Analyses
The outcome variable was ECC. Risk factors for the prevalence of ECC were evaluated by univariate and multivariate analysis using a logistic regression model. The odds ratios (ORs) and their 95% confidence intervals (CIs) were estimated with regard to risk factors for ECC. Before the multivariate logistic regression analysis was conducted, we evaluated multicollinearity among the variables using Spearman’s rank correlation test. Tests of statistical significance were based on a two-sided P-value, and the α-error was set at 5%. The SAS system (ver. 9.2; SAS Institute, Cary, NC, USA) was employed for the analysis.