R program

We explored several ordination methods available in the R program (R Foundation for Statistical Computing, Vienna, AT) through the JUICE vegetation analysis package (Tichý, 2002). Detrended Correspondence Analysis (DCA; Hill & Gauch, 1980) provided the clearest, most easily interpreted separation of plots along complex environmental gradients. Plot and species similarities were calculated using the Sørenson similarity index. Rare species were down‐weighted and the axes scaled according to the program defaults. The four main DCA axes 1, 2, 3 and 4 were correlated with continuous and ordinal environmental variables in each plot using species–environment correlations in the program CONOCO via JUICE. Only variables with p ≤ 0.002 determined by global permutation test with forward selection (number of permutations: 499) are shown in the biplot diagrams.

Categorical variables were expressed as counts and percentages and quantitative variables as means ± standard deviation. Quantitative variables were compared using Student's t test or the Mann-Whitney test, as appropriate. Categorical variables were compared using the chi-square test and Fisher's exact test. Linear regression models were used to evaluate the effect of variables on AIx@75 and PWV. Variables with p<0.20 in the univariate analysis were included in a saturated model and assessed with a stepwise strategy. The quality of fit was assessed by the R² value and residual analysis. Analyses were performed using the R program (The R Foundation, Auckland, New Zealand), version 3.3.2, and a significance level of 5% was adopted.

The RNA expression data (level 3) of CRC patients—obtained from 622 CRC cancer tissues and 51 adjacent non-tumor normal tissues (up to June 13, 2018)—were downloaded from the TCGA data portal. The expression profiles of RNA and miRNA from the 673 samples had been derived from the IlluminaHiSeq RNASeq and the IlluminaHiSeq miRNASeq sequencing platforms. The mRNAs, lncRNAs, and pseudogenes were identified based on the annotation from the Ensembl database (http://www.ensembl.org/index.html, version 93). RNAs not included in the Ensembl database were excluded from the present study. We mainly used the R program (R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/) for analysis of RNA data. Raw counts data were normalized by the edgeR package [6 (link)] and then transformed by the limma package [7 (link)].

Experimental comparison between groups was performed with one-way analysis of variance test (ANOVA) and post-hoc, with Duncan’s multiple comparison test; examined properties included microhardness, shear bond strength, antibacterial test, cell viability test, and pH cycle test. ARI was verified with the Kruskal-Wallis test. All statistical analysis was performed with the R program (version 3.5.1; R Foundation for Statistical Computing, Vienna, Austria).

Descriptive statistical analysis was applied to ascertain baseline characteristics. Categorical variables were expressed as frequencies and were analyzed using the Chi-square test or Fisher’s exact test. Continuous variables were expressed as means with standard deviations and were analyzed using Student’s T-test. The Kolmogorov–Smirnov test was used to test for the normality of data distribution in all datasets. To identify predictors of outcomes, intergroup covariates, including the binary variable of complication development, were evaluated using multivariate analysis, which was independently performed by logistic regression using the “enter” method. Age (e.g., child, adult, and older adult), surgical history, time from onset, and time to antibiotics from onset were adjusted.
We used the Jamovi statistical program (version 2.3.18, The Jamovi project, Sydney, Australia) and R program (version 4.2.2, The R Foundation for Statistical Computing, Vienna, Austria) for all statistical analyses. A p-value of <0.05 was considered statistically significant.

Continuous variables were presented as means and standard deviations (mean±SD) and compared by using Student's t-test. Categorical variables were expressed as frequencies and percentages (n, %) and compared by using Chi-square test.
Restricted cubic-spline (RCS) plots with three knots were used to explore the shape of the association of FPG or 2h-PG with the primary outcomes (Figure 2). Age-adjusted and multi-adjusted cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (95%CIs) for each tertile of or per one SD increase in baseline level of FPG and 2h-PG. Analyses were also conducted among participants within normal glycemia level and free of diabetes. Considering the possible effect modification by metabolic states, we repeated the above analysis in participants stratified by baseline level of SBP, TG or BMI, as well as the status of hypertension, dyslipidemia or obesity. Subgroup analysis were also performed in subgroups stratified by demographic factors (age, gender, ethnic group, and area).
The level of statistical significance was defined as α = 0.05 of two-side probability. All analyses and figures were performed by using R program (version 4.1.0, R Foundation for Statistical Computing, Vienna, Austria).