Study outline
This study was performed completely in accordance with the relevant guidelines and regulations of Mendelian Randomization study13.
In this study, MR method was used as a method to determine whether there is a causal association between BMI and IKD, and the aggregated data of SNPs exposure (BMI) and SNPs outcomes (IKD) based on GWAS were used. The basic conditions for a genetic variation to be an IV are summarized as: Ⅰ. The variant is associated with exposure; Ⅱ. The variant does not affect the outcome directly, only possibly indirectly via the exposure; Ⅲ. The variant is not associated with the outcome via a confounding pathway, which follows the basic principles of MR research14. We performed MR analysis by inverse-variance weighted (IVW), MR-Egger, Weighted median. In order to test the robustness of the correlation, we further conducted sensitivity analyses through Cochran’s Q test, MR-Egger intercept test and leave-one-out analyses. Statistical analyses were performed using R studio v.4.2.1 (the R studio Foundation: Open source & professional software for data science teams. https://www.rstudio.com/; the R Foundation: The R project for Statistical Computing. https://www.R-project.org/;). An R package-TwoSampleMR v.0.5.6 was used for conducting the two-sample MR.
Ethics Approval
Our research is based on the currently publicly available statistical data, and no human patients are included in the research, so it doesn't need to be approved by the ethics review committee.
Genetic Instruments For Body Mass Index
The GWAS summary dataset for BMI was extracted from the Genetic Investigation of Anthropometric Traits (GIANT) consortium ( https://portals.broadinstitute.org/collaboration/giant/index.php/GIANT_consortium_data_files/ ), and identified from the meta-analysis of 681,275 individuals of European ancestry15.
Genetic Summary Data For Internal Knee Derangement
The genetic summary data for IKD can be obtained from the UK Medical Research Council Integrative Epidemiology Unit Open GWAS Project database ( https://gwas.mrcieu.ac.uk ). Developed by the MRC Integrative Epidemiology Unit (IEU) at the University of Bristol, this resource is a manually collected complete GWAS summary datasets, which can be downloaded as an open-source files or obtained by querying the database of complete data. The relevant ethics committees approved all studies that provided data for these analyses, and all participants provided written informed consent.
Ivs Filtering And Harmonization
, we select the genetic instruments for BMI via the following steps: Ⅰ. Instruments were selected based on their genome-wide significance (p-value༜5×10− 8) and independence (linkage disequilibrium [LD] r2 of ༜0.001, and ༜1MB from the index variant)16. Ⅱ. To evaluate the strength of IVs, we calculate the F-statistic for each SNP using the following formula: F-statistic = R2×(N-2) / (1- R2), where R2 is the variance of the phenotype explained by each genetic variant in exposure, and N is the sample size. R2 was calculated using the following formula: R2 = 2×(1-EAF) × EAF × β2, where EAF is the effect allele frequency and beta is the per allele effect size of the association between each SNP and phenotype 17. F-statistic larger than the conventional value of 10, which means the instruments used strongly predict the BMI 18,19.
After selecting genetic instruments for BMI, we then extracted BMI-associated SNPs from the IKD data. Since IVs cannot be directly related to the outcome, we manually eliminate SNPs with P༜5×10− 8. In addition, to test whether confounding factors violate the significant estimate, we check in PhenoScanner ( https://www.phenoscanner.medsch1.cam.ac.uk ), a comprehensive information platform on genotype and phenotype association, to see whether these SNPs are related to the potential risk factors and remove SNPs associated with any of these potential confounders in a genome-wide sense.
Finally, we harmonized the aggregated data for exposure and outcome so that the effect alleles reflected the allele associated with exposure. When SNPs are palindromic, just like A/T or G/C, we used the allele frequency information to resolve chain ambiguity. If they were missing a P, β, or a se for the data, we excluded SNP-trait associations from the GWAS catalog. Proxy SNPs were not included in analyses 19.
Mendelian Randomization Analyses
After filtering and harmonizing IVs, to assess causal associations between BMI and IKD, we performed MR analyses. We performed MR analysis mainly by inverse-variance weighted (IVW), MR-Egger, Weighted median. The results are expressed in odd ratios (OR) and 95% confidence intervals (CI), which provided an estimate of the relative risk caused by the increase of each standard deviation increase in the BMI.
The inverse-variance weighted (IVW) model was conducted to examine the causal association, and this approach was considered as the main analysis because of the potential observed heterogeneity 20. The IVW method combines individual MR effects across SNPs to obtain an overall weighted effect of the potential causal association. MR-Egger and weighted median were used to improve the IVW estimates as they could provide more robust estimates in a broader set of scenarios 21.
In order to detect and correct the pleiotropy in the IVW analyses, MR-Egger and Weighted median analyses were then conducted. The MR-Egger allows all genetic variants to have pleiotropy but requires that the pleiotropic effects be independent of the variant-exposure association. MR-Egger methodology tests and explains the existence of unbalanced pleiotropy by introducing this biased parameter and combining outline information estimates of causative effects from multiple individual variants 22.
Weighted median method orders the estimates in MR using each instrument weighted for the inverse of their variance, and the median result is selected and show the single MR estimate with confidence intervals based on bootstrapping technique 23. The weighted median requires and assumes that at least half of the instruments are valid. Compared with MR-Egger analysis, the weighted median calculator has the advantage of maintaining larger precision within the estimates 24. When the p values༜0.05, the test is statistically significant.
Sensitivity Analyses
In order to test the robustness of the correlation, we further conducted sensitivity analyses through Cochran’s Q test, MR-Egger intercept test and leave-one-out analyses. The Cochran’s Q test was used to identify heterogeneity25. For significant estimates, we further assessed horizontal pleiotropy using the MR-Egger intercept test and leave-one-out analyses which we sequentially omitted one SNP at a time, to evaluate whether the MR estimate was driven or biased by a single SNP. Th value of the intercept provides an estimate of the degree of pleiotropy affecting the result, and the beta (slope) coefficient represents the causal effect between exposure and outcome adjusted for pleiotropy. As the intercept neared zero in the MR-Egger intercept test, horizontal pleiotropy was reduced. A funnel plot was also used to assess the probable directional pleiotropy 26. In all these three statistical methods, the threshold of statistical significance for evidence of pleiotropy is p༜0.05.
Submit Statements
All authors claim that this manuscript has not been repeatedly submitted and published in any other journals. If it is accepted, it will not be published in any other journals. All authors participated in the editing of the manuscript, strictly examined the accuracy of the data and agreed to the final manuscript.