The LEADER trial (N=9340) compared cardiovascular (CV) outcomes with liraglutide (LIRA) vs placebo (PLAC) when added to standard of care (SoC) in T2DM patients at high CV risk. Beyond the primary CV outcome, LEADER provided an opportunity to assess long-term glycaemic effectiveness with LIRA. Aim: to describe the glycaemic durability of LIRA vs PLAC both in addition to SoC for a treatment period up to 5 years. Methods: We used Fine and Gray’s proportional sub-hazards model adjusting for all-cause death as a competing risk to analyse time to glycaemic deterioration in LIRA vs SoC patients. Glycaemic deterioration was defined as a composite of HbA1c >=8.0% and reduction <0.5% since the previous visit or an increase in insulin/OAD (start of OAD or new OAD; start of insulin, increase of insulin dose >=10U or addition of fast-acting or mixed insulin to basal). This glycaemic deterioration endpoint was designed to capture increases in HbA1c and events where patients with elevated baseline HbA1cnot respond/stop responding to therapy to reach treatment targets. We examined time to the individual components of this composite endpoint using Fine and Gray model. Results: Patients (LIRA n=4668; PLAC n=4672) were well-matched between treatment groups: mean age=64 y., diabetes duration=12.8 y, HbA1c=8.7%, 45% treated with insulin at baseline. HbA1c was reduced from baseline with both LIRA and SoC, this reduction was greater with LIRA ( DHbA1c at month 36: -1.16% [LIRA] vs -0.77% [SoC], ETD at month 36: -0.40%, 95% CI: -0.45;-0.34) despite the addition of more anti-hyperglycaemic drugs in the SoC group. There were fewer episodes of hypos with LIRA vs SoC. LIRA-treated patients had a lower risk of the glycaemic deterioration composite endpoint as well as each of the individual components vs SoC-treated patients: 69% of LIRA-treated patients were estimated to experience glycaemic deterioration over 5 years compared with 85% of patients in SoC. Discussion: In LEADER trial population, patients treated with LIRA in addition to SoC experienced greater reductions in HbA1c and were at lower risk of glycaemic deterioration when compared with patients treated with PLAC and SoC despite greater use of other anti-hyperglycaemic drugs in the PLAC group. The risk of hypos was also reduced in LIRA-treated patients.