Diuretic-induced laboratory abnormalities that predict ventricular ectopy.

In order to determine which of the many diuretic-induced laboratory changes might be associated with an increased risk of ventricular ectopy (VE), we performed logistic regression analyses of patient data from a large computerized medical record system. Study variables included serum Ca2+, K+, Cl-, HCO-3, glucose, cholesterol, albumin, uric acid, and hematocrit. Controlling variables included race, use of diuretics, blood pressure, smoking history, age, and weight. (In one analysis we also included cardiac drug history and evidence of pre-existing cardiovascular disease). Separate analyses were performed for males and females. For the retrospective cohort-like design, we analyzed data for 9561 patients with complete data. For the case-control design we analyzed data from 4786 patients. Diuretic usage predicted ventricular ectopy in women, but not men. Serum uric acid and hematocrit were the only significant laboratory predictors of ventricular ectopy in each of the four analyses. Abnormalities in these variables might provide an explanation for the greater incidence in sudden (and presumably arrhythmic) deaths reported in MRFIT study patients with mild hypertension.