Statistical analysis
Data were analyzed using SPSS 20.0 software. Normally distributed continuous data were presented as mean ± standard deviation, and non-normally distributed continuous data were expressed as median and quartile interval (M [P25-P75]). Categorical data were reported as the number of cases and frequency (%). Independent sample T-test was used for intergroup comparisons of normally distributed continuous data. Non-parameter test (rank-sum) was used for non-normal distribution. The Chi-square tests (or Fisher’s exact tests, if appropriate) were used to determine the statistical significance between percentages for categorical data. A multivariate logistic regression model was used to determine odds ratios (ORs) and associated 95% confidence intervals (CIs) when comparing LBR and CLBR between two BMI groups. Models were adjusted for a series of possible confounders: maternal age, fertilization method, basal follicle-stimulating hormone (FSH), antral follicle count, and retrieved oocytes. Statistical significance was set at P < 0.05.