Statistical analysis

Continuous variates are indicated as mean ± standard deviation (SD) or median (range) according to data distribution, while categorical variates given as frequencies and percentages. The Kruskal–Wallis H-test (skewed distribution), one-way ANOVA test (normal distribution), and χ² (categorical variables) were applied to evaluate variance among different DII (tertiles). The Kaplan–Meier method was used to estimate the survival among different DII levels, and any differences in survival were evaluated with a stratified log-rank test. The study population was censored on 31 December 2015 if they survived the thorough follow-up cycle. The univariate and multivariate Cox proportional-hazards regression models were applied to analyze the relationship between DII and all-cause mortality with three different models. Therefore, no covariates were adjusted in model 1, and only socio-demographic variables were adjusted in model 2, while the covariates presented in Table 1 were fully adjusted in model 3. To test the robustness of our results, we converted DII into a categorical variable according to tertile and calculated the p for trend to confirm the results of DII as the continuous variable, along with the examination of non-linear possibility.

Baseline characteristics of the study participants with chronic kidney disease (CKD) (n = 4,554).

Values are presented as mean (standard deviation, SD) or median (Q1–Q3) for continuous variables, and as number (percentage) for categorical variables. BMI, body mass index; eGFR, estimated glomerular filtration rate.

All the statistical analyses were performed with software EmpowerStats (X&Y Solutions, Inc., Boston, MA)¹ and software R (The R Foundation, RRID:SCR_001905).² The results were affirmed statistically significant if p-values were less than 0.05 (two-sided).