Improvement in All‐Cause Mortality With Blood Pressure Control in a Group of US Veterans With Drug‐Resistant Hypertension

Abstract

The current definition of drug‐resistant hypertension includes patients with uncontrolled (URH) (taking ≥3 antihypertensive medications) and controlled hypertension (CRH; blood pressure [BP] ≤140/90 mm Hg) (taking ≥4 medications). The authors hypothesized that all‐cause mortality is reduced when URH is controlled. Qualified patients followed at the Washington DC VA Medical Center were included. BPs were averaged for each year of follow‐up. In 2006, among 2906 patients who met the criteria for drug‐resistant hypertension, 628 had URH. During follow‐up, 234 patients were controlled (group 1) and 394 patients remained uncontrolled (group 2). The mortality rate among patients with URH was 28% (110 of 394) and among patients with CRH was 13% (30 of 234), a 54% reduction (P<.01). Multivariate analysis identified independent predictors of mortality as uncontrolled HTN (hazard ratio, 2.5; 95% confidence interval, 1.67–3.75; P<.01), age (hazard ratio, 1.03; 95% confidence interval, 1.01–1.04; P<.01), and diabetes (hazard ratio, 1.46; 95% confidence interval, 1.04–2.05; P<.027). The authors conclude that controlling drug‐resistant hypertension markedly reduces all‐cause mortality.

Hypertension is the most common modifiable cardiovascular risk factor, the leading cause of cerebrovascular accidents, and the most important attributable cause of mortality in the world.1, 2, 3, 4, 5 The prevalence of hypertension in industrialized nations is estimated at approximately 25% to 30%5, 6, 7 and exceeds 74 million patients in the United States alone. In recent estimates, prevalence of hypertension worldwide is projected to exceed 1.5 billion people by 2025.3, 5

In the United States and Europe, resistant hypertension has a prevalence of 12% to 15% among patients with drug‐treated hypertension.6, 8, 9 Thus, there are approximately 130 million people worldwide who may meet the definition of having resistant hypertension. This is of great interest since uncontrolled resistant hypertension leads to an elevated risk of renal failure, myocardial infarction, heart failure, and stroke.9, 10, 11

Resistant hypertension is defined by the American Heart Association (AHA) as blood pressure (BP) that remains above goal despite concurrent use of three antihypertensive agents prescribed at maximally tolerated doses (of which one is a diuretic) or patients whose BP is controlled with four or more medications.11 It does not address the two types of resistant hypertension that exist in practice: resistant hypertension with controlled BP (at goal) and resistant hypertension with uncontrolled BP (above goal).12, 13

While cohort studies have reported increased risk of uncontrolled resistant hypertension, little is known regarding the risk of controlled as compared with uncontrolled resistant hypertension. Multiple prior studies have shown an increased risk of stroke, myocardial infarction, heart failure, and chronic kidney disease in patients with resistant hypertension compared with those without it.9 Although the risk of controlled resistant hypertension has never been estimated, it is likely much lower than that of uncontrolled resistant hypertension.6, 14

In the present study, therefore, we aimed to assess the impact of BP control on all‐cause mortality in patients who meet the criteria for drug‐resistant hypertension.

Methods

Data were extracted from the Electronic Medical Records at the Washington DC VA Medical Center.

Setting

The Veterans Affairs (VA) electronic health record system was first deployed in 1982 as the Decentralized Hospital Computer Program. Details about the system have previously been published.15 The system is searchable and allows individual medical centers to control the care of the entire patient population. Built‐in reminders notified the provider when the last BP was >140/90 mm Hg and prompted physicians to treat elevated BP with a combination of medication and lifestyle modification. Furthermore, as part of the BP‐control initiative, patients with elevated BP were seen at frequent intervals until BP was controlled, and then follow‐up visits were scheduled at 3‐month intervals. Patients with resistant or difficult to control hypertension were referred to specialized hypertension clinics.

Patients

All veteran patients receiving healthcare at the VA Medical Center in Washington, DC, with BP readings in the VA Vital Signs Database during the period of October 1999 to August 2012 were examined, but only patients with uncontrolled resistant hypertension in 2006 were included in this analysis.

Measures

BP measurements were taken by trained healthcare professionals (nurses, physician assistants, and attending physicians) who were instructed to follow standard procedures for BP measurement. Patients were seated in a comfortable position with the back supported and legs not crossed, and BP was taken after 3 to 5 minutes of rest. BP was taken at least twice if elevated and was entered into the vital sign package. Standard cuffs were used in most patients, but large cuffs were available for patients with large arm circumference. All BP readings, age, race, sex, height, and weight were recorded. Age at entry into the data set was used, ie, on the first day with BP (for normotensive patients), or on the date of the third elevated reading (for patients with hypertension). Records were reviewed anonymously.

An elevated BP reading was defined as a systolic BP >140 mm Hg or diastolic BP >90 mm Hg. Resistant hypertension was defined as BP >140/90 mm Hg (or >130/80 mm Hg for patients with diabetes and/or renal disease) despite three or more optimally dosed agents with one diuretic.11 Patients with controlled BP met the definition of resistance if they required four or more medications to maintain their BP at goal.11 Uncontrolled resistant hypertensive patients were those who remained hypertensive despite taking three or more guideline‐supported medications. Only months with readings after the patient was identified as being resistant hypertensive (ie, after the patient's third day with elevated readings) were used for assessing patients with resistant hypertension. All clinic BP measurements were averaged for each calendar year and the average for each patient represented his or her BP for that year. The number of antihypertensive medications was also assessed each calendar year and grouped into five major categories: diuretics, β‐blockers, angiotensin‐converting enzyme inhibitors/angiotensin receptor blockers, calcium channel blockers, and other (clonidine, minoxidil, hydralazine, and α‐blockers). A patient was considered treated with each medication if he or she was prescribed that medication for at least 75% of the time for that year.

For the purpose of this study, all patients who met the criteria for drug‐resistant uncontrolled hypertension in the year 2006 were included. Of those patients, a portion was controlled in subsequent years, with intensified treatment and close follow‐up.

Patients whose average yearly BP was controlled (<140/90 mm Hg) by 2009 were deemed to have controlled resistant hypertension (group 1) and those who were not were classified as having uncontrolled resistant hypertension (group 2). Patients who were not hypertensive, controlled (but not resistant) hypertensive, controlled on one or two medications, or who were missing medication data were excluded from this analysis. The outcome measured was all‐cause mortality from 2006 to 2012 among this group of resistant hypertensive patients.

Statistical Analysis

Follow‐up time was calculated from the date each patient met the criteria for resistant hypertension to the date of death for the decedents, or to December 31, 2012, for those who survived. The data are presented as median and mean±standard deviation (SD). The death rate was calculated as the ratio of the event by the person‐years of observation. Continuous variables are presented as mean±SD and categorical variables as relative frequencies (percentages). Baseline associations between categorical variables were tested using chi‐square. Univariate analysis was applied to evaluate differences of normally distributed variables between fitness categories. The analyses were adjusted for age, body mass index, and resting systolic BP when appropriate. Proportional hazard analyses were used to determine hazard ratios (HRs) associated with BP status (controlled/uncontrolled). In the fully adjusted model, the covariates were baseline age, race, sex, baseline BP, type 2 diabetes mellitus, dyslipidemia, smoking, cardiac/hypertension medications, and cardiovascular disease. All variables included in the models were based on the rationale for their clinical role on the outcome and the main factors of interest. Cox proportional hazard models were then used to compare risks between the fitness categories using the least‐fit category as the reference group. The model was adjusted for the aforementioned covariates. The assumption of proportionality for the Cox proportional hazards models was graphically tested by plotting the logarithm of the cumulative hazards with time for each covariate; the proportionality assumption was fulfilled for each model. All hypotheses were two‐sided and P values <.05 were considered statistically significant. All statistical analyses were performed using SPSS software version 21.0 (SPSS Inc, Chicago, IL).

Results

There are approximately 72,000 patients followed at the Washington DC VA Medical Center. Of those, approximately 42,000 have clinical hypertension, and 2906 hypertensive patients met the criteria for resistant hypertension (as defined by the AHA). Of these 2906 resistant hypertensive patients followed by the hypertension clinic (controlled and uncontrolled), 628 patients were found to be uncontrolled in the year 2006 (Figure 1, flow chart). Of those, by the year 2009, 234 of the patients demonstrated satisfactory BP control and remained controlled throughout. Patients were then separated into two groups: controlled (group 1, n=234) and uncontrolled (group 2, n=394) patients with resistant hypertension.

Figure 1.

The total population of patients followed at the Washington VA Medical Center were screened. Of the 72,000 patients, 42,000 were hypertensive and 2906 met the criteria of drug‐resistant hypertension. In 2006, 628 patients had uncontrolled resistant hypertension. Of those patients, 234 achieved blood pressure (BP) control (group 1) and 394 remained uncontrolled (group 2) by the year 2009. AHA indicates American Heart Association.

The Table shows the baseline characteristics of all included patients. The first column contains the characteristics of the overall 628 patients who were uncontrolled resistant hypertensive patients in 2006. Group 1 represents the 234 patients who later achieved BP control, and group 2 represents the 394 patients who remained uncontrolled in 2009. In 2006, there was no significant demographic difference between the groups. In particular, the patients were aged 66 years on average, with no statistically significant difference between the age of group 1 and group 2. There was no difference between the two groups with respect to cardiovascular disease and other cardiovascular risk factors such as chronic kidney disease, diabetes mellitus, dyslipidemia, or tobacco abuse.

Table 1.

Baseline Characteristics

Characteristics	All Patients	Group 1	Group 2	P Value
	(Resistant Uncontrolled in 2006) (N=628)	(Controlled in 2009) (n=234)	(Uncontrolled in 2009) (n=394)
Age, mean (SD), y	66.4 (12.3)	66.6	66.3	.73
Men, No. (%)	613 (98)	229 (98)	384 (98)	.75
Black race, No. (%)	259 (41)	89 (38)	170 (43)
White race, No. (%)	62 (10)	27 (12)	35 (9)
Other race, No. (%)	307 (49)	118 (50)	189 (48)
Body mass index, No.	29.8	30.2	29.6	.24
Cardiovascular disease, No. (%)	257 (41)	103 (44)	154 (39)	.22
Chronic kidney disease, No. (%)	170 (28)	62 (27)	108 (27)	.8
Diabetes mellitus, No. (%)	320 (51)	126 (54)	194 (49)	.26
Dyslipidemia, No. (%)	136 (22)	53 (23)	83 (21)	.64
Sleep apnea, No. (%)	73 (12)	32 (14)	41 (10)	.22
Tobacco use, No. (%)	82 (13)	32 (14)	50 (13)	.72
Baseline hypertension medications
Calcium channel blocker, No. (%)	456 (73)	165 (71)	291 (74)	.36
ACE inhibitor/ARB, No. (%)	514 (82)	191 (82)	323 (82)	.91
β‐Blocker, No. (%)	433 (69)	173 (74)	260 (66)	.04
All diuretics, No. (%)	628 (100)	234 (100)	394 (100)	NA
Others, No. (%)	98 (16)	31 (13)	67 (17)	.52

Abbreviations: ACE, angiotensin‐converting enzyme; ARB, angiotensin receptor blocker NA, not available; SD, standard deviation.

Antihypertensive medication regimens were similar between the two groups, with no statistical differences in use of calcium channel blockers, angiotensin‐converting enzyme inhibitors/angiotensin receptor blockers, diuretics, or others. The lone statistically significant difference was that more patients in group 1 were taking β‐blockers than in group 2 (P=.04). In addition, the number of medications between group 1 and group 2 did not differ significantly throughout the entire study follow‐up period (P=.063). The average systolic and diastolic BPs (average of 1 year) of groups 1 and 2 are displayed graphically by year in Figure 2. The mean systolic BP of all 628 uncontrolled resistant hypertensive patients in 2006 was 153.1±12.1 mm Hg and the mean diastolic BP was 81.2±10.9 mm Hg. In 2009, group 1 (controlled) had a mean systolic BP of 129.8±8.2 mm Hg during that year and a mean diastolic BP of 71.6±8.0 mm Hg, whereas group 2 (uncontrolled) had a mean systolic BP of 151.9±10.8 mm Hg and a mean diastolic BP of 79.8±10.8 mm Hg (P<.001 for the difference between both mean systolic and diastolic BPs). In 2012, group 1 had a mean systolic BP of 138.3±14.0 mm Hg and a mean diastolic BP of 73.9±9.6 mm Hg, while group 2 had a mean systolic BP of 147.5±14.2 mm Hg and a mean diastolic BP of 77.0±9.5 mm Hg (P<.001 for both systolic and diastolic BP difference).

Figure 2.

Change in systolic and diastolic blood pressure (BP) from 2006 to 2012 in the two groups. Dots also show the trend in BP for the total population.

Overall, there were 30 deaths (13%) in group 1 (controlled) and 110 deaths (28%) in group 2 (uncontrolled) over the 6‐year follow‐up (P<.001). The Cox regression curves for cumulative survival among the two groups are shown in Figure 3. Further analysis revealed that the sole multivariate predictors of mortality were age (HR, 1.03; confidence interval [CI], 1.01–1.04; P<.001), diabetes (HR, 1.46; CI, 1.04–2.07; P=.03), and uncontrolled resistant hypertension (HR, 2.48; CI, 1.64–3.76; P<.001) (Figure 4).

Figure 3.

Cox regression survival curves for patients with controlled and uncontrolled drug‐resistant hypertension. By the year 2012, mortality was 13% in the controlled group and 28% in the uncontrolled group, a relative reduction of 54%.

Figure 4.

Shows predictors of all‐cause mortality in this population. Uncontrolled hypertension was a strong predictor and associated with 2.4‐fold increase in risk of death. CVD indicates cardiovascular disease; CKD, chronic kidney disease; HTN, hypertension; CI, confidence interval.

Discussion

Our study examines the impact of BP control in patients who meet the definition of drug‐resistant hypertension as defined by the AHA. However, the AHA‐endorsed definition of resistant hypertension includes both patients with controlled hypertension and patients with uncontrolled hypertension, two populations with potentially different cardiovascular risk.9, 10, 11

The most important finding of the present study is the substantial decrease in all‐cause mortality risk among the cohort of patients with resistant but controlled hypertension. To the best of our knowledge, this is the first study to elucidate such a mortality benefit in this subpopulation of hypertension patients.

The overall study mortality rate is strikingly high at 22% (140 of 628) over the course of 6 years, which is largely driven by the mortality in the persistently uncontrolled resistant hypertensive group (group 2), with an HR of 2.48 (P<.001). Although retrospective in nature, these data strongly suggest a robust mortality benefit probably derived from BP control in this high‐risk population.

Resistant hypertension is a challenging clinical entity to treat. In addition, the lack of a uniform definition has caused a potential misclassification of patients as having resistant hypertension when, in fact, other explanations may account for the apparent treatment‐resistant hypertension (aTRH).11, 16, 17

The definition of resistant hypertension does not differentiate between patients with true resistant hypertension and those who are pseudoresistant, nor does it account for the different classes of aTRH. Pseudoresistant hypertension is defined as resistant hypertension caused by white‐coat hypertension, improper BP measurement, or medication nonadherence.6, 10 aTRH encompasses four classes of patients: those with (1) suboptimal medication adherence, (2) BP measurement artifacts, (3) suboptimal medication regimens, and (4) true treatment‐resistant hypertension,11, 16, 17, 18 and includes both patients with controlled and uncontrolled BPs, all of which meet the AHA definition of resistant hypertension.12, 13

Irvin and colleagues17 showed that poor medication adherence accounts for a small percentage of patients with aTRH, with 67.8% of an aTRH cohort of 2654 patients reporting perfect medication adherence and only 8.1% reporting low adherence.17

Since the AHA definition is based only on office BP and does not take into consideration ambulatory BP, it cannot exclude patients with white‐coat hypertension.6, 10, 11, 14, 19, 20, 21 Ambulatory BP should be used to define true resistant hypertension (and exclude pseudoresistant hypertension).6, 11, 19, 20, 21

Inappropriate regimens that include small doses of diuretics or no diuretics, coadministration of angiotensin‐converting enzyme inhibitor and angiotensin receptor blockers, or two types of calcium channel blockers are some of the reasons for “resistance.”11, 22 Daugherty and colleagues16 highlight several issues in clinical practice and the practice of many physicians who fail to intensify treatment for resistant hypertensive patients during clinic visits when BP is above goal; there was an “increase in medication class or dose” at a mere 21.6% of such visits in a cohort of 3550 patients, demonstrating that physician attitudes and management styles are to blame for poor BP control.16

Several studies have indicated an increased cardiovascular risk associated with uncontrolled resistant hypertension. A large retrospective study by Daugherty and colleagues10 found that patients with resistant hypertension had an almost 50% greater likelihood of experiencing a cardiovascular event over the course of 5 years.10 While Daugherty and colleagues examined the incidence and cardiovascular outcomes of those patients who became resistant among a group of hypertensive patients,10 our study began with patients who were already deemed resistant and followed them for a longer duration.

The most recent prospective, observational study on this subject by Tsioufis and colleagues23 revealed a 2.2‐fold increased risk of cardiovascular morbidity among patients with persistent resistant hypertension when compared with those without it.23 However, data on mortality associated with resistant hypertension, as well as data comparing outcomes in patients with controlled resistant vs uncontrolled resistant hypertension, are lacking.

There are essentially no publications on the topic of controlled vs uncontrolled resistant hypertension, with the exception of an abstract published in 2010 that studied 47 patients with uncontrolled resistant hypertension and 43 patients with controlled resistant hypertension. The study revealed that body mass index, systolic and diastolic BP, plasma aldosterone concentration, and left ventricular mass index were all significantly higher in the uncontrolled group, supporting the hypothesis that uncontrolled and controlled resistant hypertension are different conditions and that obesity and uncontrolled BP may well be linked by hyperaldosteronism.12

Regarding the pathophysiology of resistant hypertension, there is no known single mechanism that can account for this difficult to treat condition. Aldosterone has been identified as an important factor and target for therapy since elevated aldosterone levels are seen in hypertensive patients, particularly in those who are obese and insulin‐resistant. Aldosterone is involved in numerous biochemical pathways of inflammation, endothelium relaxation, vascular smooth muscle cell proliferation, salt retention, and acts via the renin‐angiotensin‐aldosterone system.24, 25

The sympathetic nervous system plays a significant role in resistant hypertension. Excess aldosterone levels may be an indirect result of activated central nervous system.26, 27 Dating as far back as 1938, surgical procedures such as radical lumbodorsal splanchnicectomies were performed for severe hypertension, and the field has progressed to pharmacologic agents and the development of renal artery sympathetic denervation and carotid baroreceptor stimulation in an attempt to target the sympathetic nervous system.26, 28, 29 The impact of the renal sympathetic nervous system has received more attention in recent years, and catheter‐based renal artery sympathetic denervation has been systematically studied based on the observation that renal sympathetic innervation overactivity contributes to resistance to pharmacologic antihypertensive treatments.28, 30 While there was initial excitement over the use of renal artery denervation for the treatment of resistant hypertension, the efficacy of such an approach is unresolved due to recent negative trials.31, 32, 33

Limitations

There are several limitations of this retrospective cohort study. As it was conducted at the Washington DC VA Medical Center, the results cannot be extrapolated to female or other underrepresented minority patients. Given the methodology, which employed the AHA definition of resistant hypertension, it is possible that some patients with uncontrolled hypertension taking fewer than three medications were inappropriately excluded from the study by definition, but may actually demonstrate frank resistant hypertension when tried on more medications.

Similarly, as previously discussed, our study was unable to separate true resistant hypertensive patients from those with pseudoresistant hypertension. This is because of our use of office‐based BP measurements, as opposed to ambulatory BP monitoring, which may lead to more accurate estimates of resistant hypertension and are reportedly more prognostic.10, 19, 20, 21 Salles and colleagues19 showed that ambulatory BP could be used to separate true resistant hypertension patients from those with white‐coat hypertension and that it was prognostic of cardiovascular morbidity and mortality in resistant hypertension patients, whereas office‐based BP was not.19 However, as noted by Daugherty and colleagues,10 office‐based measurements represent the current clinical practice in the management of hypertension, thus justifying our methodology.10

As the AHA scientific statement definition of resistant hypertension incorporates BP control as part of the criteria for resistance, it is not possible to distinguish whether the overall increase in mortality observed for patients with resistant hypertension was caused by an inherent quality of the disease state itself, or whether it was caused by poor BP control.

Conclusions

Data from this retrospective study indicate an impressive reduction in all‐cause mortality with control of BP in patients who met the AHA criteria of drug‐resistant hypertension. Furthermore, these data indicate an urgent need to intensify the efforts to identify, properly treat, and control patients with drug‐resistant hypertension worldwide.

Disclosures

None.

J Clin Hypertens (Greenwich). 2016;18: 33–39. DOI: 10.1111/jch.12672 © 2015 Wiley Periodicals, Inc.