0
We're unable to sign you in at this time. Please try again in a few minutes.
Retry
We were able to sign you in, but your subscription(s) could not be found. Please try again in a few minutes.
Retry
There may be a problem with your account. Please contact the AMA Service Center to resolve this issue.
Contact the AMA Service Center:
Telephone: 1 (800) 262-2350 or 1 (312) 670-7827  *   Email: subscriptions@jamanetwork.com
Error Message ......
Original Contribution |

Association Between Cardiovascular Outcomes and Antihypertensive Drug Treatment in Older Women FREE

Sylvia Wassertheil-Smoller, PhD; Bruce Psaty, MD, PhD; Philip Greenland, MD; Albert Oberman, MD, MPH; Theodore Kotchen, MD; Charles Mouton, MD; Henry Black, MD; Aaron Aragaki, MS; Maurizio Trevisan, MD, MS
[+] Author Affiliations

Author Affiliations: Department of Epidemiology & Population Health, Albert Einstein College of Medicine, Bronx, NY (Dr Wassertheil-Smoller); University of Washington, Seattle (Dr Psaty); Northwestern University, Chicago, Ill (Dr Greenland); University of Alabama at Birmingham (Dr Oberman); Medical College of Wisconsin, Milwaukee (Dr Kotchen); Howard University College of Medicine, Washington, DC (Dr Mouton); Rush University Medical Center, Chicago, Ill (Dr Black); Fred Hutchinson Cancer Research Center, Seattle, Wash (Mr Aragaki); and University of Buffalo, Buffalo, NY (Dr Trevisan).

More Author Information
JAMA. 2004;292(23):2849-2859. doi:10.1001/jama.292.23.2849.
Text Size: A A A
Published online

Context Diuretic-based therapy is at least as effective as newer classes of agents for hypertension. However, many patients with hypertension require treatment with more than 1 drug class to achieve blood pressure control. The relative benefits or risks of 2-drug–class combinations are not well known.

Objective To prospectively evaluate if there are differences in cardiovascular mortality among postmenopausal women with hypertension but no history of cardiovascular disease (CVD) treated with different classes of antihypertensive agents, singly or in combination.

Design, Setting, and Participants Women with hypertension enrolled in the Women’s Health Initiative Observational Study, a longitudinal multicenter cohort study of 93 676 women aged 50 to 79 years at baseline (1994-1998), assessed for a mean of 5.9 years.

Main Outcome Measures Relationship between baseline use of ACE inhibitors, β-blockers, calcium channel blockers, or diuretics, or a combination of these, and incidence of coronary heart disease, stroke, and CVD mortality.

Results Among 30 219 women with hypertension but no history of CVD, 11 294 (57%) were receiving monotherapy with an ACE inhibitor, β-blocker, calcium channel blocker, or diuretic, and 4493 (23%) were treated at baseline with a combination of diuretic plus either ACE inhibitor, β-blocker, or calcium channel blocker or ACE inhibitor plus calcium channel blocker. Monotherapy with calcium channel blockers vs diuretics was associated with greater risk of CVD death (hazard ratio, 1.55; 95% confidence interval, 1.02-2.35), controlling for multiple covariates. Women treated with a diuretic plus a calcium channel blocker had an 85% greater risk of CVD death vs those treated with a diuretic plus a β-blocker, after adjustment for age, race, smoking, high cholesterol levels requiring medication, body mass index, physical activity, use of hormone therapy, and diabetes. After exclusion of women with diabetes the hazard ratio was 2.16 (95% confidence interval, 1.16-4.03). Analyses adjusting for propensity to be receiving a particular treatment did not change the results. For morbid events of coronary heart disease or stroke, diuretics plus ACE inhibitors or calcium channel blockers did not differ from diuretics plus β-blockers.

Conclusions Among women with hypertension but no history of CVD, a 2-drug–class regimen of calcium channel blockers plus diuretics was associated with a higher risk of CVD mortality vs β-blockers plus diuretics. Risks were similar for ACE inhibitors plus diuretics and β-blockers plus diuretics. Monotherapy with diuretics was equal or superior to other monotherapy in preventing CVD complications of high blood pressure.

Figures in this Article

Diuretics or β-blockers are considered to be the first step as monotherapy for high blood pressure (BP), based on clinical trial results.1 Newer classes of drugs, such as angiotensin-converting enzyme (ACE) inhibitors and calcium channel blockers, are also effective in lowering BP and may be more effective at lowering BP in particular subgroups of patients. These newer classes of drugs have increasingly been used as first-step monotherapy.2 In 1995, of the 10 most frequently prescribed antihypertensive drugs, 4 were calcium antagonists, 3 were ACE inhibitors, and 1 was a β-blocker.3

The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) provided clinical trial data on monotherapy and indicated that diuretics were equal or superior to other antihypertensive agents as first-line therapy.4,5 However, most patients with hypertension require more than 1 drug class to control hypertension.6 It is unresolved as to which of the other drug classes added to diuretics has better effects on cardiovascular events. While this question would best be settled by a clinical trial, there are no such data available yet, and results of observational studies currently must provide the information to guide choice of combined therapies using 2 drug classes for hypertension.

This report examines the relation of different classes of antihypertensive drugs to incidence of coronary heart disease, stroke, and cardiovascular disease (CVD) mortality after a mean of 5.9 years of follow-up, focusing on women with no prior history of CVD in the Women’s Health Initiative Observational Study (WHI-OS),7 a large observational cohort study of postmenopausal women. Among women enrolled in the WHI-OS, about 38.8% had hypertension8 and approximately two thirds of those with hypertension reported receiving drug treatment for hypertension. Thus, the WHI-OS affords an excellent opportunity to examine the effects of different drug treatments for hypertension on cardiovascular morbidity and mortality.

The WHI-OS is a multicenter longitudinal cohort study to investigate risk factors for heart disease, osteoporosis, and breast and colorectal cancer and other diseases of older women. Details of the WHI design are reported elsewhere.7 Postmenopausal women aged 50 to 79 years at baseline (1994-1998) were enrolled at 40 Clinical Centers in the United States, mostly through mass mailings to age-eligible women. All women provided written informed consent. The study was approved by the institutional review boards of the participating clinical centers, the coordinating center at the Fred Hutchinson Cancer Center, and the National Institutes of Health. Exclusions were participation in other randomized trials, predicted survival of less than 3 years, alcoholism, drug dependency, diagnosed mental illness, or dementia. The WHI-OS cohort of 93 676 participants represents diverse socioeconomic groups and is multiethnic, with 0.5% American Indian/Alaskan Native, 8.2% African American, 2.9% Asian/Pacific Islander, 3.9% Hispanic, 83.3% white, and 1.4% unknown.

At the baseline visit women completed questionnaires on family and personal history of disease, race/ethnicity (using investigator-defined options), lifestyle characteristics, and psychosocial variables; had measurements taken of height, weight, pulse, and BP; and had blood drawn and stored for future studies. Race/ethnicity was assessed to ensure adequate representation of minorities and because race/ethnicity may be a risk factor for outcomes studied. Blood pressure was measured by certified staff using standardized procedures and instruments, in the right arm, with a conventional mercury sphygmomanometer and an appropriately sized cuff, after the participant was seated and resting for 5 minutes and before the blood draw or a minimum of 30 minutes after the blood draw. The average of 2 seated readings, obtained at least 30 seconds apart, was used for analyses. Women with hypertension were defined as those who had elevated BP (systolic BP ≥140 mm Hg, diastolic BP ≥90 mm Hg, or both) at the clinic and/or a self-report of being diagnosed by a physician as having hypertension and receiving medications for hypertension.

Type of antihypertensive medication was determined from medications women were asked to bring in their original bottles to the baseline visit. Pill bottle data were transcribed into the computer database and matched to the corresponding item in a pharmacy database (Master Drug Data Base [MDDB; Medi-Span, Indianapolis, Ind]) that included brand and generic drug names, national drug codes, and a therapeutic class code provided by the American Hospital Formulary Service. The computer system was updated approximately every quarter with a new pharmacy database to ensure completeness of the list of available products. Use of antihypertensive medication was classified as monotherapy with ACE inhibitor, β-blocker, calcium channel blocker, or diuretic, or as 2-drug–class combined therapy with diuretic plus ACE inhibitor, β-blocker, or calcium channel blocker; ACE inhibitor plus calcium channel blocker; or other therapy that included any other drug or combination of drugs. Among those women who reported taking medications for hypertension (n = 19 889), treatment was unknown in 4.6% because they did not bring in their pills.

End Point Ascertainment

End points were ascertained from annually mailed Medical History Update Questionnaires or direct report in the intervals between questionnaires, with telephone follow-up for hospitalization and other additional information when necessary. Potential outcomes were intensively investigated by obtaining medical records information and death certificates. Physician adjudicators at clinical sites confirmed or determined the cause of event based on the detailed information. Participant fatalities were identified through communication with proxy respondents and also through National Death Index searches. Coronary disease outcome was defined as incident angina, fatal plus nonfatal myocardial infarction, angioplasty, or coronary artery bypass graft surgery; stroke as fatal plus nonfatal stroke; and CVD death as that due to definite or possible coronary heart disease, cerebrovascular disease, pulmonary embolism, or CVD of unknown type.

Statistical Analyses

Among the 93 676 women, 35 920 (38.3%) were classified as having hypertension. Because some of these women could have been receiving treatment for indications other than hypertension (ie, certain cardiovascular conditions other than hypertension for which the class of drugs is recommended), we excluded 4979 women with history of CVD and 722 who had missing CVD information, yielding 30 219 women with no history of CVD.9 History of CVD was defined as self-reported history of angina (n = 3347), myocardial infarction (n = 1410), stroke (n = 924), angioplasty (n = 743), coronary artery bypass graft surgery (n = 583), or congestive heart failure (n = 537). These categories of history of CVD are not mutually exclusive. We did not have baseline glucose values; thus, diabetes was defined as self-report of receiving pills or injections for the treatment of diabetes. According to this definition, diabetes was present in 1746 (5.8%) of women with no history of CVD and 819 (16.5%) of those with CVD. Of the 30 219 women with hypertension but no history of CVD, 19 889 (65.8%) reported they were receiving antihypertensive pharmacological treatment; of these, 920 (4.6%) did not bring in their pills and thus had no medications recorded in the MDDB, resulting in 18 969 women with data recorded for antihypertensive therapy (Figure 1).

Figure 1. Study Flow
Graphic Jump Location

ACE indicates angiotensin-converting enzyme; MDDB, Master Drug Data Base.

Follow-up was for a mean of 5.9 years (interquartile range, 4.9-6.9 years). We had minimal loss to follow-up: 94% of our participants were known to be either deceased (4.2%) or active (89.7% of participants through August 2003 had contact within the previous 15 months). Data from those lost to follow-up were censored after their last contact date. Statistical comparisons of baseline characteristics among different drug groups were determined by χ2 test of association between the subgroups of a particular baseline characteristic and the groups receiving monotherapy or combined therapy with 2 classes of drugs. Comparison of continuous variables was performed by analysis of variance across subgroups. For each end point, prespecified comparisons of hazard ratios (HRs) from Cox proportional models were performed to determine relative risk of events for 3 different monotherapy drug classes compared with diuretic monotherapy and for 3 combined therapies compared with a diuretic plus β-blocker. Multivariate analyses adjusted for age, race/ethnicity, smoking, high cholesterol levels requiring medication (by self-report), body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared), physical activity (categorized as episodes per week of moderate or strenuous activity [defined as a metabolic equivalent tasks score of ≥4.0 as indicated by Ainsworth et al10] of ≥20 minutes’ duration), use of hormone therapy, and diabetes (defined as a physician diagnosis plus self-reported use of insulin or oral medication). Separate adjusted models were also run for women without diabetes.

We did not include time-dependent variables because BP and medications data were only collected at baseline and in year 3. Interaction between effect of treatment and race (white vs black) was not statistically significant. Proportional hazards assumptions were verified by testing the interaction of survival time and treatment and by inspecting parallelism of estimated hazard functions. Nominal P values are shown for overall tests of equality among all drug treatment effects. Confidence intervals (CIs) are included to show specific differences.

Potential confounding by indication was addressed by obtaining a propensity score from a multinomial logistic regression to predict baseline medication usage from baseline risk factors (age, race/ethnicity, smoking, high cholesterol level requiring medication, prior use of hormone therapy, diabetes, diastolic and systolic BP, and all pairwise interactions, using stepwise regression). These logistic regressions used pairwise comparisons of the drug-class combinations,11,12 with diuretic as the reference for monotherapy and diuretic plus β-blocker as the reference for combined therapy. For example, to determine propensity for ACE inhibitor plus diuretic vs β-blocker plus diuretic, the propensity score for the ith participant is the conditional probability Pr{Z = 1| Xi = xi}, where Z = 1 if ACE inhibitor plus diuretic, 0 if β-blocker plus diuretic, and xi is the participant’s baseline covariates (age, race/ethnicity, BMI, etc). Thus, these propensity scores are a composite of the individual covariates for each person. We then ran Cox regression models using quintiles of propensity scores of the form: event = f(drug class + quintiles of propensity scores) to determine the HR of the event for a given monotherapy as compared with diuretic, controlling for the propensity of that monotherapy drug in comparison with propensity of diuretic.

We conducted separate Cox regressions for each 2-drug–class combination compared with diuretic plus β-blocker, controlling for propensity quintile. These analyses were supplemental to the more standard Cox regressions controlling for the individual covariates. Evidence that there was not overfitting due to colinearity is that the fully adjusted models were in agreement with the propensity score models. Power calculations for comparing 2 monotherapies for cardiovascular death with α = .05 indicate we had 90% power to detect a true HR of 1.5, but only 40% power to detect an HR of 1.25. All analyses were performed using SAS version 9.0 (SAS Institute Inc, Cary, NC).

Baseline Characteristics

Comparisons of baseline characteristics of women with hypertension by history of CVD indicate that those with and without history of CVD differed significantly in distributions of demographic characteristics, risk factors, and use of different antihypertensive drug therapy classes (Table 1 and Table 2). Only about one third (35.9%) of women with hypertension but no history of CVD were treated and controlled, and one third had BMIs of 30 or greater.

Table Graphic Jump LocationTable 1. Baseline Demographic Characteristics of Women With Hypertension, by History of CVD*
Table Graphic Jump LocationTable 2. Baseline Lifestyle Variables, Risk Factors, and Clinical Characteristics of Women With Hypertension, by History of CVD*

Among the 18 969 women with hypertension but no prior history of CVD who were receiving pharmacological treatment recorded in the MDDB, 15 787 (83.2%) were treated with ACE inhibitors, β-blockers, calcium channel blockers, or diuretics (n = 11 294) or with 1 of the four 2-drug–class combinations of these medications (n = 4493) (Figure 1). Because the effects of antihypertensive drugs on cardiovascular events may be confounded by cardiovascular indication for prescribing these drugs,9 we assessed baseline characteristics of women with no history of CVD using the different drug classes (Table 3). A greater proportion of those receiving calcium channel blockers or diuretics were black (16.2% and 17.1%, respectively) compared with those receiving ACE inhibitors (7.4%) or β-blockers (4.9%). Baseline systolic BP was lowest in the diuretic group (135.5 mm Hg) and highest in the calcium channel blocker group (139.2 mm Hg). The drug class with longest duration of use at baseline was diuretics (7.5 years), while duration of use of ACE inhibitors (4.0 years) was similar to that of calcium channel blockers (3.6 years).

Table Graphic Jump LocationTable 3. Baseline Characteristics of Women With Hypertension but No History of CVD, by Drug Class*

The group receiving no medications is composed of a mixture of subgroups (ie, those who may not have had elevated BPs except at our clinic screening; those who had elevated BP in the physician’s office but were not prescribed medication; and those who were prescribed medications but did not bring them to their baseline visit to be entered in the medications inventory or report that they were currently receiving treatment). For this reason, we did not compare all medication combinations with the group receiving no medications, since the results would not have been interpretable. We did compare those receiving no medications with those receiving diuretics as monotherapy and found the HRs and 95% CIs for coronary disease, stroke, and CVD death to be, respectively, 0.87 (0.72-1.05), 1.05 (0.77-1.41), and 0.92 (0.63-1.33). The baseline systolic BP of the no medications group was higher (149 mm Hg) than for those receiving drug treatment, but we do not know the BPs of the women who were receiving therapy before they began the therapy.

Cardiovascular Events by Treatment

During the mean of 5.9 years of follow-up among the 30 219 women with hypertension but with no prior history of CVD, there were 1509 coronary disease events (5.0%), 607 strokes (2.0%), and 406 CVD deaths (1.3%). Among the 4979 participants with a prior history of CVD, there were 945 coronary disease events (19.0%), 234 strokes (4.7%), and 245 CVD deaths (4.9%).

Among women with hypertension but no history of CVD who were receiving monotherapy, the highest annualized rates were for those receiving calcium channel blockers (Table 4). Among women receiving combined therapy, the highest annualized rates were for those receiving diuretic plus calcium channel blocker (stroke and CVD death) and for those receiving ACE inhibitor plus calcium channel blocker (coronary disease), although there were fewer women receiving this combination (n = 477) than the other drug combinations.

Table Graphic Jump LocationTable 4. Incident Events and Annualized Rates for Women With Hypertension but No History of CVD (n=30219)
CVD Events With Monotherapy

Table 5 compares HRs among women receiving monotherapies. For coronary disease there was a slight elevation of risk for monotherapy with calcium channel blockers (HR, 1.25; 95% CI, 1.00-1.55) after adjustment for age and race/ethnicity, which became nonsignificant (HR, 1.19, 95% CI, 0.94-1.51) after adjustment for other risk factors and exclusion of women with diabetes. There were no significant differences in coronary disease events between ACE inhibitors and β-blockers compared with diuretics. For stroke, there was an elevation of risk for monotherapy with calcium channel blockers after full adjustment and exclusion of women with diabetes, but the CI overlapped 1.0 (HR, 1.32; 95% CI, 0.92-1.89).

Table Graphic Jump LocationTable 5. Risk of Cardiovascular Events Among Women With No History of CVD, by Antihypertensive Monotherapy

For total CVD mortality, the overall test of significance suggests a modest difference among therapies (P = .04 for the fully adjusted model). There was a significant 57% higher risk (HR, 1.57; 95% CI, 1.04-2.35) for women receiving monotherapy with calcium channel blockers compared with diuretics. Further adjustment for smoking, high cholesterol levels requiring medication, BMI, physical activity, use of hormone therapy, and diabetes did not alter this elevated risk (HR, 1.55; 95% CI, 1.02-2.35), nor did exclusion of women with diabetes (HR, 1.55; 95% CI, 0.99-2.43). We considered separately the women with hypertension whose systolic BP was less than 140 mm Hg at our baseline. This is the group of participants who had a BP-lowering response sufficient to be considered controlled with the particular therapy they had been receiving prior to our baseline. In this group of women with controlled hypertension but no diabetes and no history of CVD, the HR of CVD death for monotherapy with calcium channel blockers increased to 2.57 (95% CI, 1.20-5.50). Neither ACE inhibitors nor β-blockers compared with diuretics showed a significantly higher risk of total CVD mortality. Of the 3096 participants without a history of CVD who were taking calcium channel blocker monotherapy, 1488 were taking dihydropyridines and 1601 were taking nondihydropyridines (7 were taking both classes of calcium channel blockers). There were no significant differences in HRs between women taking dihydropyridines and those taking nondihydropyridines (Table 6).

Table Graphic Jump LocationTable 6. Risk of Cardiovascular Events Among Women With No History of CVD, by Antihypertensive Monotherapy With Dihydropyridine and Nondihydropyridine Calcium Channel Blockers

The primary results presented do not adjust for duration of treatment. The MDDB only records the duration receiving the baseline medication but not the participant’s total duration receiving antihypertensive agents. Although baseline BP does not provide any evidence about pretreatment BP, when a decision as to which drug therapy to use would have been made, models adjusting for the baseline BP and duration receiving baseline medication (data not shown) did not change our results.

CVD Events With Combined Therapy

Table 7 compares three 2-drug–class combinations of antihypertensive therapy (diuretic plus ACE inhibitor, diuretic plus calcium channel blocker, and ACE inhibitor plus calcium channel blocker) with diuretic plus β-blocker. There were no significant differences between any of these combinations and diuretic plus β-blocker for coronary heart disease or for stroke. For CVD death, however, an overall test suggested evidence that risk for CVD death conferred by combination therapies did differ (P = .05 and P = .002 for the fully adjusted and nondiabetic models, respectively). The combination of diuretic plus calcium channel blocker was associated with an 80% increase, which increased to 85% after adjustment for covariates. Exclusion of women with diabetes further increased the HR to 2.16 (95% CI, 1.16-4.03; P = .002). In analyses restricted to the women without diabetes but with uncomplicated hypertension whose systolic BP while receiving drug therapy was less than 140 mm Hg at baseline, the HR for the combination of diuretic plus calcium channel blocker was 2.53 (95% CI, 1.01-6.36). A propensity score (in quintiles) was included in the propensity analysis, yielding an HR of 1.83 (95% CI, 1.00-3.36) and indicating an elevated risk of CVD death with diuretic plus calcium channel blocker, after accounting for the propensity to be receiving this drug combination based on the covariates. None of the other combined-therapy comparisons were significant.

Table Graphic Jump LocationTable 7. Risk of Cardiovascular Events Among Women With No History of CVD, by Antihypertensive for 2-Drug Combinations

Figure 2 shows the estimated cumulative cardiovascular mortality rates for antihypertensive medications adjusted for age and race/ethnicity plus smoking, high cholesterol level requiring medication, BMI, physical activity, use of hormone therapy, and diabetes for the monotherapy drugs and for the 2 drug–class combinations. The cumulative cardiovascular mortality rate for calcium channel blockers begins to diverge from the survivor functions for the other drugs after year 3 from baseline, continues to diverge, and is lower than for ACE inhibitors, β-blockers, or diuretics. There is no difference between the cumulative mortality rates for ACE inhibitors or diuretics, with both having a slightly higher mortality than β-blockers. The cardiovascular mortality rates for the combination drugs begin to diverge at about year 2 after baseline, with the highest mortality for diuretic plus calcium channel blocker and the lowest mortality for diuretic plus ACE inhibitor.

Figure 2. Cumulative Probability of Cardiovascular Disease Mortality for Monotherapy and Combined Therapy
Graphic Jump Location

ACE indicates angiotensin-converting enzyme.

The WHI-OS, with 93 676 participants, represents the largest multiethnic cohort of postmenopausal women assessed prospectively with rigorously ascertained end points. At baseline in 1994-1998 we identified 35 920 women with hypertension, of whom 30 219 had no history of CVD; among the latter, 18 969 were known to be receiving pharmacological treatment for hypertension using ACE inhibitors, β-blockers, calcium channel blockers, or diuretics or a 2-drug–class combination of these.

Compared with monotherapy with diuretics, women receiving monotherapy with calcium channel blockers were more likely to experience CVD morbidity and mortality after a mean of 5.9 years of follow-up, even after controlling for multiple potential confounders. Among 2-drug–class combinations used in women with no prior history of CVD, diuretics plus calcium channel blockers were associated with twice the risk of CVD death compared with diuretics plus β-blockers (HR, 2.16; 95% CI, 1.16-4.03).

The JNC (Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure) 52 and JNC 61 guidelines recommend use of diuretics or β-blockers as monotherapy in patients with uncomplicated hypertension. The JNC 7 guideline6 recommends thiazide-type diuretics as preferred initial therapy for most patients with stage 1 hypertension (systolic BP of 140-159 mm Hg or diastolic BP of 90-99 mm Hg) or ACE inhibitors, angiotensin receptor blockers, β-blockers, or calcium channel blockers, and combined therapy—usually including thiazide diuretics plus another drug class—for patients with stage 2 hypertension (systolic BP ≥160 mm Hg or diastolic BP ≥100 mm Hg). In our cohort, of those 11 294 women with no history of CVD who were treated with monotherapy with 1 of the 4 drug classes, use of ACE inhibitors, calcium channel blockers, or diuretics were similar (28.1%, 27.4%, and 26.1%, respectively) and use of β-blockers was lower (18.4%); among the 4493 women treated with the 2-drug–class combinations considered here, use of diuretics plus ACE inhibitors, β-blockers, or calcium channel blockers was similar (around 30%) and use of ACE inhibitors plus calcium channel blockers was markedly less (approximately 11%).

Other studies have reported that prescription of calcium channel blockers has increased while use of diuretics has declined,13 in spite of the higher costs of calcium channel blockers and lack of evidence of superior tolerability or health outcomes.14,15 An important question is whether it is the BP-lowering effect of antihypertensive therapy that is the critical element in preventing CVD sequelae or whether particular drug classes offer benefits beyond their effects in lowering BP. In our study, as previously reported,8 taking a diuretic was associated with better control of BP than was taking an ACE inhibitor, β-blocker, or calcium channel blocker after adjustment for multiple covariates that might be related to choice of therapy, and this may have accounted for the better outcome. However, our findings were not altered when we controlled for baseline systolic BP as well as duration of receiving antihypertensive medication at baseline and when we restricted our analyses to those whose systolic BP was controlled with the therapy they were taking.

Staessen et al,16 in an analysis of 9 randomized treatment trials, found that calcium channel blockers showed greater benefit with regard to stroke than did β-blockers or diuretics but less benefit for myocardial infarction. The authors conclude, however, that the results of the outcome trials for antihypertensive drugs can be explained by the BP differences between randomization groups. Different conclusions were reached by Pahor et al,17 who analyzed 9 clinical trials and found that, while the reduction in BP varied across trials, there were no significant differences between calcium channel blockers and other drugs in lowering either systolic or diastolic BP. However, while BP reduction was similar, intermediate and long-acting calcium channel blockers, compared with other antihypertensive drugs, were associated with higher risk of acute myocardial infarction, congestive heart failure, and combined major CVD events. There were no significant differences for stroke or all-cause mortality. The authors conclude that long-acting calcium channel blockers cannot be recommended as first-line therapy. A review and comparison of outcome studies and meta-analyses by Opie18 concluded that although calcium channel blockers may show a small benefit with regard to stroke and a small increase in myocardial infarction, overall they are neither better nor worse than conventional therapy; ACE inhibitors may be protective, although there are not sufficient data to support this.

A placebo-controlled trial of the ACE inhibitor ramipiril was stopped early because of significant and marked reductions of cardiovascular events in high-risk patients.19 Additionally, the Perindopril Protection Against Recurrent Stroke Study (PROGRESS) Collaborative Group20 recently reported that among individuals with and without hypertension but with a history of stroke or transient ischemic attack, treatment with the ACE inhibitor perinodopril plus a diuretic offered significantly greater reductions in recurrent stroke than treatment with perinodopril alone. In the group that received a diuretic plus perinodopril, there was a 42% reduction compared with placebo; in the perinodopril-alone group there was only a 5% reduction. However, only 30% of the participants were women.20

The Valsartan Antihypertensive Longterm Use Evaluation (VALUE) randomized trial21 aimed to compare the effects on cardiac morbidity and mortality of a calcium channel blocker (amlodipine) with an angiotensin receptor blocker (valsartan) for the same degree of BP control among 15 245 patients aged 50 years or older, assessed for a mean of 4.2 years. A low-dose diuretic (hydrochlorothiazide, 12.5 mg) was added if necessary. Blood pressure was reduced more in the amlodipine-based group than in the valsartan-based group, though there was no significant difference in the composite outcome. In our study, we examined the effects of calcium channel blockers in those who had their BPs controlled at baseline with each of the drugs we considered (systolic BP <140 mm Hg) and were essentially low-risk participants, since they had no history of CVD and no diabetes, and we found that calcium channel blockers alone or in combination with a diuretic conferred a higher risk of CVD death than did diuretics alone or diuretics plus β-blockers.

One limitation of our study is that, while prospective, it is an observational rather than a clinical trial; thus, we cannot be certain that we were able to control fully for confounding by indication. Although we studied women with hypertension who had no history of CVD or diabetes and there were no substantial differences in baseline characteristics among those receiving the different drug classes, there still may have been unmeasured factors that would lead physicians to select one class of antihypertensive drug as first-line therapy rather than another. We have no way of knowing, from our data, whether physicians prescribed a diuretic first and then changed drug class for patients who did not respond to the diuretic. However, since all the drug classes were used about equally as monotherapy, it is reasonable to suppose that this was not the case and that if a patient did not have a sufficient response to the first-line drug either the dosage of that drug would be increased or a second class of drug would be added.

Furthermore, our data pertain to postmenopausal women aged 50 to 79 years and may not be generalizable to younger women or to men. The study population is not a population-based sample, similar to most observational studies of CVD; however, it is diverse with regard to race/ethnicity, income, education, and region of the country, with 40 clinical centers participating throughout the United States. Chance findings cannot be ruled out, and in subgroups with small sample sizes the CIs are wide.

Strengths of our study include the very large sample size, meticulous ascertainment of end points, and rigorous collection of baseline and medication data, enabling us to control for many potential confounders.

The ALLHAT study,22 a randomized double-blind trial among 40 000 high-risk patients with hypertension, provides clinical trials evidence of the risk/benefit of different antihypertensive drug classes compared with diuretic therapy. Subgroup analyses of women aged 50 to 79 years with no history of CVD would need to be performed to determine if effects found in the trial apply to postmenopausal women similar to those in our cohort. Early findings from that study resulted in discontinuation of the study arm treated with doxazasin23,24 because of adverse effects.

The findings of this observational study concerning monotherapy are consistent with the comparative clinical trials published thus far and support the current guidelines set forth in the report of the JNC 7, which recommend low-dose diuretic therapy for most patients with hypertension.6 Hypertension is more prevalent and greatly heightens the risk of CVD among patients with diabetes, and numerous clinical trials have demonstrated reduction in CVD among those aggressively treated.6 However, the need to attain the recommended targets of 140/90 mm Hg or lower and of 130/80 mm Hg or lower for those with diabetes most often requires 2 to 3 drugs. In our study, among the postmenopausal women with hypertension but with no history of CVD who were receiving monotherapy or combination drug therapy, 57.9% had systolic BPs lower than 140 mm Hg and 23.9% had BPs lower than 130/80 mm Hg. Among the women with hypertension and diabetes, only 21.1% had BPs lower than 130/80 mm Hg. Clearly these women are not treated aggressively enough to control their BP to recommended levels.

Our findings with regard to combined therapy with 2 drug classes indicate that calcium channel blockers combined with diuretics are associated with a higher risk of CVD death than are ACE inhibitors plus diuretics or β-blockers plus diuretics.

Corresponding Author: Sylvia Wassertheil-Smoller, PhD, Department of Epidemiology & Population Health, Albert Einstein College of Medicine, Morris Park Ave, Room 1312, Belfer Building, Bronx, NY 10461 (smoller@aecom.yu.edu).

Financial Disclosures: Dr Psaty has received honoraria from Celera Diagnostics. Dr Black has served as a consultant for Astra-Zeneca, Biovail, Bristol-Myers Squibb, Merck Sharp & Dohme, Novartis, and Pfizer and has served on the speakers bureau for Astra-Zeneca, Merck Sharp & Dohme, Novartis, and Pfizer.

Author Contributions: Dr Wassertheil-Smoller had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analyses.

Study concept and design: Wassertheil-Smoller, Psaty, Greenland, Mouton, Trevisan.

Acquisition of data: Wassertheil-Smoller, Greenland, Oberman, Kotchen, Black, Trevisan.

Analysis and interpretation of data: Wassertheil-Smoller, Psaty, Greenland, Oberman, Kotchen, Black, Aragaki, Trevisan.

Drafting of the manuscript: Wassertheil-Smoller, Oberman, Black, Aragaki, Trevisan.

Critical revision of the manuscript for important intellectual content: Wassertheil-Smoller, Psaty, Greenland, Oberman, Kotchen, Mouton, Black, Aragaki, Trevisan.

Statistical analysis: Psaty, Aragaki.

Obtained funding: Wassertheil-Smoller, Oberman, Black, Trevisan.

Administrative, technical, or material support: Trevisan.

Study supervision: Wassertheil-Smoller, Oberman, Mouton, Trevisan.

Women’s Health Initiative Program Office:National Heart, Lung, and Blood Institute, Bethesda, Md: Barbara Alving, Jacques Rossouw, Linda Pottern.

Clinical Coordinating Center:Fred Hutchinson Cancer Research Center, Seattle, Wash: Ross Prentice, Garnet Anderson, Andrea LaCroix, Ruth E. Patterson, Anne McTiernan; Wake Forest University School of Medicine, Winston-Salem, NC: Sally Shumaker, Pentti Rautaharju; Medical Research Labs, Highland Heights, Ky: Evan Stein; University of California, San Francisco: Steven Cummings; University of Minnesota, Minneapolis: John Himes; University of Washington, Seattle: Bruce Psaty.

Clinical Centers and Investigators:Albert Einstein College of Medicine, Bronx, NY: Sylvia Wassertheil-Smoller; Baylor College of Medicine, Houston, Tex: Jennifer Hays; Brigham and Women's Hospital, Harvard Medical School, Boston, Mass: JoAnn Manson; Brown University, Providence, RI: Annlouise R. Assaf; Emory University, Atlanta, Ga: Lawrence Phillips; Fred Hutchinson Cancer Research Center, Seattle, Wash: Shirley A. A. Beresford; George Washington University Medical Center, Washington, DC: Judith Hsia; Harbor-UCLA Research and Education Institute, Torrance, Calif: Rowan Chlebowski; Kaiser Permanente Center for Health Research, Portland, Ore: Evelyn Whitlock; Kaiser Permanente Division of Research, Oakland, Calif: Bette Caan; Medical College of Wisconsin, Milwaukee: Jane Morley Kotchen; MedStar Research Institute/Howard University, Washington, DC: Barbara V. Howard; Northwestern University, Chicago/Evanston, Ill: Linda Van Horn; Rush-Presbyterian St Luke's Medical Center, Chicago, Ill: Henry Black; Stanford Center for Research in Disease Prevention, Stanford University, Stanford, Calif: Marcia L. Stefanick; State University of New York, Stony Brook: Dorothy Lane; The Ohio State University, Columbus: Rebecca Jackson; University of Alabama, Birmingham: Cora Beth Lewis; University of Arizona, Tucson/Phoenix: Tamsen Bassford; State University of New York, Buffalo: Jean Wactawski-Wende; University of California at Davis, Sacramento: John Robbins; University of California at Irvine, Orange: Allan Hubbell; University of California, Los Angeles: Howard Judd; University of California at San Diego, LaJolla/Chula Vista: Robert D. Langer; University of Cincinnati, Cincinnati, Ohio: Margery Gass; University of Florida, Gainesville/Jacksonville: Marian Limacher; University of Hawaii, Honolulu: David Curb; University of Iowa, Iowa City/Davenport: Robert Wallace; University of Massachusetts/Fallon Clinic, Worcester: Judith Ockene; University of Medicine and Dentistry of New Jersey, Newark: Norman Lasser; University of Miami, Miami, Fla: Mary Jo O’Sullivan; University of Minnesota, Minneapolis: Karen L. Margolis; University of Nevada, Reno: Robert Brunner; University of North Carolina, Chapel Hill: Gerardo Heiss; University of Pittsburgh, Pittsburgh, Pa: Lewis Kuller; University of Tennessee, Memphis: Karen C. Johnson; University of Texas Health Science Center, San Antonio: Robert Brzyski; University of Wisconsin, Madison: Gloria Sarto; Wake Forest University School of Medicine, Winston-Salem, NC: Denise Bonds; Wayne State University School of Medicine/Hutzel Hospital, Detroit, Mich: Susan Hendrix.

Funding/Support: The Women’s Health Initiative (WHI) program is funded by the National Heart, Lung, and Blood Institute (NHLBI).

Role of the Sponsor: The NHLBI participated in the design, conduct, and oversight of the WHI and reviewed this report. One NHLBI representative serves as a member of the WHI Steering Committee.

Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure.  The Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.  Arch Intern Med. 1997;157:2413-2446
PubMed   |  Link to Article
Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure.  The Fifth Report of the Joint National Committee on Detection Evaluation and Treatment of High Blood Pressure (JNC V).  Arch Intern Med. 1993;153:154-183
PubMed   |  Link to Article
Siegel D, Lopez J. Trends in antihypertensive drug use in the United States.  JAMA. 1997;278:1745-1748
PubMed   |  Link to Article
ALLHAT Officers and Coordinators for the for the ALLHAT Collaborative Research Group.  Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic.  JAMA. 2002;288:2981-2997
PubMed   |  Link to Article
Turnbull F. Effects of different blood-pressure-lowering regimens on major cardiovascular events.  Lancet. 2003;362:1527-1535
PubMed   |  Link to Article
Chobanian AV, Bakris GL, Black HR.  et al.  The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.  JAMA. 2003;289:2560-2572
PubMed   |  Link to Article
WHI Study Group.  Design of the Women’s Health Initiative clinical trial and observational study.  Control Clin Trials. 1988;19:61-109WHI Study
Wassertheil-Smoller S, Anderson G, Psaty BM.  et al.  Hypertension and its treatment in postmenopausal women.  Hypertension. 2000;36:780-789
PubMed   |  Link to Article
Psaty BM, Koepsell TD, Lin D.  et al.  Assessment and control for confounding by indication in observational studies.  J Am Geriatr Soc. 1999;47:749-754
PubMed
Ainsworth BE, Haskell WL, Leon AS.  et al.  Compendium of physical activities: classification of energy costs of human activities.  Med Sci Sports Exerc. 1993;25:71-80
PubMed   |  Link to Article
Rubin DB. Estimating causal effects from large data sets using propensity scores.  Ann Intern Med. 1997;127:757-763
PubMed   |  Link to Article
 Rubin DB. Estimation from nonrandomized treatment comparisons using subclassification on propensity scores. Presented at: International Conference on Nonrandomized Comparative Clinical Studies; April 10-11, 1997; Heidelberg, Germany
Psaty BM, Koepsell TD, Yanez ND.  et al.  Temporal patterns of antihypertensive medication use among older adults, 1989 through 1992.  JAMA. 1995;273:1436-1438
PubMed   |  Link to Article
Hypertension Detection and Follow-up Program Study Group.  Five-year findings of the Hypertension Detection and Follow-up Program: reduction in mortality of persons with high blood pressure including mild hypertension.  JAMA. 1979;242:2562-2571
PubMed   |  Link to Article
SHEP Cooperative Research Group.  Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension.  JAMA. 1991;265:3255-3264
PubMed   |  Link to Article
Staessen JA, Wang JG, Thijs L. Cardiovascular protection and blood pressure reduction: a meta-analysis.  Lancet. 2001;358:1305-1315
PubMed   |  Link to Article
Pahor M, Psaty BM, Alderman MH.  et al.  Health outcomes associated with calcium antagonists compared with other first-line antihypertensive therapies.  Lancet. 2000;356:1949-1954
PubMed   |  Link to Article
Opie LH. Calcium channel blockers in hypertension.  Am J Hypertens. 2001;14:1074-1081
PubMed   |  Link to Article
Gueyffier F, Boutities F, Boissel J.  et al. the INDANA Investigators.  Effect of anti-hypertensive drug treatment on cardiovascular outcomes in men and women.  Ann Intern Med. 1997;126:761-767
PubMed   |  Link to Article
PROGRESS Collaborative Group.  Randomised trial of a perindopril-based blood-pressure-lowering regimen among 6,105 individuals with previous stroke or transient ischaemic attack.  Lancet. 2001;358:1033-1041
PubMed   |  Link to Article
Julius S, Kjeldsen SE, Weber M.  et al.  Outcomes in hypertensive patients at high cardiovascular risk treated with regimens based on valsartan or amlodipine.  Lancet. 2004;363:2022-2031
PubMed   |  Link to Article
Davis BR, Cutler JA, Gordon DJ.  et al. ALLHAT Research Group.  Rationale and design for the Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT).  Am J Hypertens. 1996;9:342-360
PubMed   |  Link to Article
Ball SG. Discontinuation of doxazosin arm of ALLHAT.  Lancet. 2000;355:1558
PubMed   |  Link to Article
Messerli FH. Implications of discontinuation of doxazosin arm of ALLHAT.  Lancet. 2000;355:863-864
PubMed   |  Link to Article

Figures

Figure 1. Study Flow
Graphic Jump Location

ACE indicates angiotensin-converting enzyme; MDDB, Master Drug Data Base.

Figure 2. Cumulative Probability of Cardiovascular Disease Mortality for Monotherapy and Combined Therapy
Graphic Jump Location

ACE indicates angiotensin-converting enzyme.

Tables

Table Graphic Jump LocationTable 1. Baseline Demographic Characteristics of Women With Hypertension, by History of CVD*
Table Graphic Jump LocationTable 2. Baseline Lifestyle Variables, Risk Factors, and Clinical Characteristics of Women With Hypertension, by History of CVD*
Table Graphic Jump LocationTable 3. Baseline Characteristics of Women With Hypertension but No History of CVD, by Drug Class*
Table Graphic Jump LocationTable 4. Incident Events and Annualized Rates for Women With Hypertension but No History of CVD (n=30219)
Table Graphic Jump LocationTable 5. Risk of Cardiovascular Events Among Women With No History of CVD, by Antihypertensive Monotherapy
Table Graphic Jump LocationTable 6. Risk of Cardiovascular Events Among Women With No History of CVD, by Antihypertensive Monotherapy With Dihydropyridine and Nondihydropyridine Calcium Channel Blockers
Table Graphic Jump LocationTable 7. Risk of Cardiovascular Events Among Women With No History of CVD, by Antihypertensive for 2-Drug Combinations

References

Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure.  The Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.  Arch Intern Med. 1997;157:2413-2446
PubMed   |  Link to Article
Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure.  The Fifth Report of the Joint National Committee on Detection Evaluation and Treatment of High Blood Pressure (JNC V).  Arch Intern Med. 1993;153:154-183
PubMed   |  Link to Article
Siegel D, Lopez J. Trends in antihypertensive drug use in the United States.  JAMA. 1997;278:1745-1748
PubMed   |  Link to Article
ALLHAT Officers and Coordinators for the for the ALLHAT Collaborative Research Group.  Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic.  JAMA. 2002;288:2981-2997
PubMed   |  Link to Article
Turnbull F. Effects of different blood-pressure-lowering regimens on major cardiovascular events.  Lancet. 2003;362:1527-1535
PubMed   |  Link to Article
Chobanian AV, Bakris GL, Black HR.  et al.  The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.  JAMA. 2003;289:2560-2572
PubMed   |  Link to Article
WHI Study Group.  Design of the Women’s Health Initiative clinical trial and observational study.  Control Clin Trials. 1988;19:61-109WHI Study
Wassertheil-Smoller S, Anderson G, Psaty BM.  et al.  Hypertension and its treatment in postmenopausal women.  Hypertension. 2000;36:780-789
PubMed   |  Link to Article
Psaty BM, Koepsell TD, Lin D.  et al.  Assessment and control for confounding by indication in observational studies.  J Am Geriatr Soc. 1999;47:749-754
PubMed
Ainsworth BE, Haskell WL, Leon AS.  et al.  Compendium of physical activities: classification of energy costs of human activities.  Med Sci Sports Exerc. 1993;25:71-80
PubMed   |  Link to Article
Rubin DB. Estimating causal effects from large data sets using propensity scores.  Ann Intern Med. 1997;127:757-763
PubMed   |  Link to Article
 Rubin DB. Estimation from nonrandomized treatment comparisons using subclassification on propensity scores. Presented at: International Conference on Nonrandomized Comparative Clinical Studies; April 10-11, 1997; Heidelberg, Germany
Psaty BM, Koepsell TD, Yanez ND.  et al.  Temporal patterns of antihypertensive medication use among older adults, 1989 through 1992.  JAMA. 1995;273:1436-1438
PubMed   |  Link to Article
Hypertension Detection and Follow-up Program Study Group.  Five-year findings of the Hypertension Detection and Follow-up Program: reduction in mortality of persons with high blood pressure including mild hypertension.  JAMA. 1979;242:2562-2571
PubMed   |  Link to Article
SHEP Cooperative Research Group.  Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension.  JAMA. 1991;265:3255-3264
PubMed   |  Link to Article
Staessen JA, Wang JG, Thijs L. Cardiovascular protection and blood pressure reduction: a meta-analysis.  Lancet. 2001;358:1305-1315
PubMed   |  Link to Article
Pahor M, Psaty BM, Alderman MH.  et al.  Health outcomes associated with calcium antagonists compared with other first-line antihypertensive therapies.  Lancet. 2000;356:1949-1954
PubMed   |  Link to Article
Opie LH. Calcium channel blockers in hypertension.  Am J Hypertens. 2001;14:1074-1081
PubMed   |  Link to Article
Gueyffier F, Boutities F, Boissel J.  et al. the INDANA Investigators.  Effect of anti-hypertensive drug treatment on cardiovascular outcomes in men and women.  Ann Intern Med. 1997;126:761-767
PubMed   |  Link to Article
PROGRESS Collaborative Group.  Randomised trial of a perindopril-based blood-pressure-lowering regimen among 6,105 individuals with previous stroke or transient ischaemic attack.  Lancet. 2001;358:1033-1041
PubMed   |  Link to Article
Julius S, Kjeldsen SE, Weber M.  et al.  Outcomes in hypertensive patients at high cardiovascular risk treated with regimens based on valsartan or amlodipine.  Lancet. 2004;363:2022-2031
PubMed   |  Link to Article
Davis BR, Cutler JA, Gordon DJ.  et al. ALLHAT Research Group.  Rationale and design for the Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT).  Am J Hypertens. 1996;9:342-360
PubMed   |  Link to Article
Ball SG. Discontinuation of doxazosin arm of ALLHAT.  Lancet. 2000;355:1558
PubMed   |  Link to Article
Messerli FH. Implications of discontinuation of doxazosin arm of ALLHAT.  Lancet. 2000;355:863-864
PubMed   |  Link to Article
CME
Also Meets CME requirements for:
Browse CME for all U.S. States
Accreditation Information
The American Medical Association is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The AMA designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 CreditTM per course. Physicians should claim only the credit commensurate with the extent of their participation in the activity. Physicians who complete the CME course and score at least 80% correct on the quiz are eligible for AMA PRA Category 1 CreditTM.
Note: You must get at least of the answers correct to pass this quiz.
Please click the checkbox indicating that you have read the full article in order to submit your answers.
Your answers have been saved for later.
You have not filled in all the answers to complete this quiz
The following questions were not answered:
Sorry, you have unsuccessfully completed this CME quiz with a score of
The following questions were not answered correctly:
Commitment to Change (optional):
Indicate what change(s) you will implement in your practice, if any, based on this CME course.
Your quiz results:
The filled radio buttons indicate your responses. The preferred responses are highlighted
For CME Course: A Proposed Model for Initial Assessment and Management of Acute Heart Failure Syndromes
Indicate what changes(s) you will implement in your practice, if any, based on this CME course.

Multimedia

Some tools below are only available to our subscribers or users with an online account.

1,757 Views
53 Citations

Related Content

Customize your page view by dragging & repositioning the boxes below.

Articles Related By Topic
Related Collections
PubMed Articles
Jobs
×