Re Hawkins RG Houston MC Is populationwide diur

Authors: Rahman Mahboob Nwachuku Chuke Whelton Paul Probstfield Jeffrey Davis Barry R

Publish Date: 2006/06/01

Volume: 19, Issue: 6, Pages: 656-657

PDF Link

Abstract

Mahboob Rahman Chuke Nwachuku Paul Whelton Jeffrey Probstfield Barry R Davis Re Hawkins RG Houston MC Is populationwide diuretic use directly associated with the incidence of endstage renal disease in the United States A hypothesis Am J Hypertens 2005 Jun1867449 American Journal of Hypertension Volume 19 Issue 6 June 2006 Pages 656–657 https//doiorg/101016/jamjhyper200509008We read with interest the article by Hawkins and Houston1 invoking a relationship between use of thiazide diuretics and end stage renal disease ESRD We agree with the concerns expressed in the accompanying editorials about the inherent pitfalls in inferences the investigators drew from their ecologic study23 Hawkins and Houston failed to address another major problem with their study namely “indication bias” In fact diuretics have a role in the management of chronic kidney disease especially the loop diuretics which they suggested may comprise as much as 46 of all diuretics prescribed Therefore individuals diagnosed with chronic kidney disease an antecedent to ESRD may be prescribed diuretics for clinical management hence the indication bias In addition although the investigators correctly point out that estimated glomerular filtration rate GFR was lower in the thiazide group compared to amlodipine in the Antihypertensive and LipidLowering Treatment to Prevent Heart Attack Trial ALLHAT this parameter may be difficult to interpret due to acute hemodynamic changes associated with initiation of amlodipine therapy4 They fail to report that the incidence of ESRD in ALLHAT was not different between the angiotensinconverting enzyme ACE inhibitor and the diuretic groups relative risk lisinopril versus chlorthalidone 112 95 confidence interval CI 89–104 P = 33 and the calcium channel blocker versus diuretic groups relative risk amlodipine versus chlorthalidone 111 95 CI 088–138 P = 038 This was consistent when stratified by baseline GFR and the presence of diabetes at baseline5 The composite end point of a 50 decline in GFR or ESRD was also not different between thiazide diuretic compared to amlodipine and lisinopril in patients with moderate/severe reduction in GFR5 Renal outcomes were prespecified secondary end points in ALLHAT and the data are robust with large numbers of patients and a total of 448 ESRD events total of 1049 renal end points including ≥50 decline in GFR events clearly sufficient to refute the notion that diuretic therapy is associated with increased risk of ESRD The question studied by the investigators is of clinical interest but the question is better answered by treatment comparisons in randomized controlled trials such as ALLHAT

Keywords: