We'd like to point out an interesting pair of articles – an editorial and an original paper – that were published in the November 13 issue of the Archives of Internal Medicine. Dr. Joshua Barzilay and his colleagues at the ALLHAT Collaborative Research Group reported the finding that thiazide diuretics for hypertension are associated with higher increases in fasting plasma glucose and incidence of diabetes than angiotensin-converting enzyme (ACE) inhibitors or calcium channel blockers. However, they found that this higher incidence of diabetes was not associated with increased mortality, cardiovascular disease, or renal outcomes over five years of follow-up, leading them to write that "there is no conclusive or consistent evidence that this diuretic-associated increase in DM risk increases the risk of clinical events." Editorialist Dr. Robert Phillips agrees with the authors of the study that thiazides should be "the cornerstone on which blood pressure-lowering treatment should be built" because new-onset diabetes is less deadly than high blood pressure. He writes that practitioners and policy makers alike should take this as evidence that diuretics should be first-line therapy for hypertension.
Frankly, we're a little surprised by these articles. In their post-hoc subgroup analysis from ALLHAT, which included a cohort of 18,000 patients not diagnosed with diabetes at baseline, Dr. Barzilay and colleagues found that patients who were randomized to receive treatment with chlorthalidone, amlodipine, and lisinopril saw rises in fasting glucose of 8.5 mg/dL, 5.5 mg/dL, and 3.5 mg/dL respectively. There was no significant association of fasting glucose level change at two years with the subsequent risk of coronary heart disease, stroke, cardiovascular disease, total mortality, or end-stage renal disease. Neither did they see any significant association between incidence of diabetes at two years and clinical outcomes… except for with coronary heart disease, for which the risk ratio was 1.64 for patients who developed diabetes. While the risk ratio of coronary heart disease for the chlorthalidone group overall was not statistically significant at 1.46, we think this is due to the relatively short follow-up in this trial. We know incident diabetes, if not elevated fasting glucose alone, is associated with higher long-term incidence of cardiovascular disease and other clinical outcomes, so we found it a little sobering to read the authors' and editorialists' dismissal of the possible risk associated with glucose-raising medications. We consider this an apt reminder that outside of the specialist community the risks associated with diabetes and high blood glucose are still often underestimated.