7/29/2019 S25.full

1/5

Thiazide Diuretics in the Treatment of Hypertension:An Update

Antonio Salvetti and Lorenzo GhiadoniDepartment of Internal Medicine, University of Pisa, Pisa, Italy

Thiazide diuretics were the first tolerated efficient antihypertensive drugs that significantly reduced cardiovascular morbidity

and mortality in placebo-controlled clinical studies. Although these drugs today still are considered a fundamental therapeutic

tool for the treatment of hypertensive patients, the following considerations should be taken into account. Although there are

some indications that chlorthalidone can offer additional advantages as compared with other compounds, a recent meta-

analysis of placebo-controlled trials suggested that the beneficial effects of thiazide diuretics could be a class effect. Thiazide

diuretics must be used at appropriate and/or optimal doses to achieve the optimal antihypertensive effect with the smallest

occurrence of side effects, including alterations in glucose and lipid profiles and hypokalemia. Moreover, because thiazide

diuretics can increase the incidence of new-onset diabetes, especially when combined with blockers, caution is advised in

using these drugs above all in patients who are at high risk for developing diabetes, in whom thiazide diuretics should be used

at the lowest active dose and possibly in combination with drugs that block the renin-angiotensin system. Finally, the current

debate on whether thiazide diuretics are the first-choice drug for most patients with uncomplicated hypertension, as stated in

the Seventh Joint National Committee Report, or are included in the major classes of antihypertensive agents that are suitablefor initiation and maintenance of therapy, as reported in the European Society of HypertensionEuropean Society of

Cardiology Guidelines, derives from different interpretations of controlled clinical trial data on drug class comparison and of

cost-benefit analyses. However, considering that the benefit of antihypertensive drugs seems to be due principally to BP

lowering per se without definitive evidence of the superiority of a particular drug class and that there is no cost-benefit

analysis showing the superiority of thiazide diuretics, it is believed that these drugs should not be considered as the only

first-choice drug but included among first-choice drugs.

J Am Soc Nephrol 17: S25S29, 2006. doi: 10.1681/ASN.2005121329

S

ince the discovery of thiazide diuretics in 1957, which

for the first time offered the possibility of efficiently

reducing BP, these drugs have represented a fundamen-

tal tool for the treatment of hypertension. Moreover, placebo-

controlled clinical studies have documented clearly the benefit

of thiazide diuretics, either given alone or combined with

blockers, in reducing cardiovascular (CV) morbidity and mor-

tality (1), a benefit similar to that achieved with other antihy-

pertensive drug classes, such as angiotensin-converting en-

zyme (ACE) inhibitors and calcium antagonists (2). In this

article, we review topics concerning the role of thiazide diuret-

ics in the treatment of hypertension.

Do All Thiazide Diuretics Give the SameBenefit?

This debate first was raised by data from the Multiple Risk

Factor Intervention Trial (3), which aimed to examine special

intervention and usual care groups. The data indicated that

coronary heart disease (CHD) and total mortality trends were

less favorable at clinics where special intervention clinicians

favored hydrochlorothiazide over chlorthalidone. This led to a

protocol change with the recommendation that all patients

switch from hydrochlorothiazide to chlorthalidone at a maxi-

mum dose from 100 to 50 mg/d, which subsequently led to a

more favorable trend in CHD and total mortality (3). Although

these results suggested a beneficial effect of chlorthalidone as

compared with hydrochlorothiazide, such conclusions are not

robust, especially considering that they are based on a post hoc

subgroup analysis. Furthermore, the data are based on a group

identifier (clinic) rather than individual patient treatments, and

the beneficial effect of the change in treatment could be due to

reduction in the chlorthalidone dose (4).

The debate has been reinforced further by the data from two

large, controlled, clinical studies on drug class comparison.

Whereas in the Antihypertensive and Lipid-Lowering Treat-

ment to Prevent Heart Attack Trial (ALLHAT) (5) chlorthali-

done as compared with lisinopril-based treatment showed no

inferiority for primary outcomes (CHD mortality and nonfatal

myocardial infarction) and a superiority for stroke and heart

failure (HF) prevention, the Second Australian National Blood

Pressure trial (ANBP2) (6) indicated a marginal superiority of

enalapril versus hydrochlorothiazide-based treatment on pri-

mary end points (all CV [first and recurrent] plus all-cause

mortality) with no substantial difference for stroke and HF.

Although these conflicting results could be dependent on the

specific thiazide diuretic used, there are other likely explana-

tions, such as samples studied, trial design, and type of statis-

tical analysis (4). Finally, a recent meta-analysis of placebo-

Address correspondence to: Prof. Antonio Salvetti, Department of Internal Med-

icine, University of Pisa, Via Roma 67, 56127 Pisa, Italy. Phone/Fax: 39-050-

553407; E-mail: a.salvetti@med.unipi.it

Copyright 2006 by the American Society of Nephrology ISSN: 1046-6673/1704-0025

7/29/2019 S25.full

2/5

controlled trials that used chlorthalidone or other thiazide-like

diuretics indicated that major health outcomes did not differ

between two treatments, a finding that suggests that the bene-

ficial effects of thiazide diuretics could be a class effect (7).

Choice of Optimal DoseThe relevance of dosage choice of thiazide diuretics can be

evaluated considering both the effect of different diuretic ther-apy doses on CV mortality and morbidity and the dose-re-

sponse curve in terms of antihypertensive efficacy and meta-

bolic effects. With regard to the first issue, a meta-analysis of

clinical studies indicated that low-dose (12.5 to 25 mg/d

chlorthalidone or hydrochlorothiazide) and high-dose (50

mg/d or more of both drugs) diuretic therapy lowered BP to a

similar degree and exerted a similar benefit in reducing stroke,

congestive HF, CV, and total mortality, but only low-dose

diuretic therapy significantly reduced CHD incidence (8). The

concept of a high dose of diuretic therapy should be replaced

by that of optimal dose, which derives from studies that eval-

uated the dose-response curve of thiazide diuretics, as recentlyreviewed (9).

Data obtained with chlorthalidone so far indicate that the

plateau of the dose-response antihypertensive curve is reached

with a daily dose of 25 mg and that increasing the dosage does

not increase the BP-lowering effect but further increases the

occurrence of negative metabolic effects, principally hypokale-

mia (10,11). Data obtained with hydrochlorothiazide, the most

widely used thiazide diuretic in clinical practice, are less certain

and suggest that the plateau of the dose-response BP-lowering

curve is reached with the dose of 50 mg/d, whereas negative

metabolic effects, principally hypokalemia, occur in a dose-

dependent manner with a dose up to 100 mg/d (9,12). There-

fore it seems rational to recommend that diuretic treatment

with chlorthalidone should start with 6.25 mg in the elderly and

12.5 mg in younger patients, with a maximum dose of 25 mg/d,

whereas the starting doses of hydrochlorothiazide should be

12.5 mg in the elderly and 25 mg in younger patients, with a

maximum dose of 50 mg/d (9). Moreover the decision to in-

crease the dosage of these drugs to achieve BP control should

be weighed against the occurrence of negative metabolic ef-

fects.

Negative Metabolic Effects of Thiazide

DiureticsMetabolic effects that can have a negative impact on CV riskprofile include alterations in lipids, hypokalemia, impairment

of glucose metabolism, and, therefore, the occurrence of type 2

diabetes (13). Although changes in lipid profile, characterized

by an increase in total and LDL cholesterol and decrease in

HDL cholesterol, have been reported to be modest and tran-

sient (13), data from ALLHAT (5) indicate that the increase in

total cholesterol is detectable up to 4 yr of follow-up.

Hypokalemia, which as already stated is a dose-dependent

phenomenon, can worsen the patients prognosis because it can

increase the risk for sudden death (14) and CV events among

diuretic-treated hypertensive patients (15) and could abolish

the benefit of treatment in elderly patients with isolated systolic

hypertension (16). Furthermore, hypokalemia can impair glu-

cose metabolism by reducing insulin secretion and insulin sen-

sitivity (13).

Data from controlled clinical trials, recently reviewed (17),

indicate that the incidence of new-onset diabetes (NOD) is

lower in patients who are treated with drugs that block the

renin-angiotensin system (RAS), such as ACE inhibitors and

angiotensin receptor blockers, as compared with those who aretreated with conventional therapy, i.e., thiazide diuretics alone

and above all combined with blockers.

These findings prompted the debate on whether this differ-

ence in the incidence of NOD is due to the beneficial effects of

RAS-blocking drugs or, conversely, to the negative action of

conventional therapy on glucose metabolism. This second hy-

pothesis seems to be a rational explanation for the following

considerations. Incidence of NOD also is reduced in patients

who receive calcium antagonists, drugs that are believed to be

neutral on glucose metabolism, when compared with conven-

tional therapy (17). A longitudinal observational study showed

that given similar fasting blood glucose values, low-dose thia-zide diuretic treatment further increased the incidence of NOD

(18). In the International Verapamil SR-Trandolapril Study

(INVEST) (19), the addition of hydrochlorothiazide dose-de-

pendently increased the risk for NOD not only in patients who

received atenolol but also in those who were treated with

verapamil. Similarly, in the Valsartan Antihypertensive Long-

term Use Evaluation (VALUE) Trial (20), the greater incidence

of NOD in the amlodipine arm can be explained tentatively by

the greater occurrence of hypokalemia induced by the associa-

tion of hydrochlorothiazide. Finally, the Antihypertensive

Treatment and Lipid Profile in a North of Sweden Efficacy

Evaluation (ALPINE) Study (21) showed that in 1 yr of fol-

low-up of previously untreated hypertensive patients, treat-

ment with hydrochlorothiazide frequently associated with

atenolol impaired glucose metabolism, whereas treatment with

candesartan frequently associated with felodipine was neutral.

Considering that NOD has a bad prognostic significance be-

cause it increases the incidence of CV events in hypertensive

diuretic users even when BP is well controlled (22) and gives

the same CV risk as pre-existing diabetes in treated hyperten-

sive patients (18), caution is advised in using thiazide diuretics,

especially when combined with blockers, above all in patients

who are at high risk for developing diabetes, such as those with

genetic (familial or ethnic) predisposition, obesity, fasting glu-

cose intolerance, and/or the metabolic syndrome. Particularly

in these patients, thiazide diuretics should be used at the lowest

active dose with close monitoring of metabolic parameters and

possibly in combination with RAS-blocking drugs.

Are Thiazide Diuretics the First-Choice Drugor One of the First-Choice Drugs?

The debate on whether thiazide diuretics are the first-choice

drug, as recommended by the Seventh Joint National Commit-

tee (JNC-7) Report (23), or are among first-choice drugs, as

recommended by European Society of HypertensionEuropean

Society of Cardiology (ESH-ESC) Guidelines (24), is principally

S26 Journal of the American Society of Nephrology J Am Soc Nephrol 17: S25S29, 2006

7/29/2019 S25.full

3/5

based on different interpretations of controlled clinical trials

and secondly on drug cost considerations.

The JNC-7 Report (23) recommendation arose from the two-

fold recognition that thiazide-type diuretics have been the basis

of antihypertensive therapy in most outcome trials and that in

these trials, including the recently published ALLHAT (5,25),

diuretics have been virtually unsurpassed in preventing the CV

complications of hypertension. Moreover, diuretics can be use-ful in achieving BP control as well as enhancing the antihyper-

tensive efficacy of multidrug regimens, and they are more

affordable than other antihypertensive agents. These recom-

mendations need to be evaluated according to these lines of

reasoning.

The strengths and limitations of ALLHAT already have been

commented on (26) and can be summarized as follows. The

main strength of this study is that with regard to primary

outcomes (coronary mortality and nonfatal myocardial infarc-

tion), chlorthalidone-based treatment was equally as effective

as treatment that was based on amlodipine or lisinopril or

doxazosin, and with regard to some secondary end points suchas prevention of stroke, it was superior when compared with

doxazosin and lisinopril. It was also more effective in preven-

tion of morbidity but not mortalityfrom congestive HF

when compared with the other three treatments (5,25). How-

ever, the difference with regard to stroke could be due to a

difference in systolic BP (26). In contrast, the difference with

regard to congestive HF might be explained by poor accuracy

and/or difficulty in diagnosis; alternatively, withdrawal of pre-

vious diuretic therapy may have unmasked congestive HF

symptoms in patients with left ventricular dysfunction (26).

Therefore, ALLHAT confirmed and strengthened the clinical

relevance of thiazide diuretics in the treatment of hypertension

but did not prove the superiority of these drugs. This conclu-

sion is in agreement with an expanded analysis of the ALLHAT

data presented at the American Society of Hypertension Meet-

ing 2004 (27), which suggests the following interpretations:

The superiority of chlorthalidone versus lisinopril was detect-

able in black but not in white patients. Therefore, it would be

reasonable to state that whereas diuretics remain the pre-

ferred first-line drugs for black patients, ACE inhibitors and

diuretics could be regarded as coequal recommendations for

initiating therapy in white patients.

The primary coronary end point was not different for amlo-

dipine compared with the other two drugs, and the othermajor end points of stroke and all-cause mortality tended

slightly in its favor. Therefore, for many patients, the excel-

lent antihypertensive efficacy and tolerability of calcium an-

tagonists continue to make them a popular and appropriate

choice.

The nonsuperiority of a particular drug class, beyond BP

reduction, is also supported by the following considerations:

Placebo-controlled clinical studies have shown that the ben-

efit of antihypertensive therapy in preventing CV events

with diuretics alone or combined with a blocker (1) is

similar to that achieved with ACE inhibitors and calcium

antagonists (2). Such a finding suggests that this benefit is

due to BP lowering per se (1,2).

Meta-analyses of controlled clinical studies, including those

on drug class comparison, indicate that reduction in CV

morbidity and mortality rates depends BP reduction and that

larger reductions in BP produce larger reductions in risk for

major CV events (2,28,29).

These conclusions are supported further by data from the

VALUE Trial (30), which indicate that early (within 6 mo)

control of systolic BP (140 mmHg) can significantly reduce

the incidence of CV events in high-risk hypertensive patients.

The overall data given above do not point to a greater benefit

of a particular drug class, beyond reduction in BP values, which

is actually the more likely explanation of the benefit of antihy-

pertensive therapy. In this setting, thiazide-like diuretics can be

viewed as efficient but not as superior to other antihypertensive

drugs. Therefore, their preferential choice should be based

principally on drug cost and, in particular, whether they are

really less expensive.Certainly thiazide-like diuretics are less expensive than

newer types of antihypertensive agents and therefore are fa-

vored in terms of cost minimization (4), but cost is not the sole

consideration, and further costbenefit analyses, announced by

the ALLHAT authors although not performed so far (4), are

awaited. We believe that their analysis also should take into

account the adverse metabolic effects of thiazide diuretics, con-

sisting of an increase in cholesterol levels, blood glucose, NOD,

and hypokalemia (4,5). Although these metabolic effects did

not translate into a greater frequency of CV events in the

relatively short-term follow-up of ALLHAT (5), they could

have a major impact on costbenefit analyses because in thelong term, they can reduce the benefit of treatment and increase

the cost owing to the need for further pharmacologic therapy

that is designed to treat the metabolic abnormalities (26).

The last point to be taken into account is that choice of

antihypertensive drugs should be based on additional consid-

erations, as follows. First, efficacy in reducing BP and tolera-

bility, the latter also including metabolic effects, must be eval-

uated in individual patients. Second, as shown in Table 1, the

presence of associated clinical conditions with compelling in-

dications (23), subclinical target organ damage (24), and other

associated (but not causally related to hypertension) clinical

conditions can indicate the choice of a particular antihyperten-

sive drug class not necessarily including thiazide diuretics.

Finally, a combination of two or more drugs, not always in-

cluding diuretics, is needed in the majority of hypertensive

patients to achieve goal BP (24). Above all, these last consider-

ations could indicate that the debate on the first-choice drug

class probably has been overemphasized (24).

ConclusionThe answer to the question of whether thiazide diuretics are

the first-choice drug for treatment of hypertension is two-fold:

No, if the diuretic is the only first-choice drug; yes, if the

diuretic is one of the first-choice drugs.

J Am Soc Nephrol 17: S25S29, 2006 Thiazide Diuretics in Hypertension S27

7/29/2019 S25.full

4/5

References1. Collins R, Peto R, MacMahon S, Hebert P, Fiebach NH,

Eberlein KA, Godwin J, Qizilbash N, Taylor JO, HennekensCH: Blood pressure, stroke, and coronary heart disease.Part 2: Short-term reductions in blood pressure: Overviewof randomised drug trials in their epidemiological context.Lancet 335: 827838, 1990

2. Turnbull F: Effects of different blood-pressure-loweringregimens on major cardiovascular events: Results of pro-spectively-designed overviews of randomised trials. Lancet362: 15271535, 2003

3. Multiple Risk Factor Intervention Trial Research Group:Mortality after 10 1/2 years for hypertensive participants

in the Multiple Risk Factor Intervention Trial. Circulation82: 16161628, 1990

4. Davis BR, Furberg CD, Wright JT Jr, Cutler JA, Whelton P:ALLHAT Collaborative Research Group: ALLHAT: Settingthe record straight. Ann Intern Med 141: 39 46, 2004

5. ALLHAT Officers and Coordinators for the ALLHATCollaborative Research Group: The Antihypertensiveand Lipid-Lowering Treatment to Prevent Heart AttackTrial: Major outcomes in high-risk hypertensive patientsrandomized to angiotensin-converting enzyme inhibitoror calcium channel blocker vs diuretic: The Antihyper-tensive and Lipid-Lowering Treatment to Prevent HeartAttack Trial (ALLHAT). JAMA 288: 29812997, 2002

6. Wing LM, Reid CM, Ryan P, Beilin LJ, Brown MA, Jen-nings GL, Johnston CI, McNeil JJ, Macdonald GJ, MarleyJE, Morgan TO, West MJ: A comparison of outcomes withangiotensin-converting-enzyme inhibitors and diureticsfor hypertension in the elderly. N Engl J Med 348: 583592,2003

7. Psaty BM, Lumley T, Furberg CD: Meta-analysis of healthoutcomes of chlorthalidone-based vs nonchlorthalidone-based low-dose diuretic therapies. JAMA 292: 4344, 2004

8. Psaty BM, Smith NL, Siscovick DS, Koepsell TD, Weiss NS,Heckbert SR, Lemaitre RN, Wagner EH, Furberg CD:Health outcomes associated with antihypertensive thera-pies used as first-line agents. A systematic review andmeta-analysis. JAMA 277: 739745, 2004

9. Carter BL, Ernst ME, Cohen JD: Hydrochlorothiazide ver-

sus chlorthalidone: Evidence supporting their interchange-ability. Hypertension 43: 49, 2004

10. Materson BJ, Oster JR, Michael UF, Bolton SM, Burton ZC,Stambaugh JE, Morledge J: Dose response to chlorthali-done in patients with mild hypertension. Efficacy of alower dose. Clin Pharmacol Ther 24: 192198, 1978

11. Salvetti A, Pedrinelli R, Bartolomei G, Cagianelli MA, CinottiG, Innocenti P, Loni C, Saba G, Saba P, Papi L: Plasma renin

activity does not predict the antihypertensive efficacy ofchlorthalidone. Eur J Clin Pharmacol 33: 221226, 1987

12. Materson BJ, Cushman WC, Goldstein G, Reda DJ, FreisED, Ramirez EA, Talmers FN, White TJ, Nunn S, ChapmanRH: Treatment of hypertension in the elderly: I. Bloodpressure and clinical changes. Results of a Department ofVeterans Affairs Cooperative Study. Hypertension 15: 348 360, 1990

13. Tuck ML: Metabolic considerations in hypertension. Am JHypertens 3: 355S365S, 1990

14. Grobbee DE, Hoes AW: Non-potassium-sparing diureticsand risk of sudden cardiac death. J Hypertens 13: 15391545, 1995

15. Cohen HW, Madhavan S, Alderman MH: High and lowserum potassium associated with cardiovascular events indiuretic-treated patients. J Hypertens 19: 13151323, 2001

16. Franse LV, Pahor M, Di Bari M, Shorr RI, Wan JY, SomesGW, Applegate WB: Serum uric acid, diuretic treatmentand risk of cardiovascular events in the Systolic Hyperten-sion in the Elderly Program (SHEP). J Hypertens 18: 11491154, 2000

17. Pepine CJ, Cooper-Dehoff RM: Cardiovascular therapiesand risk for development of diabetes. J Am Coll Cardiol 44:509512, 2004

18. Verdecchia P, Reboldi G, Angeli F, Borgioni C, GattobigioR, Filippucci L, Norgiolini S, Bracco C, Porcellati C: Ad-

verse prognostic significance of new diabetes in treatedhypertensive subjects. Hypertension 43: 963969, 200419. Pepine CJ, Handberg EM, Cooper-DeHoff RM, Marks RG,

Kowey P, Messerli FH, Mancia G, Cangiano JL, Garcia-Barreto D, Keltai M, Erdine S, Bristol HA, Kolb HR, BakrisGL, Cohen JD, Parmley WW: A calcium antagonist vs anon-calcium antagonist hypertension treatment strategyfor patients with coronary artery disease. The InternationalVerapamil-Trandolapril Study (INVEST): A randomizedcontrolled trial. JAMA 290: 28052816, 2003

20. Julius S, Kjeldsen SE, Weber M, Brunner HR, Ekman S,Hansson L, Hua T, Laragh J, McInnes GT, Mitchell L, PlatF, Schork A, Smith B, Zanchetti A: Outcomes in hyperten-

sive patients at high cardiovascular risk treated with reg-imens based on valsartan or amlodipine: The VALUE ran-domised trial. Lancet 363: 20222031, 2004

21. Lindholm LH, Persson M, Alaupovic P, Carlberg B, Svens-son A, Samuelsson O: Metabolic outcome during 1 year innewly detected hypertensives: Results of the Antihyper-tensive Treatment and Lipid Profile in a North of SwedenEfficacy Evaluation (ALPINE study). J Hypertens 21: 15631574, 2003

22. Alderman MH, Cohen H, Madhavan S: Diabetes and car-diovascular events in hypertensive patients. Hypertension33: 11301134, 1999

23. Chobanian AV, Bakris GL, Black HR, Cushman WC, GreenLA, Izzo JL Jr, Jones DW, Materson BJ, Oparil S, Wright JT

Jr, Roccella EJ: The Seventh Report of the Joint National

Table 1. Indications for antihypertensive drug classeschoice

Compelling indicationsheart failure, after myocardial infarctionsa, high

coronary risk, diabetes, chronic kidney diseasea,recurrent stroke prevention

Subclinical target organ damageleft ventricular hypertrophya, carotid atherosclerosisa,microalbuminuriaa

Associated, not causally related, clinical conditionsheadachea, Raynauds phenomenona, irritable bowela,

gouta, prostatic hypertrophya, palpitationsa,depression and panica, tremora, renal stones,osteoporosis

aDrug classes other than thiazide diuretics.

S28 Journal of the American Society of Nephrology J Am Soc Nephrol 17: S25S29, 2006

7/29/2019 S25.full

5/5

Committee on Prevention, Detection, Evaluation, andTreatment of High Blood Pressure: The JNC 7 report.JAMA 289: 25602572, 2003

24. European Society of Hypertension-European Society ofCardiology Guidelines Committee: 2003 European Societyof Hypertension-European Society of Cardiology guide-lines for the management of arterial hypertension. J Hyper-tens 21: 10111153, 2003

25. The ALLHAT: Officers and coordinators for the ALLHATCollaborative Research Group: Major cardiovascularevents in hypertensive patients randomized to doxazosinvs chlorthalidone: The antihypertensive and lipid-lower-ing treatment to prevent heart attack trial (ALLHAT).ALLHAT Collaborative Research Group. JAMA 283: 19671975, 2000

26. Salvetti A, Ghiadoni L: Guidelines for antihypertensive

treatment: An update after the ALLHAT study. J Am SocNephrol 15[Suppl]: S51S54, 2004

27. Weber MA: New results and analyses expand and modifykey interpretations of the ALLHAT Trial. Rev CardiovascMed 5: 164169, 2004

28. Staessen JA, Wang JG, Thijs L: Cardiovascular preventionand blood pressure reduction: A quantitative overviewupdated until 1 March 2003. J Hypertens 21: 10551076, 2003

29. Lawes CM, Bennett DA, Feigin VL, Rodgers A: Bloodpressure and stroke: An overview of published reviews.Stroke 35: 776785, 2004

30. Weber MA, Julius S, Kjeldsen SE, Brunner HR, Ekman S, Hans-son L, Hua T, Laragh JH, McInnes GT, Mitchell L, Plat F, SchorkMA, Smith B, Zanchetti A: Blood pressure dependent and inde-pendent effects of antihypertensive treatment on clinical eventsin the VALUE Trial. Lancet 363: 20492051, 2004

J Am Soc Nephrol 17: S25S29, 2006 Thiazide Diuretics in Hypertension S29