Determinants of, and outcomes associated with antihypertensive‐associated incident diabetes and metabolic syndrome in hypertensive patients in the ASCOT‐trial

Abstract

Controversy exists whether the adverse effects on glucose metabolism and the increased incidence of new‐onset diabetes (NOD) associated with beta‐blockers and/or diuretics translate into adverse cardiovascular (CV) events, and thereby seriously detract from the beneficial effects of the these agents. Furthermore, little is known about the determinants of NOD among hypertensive patients. Controversy also surrounds the clinical utility of the metabolic syndrome (MetS) and its role in predicting NOD and CV outcomes. In this thesis I present an evaluation of these issues using the database of the Anglo‐Scandinavian Cardiac Outcomes Trial. Increase in the baseline fasting plasma glucose (FPG), body mass index, serum triglyceride and systolic blood pressure significantly increased the risk for NOD. In contrast, randomisation to amlodipine ± perindopril treatment (in comparison with atenolol ± thiazide), high HDL ‐cholesterol, alcohol use and age over 55 years were significant protective factors. On further analysis, cumulative‐mean glucose (CMG) and FPG were found to be independent and significant risk factors for CV outcomes and death. Allocation to atenolol‐based treatment (vs. amlodipine‐based treatment) was associated with a significantly greater increase in CMG levels, and a progressive worsening of glycaemic status (normoglycaemia, impaired glycaemia and NOD) after 1 year of follow‐up. Worsening glycaemic status was found to have a significant and linear relationship with increased risk of CV outcomes and death. The MetS, after adjustment for its constituent components, was found to be an independent predictor for stroke and death, but not for coronary outcomes. Furthermore, the MetS was associated with a 22% increased risk of NOD, after adjusting for its individual components, and was found to be a significantly better predictor of NOD than impaired fasting glucose. In summary, these data suggest that among hypertensive patients, antihypertensive agents are important determinants for the development of NOD, and the antihypertensive-associated in‐trial glycaemic worsening is associated with increased risk of CV outcomes and death. I have also shown that, in routine clinical practice, the MetS has an important role as an easy‐to‐use predictor for the risk of NOD, CV outcomes and death; and that the increased risk of CV outcomes and death associated with the MetS is independent of the influence of its constituent components.