The effect of baseline characteristics on clinical efficacy of liraglutide in patients treated with high-dose insulin.

Abstract

In patients requiring high-dose insulin treatment, a randomized, double-blind, placebo-controlled study showed that liraglutide improved glycaemic control and treatment satisfaction while promoting weight loss. We performed a post hoc analysis to evaluate if patients' baseline characteristics impact the efficacy of liraglutide, and which outcomes correlate with treatment satisfaction. We used regression analysis to model the change in HbA1c and weight, with treatment assignment and baseline characteristics [HbA1c, age, body mass index (BMI), total daily dose (TDD) of insulin, duration of insulin treatment, and type of insulin regimen] as independent variables. Improvement in HbA1c was best predicted by treatment with liraglutide, followed by higher baseline HbA1c, BMI and age. Changes in weight were only associated with liraglutide treatment, independent of all baseline characteristics. Improvement in HbA1c was the only significant predictor of improvement in treatment satisfaction, while weight loss, change in TDD of insulin and rate of hypoglycaemia did not influence treatment satisfaction. In patients treated with high-dose insulin, liraglutide significantly improved glycaemic control and led to weight loss regardless of patients' baseline characteristics. Improvement in HbA1c was the most important predictor of patients' treatment satisfaction.