Abstract
AIM: Tirzepatide is a dual glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1) receptor agonist for type 2 diabetes, chronic weight management and obstructive sleep apnoea (in the US). To assess the contribution of weight loss (WL) to the observed glycaemic effect, we performed mediation analysis of three randomized, controlled, parallel, 4-arm SURPASS(S)-trials; 1, 2 and 5, for a total of 2831 participants. MATERIALS AND METHODS: WL dependent (WL-D) and WL independent (WL-IND) effects on comparator-adjusted glycated haemoglobin A1c (HbA1c) reduction at Week 40 were estimated using mediation analysis, adjusted for baseline HbA1c and study-specific stratification factors. RESULTS: The difference in mean HbA1c change from baseline at 40 weeks (total effect) between tirzepatide and the placebo group ranged from -20.0 to -14.6 mmol/mol, and that between tirzepatide and semaglutide (1 mg) ranged from -5.1 to -1.9 mmol/mol. In the placebo-controlled trials, 12%-27% of the difference in HbA1c change between tirzepatide dose arms and placebo was estimated as being mediated through WL when given as monotherapy (S1), and 25%-45% when on the background of insulin with or without metformin (S5). Compared with semaglutide, 54%-71% of the difference in HbA1c change between tirzepatide and semaglutide was estimated as being WL-D. CONCLUSION: The tirzepatide-induced HbA1c reductions from baseline, compared with placebo or semaglutide, were mediated through both WL-D and WL-IND effects. Estimated contributions of WL to the difference in glycaemic efficacy varied and were highest when tirzepatide was compared with semaglutide and less pronounced when compared with placebo. These results help to understand that other factors beyond weight reduction contribute to HbA1c reduction with tirzepatide.