Change in body weight by pre‐trial insulin dose group in the (a) DUAL V and (b) DUAL VII trials. Data are observed means based on the full analysis set using either last observation carried forward imputed data for DUAL V or observed data with no imputation for DUAL VII. In DUAL V, change in body weight was analysed using an analysis of covariance (ANCOVA) model (including the variables: subgroup, treatment, interaction between subgroup and treatment and region as fixed factors and baseline response as covariates). In DUAL VII, the mixed model of repeated measures (MMRM) included subgroup, treatment, visit and region as fixed factors and baseline response as covariate and the interactions subgroup × treatment × visit, region × visit, baseline response × visit. *P ≤ 0.0001. There was no significant interaction between pre‐trial insulin dose groups and treatment arm in the DUAL V (P = 0.17) trial; however, there was a significant interaction in DUAL VII (P = 0.03) in participants’ body weight. ETD, ETD, estimated treatment difference; IAsp, insulin aspart; IDegLira, insulin degludec/liraglutide; IGlar U100, insulin glargine 100 units/ml.