Delayed Heart Rate Recovery is Strongly Associated With Early and Late-Stage Prehypertension During Exercise Stress Testing

American Journal of Hypertension 1

ORIGINAL ARTICLE

Delayed Heart Rate Recovery is Strongly Associated With Early and Late-Stage Prehypertension During Exercise Stress TestingEhimen Aneni,1 Lara L. Roberson,1 Sameer Shaharyar,1 Michael J. Blaha,2 Arthur A. Agatston,1,3 Roger S. Blumenthal,2 Romeu S. Meneghelo,4 Raquel D. Conceiçao,4 Khurram Nasir,1,2,3,5,6 and Raul D. Santos4,7

BACKGROUNDHeart rate recovery (HRR) has been shown to predict cardiovascular dis-ease mortality. HRR is delayed in hypertension, but its association with prehypertension (PHT) has not been well studied.

METHODSThe study population consisted of 683 asymptomatic individuals (90% men, aged 47 ± 7.9 years). HRR was de"ned as peak heart rate minus heart rate after a 2-minute rest. PHT was categorized into stage I (systolic blood pressure (SBP) 120–129 mm Hg or diastolic BP (DBP) 80–84 mm Hg) or stage II (SBP 130–139 mm Hg or DBP 85–89 mm Hg). Logistic regres-sion was used to generate odds ratios (ORs) for the relationship between HRR and PHT.

RESULTSThe mean HRR was lower in the PHT groups than in those who were normotensive (60 bpm and 58 bpm in stages I and II PHT vs. 65 bpm

in normal BP; P <0.01). Persons with PHT were more likely to be in the lowest quartile of HRR compared with those with normal BP (adjusted OR, 3.80 and 95% con"dence interval [CI], 1.06, 13.56 for stage II PHT and adjusted OR, 3.01 and 95% CI 1.05, 8.66 for stage I PHT). In a fully adjusted model, HRR was still signi"cantly associated with both stages of PHT.

CONCLUSIONAmong asymptomatic patients undergoing stress testing, delayed HRR was independently associated with early and late stages of PHT. Further studies are needed to determine the usefulness of measuring HRR in the prevention and management of hypertension.

Keywords: blood pressure; cardiac autonomic function; heart rate recovery; hypertension; prehypertension.

doi:10.1093/ajh/hpt173

1Center for Prevention and Wellness Research, Baptist Health Medical Group, Miami Beach, Florida; 2The Johns Hopkins Ciccarone Center for the Prevention of Heart Disease, Baltimore, Maryland; 3Department of Medicine, Herbert Wertheim College of Medicine, Miami, Florida; 4Preventive Medicine Center Hospital Israelita Albert Einstein, Sao Paulo, Brazil; 5Department of Epidemiology, Robert Stempel College of Public Health, Florida International University, Miami, Florida; 6Baptist Cardiac and Vascular Institute, Miami, Florida; 7Heart Institute (InCor) University of São Paulo Medical School Hospital, Sao Paulo, Brazil.

Correspondence: Khurram Nasir ([email protected]).

Initially submitted April 23, 2013; date of "rst revision August 16, 2013; accepted for publication August 16, 2013.

© American Journal of Hypertension, Ltd 2013. All rights reserved. For Permissions, please email: [email protected]

Prehypertension (PHT), de!ned as systolic blood pressure (SBP) between 120 and 139 mm Hg or diastolic blood pres-sure (DBP) between 80 and 89 mm Hg, in individuals who are untreated1 is known to predict future development of hypertension and is associated with increased risk of devel-oping a cardiovascular event.2–4 In the United States, the prevalence of PHT among adults aged >20 years is about 27%,5 and the 4-year risk of progression from PHT to hyper-tension is between 17.6% in those with early PHT (SBP 120–129 mm Hg or DBP 80–84 mm Hg) and 37.3% in those with late PHT (SBP 130–139 mm Hg or DBP 85–89 mm Hg).6 PHT has been associated with several markers of car-diovascular risk including microalbuminuria,5 elevated uric acid, total cholesterol, low-density lipoprotein cholesterol

(LDL-C), fasting plasma glucose, and reduced high-density lipoprotein cholesterol (HDL-C).7–10

Findings from the Framingham Heart Study suggest that autonomic dysfunction, as measured by heart rate variability, might be implicated in the development of hypertension.11 However, the pathogenetic mechanisms underlying PHT development remain unclear. Recently, associations between autonomic dysfunction and PHT have been made.12–15 #ese include overactivity of the sympathetic nervous system and cardiac parasympathetic dysfunction. Heart rate recovery (HRR) a$er exercise is a useful marker of cardiac parasym-pathetic function.16 It is measured as the di%erence between the peak heart rate during exercise and the heart rate at an interval a$er the end of the exercise, usually 30 seconds or 1,

American Journal of Hypertension Advance Access published September 16, 2013 at Baptist H

ospital of Miam

i/Health Science Library on Septem

ber 24, 2013http://ajh.oxfordjournals.org/

Dow

nloaded from

at Baptist Hospital of M

iami/H

ealth Science Library on September 24, 2013

http://ajh.oxfordjournals.org/D

ownloaded from

at Baptist H

ospital of Miam



Dow

nloaded from


iami/H



ownloaded from

at Baptist H

ospital of Miam



Dow

nloaded from


iami/H



ownloaded from

at Baptist H

ospital of Miam



Dow

nloaded from


iami/H



ownloaded from

mailto:[email protected]

http://ajh.oxfordjournals.org/








2 American Journal of Hypertension

Aneni et al.

2, or 4 minutes. Delay in HRR a$er exercise has repeatedly been shown to be associated with all-cause cardiovascular mortality.17–22 However, studies examining the association between HRR a$er exercise and PHT are scant. #e only study we found showed that the prevalence of delayed HRR (<18 bpm in the !rst minute) was about 2 times that in pre-hypertensive patients compared with a control group with optimal BP.23

Here, we examine the association between HRR and severity of PHT. It was our hypotheses that as BP increases, the likelihood of having delayed HRR increases.

METHODS

#is cross-sectional study was conducted in a popula-tion of 683 asymptomatic individuals free of known car-diovascular disease. #ese individuals were executives at several private companies and underwent exercise stress testing as part of an obligatory health evaluation paid for by their employers. #e participants included in the study had no self-reported history of a cardiovascular event, were asymptomatic and devoid of physical !ndings suggestive of cardiovascular disease, and did not have electrocardio-gram !ndings suggestive of coronary disease. #e study was conducted at the Preventive Medical Center of the Hospital Israelita Albert Einstein, Sao Paulo, Brazil, between July 1999 and June 2003 and was approved by the institution’s review board.

Details of the methods have been described elsewhere.24 In brief, patients received exercise testing according to the Bruce protocol. Prior to commencing exercise, participants had their sitting and standing BPs taken with calibrated aneroid devices. Prior to commencing the exercise testing, 3 resting BPs were taken, the !rst being a$er 5 or more min-utes of rest, all on the same day in the sitting position using the method proposed by the American Heart Association.25 #e average of the 3 BPs was used as the baseline BP for this analysis. Appropriate-sized cu%s based on arm cir-cumference were used. A single BP measurement was also obtained immediately before commencing exercise while the participant was standing on the treadmill. Single BP measurements were obtained at exercise peak; at the end of the stress testing stages; and at 2, 4, and 6 minutes post-exercise testing using the same instruments. Weight (in kilograms) and height (in meters) were measured using a standard physician’s weight scale and a stadiometer. Body mass index (BMI) was calculated as body weight (kg) divided by the square of height (m). Percent body fat was measured by bioimpedance (Tanita Corp., Japan). Fasting samples were obtained for total cholesterol, HDL-C, and tri-glycerides. LDL-C was calculated by the Friedwald formula. A pertinent, self-reported clinical history, including current smoking, diabetes mellitus, hypertension, and medication use, was taken from each participant. Presence of cigarette smoking (referred to here as “smoking”) was only consid-ered in those who smoked at least 1 cigarette in the month preceding the examination.

All participants underwent exercise stress testing accord-ing the Bruce protocol.26 Metabolic equivalents (METs) were determined from the peak oxygen consumption at

the last stage of exercise such that 1 MET = 3.5 mL O2/kg mL. Baseline heart rate, peak heart rate during exercise, and heart rate at 2 and 4 minutes immediately a$er exercise were measured. Baseline BP, peak BP during exercise, and BP at 2 and 4 minutes a$er exercise were also measured. For this analysis, HRR was de!ned as peak heart rate minus heart rate at 2 minutes a$er exercise. HRR was categorized accord-ing to quartiles—Q1 to Q4—with Q1 being the slowest rate of HRR and Q4 being the fastest.

In the categorization of blood groups, a combination of resting BP measurements, use of antihypertensive medica-tions, and a self-reported history of hypertension were used. A history of hypertension was de!ned as reporting a diag-nosis of hypertension at the time of presentation or report-ing use of antihypertensive medication. Optimal BP was de!ned as SBP/DBP <120/80 mm Hg, not on treatment for hypertension, and with no known history of hypertension. PHT was categorized into 2 groups. Stage 1, or early PHT, was de!ned as resting (sitting) SBP 120–129 mm Hg or DBP between 80 and 84 mm Hg in persons not on antihyperten-sive medication. Stage 2, or late PHT, was de!ned as SBP between 130 and 139 mm Hg or DBP 85–89 mm Hg in those not on antihypertensive medication. Persons with blood SBP !140 mm Hg or DBP !90 mm Hg were categorized as hypertensive. Similarly, those on treatment for hyperten-sion or those with a history of hypertension regardless of measured BP were categorized as hypertensive. BP recovery was de!ned as 2-minute and 4-minute BP decline from the peak BP.

Statistical analysis

Continuous variables were graphically examined for normality. Continuous variables are presented as mean ± standard deviation or median (interquartile range); categorical variables are presented as number (percent). Analysis of variance was used to determine signi!cance in di%erences across the means of continuous variables including HRR. #e Bonferroni test of multiple compari-sons was used to compare normally distributed continu-ous variables within BP groups. #e Kruskal–Wallis test was used to compare nonnormally distributed continuous variables. #e χ2 test of independence was used to com-pare frequencies across categorical variables. Pearson and Spearman tests of correlation were conducted to deter-mine the association between BP changes during the exer-cise stress test and HRR at 2 minutes, both expressed as continuous variables. We divided HRR into quartiles. To test for the association between BP groupings and quar-tiles of HRR, we used logistic regression adjusting for age, smoking, and BMI in the !rst model and adjusting for age, smoking, BMI, diabetes mellitus, exercise capacity (METs), SBP recovery at 2 minutes, and SBP recovery at 4 minutes in the second model. Odds ratios (ORs) with 95% con!-dence intervals (CIs) were generated for the association between stages of hypertension and HRR using optimal BP and the highest quartile of HRR as references. All analy-ses were performed using Stata statistical so$ware, version 12.0 (StataCorp, College Station, TX). A P value <0.05 was considered statistically signi!cant.


Association of Delayed Heart Rate Recovery With Early and Late-Stage Prehypertension

RESULTS

A total of 683 participants who had complete data were included in this study. Approximately 90% of participants were male, with a mean age of 47 ± 8 years. #e prevalence of PHT was high (about 53%) as was the prevalence of smok-ing (50%). #e mean 10-year Framingham risk (percent) was 9.3 ± 8.0. Details of the baseline characteristics can be found in Table 1.

#ere was a reduction in baseline heart rate with increasing quartiles of HRR such that those in the fastest HRR quartile (Q4) had the slowest mean heart rate at baseline and those in the slowest quartile (Q1) had the fastest baseline heart rate on average (P < 0.001). By contrast, the peak heart rate

increased with increasing rates of HRR (P < 0.001). Baseline SBP and DBP and peak DBP also showed a decline in the mean with increasing HRR quartiles (P < 0.001). However, there was no di%erence in the peak SBP across quartiles of HRR. Table 2 details these !ndings.

Heart rate analysis showed that there was an increase in the baseline heart rate (P <0.002) and a reduction in peak heart rate (P < 0.001) with increasing BP. #ere was pro-gressive reduction in HRR at 2 minutes from optimal BP through the early (stage 1) and late (stage 2) stages of PHT to hypertension. Compared with those with optimal BP, the HRR at 2 minutes was slower in those with hypertension (P < 0.001) and in those with late PHT (stage 2; P = 0.019) but not stage 1 PHT (P = 0.19). Compared with those with

Table 1. Baseline characteristics of participants according to blood pressure category

Characteristic

Optimal

blood pressure

Stage 1

Prehypertensiona

Stage 2

Prehypertensionb Hypertensionc

All

participants

Number of patients (%) 60 (8.8) 272 (39.8) 93 (13.6) 258 (37.8) 683 (100.0)

Male sex (%) 41 (68.3) 238 (87.5) 91 (97.9) 246 (95.4) 616 (90.2)*

Smoker (%) 26 (43.3) 134 (49.3) 44 (47.3) 139 (53.9) 343 (50.2)

Diabetes (%) 0 (0.0) 5 (1.8) 2 (2.2) 11 (4.3) 18 (2.6)

On antihypertensive medication (%) None None None 101 (39.2) 101 (14.79)

Mean age (years) 44.7 ± 7.4 44.8 ± 7.1 47.5 ± 7.2 50.0 ± 7.9 47.1 ± 7.8*

Mean total cholesterol (mg/dL) 200 ± 40 201 ± 39 205 ± 34 214 ± 43 207 ± 40*

Mean low-density lipoprotein cholesterol (mg/dL)

122 ± 35 127 ± 37 129 ± 28 131 ± 35 128 ± 35

Mean high-density lipoprotein cholesterol (mg/dL)

52 ± 15 47 ± 13 44 ± 11 44 ± 14 46 ± 13*

Mean triglycerides (mg/dL) 117 ± 68 146 ± 105 162 ± 99 189 ± 138 162 ± 118*

Mean 10-year Framingham risk (%) 5.7 ± 5.9 6.8 ± 6.4 10.1 ± 8.0 12.6 ± 8.6 9.3 ± 8.0*

Mean body mass index (kg/m2) 24.9 ± 2.3 26.3 ± 3.8 27.6 ± 4.4 28.5 ± 3.8 27.1 ± 4.0*

Mean % body fat 26.7 ± 5.1 27.6 ± 6.7 28.2 ± 5.1 29.3 ± 4.6 28.2 ± 5.7*

Abbreviations: DBP, diastolic blood pressure; HR, heart rate; HRR, heart rate recovery; SBP, systolic blood pressure.aSBP 120–129 mm Hg or DBP 80–84 mm Hg.bSBP 130 mm Hg–140mmHg or DBP 85–89 mm Hg.cSBP ≥140 mm Hg or DBP ≥90 mm Hg.*Statistically significant (P ≤ 0.05) across hypertensive groups.

Table 2. Heart rate and blood pressure characteristics grouped by quartiles of heart rate recovery

Characteristic Q1 Q2 Q3 Q4 P value

Mean baseline heart rate (± SD) 75.6 ± 10.8 75.0 ± 10.9 73.1 ± 11.2 69.3 ± 10.8 <0.001

Mean peak heart rate (± SD) 150.7 ± 16.7 164.2 ± 11.6 167.7 ± 10.2 171.7 ± 11.5 <0.001

Baseline SBP (± SD) 132 ± 16 126 ± 14 126 ± 14 122 ± 13 <0.001

Peak SBP (± SD) 181 ± 27 180 ± 25 176 ± 25 176 ± 26 0.24

Baseline DBP (± SD) 84 ± 9 82 ± 8 82 ± 7 80 ± 8 <0.001

Peak DBP (± SD) 83 ± 14 80 ± 14 79 ± 12 77 ± 10 <0.001

Exaggerated BP response (%) 16.8 13.6 10.8 12.2 0.479

Q1, heart rate recovery (HRR) at 2 minutes in the first (slowest) quartile; Q2, HRR in the second quartile; Q3, HRR in the third quartile; Q4, HRR in the fourth (fastest) quartile. Baseline SBP and DBP are resting BP measured prior to commencing the stress test. Exaggerated BP response is peak exercise SBP ≥210 mm Hg.

Abbreviations: BP, blood pressure; DBP, diastolic blood pressure; SBP, systolic blood pressure; SD, standard deviation.


Aneni et al.

optimal BP, the peak heart rate was signi!cantly reduced in those with hypertension (P = 0.002) but not in the prehy-pertensive groups (P = 0.54 for stage 2 PHT and P =1.00 for stage 1 PHT). Other details can be found in Table 3. Figure 1 shows increasing frequencies of persons with HRR in the !rst quartile (lowest) and decreasing frequencies of those in the fourth quartile (highest) as BP increases from optimal through PHT to hypertension.

Peak SBP and DBP among BP groups were signi!cantly di%erent, with the lowest readings in those with optimal BP and the highest readings in those with hypertension

(P < 0.001). However, there was no signi!cant di%erence between SBP and DBP recovery at 2 or 4 minutes among BP groups. SBP rise, the di%erence between peak SBP and SBP before exercise but standing, showed a marginally signi!cant di%erence among the BP groups (P = 0.059). Other details can be found in Table 4.

HRR was found to be independent of peak SBP, SBP and DBP rise from baseline to peak, DBP decline from peak DBP at 2 and 4 minutes, and SBP decline at 2 minutes (P > 0.05; data not shown). #ere was a weak negative correla-tion between HRR and peak DBP (r = −0.14; P < 0.001),

Table 3. Heart rate characteristics according to blood pressure category

Characteristic

Optimal blood

pressure

Stage 1

Prehypertensiona

Stage 2

Prehypertensionb Hypertensionc P value

Baseline (supine) heart rate 72.2 ± 11.4 71.9 ± 11.1 73.0 ± 10.1 76.6 ± 11.6 <0.001

% with HRR ≤ population median (58 bpm)

38.3 44.1 57.0 63.1 <0.001

Mean HRR at 2 minutes (bpm) 65.0 ± 12.5 60.4 ± 13.2 57.7 ± 20.7 54.0 ± 14.4 <0.001

Mean peak heart rate (standard deviation; bpm)

167.5 ± 10.3 165.7 ± 13.7 163.3 ± 16.8 159.8 ± 16.1 <0.001

Mean HRR/peak heart rate % 38.8 ± 6.9 36.4 ± 8.0 34.5 ± 12.6 33.5 ± 8.1 <0.001

Positive stress test (%) d 15 (26.3) 65 (25.2) 24 (14.6) 61 (37.0) 0.989

Abbreviations: DBP, diastolic blood pressure; HR, heart rate; HRR, heart rate recovery; SBP, systolic blood pressure.aSBP 120 mm Hg–129 mm Hg or DBP 80 mm Hg–84 mm Hg.bSBP 130 mm Hg–140 mm Hg or DBP 85 mm Hg–89 mm Hg.cSBP ≥140 mm Hg or DBP ≥90 mm HgdThe high frequency of positive stress tests was due to the fact that we considered a 0.1-mv change independent of the morphology in order

to increase sensitivity.

Figure 1. Bar Charts demonstrating the increasing prevalence of HRR in the slowest quartile (Q1) and decreasing prevalence of HRR in the fastest quartile (Q4) as blood pressure increases.



while there was a weak positive correlation between HRR and SBP decline at 4 minutes (Spearman rho = 0.10; P = 0.016).

Table 5 shows results for unadjusted and adjusted ORs (95% CI) for the association between delayed HRR and PHT. Compared with those with optimal BP, those with stage 1 PHT were 3 times more likely to be in the slowest recov-ery quartile than in the fastest recovery quartile adjusting for sex, age, smoking, and BMI (adjusted OR, 3.01; 95% CI, 1.05, 8.66). Similarly, those with stage 2 PHT had almost 4 times the likelihood of being in the slowest quartile com-pared with those in the highest quartile a$er adjusting for age, sex, smoking, and BMI (adjusted OR, 3.8; 95% CI,1.06, 13.56). Figure 2 shows a progressive increase in the likeli-hood (OR) of being in the slowest quartile (Q1) of HRR compared with the fastest quartile (Q4) with optimal BP as a reference. In a third model controlling for age, sex, BMI, cigarette smoking, diabetes, exercise capacity, SBP decline at 2 minutes, and SBP decline at 4 minutes, HRR was still strongly associated with both stages of PHT and hyperten-sion (Table 5 and Figure 2).

DISCUSSION

In this study, both early and late PHT were strongly and independently associated with delayed HRR. #is suggests that worsening of parasympathetic function occurs early in the development of hypertension.

It has been established that sympathetic overactivity is involved in the pathogenesis of hypertension.27,28 However, several studies have shown that reduction of cardiac para-sympathetic control also plays a role in the development of hypertension.29–31 Emerging data suggest that parasympa-thetic dysfunction also contributes to the development of PHT and its transition to hypertension.12,32 In particular, one study demonstrated that parasympathetic dysfunction is

present in individuals with normotension and a family his-tory of hypertension.32

It has been demonstrated that parasympathetic activity is the principal determinant of HRR a$er exercise and that persons with delayed HRR have parasympathetic dysfunc-tion.33 Apart from our study, we are aware of only one other study in which HRR has been examined in prehypertensive patients.23 #e authors of that study noted that HRR among those with PHT was signi!cantly lower than in normoten-sive controls (P = 0.02). However, when a HRR <18 bpm in the !rst minute was used as the de!nition of delayed HRR, there was no signi!cant di%erence between those with PHT and those with normotension, probably because of low sta-tistical power. #ey also did not demonstrate delay in HRR in early PHT. We found similar associations between delayed HRR and PHT. However, we also demonstrated an increas-ing likelihood of being in the quartile with the slowest HRR (compared with the fastest recovery quartile) as progression through di%erent stages of hypertension occurs.

#e interplay among BP response, HRR, and PHT is poorly understood. However, abnormal BP response has been asso-ciated with cardiovascular events and mortality34 and with the development of hypertension in those with normal BP.35 In our study, both peak SBP and peak DBP were associated with baseline BP. Exaggerated BP response was greater in late stage (stage 2) PHT and hypertension, but there was no sig-ni!cant di%erence in frequency of exaggerated BP response across quartiles of HRR. Although the declines from peak BP at 2 and 4 minutes were not associated with baseline BP, we included SBP decline at 2 and 4 minutes in a separate model that included those in the initial model (age, sex, smoking, and BMI), exercise capacity (METs), and diabetes and still found signi!cant associations between HRR and PHT.

#e heart rate !ndings in this study are very important since measuring HRR is noninvasive, relatively simple, and does not require specialized training. It can be done without

Table 4. Blood pressure and blood pressure response to stress testing

Characteristic Optimal BP Stage 1 PHT Stage 2 PHT HTN P value

Mean resting SBP 106 ± 6 118 ± 5 129 ± 3 139 ± 14 <0.001

Mean resting DBP 68 ± 4 80 ± 2 81 ± 2 88 ± 8 <0.001

Peak SBP 160 ± 25 170 ± 23 180 ± 21 189 ± 26 <0.001

Peak DBP 72 ± 8 76 ± 10 78 ± 7 87 ± 15 <0.001

SBP rise 50 (30, 60) 50 (40, 60) 50 (40, 70) 60 (40, 70) 0.059

DBP rise 0 (−10, 0) 0 (−10, 0) 0 (−10, 0) 0 (−10, 7.5) 0.366

SBP decline at 2 min post exercise 0 (−2.5, 10) 10 (0, 20) 10 (0, 20) 0 (−10, 20) 0.076

SBP decline at 4 min post exercise 30 (20, 40) 30 (20, 40) 40 (20, 40) 30 (20, 50) 0.218

DBP decline at 2 min post exercise 0 (−10, 0) 0 (−10, 0) 0 (0, 0) 0 (−10, 0) 0.176

DBP decline at 4 min post exercise 0 (−10, 0) 0 (−10, 0) 0 (0, 0) 0 (−10, 10) 0.106

Exaggerated BP response (%) 6.3 6.0 13.3 39.3 <0.001

Peak SBP and peak DBP are presented as mean ± standard deviation. Other variables are presented as median (interquartile range). P values generated from analysis of variance (peak SBP, peak DBP) and Kruskal–Wallis tests (other variables). SBP rise = peak SBP – baseline (standing) SBP; DBP rise = peak DBP – baseline (standing) DBP. SBP decline = peak SBP – SBP at specified time interval; DBP decline = peak DBP – DBP at specified time interval.

Abbreviations: BP, blood pressure; DBP, diastolic blood pressure; HTN, hypertension; PHT, prehypertension; SBP, systolic blood pressure.


Aneni et al.

Tab

le 5

. U

nadj

uste

d an

d A

djus

ted

Odd

s R

atio

s C

ompa

ring

Hea

rt R

ate

Rec

over

y in

Hyp

erte

nsiv

es a

nd P

rehy

pert

ensi

ves

to th

ose

with

Opt

imal

Blo

od P

ress

ures

HR

R q

uar

tile

Q1

Q2

Q3

Un

adju

sted

OR

(95

% C

I)

Ad

just

ed

OR

a (9

5% C

I)

Ad

just

edb

OR

(95

% C

I)

Un

adju

sted

OR

(95

% C

I)

Ad

just

eda

OR

(95

% C

I)

Ad

just

edb

OR

(95

% C

I)

Un

adju

sted

OR

(95

% C

I)

Ad

just

eda

OR

(95

% C

I)

Ad

just

edb

OR

(95

% C

I)

Sta

ge I

pr

ehyp

erte

nsio

nc2.

77

(1.1

2, 6

.8)

3.01

(1

.05,

8.6

6)4.

02

(1.0

7, 1

4.92

)1.

44

(0.7

1,2.

93)

1.39

(0

.62,

3.1

0)1.

02

(0.3

9, 2

.68)

2.53

(1

.16,

5.5

0)4.

41

(1.6

2, 1

2.03

)4.

40

(1.3

3, 1

4.62

)

Sta

ge II

pr

ehyp

erte

nsio

nd3.

89

(1.3

9, 1

0.86

)3.

80

(1.0

6, 1

3.56

)5.

00

(1.0

2, 2

4.42

)2.

40

(1.0

4,5.

52)

2.44

(0

.90,

6.6

0)2.

32

(0.6

8, 7

.95)

2.14

(0

.83,

5.47

)5.

64

(1.6

8, 9

.00)

8.30

(1

.88,

36.

52)

Hyp

erte

nsio

ne8.

6 (3

.47,

21.

5)4.

10

(1.3

7, 1

2.24

) 3

.92

(1.0

2, 1

5.02

)2.

76

(1.3

2, 5

.76)

1.40

(0

.58,

3.3

7)1.

09

(0.3

9, 3

.13)

3.19

(1

.41,

7.2

1)4.

34

(1.4

7, 1

2.82

)3.

95

(1.0

6, 1

4.74

)

Ref

eren

ce g

roup

s ar

e op

timal

blo

od p

ress

ure

and

Q4.

Abb

revi

atio

ns: C

I, co

nfide

nce

inte

rval

; DB

P, d

iast

olic

blo

od p

ress

ure;

HR

R, h

eart

rat

e re

cove

ry; O

R, o

dds

ratio

; SB

P, s

ysto

lic b

lood

pre

ssur

e.a A

djus

ted

for

age,

sex

, bod

y m

ass

inde

x, a

nd s

mok

ing.

b Adj

uste

d fo

r ag

e, s

ex, b

ody

mas

s in

dex,

sm

okin

g, d

iabe

tes

mel

litus

, exe

rcis

e ca

paci

ty, b

lood

pre

ssur

e de

clin

e at

2 m

inut

es, a

nd b

lood

pre

ssur

e de

clin

e at

4 m

inut

es.

c SB

P 1

20 m

m H

g −

129

mm

Hg

or D

BP

80

mm

Hg

− 84

mm

Hg.

d SB

P 1

30 m

m H

g −

140

mm

Hg

or D

BP

85

mm

Hg

− 89

mm

Hg.

e SB

P ≥

140

mm

Hg

or D

BP

≥90

mm

Hg.



the need for the entire stress test protocol and can be included in readings from physical activity monitors. Our !ndings thus open the discussion of the clinical utility of HRR a$er exercise and cardiovascular risk management in the setting of PHT. In particular, the following questions need to be answered:

Hg) or PHT predict progression to hypertension?

therapeutic target?

pursuance of BP goals?Due to its cross-sectional nature, this study cannot be used to determine a causal relationship between PHT and delayed HRR. Although serial baseline BP measurements were taken, these measurements were obtained at a single visit. In addition, ascertainment of covariates such as smok-ing, history of hypertension, or use of antihypertensive or lipid-lowering drugs was through self-report. Consequently, we cannot rule out the possibility of misclassi!cation or mis-classi!cation bias. Our !ndings could have been impacted by white-coat hypertension or masked hypertension; however, we were unable to obtain home BP or ambulatory BP meas-urements to rule this out. We note that our study was con-ducted among asymptomatic participants and a largely male population. We speculate that most of the participants were male because there were more male executives at the time of the study. As such, our !ndings may not be generalizable to women or persons who have had a cardiovascular event.

In this population, stages of PHT are independently associated with delayed HRR, suggesting early autonomic

dysfunction in the development of hypertension. Further studies are needed to determine what role parasympathetic dysfunction plays in the development of PHT, how useful the measure of HRR is in the prediction of progress to clini-cal hypertension, and what the optimal management strate-gies should be in persons with or without PHT and delayed HRR.

DISCLOSURE

#e authors declared no con(ict of interest.

REFERENCES

1. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL, Jr., Jones DW, Materson BJ, Oparil S, Wright JT, Jr., Roccella EJ. #e Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 2003; 289:2560–2572.

2. Vasan RS, Larson MG, Leip EP, Evans JC, O’Donnell CJ, Kannel WB, Levy D. Impact of high-normal blood pressure on the risk of cardiovas-cular disease. New England Journal of Medicine 2001; 345:1291–1297.

3. Qureshi AI, Suri MF, Kirmani JF, Divani AA, Mohammad Y. Is pre-hypertension a risk factor for cardiovascular diseases? Stroke 2005; 36:1859–1863.

4. Franklin SS, Pio JR, Wong ND, Larson MG, Leip EP, Vasan RS, Levy D. Predictors of new-onset diastolic and systolic hypertension: the Framingham Heart Study. Circulation 2005; 111:1121–1127.

5. Ogunniyi MO, Cro$ JB, Greenlund KJ, Giles WH, Mensah GA. Racial/ethnic di%erences in microalbuminuria among adults with

Figure 2. Unadjusted and Adjusted Odds Ratios Comparing Heart Rate Recovery in Hypertensives and Prehypertensives to those with Optimal Blood Pressures. Numbers on each bar indicates odds ratios All Odds Ratios are statistically signi"cant at the P ≤ 0.05 level.


Aneni et al.

prehypertension and hypertension: National Health and Nutrition Examination Survey (NHANES), 1999–2006. Am J Hypertens 2010; 23:859–864.

6. Vasan RS, Larson MG, Leip EP, Kannel WB, Levy D. Assessment of frequency of progression to hypertension in non-hypertensive par-ticipants in the Framingham Heart Study: a cohort study. Lancet 2001; 358:1682–1686.

7. Liang J, Xue Y, Zou C, Zhang T, Song H, Qi L. Serum uric acid and pre-hypertension among Chinese adults. J Hypertens 2009; 27:1761–1765.

8. Nery AB, Mesquita ET, Lugon JR, Kang HC, Miranda VAd, Souza BGd, Andrade JA, Rosa MLG. Prehypertension and cardiovascular risk fac-tors in adults enrolled in a primary care programme. Eur J Cardiovasc Prevent Rehabil 2011; 18:233–239.

9. Syamala S, Li J, Shankar A. Association between serum uric acid and prehypertension among US adults. J Hypertens 2007; 25:1583–1589.

10. Guo X, Zou L, Zhang X, Li J, Zheng L, Sun Z, Hu J, Wong ND, Sun Y. Prehypertension: a meta-analysis of the epidemiology, risk factors, and predictors of progression. Texas Heart Institute journal/from the Texas Heart Institute of St Luke’s Episcopal Hospital, Texas Children’s Hospital 2011; 38:643–652.

11. Singh JP, Larson MG, Tsuji H, Evans JC, O’Donnell CJ, Levy D. Reduced heart rate variability and new-onset hypertension: insights into patho-genesis of hypertension: the Framingham Heart Study. Hypertension 1998; 32:293–297.

12. Pal GK, Adithan C, Amudharaj D, Dutta TK, Pal P, Nandan PG, Nanda N. Assessment of sympathovagal imbalance by spectral analysis of heart rate variability in prehypertensive and hypertensive patients in Indian population. Clin Exp Hypertens 2011; 33:478–483.

13. Pal GK, Chandrasekaran A, Hariharan AP, Dutta TK, Pal P, Nanda N, Venugopal L. Body mass index contributes to sympathovagal imbal-ance in prehypertensives. BMC Cardiovascular disorders 2012; 12:54.

14. Pal GK, Pal P, Lalitha V, Amudharaj D, Nanda N, Dutta TK, Adithan C. Increased vascular tone due to sympathovagal imbalance in nor-motensive and prehypertensive o%spring of hypertensive parents. International angiology: a journal of the International Union of Angiology 2012; 31:340–347.

15. Davis JT, Rao F, Naqshbandi D, Fung MM, Zhang K, Schork AJ, Nievergelt CM, Ziegler MG, O’Connor DT. Autonomic and hemody-namic origins of pre-hypertension: central role of heredity. J Am Coll Cardiol 2012; 59:2206–2216.

16. Arai Y, Saul JP, Albrecht P, Hartley LH, Lilly LS, Cohen RJ, Colucci WS. Modulation of cardiac autonomic activity during and immediately a$er exercise. American Journal of Physiology - Heart and Circulatory Physiology 1989; 256:H132–H141.

17. Cole CR, Blackstone EH, Pashkow FJ, Snader CE, Lauer MS. Heart-rate recovery immediately a$er exercise as a predictor of mortality. !e New England journal of medicine 1999; 341:1351–1357.

18. Cole CR, Foody JM, Blackstone EH, Lauer MS. Heart rate recovery a$er submaximal exercise testing as a predictor of mortality in a cardiovas-cularly healthy cohort. Annals of internal medicine 2000; 132:552–555.

19. Jouven X, Empana JP, Schwartz PJ, Desnos M, Courbon D, Ducimetiere P. Heart-rate pro!le during exercise as a predictor of sudden death. !e New England journal of medicine 2005; 352:1951–1958.

20. Nishime EO, Cole CR, Blackstone EH, Pashkow FJ, Lauer MS. Heart rate recovery and treadmill exercise score as predictors of mortality in

patients referred for exercise ECG. JAMA: the journal of the American Medical Association 2000; 284:1392–1398.

21. Vivekananthan DP, Blackstone EH, Pothier CE, Lauer MS. Heart rate recovery a$er exercise is apredictor of mortality, independent of the angiographic severity of coronary disease. Journal of the American College of Cardiology 2003; 42:831–838.

22. Watanabe J, #amilarasan M, Blackstone EH, #omas JD, Lauer MS. Heart rate recovery immediately a$er treadmill exercise and le$ ven-tricular systolic dysfunction as predictors of mortality: the case of stress echocardiography. Circulation 2001; 104:1911–1916.

23. Erdogan D, Gonul E, Icli A, Yucel H, Arslan A, Akcay S, Ozaydin M. E%ects of normal blood pressure, prehypertension, and hyperten-sion on autonomic nervous system function. International Journal of Cardiology 2011; 151:50–53.

24. Orakzai RH, Orakzai SH, Nasir K, Roguin A, Pimentel I, Carvalho JAM, Meneghello R, Blumenthal RS, Santos RD. Association of increased cardiorespiratory !tness with low risk for clustering of metabolic syn-drome components in asymptomatic men. Archives of Medical Research 2006; 37:522–528.

25. Perlo% D, Grim C, Flack J, Frohlich ED, Hill M, McDonald M, Morgenstern BZ. Human blood pressure determination by sphyg-momanometry. Circulation 1993; 88:2460–2470.

26. Bruce RA, Kusumi F, Hosmer D. Maximal oxygen intake and nomo-graphic assessment of functional aerobic impairment in cardiovascular disease. American heart journal 1973; 85:546–562.

27. Parati G, Esler M. #e human sympathetic nervous system: its rel-evance in hypertension and heart failure. European heart journal 2012; 33:1058–1066.

28. Julius S. Autonomic nervous dysfunction in essential hypertension. Diabetes care 1991; 14:249–259.

29. Langewitz W, Ruddel H, Schachinger H. Reduced parasympathetic cardiac control in patients with hypertension at rest and under mental stress. American heart journal 1994; 127:122–128.

30. Korner PI, Shaw J, Uther JB, West MJ, Mcritchie RJ, Richards JG. Autonomic and non-autonomic circulatory components in essential hypertension in man. Circulation 1973; 48:107–117.

31. Neumann SA, Jennings JR, Muldoon MF, Manuck SB. White-coat hypertension and autonomic nervous system dysregulation. American journal of hypertension 2005; 18:584–588.

32. Wu JS, Lu FH, Yang YC, Lin TS, Chen JJ, Wu CH, Huang YH, Chang CJ. Epidemiological study on the e%ect of pre-hypertension and family his-tory of hypertension on cardiac autonomic function. J Am Coll Cardiol 2008; 51:1896–1901.

33. Imai K, Sato H, Hori M, Kusuoka H, Ozaki H, Yokoyama H, Takeda H, Inoue M, Kamada T. Vagally mediated heart rate recovery a$er exercise is accelerated in athletes but blunted in patients with chronic heart fail-ure. J Am Coll Cardiol 1994; 24:1529–1535.

34. Schultz MG, Otahal P, Cleland VJ, Blizzard L, Marwick TH, Sharman JE. Exercise-induced hypertension, cardiovascular events, and mortal-ity in patients undergoing exercise stress testing: a systematic review and meta-analysis. Am J Hypertens 2013; 26:357–366.

35. Singh JP, Larson MG, Manolio TA, O’Donnell CJ, Lauer M, Evans JC, Levy D. Blood pressure response during treadmill testing as a risk factor for new-onset hypertension. #e Framingham heart study. Circulation 1999; 99:1831–1836.

Delayed Heart Rate Recovery is Strongly Associated With Early and Late-Stage Prehypertension During Exercise Stress Testing

Documents