*2.5. Statistical Analysis*

Data were analyzed with SPSS 17.0 (SPSS Inc., Chicago, IL, USA, 2008). The Mann–Whitney U test was used to compare the mean di fferences between men and women, while the Pearson chi-squared test was used to compare the percentage di fferences between men and women. The one-sample Kolmogorov–Smirnov test was used to examine departures from a normal distribution of continuous variables, and a base-10 logarithmic transformation was applied to the variables that showed rightward skewing. A multivariable regression model analysis was used to evaluate the association of the eGFR with independent variables including the excretion of Cd and Pb. For each regression model, the coe fficient of determination ( *R*2) value was obtained together with standardized β. A generalized linear model analysis was used to estimate the mean eGFR with adjustment for age, covariates and interactions. Logistic regression analysis was used to estimate the prevalence odds ratio (POR) for the reduced eGFR across the quartiles of excretion of Cd and Pb. The *p*-values ≤0.05 for two-sided tests were assumed to indicate statistical significance.
