Interindividual differences in the clinical effectiveness of liraglutide in Type 2 diabetes: a real‐world retrospective study conducted in Spain

• Published in: Diabetic medicine 2018-11, Vol.35 (11), p.1605-1612
• Main Authors: Gomez‐Peralta, F., Lecube, A., Fernández‐Mariño, A., Alonso Troncoso, I., Morales, C., Morales‐Pérez, F. M., Guler, I., Cadarso‐Suárez, C.
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Aged; Aged, 80 and over; Antidiabetics; Biological Variation, Population; Blood pressure; Body weight; Diabetes; Diabetes mellitus; Diabetes mellitus (non-insulin dependent); Diabetes Mellitus, Type 2 - drug therapy; Diabetes Mellitus, Type 2 - epidemiology; Electronic medical records; Female; Humans; Insulin; Lipids; Liraglutide - therapeutic use; Male; Metformin; Middle Aged; Retrospective Studies; Spain - epidemiology; Treatment Outcome; Young Adult
• ISSN: 0742-3071; 1464-5491
• DOI: 10.1111/dme.13769

Aims To study the response of clinical variables (HbA1c, body weight, lipid profile and blood pressure) over 24 months of liraglutide treatment in a real‐world clinical setting, and to describe the evolution of HbA1c and body weight reduction in response to liraglutide treatment by employing generalized additive mixed models (GAMMs). Methods We included people aged ≥ 18 years with Type 2 diabetes mellitus that initiated liraglutide treatment between November 2011 and May 2015. Demographic and clinical data were retrieved retrospectively over 24 months from electronic medical records with a median duration of observation of 7.0 (IQR 3.0–12.0) months. Results Individuals that initiated liraglutide therapy were obese (BMI 39.1 kg/m2), with inadequate HbA1c (68 mmol/mol [8.4%]), blood pressure and lipid levels. Upon liraglutide treatment, HbA1c, body weight, mean systolic and diastolic blood pressure, and lipid levels decreased gradually. GAMMs demonstrated that longer treatment with liraglutide was a predictor of improved HbA1c response, whereas higher baseline HbA1c, longer Type 2 diabetes duration and treatment with insulin were predictors of worse HbA1c response. Higher baseline weight, longer treatment with liraglutide and the interaction between metformin and time were predictors of improved weight response. Conclusions In this real‐world study, we showed the effectiveness of liraglutide in improving body weight, HbA1c, mean systolic and diastolic blood pressure, and lipid levels. GAMMs indicated that baseline HbA1c and weight, time of treatment with liraglutide, diabetes duration and the use of metformin or insulin are predictors of clinical response to liraglutide. What's new? This multicentre real‐world study is the first using innovative generalized additive mixed models (GAMMs) to predict HbA1c and weight responses upon liraglutide treatment. GAMMS demonstrated that higher baseline HbA1c, longer diabetes duration and the use of insulin predicted a worse response on HbA1c reduction, whereas longer time of treatment with liraglutide predicted a better response. Higher baseline weight, longer time of treatment with liraglutide and the use of metformin predicted a better response on weight reduction. The previous use of dipeptidyl peptidase (DPP)‐4 inhibitors did not influence the change in either HbA1c or weight. The results reported here will help in understanding the liraglutide response, i.e. how to optimize its management to improve clinical outcom