Figure 1.

Among the patients who did not have diabetes at baseline (8098 in the verapamil SR group and 8078 in the atenolol group) addition of trandolapril in verapamil SR strategy was associated with more protection from the development of new onset diabetes than that in the atenolol/hydrochlorothiazide strategy arm. When trandolapril 2 mg was added to verapamil 180 mg and trandolapril 4 mg to verapamil 240 mg the hazard ratio for development of new onset diabetes was 0.56 (0.98–1.64, confidence interval [CI] 95%), and 0.58 (0.44–0.78, CI 95%) respectively.