Journal Logo

Editorial commentaries

Bashing diuretics or failure of surrogate endpoint?

Messerli, Franz H; Kuteyeva, Olga

Author Information
doi: 10.1097/HJH.0b013e32813a32a0
  • Free

Although thiazide diuretics remain the best documented drug class for reducing morbidity and mortality in hypertension, their adverse effects are legend. Last but not least, this may reflect the unsurpassed track record of this drug class, which has been around for almost half a century and is still extensively used. In the USA alone, more than 42.7 million prescriptions for hydrochlorothiazide were written in the year 2005 [1]. The thiazide diuretics appear to be particularly efficacious agents for reducing the risk of stroke [2]. This cerebroprotective property may be related to the diuretics stimulating effect on the renin–angiotensin system, in particular the angiotensin 2 receptor [3]. Clearly, stroke remains the most devastating complication of cardiovascular hypertensive disease and its prevention should be the foremost goal of antihypertensive therapy.

With regard to safety and adverse effects, diuretics have a less stellar track record. They are known as a class to stimulate the sympathetic nervous system, increase the activity of the renin–angiotensin system, cause insulin resistance and glucose intolerance, elevate uric acid, provoke hyponatrenia, hypokalemia, hypomagnesaemia, to have a procoagulative effect and on a long-term basis, and to possibly increase the risk of renal cell carcinoma and other forms of cancer [4]. Thus, several risk factors or surrogate endpoints for cardiovascular morbidity and mortality are adversely affected by the use of thiazide diuretics. In the present issue of the journal, Jiang et al. [5] appear to have found another pebble in the mosaic of diuretic bashing. By measuring cuff blood pressure and estimating central pressure with augmentation index and index of wave reflection from the radial pulse contour, they showed that indapamide was inferior to enalapril in lowering central aortic pressure. They documented a difference of 3.3 mmHg in central aortic pressure between the enalapril and indapamide arm whereas there was little difference in the brachial pressure. The same technique was used in CAFÉ, where atenolol-based therapy lowered central systolic pressure significantly less well than amlodipine-based therapy by 4.3 mmHg [5]. This led the CAFÉ investigators to draw the conclusion that ‘central aortic pulse pressure may be a determinant of clinical outcomes, and differences in central aortic pressures may be a potential mechanism to explain the different clinical outcomes between the two blood pressure treatment arms in ASCOT' [6,7].

Systolic pressure in the aorta is not only generated by left ventricular contraction, but also is affected by the pressure wave reflection returning from the periphery. Ideally, the pressure wave should return towards the heart during diastole to augment diastolic filling. However, if the reflected wave returns early during the cardiac cycle, it may enhance the outgoing pressure wave, thereby increasing central aortic pressure. A variety of factors may affect pressure wave reflection, including peripheral vascular constriction and heart rate, both of which are affected by beta-blockers. It is therefore not surprising that atenolol, by slowing the heart rate and causing a lesser, if any, fall in peripheral resistance, exerts a pseudo-antihypertensive effect (i.e. lowers central aortic pressure less than would be expected from cuff blood pressure measurements). A similar conclusion was drawn by the authors of the present study, suggesting the superiority of enalapril compared to indapamide. However, we should remember that although the inefficacy of beta-blockers has been extensively documented in uncomplicated hypertension, particularly in the elderly, this is not the case for the thiazide diuretics. By contrast, if anything, thiazide-based antihypertensive therapy has been shown to reduce morbidity and mortality better than angiotensin-converting enzyme (ACE) inhibitor-based therapy [8].

A close look at the PROGRESS and PATS studies may throw some light on this issue [9,10] In PROGRESS, monotherapy with perindopril lowered blood pressure by 5 mmHg and resulted in a 5% non-significant reduction in the stroke rate. In PATS, monotherapy with indapamide (the same diuretic used in the present study) reduced systolic blood pressure by the same 5 mmHg, yet there was a 29% reduction in strokes. Finally, when indapamide was added to perindopril in PROGRESS, systolic blood pressure fell by additional 7 mmHg and there was an impressive 43% fall in the stroke rate (Fig. 1) This would indicate that most, if not all the stroke reduction achieved in this landmark study was actually conferred by indapamide. Not surprisingly, indapamide reduced the stroke rate by an impressive 49%, even in the normotensive population of the PATS study.

Fig. 1
Fig. 1

Jiang et al. [5] quote the results of the ANBP2 study as being supportive of their findings. However, in ANBP2, only myocardial infarction was lower in the enalapril arm compared to the thiazide arm and these findings were significant for men only [11]. A similar number of strokes occurred in each group and the rate of fatal stroke was even higher in the ACE inhibitor arm than in the thiazide arm. Stroke is a more blood pressure-dependent complication of hypertension than is myocardial infarction and it is exceedingly unlikely that this small difference in myocardial infarctions in men only could have been related to a central systolic blood pressure difference of 4.1 as estimated in the present study. In ASCOT, surrogate endpoint (estimated central aortic pressure) and true endpoint (heart attacks and strokes) moved in the same direction. By contrast, when looking at the present data and PROGRESS, as well as ANBP2 and many other studies, there is a non-sequitur between surrogate endpoint and true endpoint. This should remind us that surrogate endpoints are not infallible and, in dubio, the true endpoint always wins. The authors concluded that a ‘diuretic is not an effective therapy as an ACE inhibitor in reducing aortic systolic and pulse pressure’. This may be true when central aortic pressure is estimated by the methodology employed but it either casts some doubts on the validity of this methodology, or suggests that the thiazides reduce stroke independently of aortic systolic pressure. The thiazide diuretics may have many drawbacks but, certainly, a lack of antihypertensive efficacy and a failure to reduce strokes and heart attacks is not one of them.


Conflicts of interest: F.H.M. is an ad hoc consultant/speaker for the following organizations: Abbott, GSK, Novartis, Pfizer, AstraZeneca, Bayer, Boehringer Ingelheim, BMS, Forest, Sankyo, and Sanofi. He has also received grants for research from: Novartis, GSK, Pfizer and CVRx. O.K. has no disclosures.


1 The Top 300 Prescriptions for 2005 by Number of US Prescriptions Dispensed. Available at [accessed 1 February 2007].
2 Messerli FH, Grossman E, Lever AF. Do thiazide diuretics confer specific protection against strokes? Arch Intern Med 2003; 163:2557–2560.
3 Fournier A, Messerli FH, Achard JM, Fernandez L. Cerebroprotection mediated by angiotensin II: a hypothesis supported by recent randomized clinical trials. J Am Coll Cardiol 2004; 43:1343–1347.
4 Grossman E, Messerli FH, Boyko V, Goldbourt U. Is there an association between hypertension and cancer mortality? Am J Med 2002; 112:479–486.
5 Jiang X-J, O'Rourke MF, Zhang Y-Q, He X-Y, Liu L-S. Superior effect of an angiotensin-converting enzyme inhibitor over a diuretic for reducing aortic systolic pressure. J Hypertens 2007; 25:1095–1099.
6 The CAFE Investigators, for the Anglo-Scandinavian Cardiac Outcomes (ASCOT) Investigators. Differential impact of blood pressure-lowering drugs on central aortic pressure and clinical outcomes: principal results of the Conduit Artery Function Evaluation (CAFE) study. Circulation 2006; 113:1213–1225.
7 Dahlof B, Sever PS, Poulter NR, Wedel H, Beevers DG, Caulfield M, et al., for the ASCOT investigators. Prevention of cardiovascular events with an antihypertensive regimen of amlodipine adding perindopril as required versus atenolol adding bendroflumethiazide as required, in the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering Arm (ASCOT-BPLA): a multicentre randomised controlled trial. Lancet 2005; 366:895–906.
8 The ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker versus diuretic: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA 2003; 289:178.
9 The PROGRESS Collaborative Group. Randomised trial of a perindopril-based blood-pressure-lowering regimen among 6105 individuals with previous stroke or transient ischemic attack. Lancet 2001; 358:1033–1041.
10 The PATS Collaborating Group. Post-stroke Antihypertensive Treatment Study. Chinese Med J 1995; 108:710–717.
11 Wing LM, Reid CM, Ryan P, Beilin LJ, Brown MA, Jennings GL, Johnston CI, et al. A comparison of outcomes with angiotensin converting enzyme inhibitors and diuretics for hypertension in the elderly. N Engl J Med 2003; 48:583–592.
© 2007 Lippincott Williams & Wilkins, Inc.