Intended for healthcare professionals

Clinical Review Evidence based management of hypertension

What are the elements of good treatment for hypertension?

BMJ 2001; 322 doi: (Published 05 May 2001) Cite this as: BMJ 2001;322:1107
  1. Cynthia D Mulrow, professor of medicinea,
  2. Michael Pignone (pignone{at}, assistant professor of medicineb
  1. a Division of General Internal Medicine, University of Texas at San Antonio, San Antonio, TX 78249, USA
  2. b University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA
  1. Correspondence to: M Pignone

    This is the third in a series of five articles

    This paper summarises the drugs available for treating patients with hypertension. It is based on a book chapter for which more than 1500 trials and systematic reviews were screened. The book chapter covers dyslipidaemia, diabetes, tobacco misuse, physical inactivity, antiplatelet treatment, alcohol consumption, and vitamin supplementation in patients with hypertension.1

    Summary points

    For the initial treatment for hypertension a single agent may be appropriate or, depending on the patient's risk factors, a combination of two or more may be needed

    Treatment decreases risks of fatal and non-fatal stroke, cardiac events, and death and may improve quality of life

    Thiazide diuretics seem the best first line agents for reducing rates of stroke and death

    Angiotensin converting enzyme inhibitors, some β blockers, and some long acting calcium channel blockers are efficacious alternatives

    Short acting α antagonists should be avoided as first line agents

    Short acting calcium channel blockers should be avoided

    Benefits and harms of antihypertensive drug treatment

    General benefits

    Many large randomised placebo controlled trials consistently show that antihypertensive drug treatment decreases the risk of fatal and non-fatal stroke, cardiac events, and death in men and women with systolic or diastolic hypertension,13 without adverse effect on quality of life, which may even be improved.4 People at greater cardiovascular risk when they start treatment, such as elderly patients with other relevant risk factors, derive the most absolute benefit from drug treatment.

    Specific antihypertensive drugs as first line agents

    It is not clear whether the benefits of specific antihypertensive drugs come from their direct effects on raised blood pressure or whether they act by various other multiple indirect actions. It is difficult to assess effects of particular agents, because most large trials have used a stepped care approach in which a second or third drug is added when the first choice does not reduce blood pressure to target level. Evidence relating to first line options is provided below and in the table.

    First line antihypertensive drugs for people with hypertension

    View this table:

    Thiazide diuretics—Many large hypertension trials have compared hydrochlorothiazide, chlorthalidone, or a combination of a thiazide and a potassium-sparing agent (such as amiloride or triamterene) with placebo or no drug treatment. Both low and high dosages of thiazide decrease rates of stroke and death; but only low dosage regimens reduce coronary artery disease.5 Several different thiazides are all apparently effective, which suggests that this is a class effect.

    βBlockers—Systematic reviews and meta-analyses of several randomised trials compare β blockers as first line antihypertensive agents with placebo.610 The data are complicated: in some trials as many as 70% of participants also receive diuretics; in others large numbers of participants cross over to other regimens. Nevertheless the data suggest, but do not prove, that β blockers reduce strokes but not coronary artery disease or death. For stroke, estimates of relative risk reductions of β blockers compared with placebo range from 0 to 0.41. 6 7 9 10 β Blockers are a heterogeneous class of agents with varying degrees of cardioselectivity and variable intrinsic sympathomimetic activity; and it is doubtful whether the cardiovascular benefits of different cardioselective β blockers represent a class effect.

    Angiotensin converting enzyme inhibitors—One large randomised placebo controlled trial has shown that the angiotensin converting enzyme inhibitor ramipril reduces cardiovascular events by 22% (relative risk 0.78; 95% confidence interval 0.70 to 0.86) and death by 16% in people at high risk.11 About half the trial participants had hypertension; about half had a history of myocardial infarction; and about 40% were taking β blockers. Reductions in relative risk for cardiovascular events in hypertensive patients were equal or greater than the effect in non-hypertensive patients. A recent overview of four randomised placebo controlled trials in patients, most of whom had coronary heart disease, showed that angiotensin converting enzyme inhibitors decreased strokes by 30% (15% to 43%) and coronary heart disease by 20% (11% to 28%).12

    Calcium channel blockers—One large randomised trial compared a long acting preparation of the dihydropyridine calcium channel blocker nisoldipine with placebo in people aged 60 or more with isolated systolic hypertension.13 Rates of cardiovascular events with active treatment were reduced by 31% (14% to 45%) compared with placebo. Calcium channel blockers are a heterogeneous class of agents with various postulated mechanisms of action and they may not have class effects in hypertensive patients.

    Aims of treatment

    • Decrease the cardiovascular risk associated with hypertension

    • Decrease the risk from coexisting cardiovascular risk factors

    • Improve quality of life and encourage a healthy lifestyle

    • Choose therapeutic agents likely to do more good than harm given each patient's social circumstances, preferences, coexisting medical conditions, and risk factors

    • Minimise the adverse effects and inconveniences from what you prescribe

    α Agonists and α blockers—No large randomised trials have compared clinical outcomes of first line treatment with either α agonists, such as clonidine, or α blockers, such as terazosin or doxazosin, with placebo.

    Comparisons of different antihypertensive agents

    Angiotensin converting enzyme inhibitors, diuretics, diuretics with β blockers, and calcium channel blockers—One open long term trial in 6600 patients aged 70 to 84 reported no differences in control of blood pressure or in cardiovascular morbidity or mortality among people randomised to receive conventional treatment with diuretics, alone or with β blockers, compared with calcium channel blockers (felodipine or isradipine), and with angiotensin converting enzyme inhibitors (enalapril or lisinopril).14 A single blind long term trial in 10 985 patients aged 25-66 reported that the angiotensin converting enzyme inhibitor captopril was not more effective than conventional treatment (diuretics or β blockers) in reducing cardiovascular morbidity or mortality,15 but these results were inconclusive because a flaw in the randomisation process resulted in unbalanced groups. Two additional smaller trials compared either nisoldipine with enalapril or amlodipine with fosinopril in hypertensive patients with type 2 diabetes. 16 17 They found that angiotensin converting enzyme inhibitors and calcium channel blockers were equally effective in reducing blood pressure, but calcium channel blockers were associated with a twofold to fivefold increase in cardiovascular events compared with angiotensin converting enzyme inhibitors. In one trial comparing captopril with atenolol in hypertensive patients with type 2 diabetes, the groups did not differ significantly in blood pressures or cardiovascular events.18

    α Blockers and diuretics—One randomised trial found that the α blocker doxazosin increased the incidence of cardiovascular events, particularly congestive heart failure, compared with the diuretic chlorthalidone.19

    β Blockers and diuretics—Five trials in nearly 20 000 people directly compared thiazide diuretics with β blockers as first line treatment.8 Pooled data showed a 12% difference in cardiovascular events (relative risk 0.88 (0.78 to 1.00) thiazide v β blocker) but no significant differences in deaths (relative risk 0.97 (0.84 to 1.11)). Systematic reviews have compared trials that used diuretics as first line agents with those using β blockers.710 The summary results showed no significant differences in effect estimates between trials that tested diuretics (compared against placebo) and trials that tested β blockers (compared against placebo). However, only diuretics showed significant reductions in coronary heart disease events compared with placebo.

    Calcium channel blockers, diuretics, and β blockers—One double blind randomised trial compared the calcium channel blocker long acting nifedipine with a thiazide and anamiloride diuretic.20 The 6321 men and women in the study had hypertension and at least one additional cardiovascular risk factor. No significant differences were reported between the groups in cardiovascular events (relative risk 110 (0.91 to 1.34), nifedipine v diuretic). A second large open randomised trial compared diltiazem with diuretics, alone or with β blockers, in more than 10 000 Scandinavian men and women aged 50 to 74.21 At first a short acting form of diltiazem was used, but in the later years of the trial a long acting form was used. After four to five years cardiovascular events were similar between groups (relative risk 1.0 (0.87 to 1.15), diltiazem v diuretic or β blocker).


    It is not clear which specific antihypertensive agents are best tolerated by patients. In all but one of four long term double blind comparisons of low dose diuretics, β blockers, angiotensin converting enzyme inhibitors, and calcium channel blockers, the diuretics and β blockers tended to be more tolerable and to improve overall quality of life more than newer drugs, 19 2124 with the exception that diuretics showed more serious effects—though fewer overall—than did the long acting calcium channel blocker nifedipine.19 Serious effects were defined as “life-threatening, disabling, or leading to hospital admission.” In trials comparing thiazides with β blockers, thiazides were associated with significantly lower rates of withdrawal due to adverse effects (relative risk 0.69; 0.63, 0.76).8

    Drugs with minor adverse effects

    Adverse effects of drugs vary by drug class and between agent within classes. For example, in the trial of 6600 people aged 70-84 who were followed for five years, mentioned above, 26% of those receiving the calcium channel blockers felodipine or isradipine reported ankle oedema; 30% receiving the angiotensin converting enzyme inhibitors enalapril or lisinopril reported cough; and 9% of those receiving diuretics with or without β blockers reported cold hands and feet.14 Although such adverse effects related to specific agents are not discussed in further detail here, the book provides additional information about adverse effects, such as sexual dysfunction, attributable to specific agents.1

    Drugs with major morbid or fatal adverse effects

    Case-control, cohort, and randomised studies suggest that short and intermediate acting dihydropyridine calcium channel blockers such as nifedipine and isradipine increase cardiovascular morbidity and mortality.25 A recent overview of trials found that calcium channel blockers significantly reduced strokes by 13% (2% to 23%) compared with diuretics and β blockers but increased the incidence of coronary heart disease by 12% (0% to 26%) and possibly heart failure by 12% (−5% to 33%).12 A large trial suggests that the α agonist doxazosin increases the risk of cardiovascular events, particularly congestive heart failure, compared with chlorthalidone.19 One systematic review of nine case-control and three cohort studies reported that long term use of a diuretic about doubles the risk of renal cell carcinoma.26 Absolute risks cannot be calculated from these studies but are likely to be low, since renal cell carcinoma is uncommon.


    • Funding None

    • Competing interests MP has received funding from Pfizer Foundation for research on treating heart failure in low literacy patients.

    • The book Evidence-Based Hypertension, edited by Cynthia D Mulrow, can be purchased through the BMJ Bookshop (

      Embedded Image


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.
    19. 19.
    20. 20.
    21. 21.
    22. 22.
    23. 23.
    24. 24.
    25. 25.
    26. 26.
    View Abstract