Primary-care observational database study of the efficacy of GLP-1 receptor agonists and insulin in the UK

AIMS: We investigated use and efficacy of glucagon-like peptide-1 (GLP-1) receptor agonists in UK practice. METHODS: People starting a GLP-1 receptor agonist (exenatide, liraglutide) or insulin (glargine, detemir, NPH) after a regimen of two or three oral glucose-lowering agents were identified from The Health Information Network observational primary care database (2007–2011). Mean change in HbA(1c) and body weight were compared at 1 year between cohorts, adjusting for baseline characteristics. RESULTS: Baseline characteristics of GLP-1 receptor agonist (n = 1123) vs. insulin (n = 1842) users were HbA(1c) 78 vs. 84 mmol/mol (9.3 vs. 9.8%) and BMI 38.2 vs. 30.9 kg/m(2). The GLP-1 receptor agonist cohort was younger, had shorter diabetes duration and follow-up, less microvascular disease and heart failure, higher estimated glomerular filtration rate and more use of oral glucose-lowering agents. Lower HbA(1c) reduction on GLP-1 receptor agonist [7 vs. 13 mmol/mol (0.6 vs. 1.2%) (n = 366 vs. 892)] was not statistically significant [adjusted mean difference −1.4 (95% CI −4.1, 1.2) mmol/mol], except in the highest HbA(1c) quintile [>96 mmol/mol (>10.9%); adjusted mean difference −17.8 (−28.6, −7.0) mmol/mol]. GLP-1 receptor agonist users lost weight [−4.5 vs. +1.5 kg; adjusted mean difference 4.7 (3.7, 5.8) kg; n = 335 vs. 634]. A UK 6-month target reduction for GLP-1 receptor agonists of 11 mmol/mol (1.0%) HbA(1c) and 3% weight was reached by 24.9% of those continuing treatment. CONCLUSIONS: Those starting GLP-1 receptor agonists are heavier with better glycaemic control than those starting basal insulin. Subsequently, they have improved weight change, with similar HbA(1c) reduction unless baseline HbA(1c) is very high. The UK 6-month GLP-1 receptor agonist target is usually not reached.