ABSTRACT: AIMS:To assess whether people with type 2 diabetes transferring from higher basal insulin doses (> 20 units) to a starting dose of 16 units of insulin degludec/liraglutide (IDegLira) benefit from IDegLira with/without transient loss of glycaemic control. METHODS:Post hoc analysis of DUAL V and VII assessed fasting self-measured blood glucose (SMBG) over weeks 1-8, changes in HbA1c, body weight and mean insulin dose over 26 weeks, and percentage of participants achieving HbA1c < 53 mmol/mol (7.0%) by end of trial in participants with type 2 diabetes uncontrolled with basal insulin. IDegLira was compared with continued up-titration of insulin glargine (IGlar U100) in DUAL V, or switching to basal-bolus therapy in DUAL VII (IGlar U100 and insulin aspart), across pre-trial insulin dose groups (20-29, 30-39 and 40-50 units/day). RESULTS:In all subgroups, participants treated with IDegLira experienced significant improvements in HbA1c by end of trial, which were greater than with IGlar U100 up-titration (estimated treatment difference -5.86, -6.59 and -6.91 mmol/mol for pre-trial insulin doses of 20-29, 30-39 and 40-50 units/day, respectively) and similar to basal-bolus therapy (estimated treatment difference -0.16, -1.0 and -0.01 mmol/mol for pre-trial insulin doses of 20-29, 30-39 and 40-50 units/day, respectively). Compared with IGlar U100 and basal-bolus therapy, IDegLira participants experienced weight loss vs. weight gain, lower rates of hypoglycaemia and a lower mean end of trial total daily insulin dose. In both trials, mean fasting SMBG decreased from weeks 1 to 8 across all subgroups, despite a temporary increase in mean fasting SMBG in the 40-50 units pre-trial insulin dose group during week 1 [mean increase (sd) +1.1 (2.0) mmol/l for DUAL V and +1.1 (2.1) mmol/l for DUAL VII], which reverted to baseline by week 4. CONCLUSIONS:Regardless of pre-trial insulin dose, IDegLira resulted in improved clinical outcomes, even in participants transferring from 40-50 units of basal insulin, despite a transient (< 4 weeks), clinically non-relevant, elevation in pre-breakfast SMBG. (Clinical Trial Registry Number NCT01952145 and NCT02420262).