- Top of page
- STATISTICAL METHODS
Context. Blood pressure control (<140/90 mm Hg) rates for hypertension fall far short of the US national goal of 50% or more. Achievable control rates in varied practice settings and geographic regions and factors that predict improved blood pressure control are not well identified.
Objective. To determine the success and predictors of blood pressure control in a large hypertension trial involving a multiethnic population in diverse practice settings.
Design. The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial is a randomized, double-blind, active-controlled clinical trial with a mean follow-up of 4.9 years. Participant enrollment began in February 1994 and follow-up was completed in March 2002.
Setting. A total of 623 centers in the United States, Canada, and the Caribbean.
Participants. A total of 33,357 participants (aged ≥55 years) with hypertension and at least one other coronary heart disease risk factor.
Interventions. Participants were randomly assigned to receive (double-blind) chlorthalidone, 12.5–25 mg/d (n=15,255), amlodipine 2.5–10 mg/d (n=9048), or lisinopril 10–40 mg/d (n=9054) after other medication was discontinued. Doses were increased within these ranges and additional drugs from other classes were added as needed to achieve blood pressure control (<140/90 mm Hg).
Main Outcome Measures. The outcome measures for this report are systolic and diastolic blood pressure, the proportion of participants achieving blood pressure control (<140/90 mm Hg), and the number of drugs required to achieve control in all three groups combined.
Results. Mean age was 67 years, 47% were women, 35% black, 36% diabetic; 90% were on antihypertensive drug treatment at entry. At the first of two pre-randomization visits, blood pressure was <140/90 mm Hg in only 27.4% of participants. After 5 years of follow-up, the percent controlled improved to 66%. Systolic blood pressure was <140 mm Hg in 67% of participants, diastolic blood pressure was <90 mm Hg in 92%, the mean number of drugs prescribed was 2.0±1.0, and the percent on ≥2 drugs was 63%. Blood pressure control varied by geographic regions, practice settings, and demographic and clinical characteristics of participants.
Conclusions. These data demonstrate that blood pressure may be controlled in two thirds of a multiethnic hypertensive population in diverse practice settings. Systolic blood pressure is more difficult to control than diastolic blood pressure, and at least two antihypertensive medications are required for most patients to achieve blood pressure control. It is likely that the majority of people with hypertension could achieve a blood pressure <140/90 mm Hg with the antihypertensive medications available today.
Since the 1990's, blood pressure (BP) control in the United States has been defined as <140/90 mm Hg. Based on this definition, the BP control rate was found to be 27%–29% in a representative sample of the noninstitutionalized adult US population in the National Health and Nutrition Examination Survey of 1988–1991 (NHANES III).1,2 Corresponding data from other countries have shown even lower control rates than those in the United States.3,4 A measure from NHANES III more comparable to Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) participants—i.e., BP control among those treated—ranged generally from about 30%—45% in older adults from various sex-race/ethnicity groups.1 Control among treated patients was only 33% among persons 45 years and older in Olmstead County, Minnesota5 and less than 25% in older mostly white males at a group of New England Veterans Affairs Medical Centers.6
These low BP control rates have been difficult to explain, given the number of apparently well-tolerated medications available for the management of hypertension,2 extensive public health efforts to identify and treat individuals with hypertension, and the periodic publication of consensus guidelines, such as reports of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of Hypertension2,7 and the World Health Organization-International Society of Hypertension.8 Factors related to access to care, provider practice patterns, and patient adherence have all been proposed as barriers to BP control. Several reports strongly suggest that a major factor in inadequate BP control is the physician's failure to increase doses or numbers of antihypertensive medications when seeing a patient with uncontrolled BP.6,9
The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) is a large clinical trial conducted in the United States, Canada, and the Caribbean, and sponsored by the National Heart, Lung, and Blood Institute of the US National Institutes of Health. The primary objective of ALLHAT was to compare the coronary heart disease (CHD) event rates in 42,418 participants randomly allocated to four antihypertensive agents: the diuretic chlorthalidone, the angiotensin-converting enzyme (ACE) inhibitor lisinopril, the calcium channel blocker amlodipine, and the alpha blocker doxazosin. The BP goal for all groups was <140/90 mm Hg. The doses of the blinded drugs could be increased and additional open-label agents from other classes of antihypertensive drugs could be added and doses increased as needed to achieve this level of BP control. The purpose of this report is to describe BP control in ALLHAT over time, its relationship to the numbers and doses of drugs used, and the factors that predicted BP control in this large hypertensive population from diverse practice and geographical settings. The doxazosin participants are excluded from this report, since that arm of the trial was stopped early due in part to excess cardiovascular event rates compared with chlorthalidone.10 Those results suggested that alpha blockers should not be used as initial therapy for hypertension; therefore BP control rates with a regimen beginning with an alpha blocker were not included in the current analyses. Otherwise, the randomization arms of the trial are not distinguished in this paper; BP control by treatment group will be presented in subsequent reports.
- Top of page
- STATISTICAL METHODS
The rationale and design of ALLHAT are presented in detail elsewhere.11,12 Briefly, eligible participants for ALLHAT were men and women aged 55 years or older who had systolic BP (SBP) of at least 140 mm Hg and/or diastolic BP (DBP) of at least 90 mm Hg, or took medication for hypertension, and had at least one additional risk factor for CHD. These risk factors included previous myocardial infarction or stroke (>6 months or age indeterminate), other atherosclerotic vascular disease, ischemic changes on an electrocardiogram within the past 2 years, left ventricular hypertrophy by electrocardiogram or echocardiogram, type 2 diabetes, current cigarette smoking, and low high-density lipoprotein cholesterol level (<35 mg/dL). BP eligibility criteria were based on the participant's current antihypertensive treatment status and on the average of two, seated BP measurements at each of two pre-randomization visits (visits 1 and 2). For untreated participants (and those treated for less than 2 months), BP inclusion criteria at both visits were average SBP ≥140 and/or DBP ≥90 mm Hg but BP ≤180/110 mm Hg. Treated participants were required to have BP ≤160/100 mm Hg at visit 1 and ≤180/110 mm Hg at visit 2. The higher reading at visit 2 allowed for withdrawal of antihypertensive medication between the two visits when necessary. Eligible participants were randomized into the trial at visit 2, at which time all prior antihypertensive medications were discontinued. Details of the inclusion and exclusion criteria have been described previously.11,13 From February 1994 to January 1998, 42,418 participants were recruited in 623 centers in the United States, Canada, Puerto Rico, and the US Virgin Islands.14,15 All participants signed an informed consent form, and all centers received institutional review board approval.
Participants were assigned by a computer-generated randomization schedule to one of four treatments: chlorthalidone, amlodipine, lisinopril, or doxazosin, in a ratio of 1.7:1:1:1, respectively. Randomization was stratified by center and blocked over time to maintain the ratio. More participants were assigned to chlorthalidone in accordance with Dunnett's multiple comparison procedure for comparing three treatment groups to a single control group.16 The randomization code was held only by the ALLHAT Clinical Trials Center (CTC). The treatment goal in each of the four arms was a BP <140/90 mm Hg. (As noted above, the doxazosin arm of the trial was terminated early and will not be discussed hereafter.)
All three study drugs were encapsulated and identical in appearance so the identity of each agent was masked at each of three dosage levels: 12.5, 12.5 (sham titration), and 25 mg/d for chlorthalidone; 2.5, 5, and 10 mg/d for amlodipine; and 10, 20, and 40 mg/d for lisinopril. If participants did not meet the BP goal while taking the maximum tolerated dosage of the initial blinded medication, an open-label Step 2 agent (but not a medication from the same class as the study drugs) could be added and increased in three doses until goal was reached. The three Step 2 agents (and doses) were atenolol (25–100 mg/d), reserpine (0.05–0.2 mg/d), and clonidine (0.1–0.3 mg twice per day). If goal BP was still not achieved, an open-label Step 3 agent, hydralazine (25–100 mg twice per day), could be added. After initial titration visits, participants were seen routinely every 3 months during the first year of follow-up and every 4 months thereafter for up to 8 years of follow-up.
Medication doses were increased or additional medications were added during follow-up if BP was not at goal. Non-study open-label drugs could also be added to or substituted for Step 2 or 3 open-label medications to improve tolerance or BP control. However, use of open-label medications from one of the masked classes of drugs was to be avoided unless SBP was >160 mm Hg and/or DBP was >100 mm Hg after maximum tolerated titration of drugs from each of the three steps, or a compelling indication, such as heart failure, arose for one of the masked classes of drugs. A medication dose could be decreased or a medication stopped if it was believed to be causing adverse effects.
Trained observers using standardized techniques measured Bps during the trial. SBP was defined as the first Korotkoff sound (1st phase), and the DBP as the reading at the last Korotkoff sound (5th phase). The measurements were taken only after the participant had been seated quietly for at least 5 minutes, in a comfortable posture, with feet flat on the floor. The readings were taken in the sitting position with the arm at or as close to the level of the heart as possible. The cuff was deflated at a rate of 2 mm/sec until 10 mm Hg below the level of the diastolic reading was reached. Two measurements were taken, with at least a 30-second interval between the measurements, and recorded to the nearest even number. All BP measurements were calculated as the average of these two readings.
- Top of page
- STATISTICAL METHODS
The baseline characteristics of age, BP, heart rate, body mass index (BMI), serum creatinine, glucose, total cholesterol, low-density lipoprotein (LDL) and high-density lipoprotein (HDL) cholesterol, and triglycerides of participants in ALLHAT were tabulated for subgroups by race, sex, diagnosed diabetes, age, smoking, diagnosed atherosclerotic cardiovascular disease (ASCVD), and obesity. Blood pressure, BP control (SBP/DBP <140/90 mm Hg), use of additional (≥2, ≥3, and ≥4) drugs, and percent remaining on Step 1 drugs at baseline and follow-up were calculated. BP control and use of additional drugs were also tabulated for the aforementioned subgroups. Frequency distributions of SBP and DBP control were obtained. Logistic regression analyses were performed to assess the influence of independent baseline variables on each of two outcomes at 36 months of follow-up: BP control and use of two or more drugs. The independent baseline variables include participants' age, sex, race, diabetes, smoking, ASCVD, left ventricular hypertrophy (LVH), BMI, participants' prior use of BP drugs, and practice type setting, geographic region, and research experience of the clinic the participant visited. In addition, intensification of treatment was examined in the setting of uncontrolled BP during follow-up.
- Top of page
- STATISTICAL METHODS
The baseline characteristics of participants, overall and by subgroups, are presented in Table I. Black and nonblack participants had similar SBP, but DBP was 1–2 mm Hg higher in both black men and women. Older participants (≥70 years) had a slightly higher mean SBP (146 mm Hg) and a lower mean DBP (81 mm Hg) than younger participants (55–69 years) (144/85 mm Hg). Diabetic participants had a lower mean DBP (82 mm Hg) than nondiabetics (84 mm Hg), but the mean SBP was similar in both groups (145 mm Hg). Persons with LVH by electrocardiogram (ECG) had a higher mean SBP (148 mm Hg) than those without (145 mm Hg). Participants in the West and Midwest had the lowest mean BP (144/82 mm Hg) compared with those in Canada, Puerto Rico, and the Virgin Islands (146–147/85–87 mm Hg). Mean BP at baseline ranged from 143/81 mm Hg among Veterans Affairs participants, 93% of whom were on antihypertensive treatment, to 146/84 mm Hg among private practice participants, of whom 89% were on treatment.
Table I. Baseline Characteristics of Participants Enrolled in the ALLHAT Hypertension Trial Component
|Subgroup||Samplen||Size%||Age (Years) Mean||SBP (mm Hg) Mean||DBP (mm Hg) Mean||HR (bpm) Mean||BMI(kg/m2) Mean||Cr* (mg/dL%) Mean||Glucose (mg/dL%) Mean||TC (mg/dL) Mean||LDL (mg/dL%) Mean||HDL (mg/dL%) Mean||TG (mg/dL%) Mean|
|Women|| || || || || || || || || || || || || |
|Men|| || || || || || || || || || || || || |
|Type 2 diabetic|| || || || || || || || || || || || || |
|Age, years|| || || || || || || || || || || || || |
|Current smoker|| || || || || || || || || || || || || |
|ASCVD present|| || || || || || || || || || || || || |
|Obese (BMI≥30)|| || || || || || || || || || || || || |
|LVH by ECG ¶|| || || || || || || || || || || || || |
|Geographic region†|| || || || || || || || || || || || || |
|PR & VI||4658||14.0||67.0||145.8||86.8||73.4||28.7||0.95||127.3||215.1||133.8||47.1||173.7|
|Practice setting‡|| || || || || || || || || || || || || |
|SBP=systolic blood pressure; DBP=diastolic blood pressure; mm Hg=millimeters mercury; HR=heart rate; BMI=body mass index; Cr=serum creatinine; TC=total cholesterol; LDL=low-density lipoprotein cholesterol; HDL=high-density lipoprotein cholesterol; TG=triglycerides; ASCVD=atherosclerotic cardiovascular disease; LVH=left ventricular hypertrophy; ECG=electrocardiogram; PR=Puerto Rico; VI=US Virgin Islands; HMO=health maintenance organization; CHC=community health center; Univ=university; VAMC=Veterans Affairs Medical Center *Conversion factors for SI units (per mg/dL): serum creatinine (88.4 (imol), glucose (0.0555 mmol), cholesterol (0.0259 mmol), LDL cholesterol (0.0259 mmol), HDL cholesterol (0.0259 mmol), triglycerides (0.0113 mmol); ¶ECG LVH defined as Minnesota code 3.1 with any 4.1–4.3 or 5.1–5.3 code; †US Census Regions; ‡Private and group practices (341 sites), health maintenance organizations (18 sites), community health centers (48 sites), university-based clinics (71 sites), Veterans Affairs Medical Centers (70 sites), and any setting other than the preceding categories (75 sites); 717 participants were from sites for which practice setting was not known.|
Blood pressure control of participants at baseline, 6 months, and annual visits is displayed in Table II. Mean SBP/DBP decreased from 145/83 mm Hg at baseline to 135/75 mm Hg at 5 years. BP control increased from 27.4% at baseline to 65.6% over this period. The percent remaining on Step 1 study drug decreased to 88.7% at 6 months and further decreased to 79.0% at 5 years. The mean number of drugs increased over the course of the trial from 1.3 at 6 months to 2.0 at 5 years. An analysis of BP control in a cohort of 19,803 participants who attended follow-up visits throughout 4 years of follow-up was remarkably similar to the one which included all follow-up data, indicating that loss to follow-up did not appear to be an important source of bias in estimating BP control (data not shown).
Table II. Blood Pressure Control in ALLHAT Participants at Baseline, at 6 Months of Follow-Up, and Annually Through 5 Years of Follow-Up
|Time in Trial||Number With Visit||SBP||DBP||Number of drugs||BP <140/90||SBP <140||DBP <90||On Step 1|
| || || || || || ||mm Hg||mm Hg||mm Hg||Drug|
| || ||Mean||Mean||SD||%||%||%||%|
|SBP=systolic blood pressure; DBP=diastolic blood pressure; BP=blood pressure; SD=standard deviation; mm Hg=millimeters mercury|
Blood pressure control of participants at baseline and follow-up, stratified by subgroups, is presented in Table III. From baseline through year 5, the best control was in nonblack men (70.0%) and the lowest in black women (58.8%). Control was better throughout follow-up in younger participants and those with pre-existing ASCVD. In contrast, BP control was consistently worse in participants with other comorbid conditions (diabetes, obesity, LVH, and elevated creatinine) even though a greater proportion of such participants were on multiple drugs as compared with those without these conditions.
Table II. Blood Pressure Control (%SBP/DBP <140/90 mm Hg) and Medication Use in ALLHAT Participants, by Selected Baseline Characteristics
|Subgroup||Baseline||6 Months|| || ||Year 1|| || ||Year 3|| || ||Year 5|| |
|(Baseline||%||% On||%||No. Drugs (%)||%||No. Drugs (%)||%||No. Drugs (%)||%||No. Drugs (%)|
|Age 55–69 years||28.1||89.8||50.7||26.6||4.4||56.0||35.2||8.0||63.3||50.7||17.5||66.3||63.0||27.5|
|Age ≥70 years||26.2||90.9||48.0||26.2||4.3||54.0||34.7||8.1||60.3||49.9||18.0||64.4||61.5||27.0|
|Obese (BMI ≥30)†||27.8||91.8||47.9||28.1||5.0||53.5||37.4||8.7||60.4||53.2||19.5||64.6||65.1||30.0|
|On BP meds||30.4||100.0||49.2||27.9||4.7||54.8||36.8||8.5||62.1||52.3||18.7||65.5||64.0||28.5|
|Not on BP meds||0.0||0.0||55.1||12.4||1.5||56.5||33.3||2.9||64.2||31.4||7.7||67.3||44.3||13.4|
|LVH (by ECG)‡|| || || || || || || || || || || || || || |
|Geographic Region¶|| || || || || || || || || || || || || || |
|PR & VI||18.8||92.2||53.0||13.6||1.4||61.0||18.3||2.4||71.1||29.0||5.8||70.8||44.5||14.7|
|Practice Setting§|| || || || || || || || || || || || || || |
|Comm Hlth Ctr||28.8||90.6||45.4||25.2||4.1||49.2||35.6||9.1||60.7||52.2||20.3||64.5||61.8||30.7|
|SBP=systolic blood pressure; DBP=diastolic blood pressure; NB=nonblack race; ASCVD=atherosclerotic cardiovascular disease; BMI=body mass index; BP meds=blood pressure medications; % Cntr=percent with controlled BP; No.=number; mm Hg=millimeters mercury; LVH=left ventricular hypertrophy; ECG=electrocardiogram *The sample sizes for baseline medication status and serum creatinine level are, respectively: 30,089 on blood pressure (BP) medication; 3268 not on BP medication; 2038 with a mean serum creatinine ≥1.5 mg/dL; and 31,319 with serum creatinine below 1.5 mg/dL. For the other groups, sample sizes are given in Table I. The nonblack race-sex categories include approximately 12% nonblack Hispanics. †Body mass index units are kg/m2; serum creatinine units are mg/dL (conversion factor for SI units is 88.4 (imol per mg/dL); ¶US census regions. §See legend for Table I. ‡ECG LVH defined as Minnesota code 3.1 with any 4.1–4.3 or 5.1–5.3 code.|
The frequency distribution of SBP and DBP at baseline and at 3 years is shown in Figure 1. Compared to baseline, both BP distributions at 3 years are shifted to the left. After 3 years 35.6% of participants still had SBP ≥140 mm Hg, but only 8% were ≥160 mm Hg. Among participants with SBP ≥140 mm Hg at 3 years, more than half had SBP 140–149 mm Hg and 77% had SBP 140–159 mm Hg. Also, after 3 years only 10%, 3.6%, and 1.8% had DBP ≥90, 95, and 100 mm Hg, respectively. Figure 2 depicts the percent at goal over time along with the percent of participants on 1, 2, 3, or 4+ drugs. Participants began treatment on monotherapy with the blinded study drug. The percent of participants who remained on one drug was 71% at 6 months of follow-up but had decreased to 36% by 5 years. The percent on two drugs increased from 22% at 6 months to 35% at 5 years. Among participants who achieved a BP <140/90 mm Hg after 5 years, 40% were on one agent, 35% required two agents, and 23% required three or more (data not shown).
Figure 1. Frequency distribution of systolic blood pressure (SBP) and diastolic blood pressure (DBP) for ALLHAT participants at the initial clinic visit and after 36 months of follow-up
Figure 2. Percent of ALLHAT participants who achieved their goal blood pressure (SBP/DBP <140/90 mm Hg) among those attending follow-up visits (dashed line) and proportions at those visits who were prescribed 1,2,3, or 4+ antihypertensive medications (solid lines)
The results from logistic regression analyses are presented in Table IV. An odds ratio (OR) greater than 1.0 implies that a characteristic is associated with better BP control than the reference category. All except two factors listed were statistically significant univariate predictors of goal BP at 3 years. The exceptions were current smoking and having been on BP treatment at baseline. In the multivariable analysis, younger participants, men, nonblacks, nondiabetics, leaner individuals (BMI <30), those with lower baseline SBP, those without LVH, and those not on treatment at entry were more likely to have better BP control than their reference groups. The largest relative difference in BP control was found in the black/nonblack race comparison, where blacks were 31% less likely to be controlled (OR 0.69; 95% confidence limits: 0.65–0.73) than nonblacks. Better BP control was also observed in participants enrolled at clinics with no research experience when compared with those at clinics with experience. Participants treated at nonprivate practice clinics tended to have better BP control than those at private clinics. Participants in clinics located in Canada, Puerto Rico, or the Virgin Islands were more likely to have better BP control than clinics in the Western United States, whereas participants in clinics located in the South tended to have worse BP control.
Table IV. Multiple Logistic Analysis of Relative Odds of Blood Pressure Control and Relative Odds of Being on Two or More Drugs at 36 Months of Follow-Up for Selected Baseline Characteristics
| || ||Blood Pressure Controlled||On Two or More Drugs|
| || ||Univariate||Full Model||Univariate||Full Model|
|Baseline Characteristic||Reference Category||Odds Ratio†||95% CL||Odds Ratio†||95% CL||Odds Ratio†||95% CL||Odds Ratio†||95% CL|
|Visit 1 SBP, mm Hg||(10 mm Hg increase)||0.82*||(0.80,0.84)||0.79*||(0.77, 0.80)||1.09*||(1.07,1.11)||1.22*||(1.19,1.24)|
|Age||(10-year increase)||0.92*||(0.89,0.95)||0.93*||(0.89, 0.97)||0.91*||(0.89, 0.94)||0.84*||(0.81,0.87)|
|Male||(Female)||1.17*||(1.11, 1.24)||1.10*||(1.03,1.18)||1.40*||(1.34, 1.47)||1.15*||(1.09,1.22)|
|Black||(Nonblack)||0.69*||(0.65, 0.73)||0.75*||(0.71,0.80)||0.90*||(0.86,0.95)||0.89*||(0.84, 0.94)|
|Type 2 diabetic||(No)||0.84*||(0.79,0.88)||0.86*||(0.80,0.91)||1.02||(0.97,1.07)||0.98||(0.93,1.04)|
|Smoker||(No)||1.03||(0.97,1.10)||0.96||(0.89,1.04)||0.84*||(0.80, 0.89)||0.80*||(0.75, 0.86)|
|Prior treatment||(No)||0.91||(0.84,1.00)||0.68*||(0.61, 0.75)||2.17*||(1.99,2.37)||2.94*||(2.66, 3.25)|
|Creatinine ≥1.5 mg/dL||(No)||0.79*||(0.70,0.88)||0.88||(0.78,1.00)||1.14*||(1.04,1.25)||1.06||(0.96,1.18)|
|ECG LVH||(No)||0.69*||(0.60, 0.80)||0.85*||(0.73, 0.99)||1.25*||(1.11,1.41)||1.20*||(1.05,1.36)|
|Clinic research exp||No)||0.93*||(0.88,0.99)||0.90*||(0.83,0.96)||1.23*||(1.17,1.29)||0.95||(0.89,1.01)|
|Comm Hlth Ctr||(Private)||0.94||(0.85,1.03)||1.14*||(1.02,1.28)||1.16*||(1.07,1.26)||0.98||(1.02, 1.28)|
|Geographic region|| || || || || || || || || |
|South||(West)||0.73*||(0.66,0.80)||0.84*||(0.76, 0.92)||0.84*||(0.78,0.91)||0.84*||(0.78, 0.91)|
|Puerto Rico & VI||(West)||1.32*||(1.18,1.49)||1.42*||(1.24,1.61)||0.27*||(0.24, 0.30)||0.29*||(0.26, 0.32)|
|CL=confidence limits; SBP=systolic blood pressure; DBP=diastolic blood pressure; SD=standard deviation; ASCVD=atherosclerotic cardiovascular disease; BMI=body mass index; VI=US Virgin Islands; HMO=health maintenance organization; Comm Hlth Ctr=community health center; VAMC=Veterans Affairs Medical Centers; LVH=left ventricular hypertrophy; ECG=electrocardiogram||*p-value <0.05. †An odds ratio >1.0 implies characteristic is associated with better BP control than the reference category; odds ratio <1.0 is associated with worse BP control. ‡An odds ratio >1.0 implies participants with characteristic or in group are more likely to have been prescribed two or more antihypertensive drugs than the reference category; odds ratio <1.0 implies participants are less likely to have been prescribed two or more drugs.|
In the multivariable analysis for treatment with two or more antihypertensive agents at 36 months of follow-up, the following characteristics were associated with a significantly greater chance (OR >1.0) of being on more medications: higher baseline SBP, male sex, atherosclerotic vascular disease, obesity, prior treatment, ECG LVH, and university or Veteran Affairs medical center sites compared with private or health maintenance organization (HMO) sites. The following were associated with a significantly lower chance (OR <1.0) of being on two or more drugs at 36 months compared with their corresponding reference groups: older age, black race, and smokers. Similarly, participants at sites in the Eastern or Southern United States, or Puerto Rico and the Virgin Islands were less likely to be on two or more drugs than those at sites in the Western United States.
Table V displays BP control status for participants prescribed two or more antihypertensive drugs, the percentage of uncontrolled participants who had or did not have their drug regimen “stepped up” at the visit, and the mean Bps for those not controlled, according to whether or not there was a step up in drug regimen. A “step up” in drug regimen was defined as either an increase in the blinded (Step 1) drug, an increase in the number of drugs prescribed, or a change in prescription without a change in number of drugs. For participants whose BP was uncontrolled at a given visit, the percentage stepped up decreased from 64% at 6-month visits to only about 30% at the second through fifth year visits. The mean SBP for those stepped up was generally about 4 mm Hg higher than those not stepped up; the DBP was about 1 mm Hg higher. Participants not stepped up had mean SBP in the 151–154 mm Hg range.
Table V. Blood Pressure Control† Status Among Participants Prescribed Two or More Antihypertensive Drugs During Follow-Up, and the Percentage of Uncontrolled Participants Who Had Their Drug Regimen Stepped Up‡ at the Visit
|Follow-Up Visit||Number With Visit||Prescribed Two or More Antihypertensive Drugs|
| || || || || ||BP Uncontrolled|
| || ||All||BP Controlled†||Not Stepped Up‡||Stepped Up‡|
| || ||n||(%)¶||n||(%)||n||(%)||SBP||DBP||n||(%)||SBP||DBP|
|†Blood pressure (BP) control defined as systolic BP(SBP)/diastolic BP(DBP)<140/90 mm Hg; ‡A “step up” in drug regimen was defined as either an increase in the dose of the blinded (Step 1) drug, an increase in the number of drugs prescribed, or a change in prescription without a change in number of agents. No information was available on change in doses of drugs other than the blinded drug, so it is likely the percent stepped up is underestimated. ¶Percentage of those with visit.|
- Top of page
- STATISTICAL METHODS
The BP control rate (66%) and the mean BP levels (135/75 mm Hg) achieved after 5 years of participant follow-up in ALLHAT should be considered an example of what may be achieved in the management of most hypertensive patients, since the majority of study sites in ALLHAT were in community practice settings. This high rate of BP control was achieved despite recruiting a population enriched with subgroups more difficult to control, such as blacks, the elderly, and diabetics, and despite the limitation on the selection of drug classes to add to the blinded therapy. Comparisons with other studies must be qualified by the potential effects of selective nonresponse, probably a larger issue for missed visits in follow-up studies such as ALLHAT than with surveys such as NHANES. In ALLHAT, it is possible that some participants who stopped attending clinics had easily controlled BP and consequently discontinued medication, while others had difficult-to-control BP and left ALLHAT clinics due to dissatisfaction with care.
ALLHAT is the largest prospective randomized trial ever completed to examine the control of hypertension and the effects of antihypertensive drugs on clinical outcomes.13 It was conducted in a variety of clinical practice settings in North America and the Caribbean and in a diverse population of older hypertensive patients at high risk for cardiovascular events by virtue of comorbid conditions or previous cardiovascular events.13,14 ALLHAT has the largest number of blacks and the largest number of diabetics of any hypertension treatment trial conducted. We believe that the ALLHAT population is representative of the middle-aged and older hypertensive population frequently seen in practice settings. Although participants with stage 3 or more severe hypertension were not included, they represent a small minority of hypertensives in the uncontrolled category.
The ALLHAT participants were older on average and more likely to be on antihypertensive drug treatment at baseline than the 1988–1991 NHANES III population, who were surveyed in the years prior to the initiation of recruitment for ALLHAT.1 Nevertheless, the proportion of the ALLHAT participants with BP controlled (<140/90 mm Hg) prior to the study at screening (27.4%) was virtually identical to the BP control rate in the adult noninstitutionalized representative population in NHANES III. Therefore, the rapid improvement in the BP control rates in ALLHAT demonstrates what can be achieved by setting a BP goal that is consistently reinforced, monitoring control rates, providing feedback to physicians about their performance in meeting the BP goal, and using a specific protocol for increasing doses and addition of antihypertensive medications to the regimen.
The ALLHAT regimens were more restrictive than what may be used in clinical practice, since the investigators could not routinely add any of the four classes of agents to which participants were allocated at randomization: a diuretic, an ACE inhibitor, a calcium channel blocker, or an alpha blocker. The first three of these are among the most commonly used classes for the management of hypertension. Only atenolol, clonidine, reserpine, and/or hydralazine were provided in the trial to be added to the blinded drugs to attempt to achieve BP control. For example, the addition of a β blocker in a participant randomized to an ACE inhibitor might be expected to lower BP less than the addition of a diuretic, which was not routinely permitted. The proportion of participants treated with two or more drugs increased from 26% at 6 months to 62% at 5 years (Table V). However, among participants with uncontrolled BP, the proportion whose regimens were stepped up leveled off at around 30% by year 2, suggesting that even better BP control rates could have been achieved with further medication titration as prescribed by the protocol.
The BP control achieved in ALLHAT was superior to the control rates among treated hypertensives surveyed in NHANES III, where BP <140/90 mm Hg was found in 30%–45% of those who had hypertension and reported taking antihypertensive medications.1 ALLHAT BP control rates also greatly exceeded control rates (25%) in treated hypertensive patients who were regularly being seen by physicians in primary care medicine clinics in several Veterans Affairs Medical Centers in the New England area.6
Several other large hypertension trials begun after ALLHAT was initiated have had the same BP goals.17,18 In the Losartan Intervention For Endpoint Reduction in Hypertension (LIFE) trial,18,19 an angiotensin receptor blocker (losartan)-based regimen was compared with a β blocker (atenolol)-based regimen in 9193 participants. A diuretic was routinely added to either regimen to achieve BP control, and any other antihypertensive drug class could be added except for another angiotensin receptor blocker, a β blocker, or an ACE inhibitor. After 4.8 years of follow-up, BP control (≤140/90 mm Hg rather than <140/90 mm Hg) was only 49% and 46% in the losartan and atenolol groups, respectively. In the Controlled Onset Verapamil Investigation of Cardiovascular Endpoints (CONVINCE) trial,17 a regimen beginning with a calcium channel blocker (controlled-onset, extended release verapamil) was compared with a regimen begun with the investigator's choice of either a diuretic (hydrochlorothiazide) or a β blocker (atenolol). After 30 months of follow-up, 67% of CONVINCE participants achieved BP <140/90 mm Hg, very similar to what was achieved in ALLHAT with a more restrictive regimen. The lower control rates in LIFE may be related to: 1) higher entry BP criteria and baseline BP; 2) ECG LVH in all participants, which predicted a lower likelihood of controlling BP in ALLHAT; 3) fewer medications available to add to the BP regimen; and 4) the different thresholds for treatment and treatment goals among European/Scandinavian investigators in LIFE.19
In all of these trials, SBP was the primary determinant of the control rates, since DBP was usually controlled in over 90% of the trial participants. In ALLHAT DBP was <90 mm Hg in 92% of participants while SBP was <140 mm Hg in 67%. A major factor that contributes to this difference is the preponderance of isolated systolic hypertension in this older study population: over 80% of untreated or treated but uncontrolled hypertensive individuals over age 60 years have isolated systolic hypertension.20 Other contributing factors may include the historical emphasis on DBP control throughout most of the 20th century in clinical practice and in drug development, the higher doses and numbers of drugs needed to achieve the SBP target, and the reluctance of investigators to prescribe, and participants to take, the large doses and number of medications often required to achieve SBP control. Additionally, a greater proportion of older hypertensive individuals have much farther for SBP to fall compared with DBP to reach the 140/90 mm Hg goal.
An average of two drugs was required to achieve BP control in two thirds of participants, primarily to control SBP. The number of drugs needed to control the majority of participants was greater than was anticipated when ALLHAT was designed. Although medications were often increased for lack of BP control, intensification of therapy was not carried out for many participants with controlled DBP but SBP between 140–159 mm Hg. However, after 36 months of follow-up, approximately twice as many participants had SBP levels of 140–149 mm Hg compared with 150–159 mm Hg (Figure 1). It is very likely that better BP control rates could have been achieved if therapy were intensified more consistently for persistent SBP ≥140 mm Hg. Judged by levels at which treatment was or was not stepped up, some ALLHAT clinicians seemed satisfied with a typical SBP in the low 150s, similar to what Hyman and Pavlik9 observed in NHANES III data. With the greater variety of medications available in clinical practice, it is reasonable to expect that even higher rates of control could be achieved in a “real world” practice setting. These data suggest that the majority of hypertensive patients will require at least two antihypertensive medications to achieve a BP goal of <140/90 mm Hg. For lower goals, such as is recommended in patients with diabetes, even more drugs likely may be needed.
In multiple logistic regression analysis, higher baseline SBP, older age, black race, and living in the Southeastern United States predicted a lower likelihood of achieving BP control in ALLHAT, confirming findings of some previous studies.19,21 Baseline level of BP, black race, older age, and living in the Southeastern United States were the strongest predictors of lack of control with some antihypertensive drugs used as monotherapy in the Veterans Affairs Single-Drug Therapy Study.21 ALLHAT identified additional factors associated with lower BP control rates. These included female sex, diabetes, obesity, prior antihypertensive therapy, ECG LVH, and care in a private practice setting or a clinic where the investigator had prior clinical research experience. This analysis cannot explain these associations. In some cases, they may result from physiological or lifestyle differences that might be expected to result in poorer BP control. In other cases, they may result from differences in practice patterns by the investigators or medication adherence by the participants. Some of the groups with lower BP control rates, e.g., blacks, women, and older-aged participants, were less likely to be on two or more antihypertensive medications. There is evidence that arteries are stiffer in type 2 diabetics, resulting in higher SBP levels and pulse pressure for age.22,23 This may have contributed to greater resistance in lowering SBP in diabetics. The presence of target organ damage, such as LVH and prior antihypertensive therapy, most likely represents more severe hypertension. There are likely unmeasured factors that contribute to the differences seen, and we did not formally model the undoubtedly complex reciprocal relationship between BP control and changes in medication over time. Our study cannot assess the potential contributions to BP control of availability of medical care or cost of care, including medications. All of the antihypertensive medications that were prescribed by the ALLHAT protocol were provided to the participants without cost.
In summary, the ALLHAT trial provides compelling evidence that BP control rates can be markedly increased to at least two thirds of the treated hypertensive population. Most of the participants who did not achieve goal Bps had persistent elevation of SBP. Many of these were not titrated to the maximum doses or numbers of drugs permitted in the trial, and more aggressive treatment of SBP could have yielded even better BP control rates in ALLHAT. Various factors that are associated with lower BP control rates were identified. Nevertheless, it is likely that the overwhelming majority of hypertensive patients could achieve BP control with the use of multiple antihypertensive medications available today and the use of tracking and feedback systems to monitor patient progress.
Acknowlegement: This study is supported by contract with the National Heart Lung and Blood Institute (NHLBI). The ALLHAT investigators acknowledge contributions of study medications supplied by Pfizer, Inc. (amlodipine and doxazosin), AstraZeneca (atenolol and lisinopril), and Bristol-Myers Squibb (pravastatin), and financial support provided by Pfizer.