Abstract
Associations between hypertension and some cardiovascular diseases are stronger in black vs white adults. We examined associations of hypertension, hypertension duration, and control with incident heart failure (HF) in black and white REasons for Geographic And Racial Differences in Stroke study participants (n = 25 770) who were followed for incident HF hospitalization (n = 947) from enrollment in 2003‐2007 through 2015. Hypertension was defined, using updated US guidelines, as systolic or diastolic blood pressure (BP) ≥130/80 mm Hg or antihypertensive medication use. Duration was assessed at baseline, and control was defined as treated BP < 130/80 mm Hg. Compared with no hypertension, hypertension was associated with higher risk of incident HF (HRwhites 1.90 [95% CI 1.49, 2.41], HRblacks 2.36 [95% CI 1.53, 3.65]), HF with preserved ejection fraction (HRwhites 2.01 [95% CI 1.34, 3.01], HRblacks 2.70 [95% CI 1.25, 2.53]), and HF with reduced/mid‐range ejection fraction (HRwhites 1.69 [95% CI 1.23, 2.33], HRblacks 2.29 [95% CI 1.26, 4.15]). Hypertension duration <10 years and ≥10 years were associated with higher risk for incident HF compared with no hypertension. Although risk of incident HF was highest among participants with uncontrolled BP, even controlled BP vs no hypertension was associated with increased risk of HF (HRwhites 1.93 [95% CI 1.44, 2.58], HRblacks 2.01 [95% CI 1.22, 3.29]). Interactions with race were not statistically significant. The risk of HF associated with hypertension, even with shorter duration or controlled BP, suggests that both prevention and therapeutic management of hypertension are important in reducing HF risk.
Keywords: epidemiology, heart failure, hypertension
What is known about this topic?
Hypertension is a major preventable and modifiable risk factor for the development of heart failure (HF).
The association between hypertension and some cardiovascular diseases (CVD) is stronger for black compared with white individuals
What this study adds?
Defining hypertension as blood pressure (BP) ≥130/80 mm Hg, the risk between hypertension and incident HF was not significantly different for black and white individuals.
The strongest risk for incident HF was among those with a long duration of hypertension and with uncontrolled BP.
Even those with a short duration of hypertension and controlled BP had an increased risk of incident HF compared to those with no hypertension.
1. INTRODUCTION
Hypertension is a major preventable and modifiable risk factor for the development of heart failure (HF). 1 , 2 For example, among 13 643 adults in the National Health And Nutrition Examination Survey I, the relative risk for incident HF associated with hypertension was 1.4 (95% confidence interval [CI] 1.24, 1.59). 2 Higher blood pressure (BP) over an individual's lifetime increases the risk of HF later in life. 3 Additionally, individuals taking antihypertensive medication who have uncontrolled BP compared with controlled BP are at an increased risk of developing HF. 4 , 5
Prior cohort studies examining incident HF have been conducted in predominately white populations with limited geographic representation. 6 , 7 , 8 , 9 Additionally, hypertension guidelines in the United States (US) were recently revised to define hypertension as BP ≥130/80 mm Hg and hypertension control as BP <130/80 mm Hg. 10 There are well‐described differences in the risk of hypertension between blacks and whites. 11 , 12 Hypertension develops earlier in life for blacks compared to whites, 13 with the higher risk of hypertension in blacks extending to older ages. Although there is a higher prevalence of hypertension awareness and treatment in blacks compared with whites, among adults taking antihypertensive medication blacks are less likely than whites to have their BP controlled to guideline‐recommended levels. 14 , 15 , 16 The Atherosclerotic Risk In Communities study, 17 Multi‐Ethnic Study of Atherosclerosis, 18 and Women's Health Initiative 19 suggest that HF risk associated with hypertension was not different for black and white participants. However, less is known about differences in the associations between hypertension and incident HF subtypes by race. We examined the association of hypertension with incident HF and HF subtypes among black and white adults enrolled in the REasons for Geographic And Racial Differences in Stroke (REGARDS) study, a contemporary nationwide cohort. Additionally, we examined the association of hypertension duration and, separately, BP control among participants taking antihypertensive medication with incident HF and HF subtypes.
2. METHODS
REGARDS is a US prospective cohort study that enrolled 12 514 black and 17 669 white adults ≥45 years of age. 20 Eligible participants were community‐dwelling and self‐identified as black or white race. Exclusion criteria included active treatment for cancer, cognitive impairment as determined by an interviewer, and inability to communicate in English. By design, the study oversampled blacks had approximately equal representation of men and women and 56% (goal 50%) of the cohort was recruited from the eight Southern US states, commonly referred to as the “stroke buckle” (coastal plain of North Carolina, South Carolina, and Georgia) and “stroke belt” (remainder of North Carolina, South Carolina, and Georgia as well as Alabama, Mississippi, Tennessee, Arkansas and Louisiana), with the remaining 44% of the sample recruited from the rest of the continental United States. Participant enrollment and baseline data collection occurred from January 1, 2003 through October 31, 2007. Follow‐up of study participants is on‐going. The Institutional Review Boards of all participating institutions approved this study. All participants provided written informed consent for participation.
For this analysis, we excluded participants missing follow‐up, missing data on baseline medication inventory or information to determine hypertension status. Additionally, we excluded participants with evidence of HF, based on use of medications with indications for the treatment of HF. Specifically, participants taking: (a) digoxin with no documentation of atrial fibrillation, (b) carvedilol, (c) angiotensin converting enzyme inhibitors (ACE‐I) or angiotensin receptor blockers (ARB) plus beta‐blockers in the absence of hypertension, (d) loop diuretics, or (e) hydralazine plus isosorbide mononitrate or isosorbide dinitrate were considered to have evidence of HF at baseline. In a prior report, this approach for defining a HF‐free cohort in REGARDS resulted in a negative predictive value >95% (in press BMC Medical Research Methodology). These criteria excluded 17.7% (n = 2216) of black participants and 12.4% (n = 2197) white participants. Analyses included 25 770 REGARDS participants (Figure S1).
2.1. HF outcomes
Participants or their proxies are contacted at 6‐month intervals for surveillance of incident events and hospitalizations. Medical records associated with suspected heart disease events or hospitalizations (including HF) are retrieved and adjudicated by pairs of trained clinician investigators using information collected as part of routine clinical care. Standardized reporting forms were used to adjudicate the presence of HF based on signs, symptoms, and biomarkers (including b‐type natriuretic peptide) consistent with HF. For adjudicated events, data on ejection fraction were also collected.
The primary outcome was incident HF (n = 947). The first adjudicated HF event for each participant between the date of their in‐home examination and December 31, 2015 was included. As a secondary outcome, incident HF events were grouped according to HF subtype based on ejection fraction. HFrEF hospitalizations (n = 413) were defined as those with documented ejection fraction <40%, 21 a qualitative report of low ejection fraction, or, in the absence of information about ejection fraction during the index hospitalization, ejection fraction <40% or qualitative report of low ejection fraction during a hospitalization within 18 months after the initial HF hospitalization. HFpEF hospitalizations (n = 354) were defined as those with documented ejection fraction ≥50%, a qualitative report of normal ejection fraction or, in the absence of information about ejection fraction during the index hospitalization, ejection fraction ≥50% or qualitative report of normal ejection fraction during a hospitalization within 18 months after the initial HF hospitalization. HF with a documented ejection fraction between 41% and 49% was categorized as mid‐range ejection fraction (HFmEF, n = 71). Due to the small number of participants with mid‐range ejection fraction and the similarities of individuals with HFrEF and HFmEF, 22 HFrEF and HFmEF were combined into a single category (HFrEF/HFmEF, n = 484). For 109 HF events, information on ejection fraction was not available.
2.2. Hypertension, duration, and BP control
At baseline, BP was measured two times by trained technicians during the in‐home examination with an aneroid sphygmomanometer following standardized protocols and averaged for analysis. Hypertension was defined as systolic BP (SBP) ≥130 mm Hg, diastolic BP (DBP) ≥80 mm Hg, 10 or self‐reported antihypertensive medication use. Self‐reported antihypertensive medication use was defined as a positive response to the question “Are you now taking any medicine for high BP?” Hypertension duration was defined based on questions from the baseline telephone interview asking participants to estimate their exact age or the 10‐year age interval (eg, 30‐39 years, 40‐49 years, etc) at which they were first told they had high BP. For participants who reported an exact age, hypertension duration was determined by subtracting the reported age at hypertension diagnosis from the age at study enrollment. For participants reporting a 10‐year age interval for hypertension diagnosis, the midpoint of this interval was subtracted from the age at study enrollment to determine duration. We grouped hypertension duration into categories of ≥10 years or <10 years. Participants with hypertension at baseline who were unaware or unsure when they were first told they had hypertension were considered to have an undetermined hypertension duration.
Participants who self‐reported they were not taking antihypertensive medication at baseline but with SBP/DBP ≥130/80 mm Hg were considered to have untreated hypertension. Among participants taking antihypertensive medication at baseline, those with SBP/DBP ≥130/80 mm Hg were considered to have uncontrolled BP and those with SBP/DBP <130/80 mm Hg were considered to have controlled BP. Untreated hypertension and uncontrolled BP were categorized separately as they may pertain to different aspects of hypertension (ie, awareness of hypertension and management of hypertension).
2.3. Covariates
Covariates were chosen based on known and potential confounders of the association between hypertension and HF and included sociodemographic and health behavior characteristics and comorbidities. At baseline, interviewers conducted computer‐assisted telephone interviews and an in‐home examination to obtain information on participant's demographics (age, gender, region of residence, income, and health insurance) and health behaviors (cigarette smoking, alcohol consumption, and physical activity). Body mass index (BMI) was calculated as weight in kilograms divided by height in centimeters squared. Blood samples were used to measure total cholesterol and blood glucose. Diabetes was defined as serum glucose ≥126 mg/dL for participants who fasted ≥8 hours prior to blood collection (or serum glucose ≥200 mg/dL for participants who failed to fast) or use of insulin or an oral hypoglycemic medication. The Chronic Kidney Disease Epidemiology Collaboration equation was used to calculate estimated glomerular filtration rate (eGFR). 23 Reduced eGFR was defined as ≤60 mL/min/1.73 m2. Perceived stress was measured using the seven‐item Perceived Stress Scale. 24 The presence of depressive symptoms was assessed using the Center for Epidemiologic Studies Depression‐4 scale. History of coronary heart disease (CHD) was defined as a self‐report or electrocardiogram evidence of CHD during the baseline examination. Antihypertensive and antidiabetes medication use was determined by self‐report. Statin medication use was identified during a medication inventory.
2.4. Statistical analysis
We calculated baseline characteristics of REGARDS participants with and without hypertension by race. We used t tests for continuous variables and chi‐square statistics for categorical variables to test for statistically significant differences across hypertension status. To examine the association between hypertension and incident HF, we first calculated incidence rates and 95% CIs for white and black participants, separately. Hazard ratios (HRs) and 95% CIs were calculated for the association of hypertension with incident HF using Cox proportional hazards regression sequentially adjusting for demographics (models 1), clinical features (model 2), and additional potential risk factors chosen a priori based on clinical judgment (model 3). Model 1 adjusted for age and gender. Model 2 adjusted for age, gender, BMI, smoking, alcohol intake, diabetes, reduced eGFR, total cholesterol, and statin use. Model 3 adjusted for the covariates in model 2 and income, region of residence, health insurance, physical activity, perceived stress, and depressive symptoms. In a sensitivity analysis, we additionally adjusted model 3 for antihypertensive medication use (ACE‐I/angiotensin receptor blocker, diuretic, calcium channel blocker, beta blocker, and other). The proportionality of hazards assumption was evaluated with a log(−log(survival)) plot. The likelihood ratio test was used to determine the statistical significance of differences in the associations of hypertension and HF between black and white participants.
Using a cause‐specific regression model described by Lunn and McNeil, 25 HRs and 95% CIs were calculated for the association between hypertension and incident HFpEF, and separately incident HFrEF/HFmEF. Participants who experienced HF events where EF data were unavailable were censored (n = 109). Next, we examined the associations of hypertension duration and separately, hypertension control with incident HF and HF subtypes using Cox proportional hazard regression with progressive adjustment as previously described. Statistical differences between HRs for levels of hypertension duration and, separately, hypertension control were examined using Wald tests. As the risk for developing incident HF associated with hypertension could differ according to age, in a secondary analysis, we repeated the analyses of hypertension and hypertension duration with incident HF stratified by baseline age (<65 years, 65‐74 years, and ≥75 years). REGARDS participants were enrolled prior to the publication of the 2017 American College of Cardiology/American Heart Association definition of hypertension. 10 In a sensitivity analysis, we repeated the analyses for hypertension, hypertension duration, and BP control, separately, with incident HF defining hypertension as SBP/DBP ≥140/90 mm Hg or self‐reported antihypertensive medication use and defining hypertension control as SBP/DBP <140/90 mm Hg. Missing data on covariates (Table S1) were imputed using fully conditional specification and 10 data sets. 26 All statistical tests were two‐sided with a P < .05 considered statistically significant. SAS 9.4 was used for all analyses.
3. RESULTS
3.1. Population characteristics
Overall, 72.7% of participants had hypertension (Table 1). The prevalence of hypertension was 65.9% and 82.9% among white and black participants, respectively. Among the 18 740 participants with hypertension, 12 601 (67.2%) self‐reported antihypertensive medication use, and 6139 had a measured BP ≥130/80 mm Hg in the absence of antihypertensive medication use. Of the participants with SBP/DBP ≥130/80 mm Hg and no antihypertensive medication use, 4032 (65.7%) had SBP or DBP s ranging between 130/80 and 139/89 mm Hg. Both white and black participants with hypertension were more likely to have lower income, BMI ≥ 30 kg/m2, reduced eGFR, diabetes, history of CHD, and were more likely to be taking a statin than their counterparts without hypertension. Those with hypertension were less likely to drink alcohol and exercise. White participants with hypertension were more likely to be male than their white counterparts without hypertension, but a gender difference was not present among black participants. Compared to white participants with hypertension, use of ACE‐I or ARB, diuretics, and calcium channel blockers was more common and use of beta‐blockers was less common among black participants with hypertension (P < .05) (Table S2).
TABLE 1.
Baseline characteristics a of REGARDS study participants without a history of heart failure by race and hypertension status
| White | Black | |||||
|---|---|---|---|---|---|---|
| Hypertension | P‐value | Hypertension | P‐value | |||
| No (n = 5271) | Yes (n = 10 201) | No (n = 1759) | Yes (n = 8539) | |||
| Age, y, mean (SD) | 62.7 (9.3) | 66.2 (9.2) | <.001 | 61.5 (9.4) | 64.1 (9.1) | <.001 |
| Age, n (%) | ||||||
| <60 y | 3186 (60.4) | 4548 (44.6) | <.001 | 1126 (64.0) | 4624 (54.2) | <.001 |
| 60‐74 y | 1436 (27.3) | 3617 (35.5) | 460 (26.2) | 2689 (31.5) | ||
| ≥75 y | 649 (12.3) | 2036 (19.9) | 173 (9.8) | 1226 (14.3) | ||
| Female, n (%) | 2977 (56.5) | 4827 (47.3) | <.001 | 1102 (62.7) | 5260 (61.6) | .41 |
| Region, n (%) | ||||||
| Stroke belt | 1806 (34.2) | 4186 (41.0) | .03 | 582 (33.1) | 2798 (32.8) | .67 |
| Stroke buckle | 1192 (22.6) | 3671 (36.0) | 323 (18.4) | 1506 (17.6) | ||
| Non‐belt/buckle | 2237 (43.2) | 2344 (23.0) | 854 (48.5) | 4235 (49.6) | ||
| Income, n (%) | ||||||
| <$20 000 | 470 (10.2) | 1242 (13.8) | <.001 | 371 (23.7) | 2233 (29.8) | <.001 |
| $20 000 to <$35 000 | 994 (21.6) | 2427 (26.9) | 438 (28.0) | 2287 (30.5) | ||
| $35 000 to <$75 000 | 1749 (38.0) | 3394 (37.6) | 536 (34.3) | 2183 (29.2) | ||
| ≥$75 000 | 1396 (30.2) | 1961 (21.7) | 219 (14.0) | 786 (10.5) | ||
| Health insurance, n (%) | 5020 (95.3) | 9748 (95.6) | .31 | 1555 (88.5) | 7684 (90.1) | .05 |
| Body mass index, kg/m2, n (%) | ||||||
| <25 | 2224 (44.9) | 2474 (26.3) | <.001 | 556 (33.9) | 1428 (18.1) | <.001 |
| 25‐29 | 1787 (36.1) | 3306 (35.2) | 556 (33.9) | 2286 (29.0) | ||
| ≥30 | 945 (19.0) | 3621 (38.5) | 528 (32.3) | 4162 (52.8) | ||
| Current smoker, n (%) | 719 (13.7) | 1220 (12.0) | .003 | 352 (20.1) | 1519 (17.9) | .03 |
| Alcohol use, n (%) | ||||||
| None | 2733 (52.5) | 5599 (55.8) | <.001 | 1189 (69.3) | 5946 (71.5) | .03 |
| Moderate | 2238 (43.0) | 3845 (38.3) | 492 (28.7) | 2149 (25.8) | ||
| Heavy | 236 (4.5) | 592 (5.9) | 36 (2.0) | 223 (2.7) | ||
| Depressive symptoms, n (%) | 416 (8.0) | 872 (8.6) | .16 | 220 (12.6) | 1095 (12.9) | .76 |
| High perceived stress, n (%) | 1320 (25.1) | 2460 (24.1) | .20 | 584 (33.2) | 2857 (33.5) | .84 |
| Physical activity, n (%) | ||||||
| ≥4 times per wk | 1797 (34.6) | 3182 (31.6) | <.001 | 508 (29.3) | 2260 (26.8) | .005 |
| 1‐3 times per wk | 1991 (38.4) | 3628 (36.1) | 668 (38.5) | 3119 (37.0) | ||
| 0 times per wk | 1401 (27.0) | 3248 (32.3) | 558 (32.2) | 3050 (36.2) | ||
| eGFR < 60 mL/min/1.73 m2, n (%) | 210 (4.1) | 1143 (11.6) | <.001 | 59 (3.6) | 869 (10.7) | <.001 |
| SBP, mm Hg, mean (SD) | 114 (9) | 131 (15) | <.001 | 115 (9) | 134 (17) | <.001 |
| DBP, mm Hg, mean (SD) | 70 (6) | 78 (9) | <.001 | 70 (6) | 80 (10) | <.001 |
| Total cholesterol, mg/dL, mean (SD) | 196 (38) | 192 (40) | <.001 | 199 (40) | 194 (41) | <.001 |
| Diabetes, n (%) | 376 (7.4) | 1636 (16.6) | <.001 | 252 (15.1) | 2427 (29.6) | <.001 |
| History of CHD a , n (%) | 520 (9.9) | 941 (19.2) | <.001 | 122 (7.0) | 1133 (13.4) | <.001 |
| Statin use, n (%) | 1159 (22.0) | 3638 (35.7) | <.001 | 277 (15.8) | 2468 (28.9) | <.001 |
| Antihypertensive medication use, n (%) | — | 6356 (64.8) | — | — | 6245 (75.6) | — |
| Antihypertensive medication classes, n (%) | ||||||
| ACE‐I/ARB | — | 4240 (41.6) | — | — | 3819 (44.7) | — |
| Diuretic | — | 2733 (26.8) | — | — | 3243 (38.0) | — |
| Calcium channel blocker | — | 1954 (19.2) | — | — | 2706 (31.7) | — |
| Beta blocker | — | 2464 (24.2) | — | — | 1930 (22.6) | — |
| Other antihypertensive medications, n (%) | — | 481 (4.7) | — | — | 406 (4.8) | — |
Abbreviations: ACE‐I, angiotensin convertase enzyme inhibitor; ARB, angiotensin receptor blocker; CHD, coronary heart disease; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; REGARDS, REasons for Geographic and Racial Differences in Stroke; SBP, systolic blood pressure; SD, standard deviation.
Missing data for some participants (Table S1).
3.2. Hypertension and incident HF
During a mean 8.3 (SD 3.3) years of follow‐up, 3.7% of participants (n = 947) had an incident HF event. Unadjusted HF rates were higher among those with vs without hypertension for both white and black participants (Table 2). After multivariable adjustment, participants with vs without hypertension had a higher risk for HF irrespective of race (White: HR 1.90 [95% CI 1.49, 2.41] and Black: HR 2.36 [95% CI 1.53, 3.65], respectively) (Table 2). The association between hypertension and HF was not statistically different by race (likelihood ratio P = .43). Further adjusting for antihypertensive medication use attenuated the associations (White: HR 1.63 [95% CI 1.26, 2.10] and Black: HR 1.86 [95% CI 1.18, 2.93], P for difference by race = .56) (Table S3). Hypertension was associated with a higher risk for both incident HFpEF and incident HFrEF/HFmEF (Table 3). After multivariable adjustment, the HR for incident HFpEF for those with vs without hypertension was 2.01 (95% CI 1.34, 3.01) for white participants and 2.70 (95% CI 1.25, 5.83) for black participants (likelihood ratio P‐value for difference by race = .49). The HR for incident HFrEF/HFmEF was 1.69 (95% CI 1.23, 2.33) for white participants and 2.29 (95% CI 1.26, 4.15) for black participants (likelihood ratio P‐value for difference by race = .43).
TABLE 2.
Association between hypertension and incident heart failure, by race
| White | Hypertension | P‐value | |
|---|---|---|---|
| No (n = 5271) | Yes (n = 10 201) | ||
| Heart failure events, n | 83 | 482 | |
| Incidence rate per 1000 person‐years (95% CI) | 1.79 (1.41, 2.18) | 5.60 (5.10, 6.10) | |
| Hazard ratio (95% CI) | |||
| Model 1 | 1 (ref) | 2.38 (1.88, 3.01) | <.001 |
| Model 2 | 1 (ref) | 1.97 (1.55, 2.50) | <.001 |
| Model 3 | 1 (ref) | 1.90 (1.49, 2.41) | <.001 |
| Black | Hypertension | P‐value | |
|---|---|---|---|
| No (n = 1759) | Yes (n = 8539) | ||
| Heart failure events, n | 22 | 360 | |
| Incidence rate per 1000 person‐years (95% CI) | 1.55 (0.90, 2.19) | 5.28 (4.73, 5.82) | |
| Hazard ratio (95% CI) | |||
| Model 1 | 1 (ref) | 3.10 (2.02, 4.78) | <.001 |
| Model 2 | 1 (ref) | 2.42 (1.56, 3.74) | <.001 |
| Model 3 | 1 (ref) | 2.36 (1.53, 3.65) | <.001 |
Hypertension defined as SBP/DBP ≥ 130/80 mm Hg or self‐reported antihypertensive medication use. Incidence rates are per 1000 person‐years. Model 1 adjusted for age and sex. Model 2 adjusted for Model 1 covariates + body mass index, smoking, diabetes, reduced estimated glomerular filtration rate, total cholesterol and statin use. Model 3 adjusted for Model 2 + income, region of residence, health insurance, physical activity, perceived stress, and depressive symptoms.
Abbreviations: CI, confidence interval; DBP, diastolic blood pressure; SBP, systolic blood pressure.
TABLE 3.
Association between hypertension and incident heart failure subtypes, by race
| Hypertension | P‐value | ||
|---|---|---|---|
| No | Yes | ||
| White | |||
| HFpEF events, n | 29 | 178 | |
| Hazard ratio (95% CI) | |||
| Model 1 | 1 (ref) | 2.53 (1.71, 3.76) | <.001 |
| Model 2 | 1 (ref) | 2.09 (1.39, 3.12) | <.001 |
| Model 3 | 1 (ref) | 2.01 (1.34, 3.01) | .001 |
| HFrEF/HFmEF events, n | 48 | 244 | |
| Hazard ratio (95% CI) | |||
| Model 1 | 1 (ref) | 2.09 (1.53, 2.85) | <.001 |
| Model 2 | 1 (ref) | 1.75 (1.27, 2.41) | .001 |
| Model 3 | 1 (ref) | 1.69 (1.23, 2.33) | .001 |
| Black | |||
| HFpEF events, n | 7 | 140 | |
| Hazard ratio (95% CI) | |||
| Model 1 | 1 (ref) | 3.72 (1.74, 7.97) | .001 |
| Model 2 | 1 (ref) | 2.77 (1.28, 5.96) | .01 |
| Model 3 | 1 (ref) | 2.70 (1.25, 5.83) | .01 |
| HFrEF/HFmEF events, n | 12 | 180 | |
| Hazard ratio (95% CI) | |||
| Model 1 | 1 (ref) | 2.92 (1.63, 5.23) | <.001 |
| Model 2 | 1 (ref) | 2.33 (1.29, 4.22) | .01 |
| Model 3 | 1 (ref) | 2.29 (1.26, 4.15) | .01 |
Hypertension defined as SBP/DBP ≥ 130/80 mm Hg or self‐reported antihypertensive medication use. Model 1 adjusted for age and sex. Model 2 adjusted for Model 1 covariates + body mass index, smoking, diabetes, reduced estimated glomerular filtration rate, total cholesterol, and statin use. Model 3 adjusted for Model 2 + income, region of residence, health insurance, physical activity, perceived stress, and depressive symptoms. Individuals who experienced HF events where ejection fraction was not available were censored at the time of the HF event.
Abbreviations: CI, confidence interval; DBP, diastolic blood pressure; HF, heart failure; HFmEF, heart failure with mid‐range ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; SBP, systolic blood pressure.
3.3. Hypertension duration and incident HF
Among participants with hypertension, 32.9%, 33.6%, and 33.5% of white participants and 23.5%, 30.8%, and 45.7% of black participants had an undetermined hypertension duration, duration <10 years prior to baseline, and duration ≥10 years prior to baseline, respectively. Among white participants with an undetermined hypertension duration, hypertension duration <10 years, and hypertension duration ≥10 years, multivariable‐adjusted HRs for incident HF were 1.59 (95% CI 1.19, 2.11), 1.87 (95% CI 1.42, 2.46), and 2.21 (95% CI 1.70, 2.89), respectively, compared to participants without hypertension (Figure 1; Table S4). Among black participants with an undetermined hypertension duration, hypertension duration <10 years, and hypertension duration ≥10 years, multivariable‐adjusted HRs for incident HF were 1.55 (95% CI 0.93, 2.58), 2.13 (95% CI 1.33, 3.42), and 2.96 (95% CI 1.89, 4.64), respectively, compared to participants without hypertension. Results for the association of hypertension duration with incident HF were not statistically different for whites and blacks (likelihood ratio P = .21). Associations of hypertension duration with incident HFpEF, and separately, incident HFrEF/HFmEF were consistent with the results for all HF (Table S5). Wald test p‐values comparing HR estimates for the association of hypertension duration with incident HF (overall), HFrEF/HFmEF, and HFpEF are listed in Table S6. The HR for hypertension duration ≥10 years was significantly higher than the HR for hypertension duration <10 years for blacks (P = .01), but not whites (P = .12).
FIGURE 1.
Multivariable‐adjusted hazard ratios (95% confidence intervals) for the association of hypertension duration and hypertension control with incident heart failure, by race
3.4. BP control and incident HF
Among participants with data on antihypertensive medication use at baseline (n = 24 787), 34.0%, 23.2%, 17.6%, and 25.2% of whites and 16.8%, 20.4%, 20.3%, and 42.5% of blacks had no hypertension, untreated hypertension, controlled BP, and uncontrolled BP, respectively. Among white participants with untreated hypertension, controlled BP, and uncontrolled BP, multivariable‐adjusted HRs for incident HF were 1.52 (95% CI 1.14, 2.04), 1.93 (95% CI 1.44, 2.58), and 2.27 (95% CI 1.74, 2.97), respectively, when compared to those without hypertension (Figure 1; Table S7). Among black participants with untreated hypertension, controlled BP, and uncontrolled BP, multivariable‐adjusted HRs for incident HF were 1.50 (95% CI 0.89, 2.54), 2.01 (95% CI 1.22, 3.29), and 2.93 (95% CI 1.86, 4.63), respectively, when compared to those without hypertension. Results for white and black participants were not statistically different by race (likelihood ratio P = .17). Associations of hypertension control with incident HFpEF and, separately, incident HFrEF/HFmEF were consistent with the results for all HF (Table S8). Wald test p‐values comparing HR estimates for the association of BP control with incident HF (overall), HFrEF/HFmEF, and HFpEF are listed in Table S9.
In a secondary analysis, there was no evidence for differences in the association of hypertension and hypertension duration with incident HF among baseline age strata for black and white participants (Tables S10‐S13). Associations of hypertension, hypertension duration, and BP control with incident HF were similar when we defined hypertension as SBP/DBP ≥140/90 mm Hg or use of antihypertensive medications (Tables S14‐S16).
4. DISCUSSION
In a contemporary population of middle‐aged and older adults with approximately equal proportions of white and black participants, hypertension was associated with an increased risk for incident HF. The strength of associations was not statistically different between black and white participants. As expected, long‐standing hypertension and uncontrolled BP were associated with a higher risk of incident HF. However, even well‐controlled and shorter hypertension duration was associated with a higher risk of HF compared with no hypertension. The results were consistent when examining these associations with incident HFpEF and incident HFrEF/HFmEF. This suggests that the risk for HF after developing hypertension remains elevated regardless of therapeutic BP control and underlines the importance of primordial prevention (ie, preventing the development of risk factors such as hypertension). 27
In the current study, hypertension was associated with an increased risk for incident HF, but was not statistically different between black and white participants. Similar results have been found in prior research examining levels of SBP and HF risk. For example, in a combined analysis of the Cardiovascular Health Study (CHS) and the Health, Aging, and Body Composition study, high SBP was associated with an increased risk for incident HF and there was no effect modification present by race (P > .7). 28 In the current study, there was an increased risk for both incident HFpEF and HFrEF/HFmEF associated with hypertension among both black and white participants. This is consistent with previous research that identified hypertension as a risk factor for HF regardless of subtype. 29 , 30 The current study adds to the findings in prior observational study by providing evidence in a cohort of black and white adults using a recently updated definition of hypertension.
Long‐term levels of high SBP are also associated with an increased risk for CVD outcomes. For example, an analysis of the Framingham Offspring cohort found a greater risk for incident CVD associated with a 1‐SD increase in BP averaged 11‐20 years before baseline enrollment age compared with BP at enrollment age (HR 1.97 [95% CI 1.45, 2.67] vs HR 1.35 [95% CI 1.08, 1.68] for participants with a baseline age of 80 years). 31 In the current study, REGARDS participants with hypertension <10 years and ≥10 years compared to those with no hypertension had an increased risk of incident HF. There was evidence of a higher HR for participants with a longer vs shorter hypertension duration that was statistically significant among black but not white participants. Because hypertension awareness is greater among blacks than whites in this population, 15 whites may have been more likely to underestimate their duration of hypertension. Differences between blacks and whites in the impact of hypertension duration may also be due to important race‐related differences in observed left ventricular responses to hypertension. 32 , 33 Regardless, hypertension develops earlier in life for blacks compared with whites. 13 Preventing and controlling hypertension through diet and lifestyle modification is effective and recommended by US BP management guidelines. 10 , 34 These strategies should be emphasized, particularly among younger black individuals.
Placebo‐controlled trials and meta‐analyses have demonstrated the benefit of hypertension control in reducing CVD risk. 35 A meta‐analysis of 12 hypertension trials found that lowering BP with pharmacological therapy over a 3‐ to 5‐year period reduced the risk of HF. 4 Uncontrolled BP compared with controlled BP and no hypertension increases the risk of HF. In an analysis of the CHS participants ≥65 years old, uncontrolled vs controlled BP was associated with a 39% increased relative risk for incident HF. 5 Among all adults in the current study, uncontrolled and controlled BP compared with no hypertension were associated with an increased risk for incident HF. In the current study, untreated hypertension was not associated with a large risk for incident HF compared with normotension. There are several potential explanations for this apparently contradictory finding. Two‐thirds of participants with untreated hypertension at baseline had relatively mild hypertension with SBP/DBP measured at a single point in time ranging between 130/80 mm Hg and 139/89 mm Hg and would not have been considered hypertensive under pre‐2017 guidelines. Additionally, participants with treated hypertension may have had hypertension for a longer period relative to those with untreated hypertension. Hypertension, even well‐controlled, is a risk factor for multiple cardiovascular diseases especially when present for many years. Because the risk for HF after developing hypertension remains elevated relative to normotensive individuals regardless of therapeutic BP control, preventing development of hypertension is an important strategy to reduce HF risk. 27
There are several strengths to the current study. The REGARDS study enrolled a geographically diverse general population sample with approximately equal proportions of black and white participants, making it ideally suited to examine racial differences related to HF. HF events were adjudicated by pairs of clinicians using standardized adjudication forms that recorded information on left ventricular systolic function, allowing us to determine HF subtypes. Although prior studies, like the Multi‐Ethnic Study of Atherosclerosis, have examined racial differences in incident HF, 18 our study identified a large number of HF events—this facilitated a robust explanatory multivariable regression model and permitted an examination of HFrEF/HFmEF and HFpEF separately among white and black individuals. Finally, few observational studies have examined the associations between hypertension and CVD outcomes using the 2017 American College of Cardiology/American Heart Association definition of hypertension (ie, SBP/DBP ≥130/80 mm Hg). 10 We also acknowledge some limitations. HF hospitalizations were adjudicated by trained clinician investigators, but the completeness and documentation of care for potential HF hospitalizations were variable. Adjudicated HF hospitalizations do not capture HF treated in the outpatient setting. There is no gold standard method of determining HF at baseline and the REGARDS study did not measure biomarkers such as b‐type natriuretic peptide in all participants or conduct echocardiograms as part of study protocols; therefore, medications were used as a proxy to determine HF at baseline. These medications may have been used to treat conditions other than HF such as hypertension, resulting in the exclusion of individuals with hypertension from the study population. We did not capture HF at early ages, and racial differences in HF risk may start earlier in life. 36 Another important limitation is that self‐reported antihypertensive medication use and BP were measured at a single time point. Although not feasible for most observational studies, guidelines recommend measurement of BP on multiple occasions to diagnose hypertension. This may have resulted in misclassification of hypertension, hypertension duration, and BP control. In particular, misclassification may have contributed to the weaker than expected association between untreated hypertension and incident HF. Onset of hypertension is often subtle and difficult to diagnose, and our assessment of hypertension duration relied on participants to report the age at which they were first told they had high BP. Additionally, many participants only reported the age they developed hypertension in a 10‐year category, limiting the granularity of the hypertension duration measurement. Finally, residual or unmeasured confounding may be present after adjustment for covariates.
In conclusion, hypertension was associated with an increased risk for incident HF and HF subtypes in the REGARDS study, and there was no evidence of differences between black and white participants. The highest risk of HF was present among participants with uncontrolled BP and long hypertension duration. However, the high risk of HF among participants with hypertension <10 years or with well‐controlled BP indicates that both primordial prevention and therapeutic management of hypertension are important in reducing HF risk. This is particularly important among black individuals who have a very high prevalence of hypertension, hypertension of long duration, and uncontrolled BP.
AUTHOR CONTRIBUTIONS
Matthew Mefford, study conception and design, data cleaning, analysis and interpretation of the data, drafting of the manuscript, and critical revisions. Parag Goyal, George Howard, Raegan W. Durant, Nancy E. Dunlap, Monika M. Safford and Paul Muntner, Interpretation of the data and critical revision of the manuscript. Emily B. Levitan, study conception and design, review of preliminary data. Interpretation of the data and critical revision of the manuscript.
DISCLOSURES
MTM has no disclosures. PG receives research support from Amgen, Inc unrelated to the topic of this manuscript. GH, RWD, and NED have no disclosures. MMS receives grant support from Amgen, Inc unrelated to the topic of this manuscript. PM receives grant support from Amgen Inc unrelated to the topic of this manuscript. EBL receives grant support from Amgen Inc unrelated to the topic of this manuscript, has consulted for a research project funded by Novartis unrelated to the topic of this manuscript, and has served on Amgen advisory boards unrelated to the topic of this manuscript.
Supporting information
Supplementary Material
Mefford MT, Goyal P, Howard G, et al. The association of hypertension, hypertension duration, and control with incident heart failure in black and white adults. J Clin Hypertens. 2020;22:857–866. 10.1111/jch.13856
Funding information
This research project is supported by cooperative agreement U01 NS041588 co‐funded by the National Institute of Neurological Disorders and Stroke (NINDS) and the National Institute on Aging (NIA), National Institutes of Health, Department of Health and Human Service. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NINDS or the NIA. Representatives of the NINDS were involved in the review of the manuscript but were not directly involved in the collection, management, analysis or interpretation of the data. The authors thank the other investigators, the staff, and the participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at: https://www.uab.edu/soph/regardsstudy/. Additional support was provided by R01 HL80477 from the National Heart Lung and Blood Institute for HF adjudication. Additional support for MTM was received through grant 5T32 HS013852‐15 from the National Institutes of Health. PM received support from the American Heart Association grant 15SFRN2390002.
REFERENCES
- 1. Dunlay SM, Weston SA, Jacobsen SJ, Roger VL. Risk factors for heart failure: a population‐based case‐control study. Am J Med. 2009;122(11):1023‐1028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. He J, Ogden LG, Bazzano LA, Vupputuri S, Loria C, Whelton PK. Risk factors for congestive heart failure in US men and women: NHANES I epidemiologic follow‐up study. Arch Intern Med. 2001;161(7):996‐1002. [DOI] [PubMed] [Google Scholar]
- 3. Kishi S, Teixido‐Tura G, Ning H, et al. Cumulative blood pressure in early adulthood and cardiac dysfunction in middle age: the CARDIA study. J Am Coll Cardiol. 2015;65(25):2679‐2687. [DOI] [PubMed] [Google Scholar]
- 4. Moser M, Hebert PR. Prevention of disease progression, left ventricular hypertrophy and congestive heart failure in hypertension treatment trials. J Am Coll Cardiol. 1996;27(5):1214‐1218. [DOI] [PubMed] [Google Scholar]
- 5. Iyer AS, Ahmed MI, Filippatos GS, et al. Uncontrolled hypertension and increased risk for incident heart failure in older adults with hypertension: findings from a propensity‐matched prospective population study. J Am Soc Hypertens. 2010;4(1):22‐31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Chae CU, Pfeffer MA, Glynn RJ, Mitchell GF, Taylor JO, Hennekens CH. Increased pulse pressure and risk of heart failure in the elderly. JAMA. 1999;281(7):634‐639. [DOI] [PubMed] [Google Scholar]
- 7. Schnabel RB, Rienstra M, Sullivan LM, et al. Risk assessment for incident heart failure in individuals with atrial fibrillation. Eur J Heart Fail. 2013;15(8):843‐849. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Senni M, Tribouilloy CM, Rodeheffer RJ, et al. Congestive heart failure in the community: a study of all incident cases in Olmsted County, Minnesota, in 1991. Circulation. 1998;98(21):2282‐2289. [DOI] [PubMed] [Google Scholar]
- 9. Psaty BM, Kuller LH, Bild D, et al. Methods of assessing prevalent cardiovascular disease in the Cardiovascular Health Study. Ann Epidemiol. 1995;5(4):270‐277. [DOI] [PubMed] [Google Scholar]
- 10. Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines. Hypertension. 2018;71(6):e13‐e115. [DOI] [PubMed] [Google Scholar]
- 11. Howard G, Cushman M, Kissela BM, et al. Traditional risk factors as the underlying cause of racial disparities in stroke: lessons from the half‐full (empty?) glass. Stroke. 2011;42(12):3369‐3375. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Rostand SG, Kirk KA, Rutsky EA, Pate BA. Racial differences in the incidence of treatment for end‐stage renal disease. N Engl J Med. 1982;306(21):1276‐1279. [DOI] [PubMed] [Google Scholar]
- 13. Muntner P, He J, Cutler JA, Wildman RP, Whelton PK. Trends in blood pressure among children and adolescents. JAMA. 2004;291(17):2107‐2113. [DOI] [PubMed] [Google Scholar]
- 14. Hyman DJ, Pavlik VN. Characteristics of patients with uncontrolled hypertension in the United States. N Engl J Med. 2001;345(7):479‐486. [DOI] [PubMed] [Google Scholar]
- 15. Howard G, Prineas R, Moy C, et al. Racial and geographic differences in awareness, treatment, and control of hypertension: the REasons for Geographic And Racial Differences in Stroke study. Stroke. 2006;37(5):1171‐1178. [DOI] [PubMed] [Google Scholar]
- 16. Hertz RP, Unger AN, Cornell JA, Saunders E. Racial disparities in hypertension prevalence, awareness, and management. Arch Intern Med. 2005;165(18):2098‐2104. [DOI] [PubMed] [Google Scholar]
- 17. Rodriguez CJ, Swett K, Agarwal SK, et al. Systolic blood pressure levels among adults with hypertension and incident cardiovascular events: the atherosclerosis risk in communities study. JAMA Intern Med. 2014;174(8):1252‐1261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Bahrami H, Kronmal R, Bluemke DA, et al. Differences in the incidence of congestive heart failure by ethnicity: the multi‐ethnic study of atherosclerosis. Arch Intern Med. 2008;168(19):2138‐2145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Breathett K, Leng I, Foraker RE, et al. Risk factor burden, heart failure, and survival in women of different ethnic groups: insights from the women's health initiative. Circ Heart Fail. 2018;11(5):e004642. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Howard VJ, Cushman M, Pulley LeaVonne, et al. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology. 2005;25(3):135‐143. [DOI] [PubMed] [Google Scholar]
- 21. Yancy CW, Jessup M, Bozkurt B, et al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association task force on practice guidelines. J Am Coll Cardiol. 2013;62(16):e147‐239. [DOI] [PubMed] [Google Scholar]
- 22. Butler J, Anker SD, Packer M. Redefining heart failure with a reduced ejection fraction. JAMA. 2019;322(18):1761. [DOI] [PubMed] [Google Scholar]
- 23. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604‐612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Redmond N, Richman J, Gamboa CM, et al. Perceived stress is associated with incident coronary heart disease and all‐cause mortality in low‐ but not high‐income participants in the Reasons for Geographic And Racial Differences in Stroke study. J Am Heart Assoc. 2013;2(6):e000447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Lunn M, McNeil D. Applying Cox regression to competing risks. Biometrics. 1995;51(2):524‐532. [PubMed] [Google Scholar]
- 26. Lee KJ, Carlin JB. Multiple imputation for missing data: fully conditional specification versus multivariate normal imputation. Am J Epidemiol. 2010;171(5):624‐632. [DOI] [PubMed] [Google Scholar]
- 27. Weintraub WS, Daniels SR, Burke LE, et al. Value of primordial and primary prevention for cardiovascular disease: a policy statement from the American Heart Association. Circulation. 2011;124(8):967‐990. [DOI] [PubMed] [Google Scholar]
- 28. Butler J, Kalogeropoulos AP, Georgiopoulou VV, et al. Systolic blood pressure and incident heart failure in the elderly. The Cardiovascular Health Study and the health, ageing and body composition study. Heart. 2011;97(16):1304‐1311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Eaton CB, Pettinger M, Rossouw J, et al. Risk factors for incident hospitalized heart failure with preserved versus reduced ejection fraction in a multiracial cohort of postmenopausal women. Circ Heart Fail. 2016;9(10):e002883. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Ho JE, Enserro D, Brouwers FP, et al. Predicting heart failure with preserved and reduced ejection fraction: the international collaboration on heart failure subtypes. Circ Heart Fail. 2016;9(6):e003116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Vasan RS, Massaro JM, Wilson PWF, et al. Antecedent blood pressure and risk of cardiovascular disease: the Framingham Heart study. Circulation. 2002;105(1):48‐53. [DOI] [PubMed] [Google Scholar]
- 32. Kizer JR, Arnett DK, Bella JN, et al. Differences in left ventricular structure between black and white hypertensive adults: the Hypertension Genetic Epidemiology Network study. Hypertension. 2004;43(6):1182‐1188. [DOI] [PubMed] [Google Scholar]
- 33. Sharp A, Tapp R, Francis DP, et al. Ethnicity and left ventricular diastolic function in hypertension an ASCOT (Anglo‐Scandinavian Cardiac Outcomes Trial) substudy. J Am Coll Cardiol. 2008;52(12):1015‐1021. [DOI] [PubMed] [Google Scholar]
- 34. James PA, Oparil S, Carter BL, et al. 2014 evidence‐based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311(5):507‐520. [DOI] [PubMed] [Google Scholar]
- 35. Hebert PR, Moser M, Mayer J, Glynn RJ, Hennekens CH. Recent evidence on drug therapy of mild to moderate hypertension and decreased risk of coronary heart disease. Arch Intern Med. 1993;153(5):578‐581. [PubMed] [Google Scholar]
- 36. Bibbins‐Domingo K, Pletcher MJ, Lin F, et al. Racial differences in incident heart failure among young adults. N Engl J Med. 2009;360(12):1179‐1190. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material