ESRD) was higher in relatively young individuals [41].MTX and some DMARDs, the most commonly used drug for the clinical treatment of RA, as mentioned earlier, has been found to have a risk of kidney damage in some studies [42–44], but other studies have shown that low-dose MTX (LD-MTX) is safe in mild to moderate CKD. Before the emergence of MTX and targeted biologics, the occurrence rate of CKD in RA patients is quite high, and most RA patients with CKD in recent years are concentrated in Asia, which may be due to the difference in drug selection of DMARDs in Asia and Western countries, which also indicates the great influence of drugs on the outcome.Daniel H Solomon et al. found in a randomized trial conducted in patients without rheumatic disease that the use of LD-MTX was associated with a mild to moderate increase in the risk of skin cancer and gastrointestinal, infectious, pulmonary, and hematological adverse events, while the risk of renal adverse events was decreased [45–46]. This may indicate that the use of MTX is a more potent confounding factor. Therefore, relevant factors have been filtered out before screening for appropriate SNPs, and the final result indicates that RA and kidney function are not related. Based on this, we can speculate that RA itself does cause kidney damage, but mesangial proliferative glomerulonephritis(MPGN) and membranous nephropathy(MN), which account for a relatively large proportion of renal lesions, have no significant damage to GFR.At the same time, the negative result of MR Analysis does not directly mean that RA and renal function have no causal relationship and are independent of each other.As mentioned above, MPGN is the most common in the course of RA disease, followed by amyloidosis and MN. As the disease is relatively mild, eGFR measured by blood creatinine level is normal in most RA patients with mesangial and membranous GN, and when amyloidosis occurs in patients' kidneys, it indicates that the lesion is more serious.Its typical manifestation is complicated by progressive renal failure [16], which must be accompanied by a decline in eGFR.It was found that eGFR decline, an outcome event, was not sensitive enough to expose RA.What's more,it is worth noting that eGFR is related to factors such as age and environment [47, 48], and environmental factors related to RA such as smoking, obesity, and gut microbiota have also been shown to have a clear correlation with eGFR [49–54]. We have reason to believe that the observation in most studies that RA patients often have a decline in kidney function should be attributed to the complex interactions between RA and various systems in the body and cannot be summarized by causal relationships.

Although TSMR analysis was used and instrumental variables were rigorously screened, this study still has limitations, such as the unsatisfactory results of instrumental variables in heterogeneity testing. Although the rigorous screening and testing of SNPs made the results robust,but exposure and outcome were selected from the same database,bias may still be caused by a small number of genetic instruments, small sample size, or potential sample overlap between exposure and outcome. Secondly, using eGFR instead of kidney function.The indicator cannot fully represent kidney function. Because Scr and BUN based eGFR sensitivity is relatively low and is influenced by many non-GFR factors such as muscle mass, diet, age, etc.Moreover, factors such as race and genetics do have an impact on eGFR [55, 56]. This study only covered the East Asian population, and in the future, GWAS databases from Europe, Africa, and other regions should be included to more clearly understand the relationship between RA and kidney function.

If we want to further comprehensively study the relationship, human metabolic products such as uric acid and creatinine,serum cystatin C,β2-microglobulin ,wihch can be sensitive to the indicators of early renal damage,should be included, and previous studies have pointed out that the creatinine-cystatin C equation has unique advantages in measuring kidney function [57]. CysC is produced by nucleated cells in vivo, and circulatory CysC is only cleared by the kidney, which is an ideal indicator of GFR in patients with normal to moderately impaired renal function [58–59].Moreover, since GWAS were aggregated data, we could not conduct additional analysis, for example, taking RA patients' age and duration of disease or drugs used as exposure, and taking more refined metabolic indicators as outcome.The hope is that the current study can be refined when the data are more detailed.

A)Ethical Approval

The manuscript does not contain clinical studies or patient data.

B) Consent for publication

Not Applicable.

C)Availability of data and materials

The datasets generated and/or analysed during the current study are available in the https://biobankjp.org/en/info/index.html

Also, all data are accessible from the Integrative Epidemiology Unit (IEU) database (https://gwas.mrcieu.ac.uk/).

D)Competing interests

The authors declare that they have no competing interests.

E)Funding

No specific funding was received from any funding bodies in the public, commercial or not-for-profit sectors to carry out the work described in this manuscript.

F)Authors' contributions

Yuheng Zhang and Yuechun Zhao designed the study. Yuheng Zhang and Zhijie Xu performed statistical analyses. All authors interpreted the results. Yuheng Zhang, Yuechun Zhao, Yongzhong Cheng and Zhicheng Sang drafted the manuscript. All authors edited the manuscript for intellectual content. All authors read and approved the final manuscript.

G)Acknowledgements

We would like to sincerely thank the original GWASs and the related consortiums for sharing and managing the summary statistics.

H)Authors’ information

Yuheng Zhang,master candidate,Wangjing Hospital of China Academy of Chinese Medical Sciences,China.

Yuechun Zhao,master candidate,Beijing University of Chinese Medicine,China.

Yongzhong Cheng,Wangjing Hospital of China Academy of Chinese Medical Sciences,China.

Zhijie Xu,master candidate,Beijing University of Chinese Medicine,China.

Zhicheng Sang*,chief physician,Wangjing Hospital of China Academy of Chinese Medical Sciences,China.

Smolen JS, Aletaha D, McInnes IB. Rheumatoid arthritis. Lancet. 2016 Oct 22;388(10055):2023-2038. doi: 10.1016/S0140-6736(16)30173-8. Epub 2016 May 3. Erratum in: Lancet. 2016 Oct 22;388(10055):1984. PMID: 27156434.
Croia C, Bursi R, Sutera D, et al.One year in review 2019: pathogenesis of rheumatoid arthritis. Clin Exp Rheumatol. 2019 May-Jun;37(3):347-357. Epub 2019 May 10. PMID: 31111823.
Malmström, V., Catrina, A. & Klareskog, L. The immunopathogenesis of seropositive rheumatoid arthritis: from triggering to targeting. Nat Rev Immunol 17, 60–75 (2017). https://doi.org/10.1038/nri.2016.124
Radu AF, Bungau SG. Management of Rheumatoid Arthritis: An Overview. Cells. 2021 Oct 23;10(11):2857. doi: 10.3390/cells10112857. PMID: 34831081; PMCID: PMC8616326.
Alivernini S, Firestein GS, McInnes IB. The pathogenesis of rheumatoid arthritis. Immunity. 2022 Dec 13;55(12):2255-2270. doi: 10.1016/j.immuni.2022.11.009. PMID: 36516818.
O'Dell JR. Treating rheumatoid arthritis early: a window of opportunity? Arthritis Rheum. 2002 Feb;46(2):283-5. doi: 10.1002/art.10092. PMID: 11840429.
Soliman MM, Ashcroft DM, Watson KD, et al. Impact of concomitant use of DMARDs on the persistence with anti-TNF therapies in patients with rheumatoid arthritis: results from the British Society for Rheumatology Biologics Register. Ann Rheum Dis. 2011 Apr;70(4):583-9. doi: 10.1136/ard.2010.139774. Epub 2011 Feb 17. PMID: 21330639; PMCID: PMC3048625.
Levey AS, Inker LA, Coresh J. GFR estimation: from physiology to public health. Am J Kidney Dis. 2014 May;63(5):820-34. doi: 10.1053/j.ajkd.2013.12.006. Epub 2014 Jan 28. PMID: 24485147; PMCID: PMC4001724.
Levey AS, Becker C, Inker LA. Glomerular filtration rate and albuminuria for detection and staging of acute and chronic kidney disease in adults: a systematic review. JAMA. 2015 Feb 24;313(8):837-46. doi: 10.1001/jama.2015.0602. PMID: 25710660; PMCID: PMC4410363.
Levey AS, Coresh J, Tighiouart H,et al. Strengths and limitations of estimated and measured GFR. Nat Rev Nephrol. 2019 Dec;15(12):784. doi: 10.1038/s41581-019-0213-9. PMID: 31578495.
Kennedy OJ, Pirastu N, Poole R, et al.Coffee Consumption and Kidney Function: A Mendelian Randomization Study. Am J Kidney Dis. 2020 May;75(5):753-761. doi: 10.1053/j.ajkd.2019.08.025. Epub 2019 Dec 11. PMID: 31837886.
Ponte B, Sadler MC, Olinger E, et al.Mendelian randomization to assess causality between uromodulin, blood pressure and chronic kidney disease. Kidney Int. 2021 Dec;100(6):1282-1291. doi: 10.1016/j.kint.2021.08.032. Epub 2021 Oct 9. PMID: 34634361.
Park S, Lee S, Kim Y, et al.A Mendelian randomization study found causal linkage between telomere attrition and chronic kidney disease. Kidney Int. 2021 Nov;100(5):1063-1070. doi: 10.1016/j.kint.2021.06.041. Epub 2021 Jul 30. PMID: 34339747.
Kjaergaard AD, Ellervik C, Witte DR,et al.Kidney function and risk of dementia: Observational study, meta-analysis, and two-sample mendelian randomization study. Eur J Epidemiol. 2022 Dec;37(12):1273-1284. doi: 10.1007/s10654-022-00923-z. Epub 2022 Nov 4. PMID: 36333541.
Fraenkel L, Bathon JM, England BR, et al. 2021 American College of Rheumatology Guideline for the Treatment of Rheumatoid Arthritis. Arthritis Care Res (Hoboken). 2021 Jul;73(7):924-939. doi: 10.1002/acr.24596. Epub 2021 Jun 8. PMID: 34101387; PMCID: PMC9273041.
Helin HJ, Korpela MM, Mustonen JT, et al. Renal biopsy findings and clinicopathologic correlations in rheumatoid arthritis. Arthritis Rheum. 1995 Feb;38(2):242-7. doi: 10.1002/art.1780380213. PMID: 7848315.
Lopez-Olivo MA, Siddhanamatha HR, Shea B,et al.Methotrexate for treating rheumatoid arthritis. Cochrane Database Syst Rev. 2014 Jun 10;2014(6):CD000957. doi: 10.1002/14651858.CD000957.pub2. PMID: 24916606; PMCID: PMC7047041.
Kapoor T, Bathon J. Renal Manifestations of Rheumatoid Arthritis. Rheum Dis Clin North Am. 2018 Nov;44(4):571-584. doi: 10.1016/j.rdc.2018.06.008. Epub 2018 Sep 7. PMID: 30274624.
Lee JS, Oh JS, Kim YG, et al.Methotrexate-related toxicity in patients with rheumatoid arthritis and renal dysfunction. Rheumatol Int. 2020 May;40(5):765-770. doi: 10.1007/s00296-020-04547-y. Epub 2020 Mar 13. PMID: 32170389.
Couderc M, Tatar Z, Pereira B, et al.Prevalence of Renal Impairment in Patients With Rheumatoid Arthritis: Results From a Cross-Sectional Multicenter Study. Arthritis Care Res (Hoboken). 2016 May;68(5):638-44. doi: 10.1002/acr.22713. PMID: 26314697.
Märker-Hermann E. Nierenmanifestationen bei rheumatoider Arthritis und Spondyloarthritiden [Renal manifestations in rheumatoid arthritis and spondylarthritis]. Z Rheumatol. 2022 Dec;81(10):845-850. German. doi: 10.1007/s00393-022-01279-1. Epub 2022 Oct 20. PMID: 36264330.
Yang PD, Zhu KJ, Lin SM, et al.Association between neutrophils and renal impairment of rheumatoid arthritis: A retrospective cross-sectional study. Immun Inflamm Dis. 2021 Sep;9(3):1000-1008. doi: 10.1002/iid3.459. Epub 2021 May 25. PMID: 34033704; PMCID: PMC8342230.
Zhang T, Liang S, Feng X,et al.ang J, Cheng Z. Spectrum and prognosis of renal histopathological lesions in 56 Chinese patients with rheumatoid arthritis with renal involvement. Clin Exp Med. 2020 May;20(2):191-197. doi: 10.1007/s10238-019-00602-6. Epub 2020 Feb 11. PMID: 32048072.
Kuroda T, Tanabe N, Kobayashi D, et al. Significant association between renal function and area of amyloid deposition in kidney biopsy specimens in reactive amyloidosis associated with rheumatoid arthritis. Rheumatol Int. 2012 Oct;32(10):3155-62. doi: 10.1007/s00296-011-2148-8. Epub 2011 Sep 27. PMID: 21947375.
Smith GD, Ebrahim S. Mendelian randomization: prospects, potentials, and limitations. Int J Epidemiol. 2004 Feb;33(1):30-42. doi: 10.1093/ije/dyh132. PMID: 15075143.
Lawlor DA, Harbord RM, Sterne JA, et al.Mendelian randomization: using genes as instruments for making causal inferences in epidemiology. Stat Med. 2008 Apr 15;27(8):1133-63. doi: 10.1002/sim.3034. PMID: 17886233.
Zheng J, Baird D, Borges MC,et al.Recent Developments in Mendelian Randomization Studies. Curr Epidemiol Rep. 2017;4(4):330-345. doi: 10.1007/s40471-017-0128-6. Epub 2017 Nov 22. PMID: 29226067; PMCID: PMC5711966.
Bowden J, Holmes MV. Meta-analysis and Mendelian randomization: A review. Res Synth Methods. 2019 Dec;10(4):486-496. doi: 10.1002/jrsm.1346. Epub 2019 Apr 23. PMID: 30861319; PMCID: PMC6973275.
Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genet Epidemiol. 2016 May;40(4):304-14. doi: 10.1002/gepi.21965. Epub 2016 Apr 7. PMID: 27061298; PMCID: PMC4849733.
Slob EAW, Groenen PJF, Thurik AR, Rietveld CA. A note on the use of Egger regression in Mendelian randomization studies. Int J Epidemiol. 2017 Dec 1;46(6):2094-2097. doi: 10.1093/ije/dyx191. PMID: 29025040.
Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017 May;32(5):377-389. doi: 10.1007/s10654-017-0255-x. Epub 2017 May 19. Erratum in: Eur J Epidemiol. 2017 Jun 29;: PMID: 28527048; PMCID: PMC5506233.
Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018 May;50(5):693-698. doi: 10.1038/s41588-018-0099-7. Epub 2018 Apr 23. Erratum in: Nat Genet. 2018 Aug;50(8):1196. PMID: 29686387; PMCID: PMC6083837.
Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015 Apr;44(2):512-25. doi: 10.1093/ije/dyv080. Epub 2015 Jun 6. PMID: 26050253; PMCID: PMC4469799.
Burgess S, Thompson SG. Bias in causal estimates from Mendelian randomization studies with weak instruments. Stat Med. 2011 May 20;30(11):1312-23. doi: 10.1002/sim.4197. Epub 2011 Mar 22. PMID: 21432888.
Brion MJ, Shakhbazov K, Visscher PM. Calculating statistical power in Mendelian randomization studies. Int J Epidemiol. 2013 Oct;42(5):1497-501. doi: 10.1093/ije/dyt179. PMID: 24159078; PMCID: PMC3807619.
Davies NM, Holmes MV, Davey Smith G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ. 2018 Jul 12;362:k601. doi: 10.1136/bmj.k601. PMID: 30002074; PMCID: PMC6041728.
Bowden J, Del Greco M F, Minelli C, et al.Assessing the suitability of summary data for two-sample Mendelian randomization analyses using MR-Egger regression: the role of the I2 statistic. Int J Epidemiol. 2016 Dec 1;45(6):1961-1974. doi: 10.1093/ije/dyw220. PMID: 27616674; PMCID: PMC5446088.
Gronau QF, Wagenmakers EJ. Limitations of Bayesian Leave-One-Out Cross-Validation for Model Selection. Comput Brain Behav. 2019;2(1):1-11. doi: 10.1007/s42113-018-0011-7. Epub 2018 Sep 27. PMID: 30906917; PMCID: PMC6400414.
Evans DM, Davey Smith G. Mendelian Randomization: New Applications in the Coming Age of Hypothesis-Free Causality. Annu Rev Genomics Hum Genet. 2015;16:327-50. doi: 10.1146/annurev-genom-090314-050016. Epub 2015 Apr 22. PMID: 25939054.
Hickson LJ, Crowson CS, Gabriel SE, et al. Development of reduced kidney function in rheumatoid arthritis. Am J Kidney Dis. 2014 Feb;63(2):206-13. doi: 10.1053/j.ajkd.2013.08.010. Epub 2013 Oct 4. PMID: 24100126; PMCID: PMC3944015.
Suh SH, Jung JH, Oh TR, et al. Rheumatoid arthritis and the risk of end-stage renal disease: A nationwide, population-based study. Front Med (Lausanne). 2023 Feb 2;10:1116489. doi: 10.3389/fmed.2023.1116489. PMID: 36817794; PMCID: PMC9932810.
Yang Y, Wang X, Tian J, Wang Z. Renal Function and Plasma Methotrexate Concentrations Predict Toxicities in Adults Receiving High-Dose Methotrexate. Med Sci Monit. 2018 Oct 29;24:7719-7726. doi: 10.12659/MSM.912999. PMID: 30369593; PMCID: PMC6216510.
Yang SL, Zhao FY, Song H, et al.. Methotrexate Associated Renal Impairment Is Related to Delayed Elimination of High-Dose Methotrexate. ScientificWorldJournal. 2015;2015:751703. doi: 10.1155/2015/751703. Epub 2015 Jun 21. PMID: 26185782; PMCID: PMC4491404.
Durando M, Tiu H, Kim JS. Sulfasalazine-Induced Crystalluria Causing Severe Acute Kidney Injury. Am J Kidney Dis. 2017 Dec;70(6):869-873. doi: 10.1053/j.ajkd.2017.05.013. Epub 2017 Jun 29. PMID: 28669550.
Sparks JA, Vanni KMM, Sparks MA, et al. Effect of Low-Dose Methotrexate on eGFR and Kidney Adverse Events: A Randomized Clinical Trial. J Am Soc Nephrol. 2021 Dec 1;32(12):3197-3207. doi: 10.1681/ASN.2021050598. Epub 2021 Dec 1. PMID: 34551998; PMCID: PMC8638389.
Solomon DH, Glynn RJ, Karlson EW, et al.Adverse Effects of Low-Dose Methotrexate: A Randomized Trial. Ann Intern Med. 2020 Mar 17;172(6):369-380. doi: 10.7326/M19-3369. Epub 2020 Feb 18. PMID: 32066146; PMCID: PMC7229518.
Otero Gonzalez A, Prol MP, Caride MJ,et al.Estimated glomerular filtration rate (eGFR), 25(OH) D3, chronic kidney disease (CKD), the MYH9 (myosin heavy chain 9) gene in old and very elderly people. Int Urol Nephrol. 2015 Aug;47(8):1403-8. doi: 10.1007/s11255-015-1041-x. Epub 2015 Jul 8. PMID: 26152646.
Jha V, Garcia-Garcia G, Iseki K, et al.Chronic kidney disease: global dimension and perspectives. Lancet. 2013 Jul 20;382(9888):260-72. doi: 10.1016/S0140-6736(13)60687-X. Epub 2013 May 31. Erratum in: Lancet. 2013 Jul 20;382(9888):208. PMID: 23727169.
Takeno M, Kitagawa S, Yamanaka J, et al.5-Hydroxy-2-methylpyridine Isolated from Cigarette Smoke Condensate Aggravates Collagen-Induced Arthritis in Mice. Biol Pharm Bull. 2018;41(6):877-884. doi: 10.1248/bpb.b17-00982. PMID: 29863076.
Li F, Wang M, Wang J, et al. Alterations to the Gut Microbiota and Their Correlation With Inflammatory Factors in Chronic Kidney Disease. Front Cell Infect Microbiol. 2019 Jun 12;9:206. doi: 10.3389/fcimb.2019.00206. PMID: 31245306; PMCID: PMC6581668.
Muñiz Pedrogo DA, Chen J, Hillmann B,et al.An Increased Abundance of Clostridiaceae Characterizes Arthritis in Inflammatory Bowel Disease and Rheumatoid Arthritis: A Cross-sectional Study. Inflamm Bowel Dis. 2019 Apr 11;25(5):902-913. doi: 10.1093/ibd/izy318. PMID: 30321331; PMCID: PMC6458525.
Jubair WK, Hendrickson JD, Severs EL,et al.Modulation of Inflammatory Arthritis in Mice by Gut Microbiota Through Mucosal Inflammation and Autoantibody Generation. Arthritis Rheumatol. 2018 Aug;70(8):1220-1233. doi: 10.1002/art.40490. Epub 2018 Jul 2. PMID: 29534332; PMCID: PMC6105374.
Hammer Y, Cohen E, Levi A, Krause I. The Relationship between Cigarette Smoking and Renal Function: A Large Cohort Study. Isr Med Assoc J. 2016 Sep;18(9):553-556. PMID: 28471605.
Tsuboi N, Okabayashi Y. The Renal Pathology of Obesity: Structure-Function Correlations. Semin Nephrol. 2021 Jul;41(4):296-306. doi: 10.1016/j.semnephrol.2021.06.002. PMID: 34715960.
Tsao YC, Chen JY, Yeh WC, Li WC. Gender- and Age-Specific Associations between Visceral Obesity and Renal Function Impairment. Obes Facts. 2019;12(1):67-77. doi: 10.1159/000496626. Epub 2019 Feb 6. PMID: 30726849; PMCID: PMC6465737.
Hsu CY, Yang W, Parikh RV, et al.Race, Genetic Ancestry, and Estimating Kidney Function in CKD. N Engl J Med. 2021 Nov 4;385(19):1750-1760. doi: 10.1056/NEJMoa2103753. Epub 2021 Sep 23. PMID: 34554660; PMCID: PMC8994696.
Tong Y, Liu X, Guan M, et al. Evaluation of Serological Indicators and Glomerular Filtration Rate Equations in Chinese Cancer Patients. Med Sci Monit. 2017 Jun 17;23:2949-2960. doi: 10.12659/msm.902138. PMID: 28623247; PMCID: PMC5486681.
Inker LA, Schmid CH, Tighiouart H, et al.Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med. 2012 Jul 5;367(1):20-9. doi: 10.1056/NEJMoa1114248. Erratum in: N Engl J Med. 2012 Aug 16;367(7):681. Erratum in: N Engl J Med. 2012 Nov 22;367(21):2060. PMID: 22762315; PMCID: PMC4398023.
Allinovi M, Sessa F, Villa G, et al. Novel Biomarkers for Early Detection of Acute Kidney Injury and Prediction of Long-Term Kidney Function Decline after Partial Nephrectomy. Biomedicines. 2023 Mar 28;11(4):1046. doi: 10.3390/biomedicines11041046. PMID: 37189664; PMCID: PMC10135876.

Tables 1 to 4 are available in the Supplementary Files section.

No competing interests reported.

Table1.png
Table 1:RA and its association with eGFR in the MR analyses. Including 5 ways analysing SNPs.
Table2.png
Table2. We can find the results of heterogeneity statistics and pleiotropy test. The exist of heterogeneity is acceptable, while the exist of pleiotropy is not allowed. The result was eligible.
Table3Sheet1.jpg
Table3. The detailed information of SNPs in the forward MR analysis.
Table4Sheet1.jpg
Table4.The detailed information of SNPs in the reverse MR analysis.

Download PDF

Version 1

posted

You are reading this latest preprint version

Association between Rheumatoid Arthritis and Renal Function: A Bidirectional Mendelian Randomization Study

Status:

Version 1

Abstract

Figures

1. Introduction

2. Methods

3. Materials and Statistical Analysis

3.1. Study Design

3.2. Data Source

3.3. Selection of Instrumental Variables

4. Results

4.1 TSMR analysis

4.2 Sensitivity Analysis

5. Discussion

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1