In the provocative review by the Nateglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research (NAVIGATOR) investigators, a single daily dose of valsartan reduced the risk for diabetes by 14% compared with placebo, a relative risk reduction lower than that see even with lifestyle modification in some trials. However, approximately 40% of patients were taking β blockers and 40% diuretics, agents that have been consistently shown to have adverse metabolic effects. In a meta-analysis of randomized trials of β blockers in patients with hypertension, treatment of 1,000 patients for 4.4 years resulted in 14 excess cases of diabetes. Similarly, thiazide diuretics increased the risk for diabetes by 32% compared with placebo or non-β-blocker agents. Conceivably, angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, by mitigating hypokalemia associated with thiazide diuretics and by other effects, may abolish some of these adverse glycemic effects if given concomitantly with β blockers or diuretics. However, in the Study of Trandolapril/Verapamil SR and Insulin Resistance (STAR), in which patients with the metabolic syndrome were randomized to either verapamil and trandolapril or to losartan and hydrochlorothiazide, at the end of 1 year, losartan and hydrochlorothiazide increased plasma glucose significantly more than verapamil and trandolapril after all oral glucose tolerance testing. Similar results were documented in the International Verapamil-Trandolapril (INVEST) study and the Effects in Diabetes Mellitus: Carvedilol-Metoprolol Comparison in Hypertensives (GEMINI) study.

In the NAVIGATOR study, a significantly higher proportion of patients were taking β blockers or diuretics in the placebo arm than in the valsartan arm. One is left to wonder whether the paltry effect of valsartan was due to the drug itself or to the diabetogenic potential of these concomitant agents used in a substantial number of patients. We urge the investigators to present an analysis (albeit post hoc) evaluating the interaction effect for the end point of diabetes after stratifying the cohort into those who were taking either β blockers or diuretics compared with neither at follow-up.