Prediction of basal glycaemia dynamics during treatment with 6-month lipid-lowering therapy in patients at high risk of cardiovascular disease

• Main Authors: Olga A. Koshelskaya, Anastasiya S. Sushkova, Olga A. Zhuravleva, Irina V. Vinnizkaya, Nataliya G. Brazovskaya, Elena S. Kravchenko, Tatyana E. Suslova, Rostislav S. Karpov
• Format: Article
• Language: English
• Published: Endocrinology Research Centre 2017-12-01
• Series: Сахарный диабет
• Subjects: lipid-lowering therapy; statins; ezetimibe; basal glycemia; coronary artery disease; diabetes mellitus
• Online Access: https://dia-endojournals.ru/dia/article/viewFile/8633/7041

Background. A major meta-analysis has confirmed the ability of statins to exert both diabetogenic and hyperglycaemic effects. To date, practical recommendations for predicting glucose dynamics during lipid-lowering therapy have not been developed. Aims. Identify the combination of factors that can predict changes in basal glycaemia during 6-month lipid-lowering therapy in patients at high risk of cardiovascular disease. Methods. This study reports on 50 patients with diabetes or impaired glucose tolerance, and 18 patients with coronary artery disease without disorders of carbohydrate metabolism. Of note, 29 of the 50 diabetic or glucose intolerant patients had documented ischaemic heart disease (stable angina). Patients were randomised into three groups: Gr.1 (n=33, atorvastatin therapy), Gr.2 (n=17, atorvastatin in combination with ezetimibe) and Gr.3 (n=18, rosuvastatin therapy). After treatment for 24 weeks, we assessed lipid profile dynamics, metabolism of glucose/insulin and the HOMA-IR index. Multivariate analysis was then performed to identify factors that predicted increases in basal glycaemia. Results. All of the included patients completed 24 weeks of treatment (N=68). Lipid-lowering effect was significant in all three groups, and overall, target LDL cholesterol level was achieved in 50% of patients (n=34). In Gr.2, basal glucose level increased from 5.5(5.36.6) to 6.3(5.67.8) mmol/l (p=0.0014), which was accompanied by an increase in HOMA-IR (p=0.024). No significant change in basal glycaemia was observed in Grs.1 and 3. Moreover, an increase in the basal glycaemia was observed in 48.5% of patients in Gr.1, 70.6% in Gr.2 and 44.4% in Gr.3. Multivariate discriminant analysis across all patient groups revealed a canonical linear discriminant function that included the following factors: baseline basal glucose levels, total cholesterol levels, triglycerides and ratio of LDL/HDL cholesterol. Sensitivity and specificity of the model accounted for 75%; 51 out of the 68 cases were correctly classified when predicting the dynamics of basal glucose during lipid-lowering therapy. Conclusions. Our data demonstrate the ability to predict the dynamics of the basal glycaemia during lipid-lowering therapy. This may allow for a new way to identify patients at high risk of statin-related increases in glycaemia.