Data analysis
SPSS Statistics 21.0 software was used for statistical analysis. Chi square test was used to compare the count data. The significant variables in univariate analysis were used as independent variables, and multiple linear regression was performed. By using Delta method, the excel table compiled by Andersson et al. [35] was introduced to calculate the related indexes of interaction, and the β1, β2, (β1+β2+β3) and the variance and covariance between factors in logistic regression analysis were input into it. The OR value obtained by logistic regression model in the interaction calculation process was used as the estimation value of relative risk (RR). Interaction index: (1) the relative excess risk due to interact (RERI) was used to evaluate the difference between the combined effect of factor A and factor B and the sum of factors A and B alone; (2) the attributable proportion due to interaction (AP) was used to evaluate the proportion of interaction between two factors when two factors A and B exist at the same time; (3) the synergy index (S): when the interaction does not exist, the confidence interval of RERI and AP should contain 0, and the confidence interval of S should include [36-38]. Correlations among variables were examined by Pearson’s correlation. Hierarchical regression analysis was used to prove the relationship of variables and to examine the moderating effect. Finally, the simple slope analysis was conducted to visualize the interaction term [39]. Significance level was α=0.05, and a two-tailed P<0.05 was considered to have statistical significance.