Statistical analysis

Data were analyzed using STAT/IC 16.1(StataCorp LLC, Texas, US) software, and figures were created using R Core Team (2018, R: A language and environment for statistical computing; R Foundation for Statistical Computing, Vienna, Austria; URL https://www.R-project.org/) software. Data were examined for normality; normally distributed variables are presented as mean ± standard deviation and non-normally distributed as median (interquartile range). Paired t-test and Wilcoxon signed-rank test were used to compare the change in the continuous variables at baseline and 3 months after surgery. Chi-square analysis was done to compare the categorical variables. Since the reference range for HDL-C and apo A-I differs for male and female, we have presented gender specific HDL-C and apo A-I values for control and individuals with obesity before and after surgery in the study. The comparison for fraction change in variables between the two types of surgery was examined using a t-test. Pearson correlation testing was used to assess the correlation between continuous variables. Univariate linear regression analysis was performed to identify the variables associated with the change in cholesterol efflux capacity. Multivariable linear regression analysis was performed to predict the level of effect the independent variables have on the outcome variable (cholesterol efflux capacity). Multivariable linear regression analysis was done by including the variables that were found to be associated (p < 0.2) with cholesterol efflux capacity during univariate analysis. P values of < 0.05 were considered significant.