paraphrase, "Weighted prevalences, means and standard deviations stratified according to DSME-participation status were calculated, using weighting factors computed based on the national population statistics of 31.12.2014. This procedure took the multistage sampling of the GEDA 2014/2015-EHIS survey into account and allowed us to address potential sample-selection bias with regard to sex, age, community type and educational level [35], [36]. We performed weighted logistic regression analyses, in which we used DSME-participation as a predictor for all examined SMB and clinical care variables, which served as outcomes. These analyses were expanded by a stepwise inclusion of the above named sociodemographic and disease-related confounders. All data was interpreted with reference to the final multivariable models. All analyses were performed using STATA version 16.1