Objective. —
To determine whether the high mortality among diabetic patients receiving treatment for hypertension can be explained by associated risk factors or must be attributed to a deleterious effect of antihypertensive treatment.
Design. —
Cohort analytic study with a median follow-up of 4.5 years.
Setting. —
Outpatients with diabetes and severe retinopathy who were enrolled in a multicenter, randomized clinical trial of laser treatment to prevent blindness had ophthalmologic examinations every 4 months and annual medical examinations that included measurement of blood pressure and recording of anti-hypertensive treatment. Only 5.5% of the patients were unavailable for follow-up. When a patient died, the circumstances surrounding the death were reviewed and classified by a mortality review committee.
Participants. —
There were 759 participants in the study; they were white, were aged 35 to 69 years, and had normal serum creatinine levels at the baseline examination.
Measurements and Main Results. —
Patients were classified into five groups according to information recorded at the baseline and first annual follow-up examinations: normotensive (diastolic blood pressure <90 mm Hg), untreated hypertensive, hypertensive treated by diuretics alone, hypertensive treated by other agents alone, and hypertensive treated by both agents. Cardiovascular mortality was higher in patients treated for hypertension than in patients with untreated hypertension. The excess was primarily found in patients treated with diuretics alone, although that group had the lowest blood pressure with treatment. After adjusting for differences in risk factors, cardiovascular mortality was 3.8 times higher in patients treated with diuretics alone than in patients with untreated hypertension (P<.001).
Conclusions.—
In individuals with diabetes, intervention with diuretics to reduce hypertension is associated with excess mortality. Until there is a clinical trial showing a beneficial effect of diuretic treatment in diabetic patients, there is urgent need to reconsider its continued usage in this population.(Arch Intern Med. 1991;151:1350-1356)