Assessing The Relationship Between High-sensitivity C-reactive Protein And Kidney Function Employing Mendelian Randomization in The Japanese Community-based J-MICC Study

ABSTRACT

Background: Inflammation is thought to be a risk factor for kidney disease. However, whether inflammatory status is either a cause or an outcome of chronic kidney disease remains controversial. We aimed to investigate the causal relationship between high-sensitivity C-reactive protein (hs-CRP) and estimated glomerular filtration rate (eGFR) using Mendelian randomization (MR) approaches. 

Methods: A total of 10,521 participants of the Japan Multi-institutional Collaborative Cohort Study were analyzed in this study. We used two-sample MR approaches (the inverse-variance weighted (IVW), the weighted median (WM), and the MR-Egger method) to estimate the effect of genetically determined hs-CRP on kidney function. We selected four and three hs-CRP-associated single nucleotide polymorphisms (SNPs) as two instrumental variables (IV): IVCRP and IVAsian, based on SNPs previously identified in European and Asian populations. IVCRP and IVAsian explained 3.4% and 3.9% of the variation in hs-CRP, respectively. Results: Using the IVCRP, genetically determined hs-CRP was not significantly associated with eGFR in the IVW and the WM methods (estimate per 1 unit increase in ln(hs-CRP), 0.000; 95% confidence interval [CI], −0.019 to 0.020 and −0.003; 95% CI, −0.019 to 0.014, respectively). For IVAsian, we found similar results using the IVW and the WM methods (estimate, 0.005; 95% CI, −0.020 to 0.010 and −0.004; 95% CI, −0.020 to 0.012, respectively). The MR-Egger method also showed no causal relationships between hs-CRP and eGFR (IVCRP: −0.008; 95% CI, −0.058 to 0.042; IVAsian: 0.001; 95% CI, −0.036 to 0.036). 

Conclusions: Our two-sample MR analyses with different IVs did not support a causal effect of hs-CRP on eGFR.

Keywords: hs-CRP; eGFR; Mendelian randomization study; genetic epidemiology; inflammation

CLICK HERE TO GET NATURAL ORGANIC CISTANCHE EXTRACT WITH 25% ECHINACOSIDE AND 9% ACTEOSIDE FOR KIDNEY FUNCTION

Supportive Service Of Wecistanche-The largest cistanche exporter in the China:

Email:wallence.suen@wecistanche.com 

Whatsapp/Tel:+86 15292862950

Shop For More Specifications Details:

https://www.xjcistanche.com/cistanche-shop

INTRODUCTION 

Systemic inflammation is considered one of the risk factors for common chronic diseases, including diabetes mellitus,1 hypertension,2 cardiovascular diseases,3 and chronic kidney disease (CKD).4 Generally, C-reactive protein (CRP) has been used as a biomarker of systemic inflammation in clinical and basic research. Although the previous longitudinal studies have examined the association between CRP levels and CKD in different populations, evidence of the causality of this association remains controversial.5–7 However, some researchers demonstrated the effect of CRP-oriented biological functions on kidney function.8,9 One researcher also has published a meta-analysis suggesting that vitamin D supplementations could lower circulating CRP levels.10 Taken together, these studies suggest that interventions on CRP may help to improve renal function.

In recent years, the Mendelian randomization (MR) approach has attracted much attention in genetic epidemiology. The biggest advantage of this method is to investigate a causal relationship between an exposure (X) and an outcome (Y) from an observational dataset using genetic variants as instrumental variables (G: IV).11 The development of MR analysis consecutively occurred after identifying single nucleotide polymorphisms (SNP) in genome-wide association studies (GWAS). As with other health outcomes, previous GWAS identified SNPs associated with CRP levels, including the CRP gene in chromosome 1.12,13 Interestingly, it is known that serum CRP levels are affected by genetic polymorphisms,14 which indicates that SNPs associated with CRP levels may reflect the long-time exposure to higher=lower CRP level. Therefore, SNPs associated with serum CRP levels are suitable for IVs to investigate the causal relationships between CRP and several pathophysiological conditions, and used in previous MR studies among adults in European countries.15–17

In Asian countries, large-scale cohort studies have collected human genomes and performed genotyping in the past several decades. Several researchers conducted GWAS and found novel loci associated with CRP levels in Asian populations.18–20 These studies enable researchers to conduct MR studies using CRP-associated SNPs in Asian populations, which seems to be important in terms of ethnic differences. Therefore, we investigated whether genetically determined hs-CRP levels using two different IVs, based on SNPs identified in European and Asian populations, were causally related to kidney function in a Japanese population using MR approaches.

METHODS

Study subjects 

The study subjects were participants of the Japan Multi-institutional Collaborative Cohort (J-MICC) Study which was conducted in 14 study areas throughout Japan. The purpose of the J-MICC study was to find out the risk factors of cancer and other diseases by examining the relationship between genetic variants, lifestyle habits, blood components, and disease. The eligibility for the J-MICC Study was adults aged 35–69 years living in each study area. The details of the J-MICC Study have been described previously elsewhere and the latest information is available on its website (http:==www.jmicc.com).21,22 The selection process of participants is shown in Figure 1. From the genotyped 14,539 subjects, 26 samples with inconsistent sex information between the questionnaire and an estimate from genotype were excluded. The identity-by-descent method implemented in the PLINK 1.9 software (https:==www.cog-genomics.org=plink2) identified 388 relative pairs (pi-hat >0.1875) and one sample of each pair was excluded. Principal component analysis with a 1,000 Genomes reference panel (phase 3) (http:==www. international genome.org= category=phase-3=) detected 34 subjects whose estimated ancestries were outliers from the Japanese population. The 34 samples were excluded. Among all the remaining 14,091 samples, five subjects withdrew their consent to participate, leaving 14,086 subjects for the final analyses. Of these, the values of serum hs-CRP were available only at three study sites. Therefore, we decided to use a two-sample MR study design, rather than a single sample MR for a smaller dataset. We divided the participants into two groups; 1) 2,503 participants (available for hs-CRP) and 2) 12,501 participants (non-overlapping participants), which are by a basic principle of two-sample MR (nonoverlapping populations with the same ethnicity, similar sex, and age distribution).23 After excluding participants who had an extremely high value for hs-CRP (hs-CRP >3.0 mg=dL, n = 828) and eGFR (eGFR >120 mL=min=1.73 m2, n = 3,647), and lower than the limit of quantification for hs-CRP (n = 8), a total of 10,521 Japanese (1,667 for genetic association with hs-CRP [called as CRP dataset] and 8,854 for genetic association with eGFR [called as eGFR dataset]) were analyzed in the two-sample MR of this study. Written informed consent was obtained from all participants of this study. The J-MICC Study was conducted with adherence to the Ethical guidelines for the Human Genome and Genetic Sequencing Research. The procedure of this study was approved by the Ethics Review Committee of the Nagoya University Graduate School of Medicine (939-14), Aichi Cancer Center, and all research institutes. We performed analyses using the dataset of version 20190728.

Measurement of hs-CRP and eGFR 

Serum samples were collected from all participants. We measured hs-CRP using latex-enhanced nephelometry. Serum creatinine was measured using an enzymatic method. Some institutes measured serum creatinine using the Jaffe method and then transformed it to the equivalent value of the enzymatic method. eGFR was calculated using the Japanese equation proposed by the Japanese Society of Nephrology: eGFR (mL= min=1.73 m2 ) = 194 × serum creatinine (mg=dL)−1.094 × age−0.287 (× 0.739 for women).24

Selection of instrumental variables 

The list of candidate SNPs for IVs is shown in table 1. First, we selected four SNPs (rs3093077, rs1205, rs1130864, and rs1800947) within the CRP gene which was used as IVs in previous MR studies.15 In this study, These SNPs were selected as a minimum subset to obtain diversity at the CRP gene in European populations and called IVCRP. Next, we considered that it is necessary to select SNPs and develop original IVs in an Asian population because the IVCRP was developed based on SNPs identified in people of European descent. Therefore, we searched the word 'CRP' in the GWAS catalog (https:==www.ebi.ac.uk= gwas=), and narrowed down to the studies according to the following criteria: 1) a study conducted in an Asian population, and 2) a study with both of the discovery and the replication phase. After the web-based selection, we finally selected 13 SNPs. For 151233628, due to low imputation quality (MAF <0.05 and r2 < 0.3), this SNP was not included in the original J-MICC dataset. Of the remaining 12 SNPs, 6 SNPs (rs12133641, rs9375813, rs2097677, rs79802086, rs2393791, and rs1169284) were excluded because these SNPs were not significantly associated with hs-CRP in our dataset (P > 0.0042 = 0.05=12). Next, rs814295 (GCKR) and rs429358 (APOE) were likely to have pleiotropic effects on kidney function. rs3093059 was excluded due to the high linkage disequilibrium (LD) with rs3093068 in the CRP dataset (r2 > 0.9). Finally, three SNPs (rs30933068, rs7553007, and rs7310409) were included in our analysis and were called as IVAsian (Table 2).

Statistical analysis

To confirm the cross-sectional association between hs-CRP and eGFR, multiple linear regression analysis was performed with adjustments for sex, age, and study sites. We performed two-sample MR approaches after dividing the participants into two datasets (CRP and eGFR datasets) as described above. Methods for two-sample MR were different from the one-sample MR method, which was described in previous methodological papers.25–27 The inversevariance weighted method (IVW) is a conventional approach to estimating a causal effect on a study outcome from different studies in meta-analysis.25 In the setting of MR analysis, the IVW method can provide a combined estimate weighted using the inverse variances of the causal effect of per-allele. However, this method can be biased when a genetic variant violates the assumptions of MR (eg, pleiotropic effect).27 Therefore, we also performed two other methods (the weighted median (WM) and the MR-Egger method) which can provide consistent estimates even under the weaker assumption.25 The MR-Egger analysis is also useful to detect either both directional pleiotropy or violation of the Instrument Strength Independent of Direct Effect (InSIDE) assumption. Additionally, the F-statistic was calculated for each IV from linear regression analyses to test whether IVs are strongly associated with exposure (referred to as relevance assumption).28 We performed linear regression analyses using the lm function in R and included all SNPs used in each IV in models. An arbitrary threshold of F-statistic >10 was used to avoid using weak genetic instruments in this study.29 All statistical analyses were performed using the software R version 3.5.0 (R Foundation for Statistical Computing, Vienna, Austria). In particular, an R package of "MendelianRandomization" was used for two-sample MR analyses.30