Abstract 4895: Competing Effects of Hypokalemia and Hydrochlorothiazide Treatment on Regression of Cornell Product Left Ventricular Hypertrophy in Hypertensive Patients: Implications for Development of Potassium-Sparing Diuretics

Background: Hydrochlorothiazide (HCTZ) treatment is associated with blood pressure reduction and regression of left ventricular hypertrophy (LVH). HCTZ is also associated with hypokalemia (hypoK), which increases blood pressure and is associated with a greater likelihood and severity of electrocardiographic (ECG) LVH. However, the competing effects of HCTZ use and concomitant hypoK on LVH regression have not been examined. Methods: Baseline and yearly Cornell product (CP) ECG LVH levels were examined in relation to hypoK (serum K ≤ 3.90, the lowest quartile) and HCTZ use in 7816 patients in the LIFE study with baseline and year-1 K levels. Patients were randomized to losartan vs atenolol-based treatment and additional HCTZ as needed. Results: Patients on HCTZ had lower serum K levels at year 1 (4.05 ± 0.38 vs 4.24 ± 0.38), year 2 (4.04 ± 0.38 vs 4.25 ± 0.38), year 3 (4.04 ± 0.39 vs 4.27 ± 0.39) and year 4 (4.05 ± 0.41 vs 4.26 ± 0.38) of the study (all p < 0.001). In 2-way analysis of covariance adjusting for age, sex, race, prior and randomized treatment, yearly body mass index, serum glucose and creatinine, and for baseline and change in diastolic and systolic pressure, hypoK was associated with less mean reduction of CP LVH whereas HCTZ use was associated with greater regression of CP LVH between baseline and years 1 to 4. Multivariate logistic regression analyses with the same covariates revealed that hypoK was associated with a statistically significant 15 to 19% lower likelihood of ≥ median (236 mm·ms) reduction in CP LVH while HCTZ use was associated with an 18 to 33% greater likelihood of CP LVH regression of ≥ 236 mm·ms between baseline and years 1 to 4. Conclusions: HCTZ therapy is independently associated with a greater likelihood and magnitude of LVH regression whereas concomitant hypoK is associated with a competing lower likelihood and magnitude of LVH regression during antihypertensive therapy. These findings suggest that hypoK may blunt regression of LVH during treatment. Regression of Cornell Product LVH in Relation to HCTZ Use and Hypokalemia