- Email: [email protected]
Exaggerated exercise blood pressure response and risk of stroke in patients referred for stress testing
European Journal of Internal Medicine 25 (2014) 533–537
Contents lists available at ScienceDirect
European Journal of Internal Medicine journal homepage: www.elsevier.com/locate/ejim
Original Article
Exaggerated exercise blood pressure response and risk of stroke in patients referred for stress testing María del Carmen Bouzas-Mosquera a, Alberto Bouzas-Mosquera b,⁎, Jesús Peteiro b, Francisco J. Broullón c, Nemesio Álvarez-García b, Alfonso Castro-Beiras b a b c
Faculty of Health Sciences, European University, Madrid, Spain Department of Cardiology, Hospital Universitario A Coruña, A Coruña, Spain Department of Health Information Technology, Hospital Universitario A Coruña, A Coruña, Spain
a r t i c l e
i n f o
Article history: Received 2 March 2014 Received in revised form 22 May 2014 Accepted 27 May 2014 Available online 11 June 2014 Keywords: Exercise test Blood pressure Stroke Prognosis
a b s t r a c t Background/objectives: There is some evidence to suggest that exercise systolic blood pressure (SBP) may be associated with future risk of stroke in subjects without a history of coronary artery disease (CAD). However, the value of an exaggerated exercise SBP response (EESBPR) for predicting stroke in patients referred for stress testing for clinical reasons has not been investigated. Methods: We evaluated a community-based sample of 10,047 patients with known or suspected CAD who underwent treadmill exercise echocardiography. An EESBPR was defined as a peak exercise SBP of N 220 mmHg. The ratio of the increase in SBP during exercise to exercise workload (ΔSBPeEW) was also estimated. The endpoints were stroke of any type, ischemic stroke and hemorrhagic stroke. Median follow-up was 3.5 years. Results: Annualized rates of stroke of any type, ischemic stroke and hemorrhagic stroke were 0.6% (95% CI 0.53– 0.67), 0.49% (95% CI 0.42–0.56) and 0.12% (95% CI 0.09–0.15) in patients without EESBPR vs. 0.69% (95% CI 0.37– 1), 0.49% (95% CI 0.23–0.76) and 0.19% (95% CI 0.02–0.35) in those with EESBPR (p = 0.68, 0.90 and 0.39, respectively). Similarly, there was no significant univariate association between ΔSBPeEW and the occurrence of any endpoint. In multivariate analysis, hypertension, male sex, age, diabetes mellitus and resting SBP remained predictors of stroke of any type. EESBPR and ΔSBPeEW were not predictors of any of the endpoints evaluated. Conclusion: We did not observe any significant association between exercise SBP and the future occurrence of stroke in patients with known or suspected CAD referred for exercise echocardiography. © 2014 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
1. Introduction Although resting hypertension is a well-established risk factor for cardiovascular events, the prognostic value of exercise hypertension remains controversial. While some studies reported an association of an exaggerated exercise systolic blood pressure response (EESBPR) with future hypertension [1–7] and cardiovascular events [4,8–10] in healthy subjects, others evaluating patients with known or suspected coronary artery disease (CAD) disagree [11–14]. There is also some evidence to suggest that an EESBPR may be associated with future risk of stroke in middle-aged men without a history of CAD [15]. However, this issue has not been investigated in patients with known or suspected CAD, who represent the majority of subjects referred for noninvasive stress testing in clinical practice.
⁎ Corresponding author at: Department of Cardiology, Hospital Universitario A Coruña, As Xubias 84, 15006 A Coruña, Spain. Tel.: +34 981178184; fax: +34 981178258. E-mail address: [email protected] (A. Bouzas-Mosquera).
Our aim was to assess the value of an EESBPR for predicting stroke in patients with known or suspected CAD referred for exercise echocardiography.
2. Methods 2.1. Patients A total of 13,328 adult patients with known or suspected CAD who underwent treadmill exercise echocardiography at our hospital were initially considered for inclusion. To ensure comprehensive follow-up data, 1113 patients who did not belong to the geographic area of reference of our hospital were excluded. Given that beta-blocker therapy may blunt an EESBPR [16], 1218 patients who received betablockers within 48 h before the tests were also excluded; lastly, because of the well-known association of an exercise hypotensive or flat blood pressure response with poorer outcome [17], 950 patients in whom systolic blood pressure (SBP) failed to increase with exercise above the baseline values were also excluded. The remaining 10,047 patients constituted
http://dx.doi.org/10.1016/j.ejim.2014.05.013 0953-6205/© 2014 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
534
M.C. Bouzas-Mosquera et al. / European Journal of Internal Medicine 25 (2014) 533–537
Table 1 Baseline characteristics of the patients.
Male, n (%) Age, years Current smokers, n (%) Diabetes, n (%) Hypertension, n (%) Hypercholesterolemia, n (%) Family history of CAD, n (%) History of CAD, n (%) Prior myocardial infarction, n (%) ≤30 days before EE, n (%) N30 days before EE, n (%) Prior coronary revascularization, n (%) PCI, n (%) CABG, n (%) Chest pain, n (%) Typical angina, n (%) Atypical/probable angina, n (%) Nonanginal chest pain, n (%) Dyspnea, n (%) Uninterpretable resting ECG, n (%) Left bundle branch block, n (%) Atrial fibrillation, n (%) ACE-i/ARB, n (%) Nitrates, n (%) Calcium channel blockers, n (%)
All patients (n = 10047)
No EESBPR (n = 9645)
EESBPR (n = 402)
p
6105 (60.8) 62.0 ± 12.2 2504 (24.9) 1810 (18) 5227 (52) 4759 (47.4) 1305 (13) 2519 (25.1) 1900 (18.9) 850 (8.5) 1096 (10.9) 1534 (15.3) 1046 (10.4) 795 (7.9)
5861 (60.8) 62.0 ± 12.1 2408 (25) 1722 (17.9) 4935 (51.2) 4560 (47.3) 1237 (12.8) 2447 (25.4) 1846 (19.1) 829 (8.6) 1063 (11) 1486 (15.4) 1014 (10.5) 770 (8)
244 (60.7) 61.7 ± 10.3 96 (23.9) 88 (21.9) 291 (72.4) 199 (49.5) 68 (16.9) 72 (17.9) 54 (13.4) 21 (5.2) 33 (8.2) 48 (11.9) 32 (8) 25 (6.2)
0.97 0.51 0.62 0.04 b0.001 0.38 0.02 0.001 0.004 0.02 0.08 0.06 0.10 0.20
778 (7.7) 4251 (42.3) 2302 (22.9) 599 (6) 2354 (23.4) 669 (6.7) 453 (4.5) 3381 (33.7) 2207 (22) 900 (9)
758 (7.9) 4049 (42) 2229 (23.1) 583 (6) 2256 (23.4) 643 (6.7) 448 (4.6) 3235 (33.5) 2133 (22.1) 861 (8.9)
20 (5) 202 (50.2) 73 (18.2) 16 (4) 98 (24.4) 26 (6.5) 5 (1.2) 146 (36.3) 75 (18.7) 39 (9.7)
0.03 0.001 0.02 0.09 0.65 0.88 0.001 0.24 0.11 0.59
ACE-i denotes angiotensin converting enzyme inhibitors; ARB, angiotensin receptor blockers; CABG, coronary artery bypass grafting; CAD, coronary artery disease; ECG, electrocardiogram; EE, exercise echocardiography; EESBPR, exaggerated exercise systolic blood pressure response; PCI, percutaneous coronary intervention.
our sample. This study was approved by the Clinical Investigation Ethics Committee of Galicia. 2.2. Data collection Demographics, clinical data and stress testing results were entered in a dedicated database at the time of testing. Hypertension, hypercholesterolemia, and diabetes mellitus were defined on the basis of history and antihypertensive, lipid-lowering, or antidiabetic treatment, respectively. Patients referred for evaluation of chest pain were classified as having typical angina, atypical/probable angina or nonischemic chest pain [18]. A history of CAD was defined as previous myocardial
infarction, previous coronary revascularization or prior angiographic documentation of any ≥50% coronary stenosis. Resting ECG was considered uninterpretable in the presence of left bundle branch block, paced rhythm, left ventricular hypertrophy with strain, treatment with digoxin, preexcitation, or other repolarization abnormalities. All patients underwent symptom-limited treadmill exercise echocardiography. Protocols employed included the Bruce protocol in 8676 patients (86.4%), the modified Bruce protocol in 464 patients (4.6%), the Naughton protocol in 104 patients (1%), and other protocols in 803 patients (8%). Heart rate, blood pressure, and a 12-lead electrocardiogram were recorded at rest and at each stage of the protocol. Peak SBP was defined as the maximum value of SBP obtained during exercise.
Table 2 Data provided by exercise stress testing.
Systolic blood pressure, mm Hg Rest Peak Heart rate, beats/min Rest Peak RPP, ×103 mm Hg beats/min Rest Peak % of MAPHR Submaximal test, n (%) Exercise-induced chest pain, n (%) Ischemic ECG changes, n (%) Exercise workload, METs Left ventricular ejection fraction, % Rest Peak Resting wall motion abnormalities, n (%) Echocardiographic ischemia, n (%) Wall motion score index Rest Peak
All patients (n = 10047)
No EESBPR (n = 9645)
EESBPR (n = 402)
p
133 ± 19 171 ± 30
132 ± 18 168 ± 25
161 ± 23 243 ± 56
b0.001 b0.001
79 ± 15 149 ± 20
79 ± 14.8 149 ± 21
80 ± 14 148 ± 18
0.19 0.22
10.5 ± 2.6 25.6 ± 5.8 94.1 ± 11.5 1831 (18.2) 1368 (13.6) 1478 (14.7) 9.5 ± 3.1
10.4 ± 2.5 25.1 ± 5.3 94.1 ± 11.5 1747 (18.1) 1335 (13.8) 1413 (14.7) 9.5 ± 3.2
12.9 ± 3.1 34.7 ± 7.7 93.1 ± 10.3 84 (20.9) 33 (8.2) 65 (16.2) 8.3 ± 2.5
b0.001 b0.001 0.08 0.15 0.001 0.4 b0.001
58.4 ± 8.2 63.2 ± 12.3 2055 (20.5) 2744 (27.3)
58.3 ± 8.2 63.1 ± 12.3 1998 (20.7) 2653 (27.5)
60.3 ± 6.7 65.1 ± 10.5 57 (14.2) 91 (22.6)
b0.001 0.002 0.001 0.03
1.09 ± 0.24 1.18 ± 0.32
1.09 ± 0.23 1.19 ± 0.24
1.05 ± 0.18 1.14 ± 0.27
0.001 0.007
ECG denotes electrocardiographic; EESBPR, exaggerated exercise systolic blood pressure response; MAPHR, maximum age-predicted heart rate; METs, metabolic equivalents; RPP, ratepressure product.
M.C. Bouzas-Mosquera et al. / European Journal of Internal Medicine 25 (2014) 533–537 Table 3 Predictors of exaggerated exercise systolic blood pressure response in patients with known or suspected coronary artery disease.
Hypertension Typical angina Atrial fibrillation Resting systolic blood pressure Resting LVEF
Adjusted odds ratio (95% CI)
p
1.44 (1.13–1.84) 0.56 (0.35–0.92) 0.24 (0.09–0.61) 1.07 (1.06–1.08) 1.04 (1.02–1.05)
0.003 0.02 0.003 b0.001 b0.001
CI denotes confidence interval; LVEF, left ventricular ejection fraction.
An EESBPR was defined as a peak SBP N220 mm Hg [13,14], which is equivalent to the 95th percentile of peak SBP within the population under study. The increase in SBP during exercise (ΔSBPe) was calculated as the difference between peak and resting SBP. Since the degree of increase in SBP during exercise may depend on exercise capacity, we also estimated the ratio of the ΔSBPe to exercise workload in METs (ΔSBPeEW). Significant ischemic ST-segment changes during the tests were defined as the development of ST-segment deviation of ≥ 1 mm which was horizontal or sloping away from the isoelectric line 80 ms after the J point. A submaximal test was defined as failure to achieve 85% of maximum age-predicted heart rate. Echocardiographic images were acquired and analyzed as previously described [19]. Left ventricular ejection fraction and wall motion score index were estimated at rest and at peak exercise. Echocardiographic myocardial ischemia was defined as the appearance of new or worsening wall motion abnormalities with exercise, except worsening from akinesia to dyskinesia and isolated hypokinesia of the inferobasal segment. 2.3. Follow-up and end-points Follow-up data were retrieved from electronic medical records and discharge code databases through the Department of Health Information Technology of our institution. Data sources were linked using unique identifiers. End points were the occurrence of ischemic stroke (International Classification of Diseases, Ninth Revision [ICD-9] codes 433.x1 and 434.x1), hemorrhagic stroke (ICD-9 codes 430–432) or stroke of any type. Patients were censored at the last date in which vital status could be ascertained. Median follow-up was 3.5 years (interquartile range 0.2–8.0 years). 2.4. Statistical analysis Categorical variables were reported as percentages and comparison between groups based on the chi-square test. Continuous variables were reported as mean ± standard deviation and differences were assessed with the unpaired t-test or Mann–Whitney U test as
535
appropriate. Logistic regression analysis was performed to determine the predictors of EESBPR, and linear regression analysis was employed to assess the predictors of ΔSBPe and ΔSBPeEW. Annualized event rates were calculated by dividing the number of events by the total number of person-years at risk. Survival free of the end point of interest was estimated by the Kaplan–Meier method, and survival curves were compared with the log-rank test. Univariable and multivariable associations of clinical and exercise variables with stroke were assessed with Cox's proportional hazard models. Hazard ratios (HRs) with 95% confidence intervals (CI) were estimated. Different multivariate models were constructed forcing the inclusion of EESPBR, ΔSBPe and ΔSBPeEW. Interactions with sex, antihypertensive drug use, history of CAD or exercise echocardiography results were tested. A sensitivity analysis was also performed after exclusion of patients with either a history of CAD, history of atrial fibrillation, antihypertensive medication use, protocols other than the Bruce protocol, abnormal exercise echocardiography (i.e., resting or exercise-induced wall motion abnormalities) or reasons for stopping the tests other than physical exhaustion. Statistical analyses were performed using SPSS software, version 15.0 (SPSS, Chicago, IL). 3. Results Mean age was 62 ± 12 years, and 6105 patients (60.8%) were male. Demographic information, cardiac risk factors, history of cardiac events, chest pain symptoms, electrocardiographic features and medication use are listed in Table 1. A total of 402 patients (4%) developed an EESBPR during the tests. There were no significant differences in age or gender between patients with and without EESBPR, but the former had a higher prevalence of diabetes and hypertension, and a lower likelihood of prior history of CAD. There were no significant differences regarding antihypertensive medications between both groups (Table 1). Data obtained during the tests are shown in Table 2. Patients developing EESBPR had higher resting SBP, lower estimated functional capacity, and higher resting and peak left ventricular ejection fraction. These patients also developed less frequently exercise-induced chest pain. There were no significant differences between both groups regarding resting heart rate or the likelihood of submaximal tests (Table 2). Predictors of EESBPR are shown in Table 3. ΔSBPe was significantly correlated with age (R = − 0.138, p b 0.001), resting left ventricular ejection fraction (R = 0.119, p b 0.001), peak left ventricular ejection fraction (0.155, p b 0.001), the increase in left ventricular ejection fraction from rest to peak exercise (R = 0.120, p b 0.001), and estimated exercise capacity (R = 0.240, p b 0.001). In linear regression analysis, age (p b 0.001), sex (p b 0.001), hypertension (p b 0.001), estimated functional capacity (p b 0.001), resting left ventricular ejection fraction (p b 0.001) and the increase in left ventricular ejection fraction from rest to peak exercise (p b 0.001) were predictors of ΔSBPe, while age
Fig. 1. Event curves for (A) stroke of any type, (B) ischemic stroke and (C) hemorrhagic stroke in patients with and without exaggerated exercise systolic blood pressure response (EESBPR).
536
M.C. Bouzas-Mosquera et al. / European Journal of Internal Medicine 25 (2014) 533–537
Table 4 Multivariate predictors of stroke of any type, ischemic stroke and hemorrhagic stroke in patients with known or suspected coronary artery disease referred for stress testing. Adjusted hazard ratio (95% confidence interval) Any stroke Age Sex Hypertension Diabetes mellitus Hypercholesterolemia Smoking Family history of CAD Atrial fibrillation History of CAD Antihypertensive drugs Rest SBP (per 10 mm Hg) Rest LVEF EESBPR
1.04 (1.03–1.06), p 1.57 (1.19–2.06), p 1.56 (1.19–2.04), p 1.35 (1.02–1.78), p 0.99 (0.77–1.27), p 1.09 (0.80–1.48), p 0.88 (0.59–1.32), p 0.91 (0.53–1.54), p 1.10 (0.85–1.42), p 0.95 (0.74–1.23), p 1.07 (1.01–1.14), p 0.99 (0.98–1.01), p 0.88 (0.53–1.47), p
Ischemic stroke b 0.001 = 0.001 = 0.001 = 0.037 = 0.94 = 0.58 = 0.54 = 0.72 = 0.47 = 0.71 = 0.03 = 0.29 = 0.62
1.05 (1.03–1.06), p 1.68 (1.24–2.27), p 1.62 (1.20–2.18), p 1.49 (1.10–2.02), p 0.99 (0.75–1.30), p 1.13 (0.81–1.59), p 1.01 (0.66–1.55), p 0.85 (0.47–1.54), p 1.01 (0.76–1.34), p 1.00 (0.76–1.32), p 1.05 (0.98–1.13), p 0.99 (0.98–1.01), p 0.80 (0.44–1.45), p
Hemorrhagic stroke b 0.001 = 0.001 = 0.002 = 0.01 = 0.93 = 0.46 = 0.95 = 0.61 = 0.97 = 0.99 = 0.18 = 0.20 = 0.46
1.03 (1.00–1.06), p 1.13 (0.64–2.00), p 1.27 (0.71–2.26), p 1.02 (0.53–1.99), p 0.95 (0.55–1.65), p 0.91 (0.45–1.84), p 0.40 (0.12–1.29), p 1.31 (0.47–3.71), p 1.50 (0.87–2.60), p 0.86 (0.49–1.48), p 1.09 (0.95–1.24), p 1.00 (0.97–1.03), p 1.25 (0.46–3.44), p
= = = = = = = = = = = = =
0.05 0.68 0.42 0.94 0.87 0.79 0.12 0.61 0.14 0.58 0.24 0.92 0.66
CAD denotes coronary artery disease; EESBPR, exaggerated exercise systolic blood pressure response; ΔSBPe, increase in systolic blood pressure during exercise; ΔSBPeEW, ratio of increase in systolic blood pressure during exercise to exercise workload; LVEF, left ventricular ejection fraction; SBP, systolic blood pressure.
(p b 0.001), hypertension (p b 0.001), resting left ventricular ejection fraction (p = 0.026) and resting heart rate (p b 0.001) were predictors of ΔSBPeEW. During follow-up there were 232 ischemic strokes and 61 hemorrhagic strokes, and 285 patients had at least one stroke of any type. Annualized rates of stroke of any type, ischemic stroke and hemorrhagic stroke were 0.6% (95% CI 0.53–0.67), 0.49% (95% CI 0.42–0.56) and 0.12% (95% CI 0.09–0.15) in patients without EESBPR vs. 0.69% (95% CI 0.37–1), 0.49% (95% CI 0.23–0.76) and 0.19% (95% CI 0.02–0.35) in those with EESBPR (p = 0.68, 0.90 and 0.39, respectively) (Fig. 1). Similarly, there was no significant univariate association between ΔSBPeEW and the occurrence of any stroke (HR 1.02, 95% CI 0.99–1.05, p = 0.25), ischemic stroke (HR 1.00, 95% CI 0.97–1.05, p = 0.8) or hemorrhagic stroke (HR 1.05, 95% CI 0.99–1.10, p = 0.08). In Cox regression analysis, a history of hypertension at rest and resting SBP remained predictors of stroke of any type, along with gender, age and diabetes mellitus (Table 4). EESBPR, ΔSBPe or ΔSBPeEW were not predictors of any endpoint in their respective models after multivariate adjustment (Tables 4 and 5). There were no significant interactions with sex, antihypertensive drug use, history of atrial fibrillation or history of known CAD in any of these models. The lack of association between EESBPR, ΔSBPe or ΔSBPeEW with any of the endpoints in multivariate analyses remained consistent even after exclusion of patients with either a history of CAD, history of atrial fibrillation, antihypertensive medication use, protocols other than the Bruce protocol, abnormal exercise echocardiography results or reasons for stopping the tests other than physical exhaustion (Table 6). 4. Discussion In our study, we did not find any significant association between the degree of increase in SBP during exercise and the subsequent risk of stroke in a series of patients with known or suspected CAD. The prognostic value of a hypertensive response with exercise remains controversial. Some studies performed in asymptomatic healthy subjects found that exercise blood pressure response was associated with higher risk of cardiovascular events [4,8–10], while others evaluating patients with known or suspected CAD reported an association with improved outcome [11,12,14]. However, there is limited information regarding the value of an EESBPR for predicting stroke. Kurl and colleagues [15] found that a higher increase in SBP during exercise with bicycle ergometer was associated with higher risk of subsequent stroke in middle-aged men. Although this contrasts with our results, there are important differences in the characteristics of the patients, in the definition of the variables evaluated and in the technique employed between both studies. In this regard, patients with a history of CAD were excluded in Kurl and colleagues' study; in addition, there may be significant
differences in the magnitude of exercise-induced increase in SBP between bicycle ergometer and treadmill stress testing [20]. Exercise SBP depends on the balance between changes in cardiac output and systemic vascular resistance during exercise. In normal conditions, these two parameters change in opposite directions, with cardiac output increasing and systemic vascular resistance decreasing with exercise. An EESBPR may result from an increase in cardiac output and/or a blunted decrease in systemic vascular resistance due to pathological changes in arteries and arterioles. In this regard, our study showed a significant correlation between the increase in SBP during exercise and the increase in left ventricular ejection fraction from rest to peak exercise. While an exercise-induced increase in cardiac output may be favorable from a prognostic point of view, a blunted decrease in systemic vascular resistance is likely detrimental. These differences in the prognostic effect of both determinants of SBP response during exercise might explain, at least in part, the inconsistencies between different studies regarding the prognostic significance of an EESBPR. Other possible explanations include differences in the characteristics of the populations and in the definition of EESBPR. Furthermore, there is no consensus on normal values for exercise SBP response. The threshold employed in our study for defining an EESBPR was higher than those used in most previous reports; this threshold was selected because the prognostic value of moderate increases in blood pressure during exercise has been well established [21,22]. Also, in contrast with previous studies [11], we did not employ different cut-offs for defining EESBPR in men and women; nonetheless, the prevalence and prognostic value of EESBPR did not differ between both sexes using this definition. Finally, the degree of increase in SBP during exercise was correlated with estimated exercise capacity, which in turn may be a predictor of outcome, hence the rationale for controlling for exercise workload. Our study has the limitations inherent to an observational, retrospective, single-center study. These results only apply to patients undergoing treadmill exercise testing for clinical reasons; although they represent the vast majority of patients referred for noninvasive testing in clinical practice, these results cannot be extrapolated to other subgroups of subjects, such as healthy asymptomatic individuals. Important variables, such as body weight [23,24], anticoagulant therapy or left Table 5 Prognostic value of the increase in systolic blood pressure during exercise (ΔSBPe) and the ratio of ΔSBPe to exercise workload (ΔSBPeEW) for predicting stroke. Adjusted hazard ratio (95% confidence interval)a
Stroke of any type Ischemic stroke Hemorrhagic stroke a
ΔSBPe
ΔSBPeEW
1.00 (0.99–1.00), p = 0.25 1.00 (0.99–1.00), p = 0.18 1.00 (0.99–1.01), p = 0.90
1.01 (0.97–1.04), p = 0.73 0.99 (0.95–1.03), p = 0.62 1.04 (0.99–1.09), p = 0.15
Adjusted by all variables listed in Table 4 except EESBPR.
M.C. Bouzas-Mosquera et al. / European Journal of Internal Medicine 25 (2014) 533–537
537
Table 6 Prognostic value of exercise systolic blood pressure for predicting stroke after exclusion of patients with either a history of CAD, history of atrial fibrillation, antihypertensive medication use, protocols other than the Bruce protocol, abnormal exercise echocardiograms or reasons for stopping the tests other than physical exhaustion. Adjusted hazard ratio (95% confidence interval)a
Stroke of any type Ischemic stroke Hemorrhagic stroke
EESBPR
ΔSBPe
ΔSBPeEW
1.10 (0.37–3.23), p = 0.87 1.01 (0.29–3.48), p = 0.99 1.60 (0.17–15.2), p = 0.68
0.99 (0.98–1.00), p = 0.17 0.99 (0.98–1.01), p = 0.20 0.99 (0.96–1.02), p = 0.49
0.93 (0.83–1.05), p = 0.25 0.91 (0.80–1.05), p = 0.19 0.97 (0.75–1.26), p = 0.83
EESBPR denotes exaggerated exercise systolic blood pressure response; ΔSBPe, increase in systolic blood pressure during exercise; ΔSBPeEW, ratio of increase in systolic blood pressure during exercise to exercise workload. a Adjusted by age, sex, diabetes mellitus, hypertension, hypercholesterolemia, smoking, family history of coronary artery disease, resting systolic blood pressure and resting left ventricular ejection fraction.
atrial size, were not available for analysis. However, exclusion of patients with a history of atrial fibrillation did not change the results. Notwithstanding the exclusion of patients taking betablockers within 48 h before the tests, we cannot rule out residual beta-blocking effect in patients who had previously been prescribed betablockers and were instructed to withdraw them temporarily before testing. The number of hemorrhagic strokes during follow-up was relatively low; thus, the possibility of a type II error cannot be excluded and the results should be interpreted with caution; whether longer follow-up studies with higher statistical power might lead to different conclusions is unknown. Lastly, we did not evaluate the prognostic value of exercise diastolic blood pressure because only SBP can be measured reliably during exercise [25], but even the latter may be difficult to assess during treadmill stress testing. However, difficulties in real-world clinical practice should also be reflected in the prognostic value of testing. 5. Conclusions In our series of patients with known or suspected CAD referred for exercise echocardiography, we did not find any significant association between exercise SBP and stroke during follow-up. Thus, our study puts into question the clinical usefulness of blood pressure measurements during treadmill exercise testing for predicting the future occurrence of stroke in patients with known or suspected CAD, although further studies with higher statistical power are warranted. Conflict of interests The authors state that they have no conflicts of interest. References [1] Manolio TA, Burke GL, Savage PJ, Sidney S, Gardin JM, Oberman A. Exercise blood pressure response and 5-year risk of elevated blood pressure in a cohort of young adults: the CARDIA study. Am J Hypertens 1994;7:234–41. [2] Miyai N, Arita M, Morioka I, Miyashita K, Nishio I, Takeda S. Exercise BP response in subjects with high-normal BP: exaggerated blood pressure response to exercise and risk of future hypertension in subjects with high-normal blood pressure. J Am Coll Cardiol 2000;36:1626–31. [3] Singh JP, Larson MG, Manolio TA, O'Donnell CJ, Lauer M, Evans JC, et al. Blood pressure response during treadmill testing as a risk factor for new-onset hypertension. The Framingham heart study. Circulation 1999;99:1831–6. [4] Allison TG, Cordeiro MA, Miller TD, Daida H, Squires RW, Gau GT. Prognostic significance of exercise-induced systemic hypertension in healthy subjects. Am J Cardiol 1999;83:371–5. [5] Sharabi Y, Ben-Cnaan R, Hanin A, Martonovitch G, Grossman E. The significance of hypertensive response to exercise as a predictor of hypertension and cardiovascular disease. J Hum Hypertens 2001;15:353–6. [6] Majahalme S, Turjanmaa V, Tuomisto M, Lu H, Uusitalo A. Blood pressure responses to exercise as predictors of blood pressure level after 5 years. Am J Hypertens 1997;10:106–16.
[7] Dlin RA, Hanne N, Silverberg DS, Bar-Or O. Follow-up of normotensive men with exaggerated blood pressure response to exercise. Am Heart J 1983;106:316–20. [8] Laukkanen JA, Kurl S, Rauramaa R, Lakka TA, Venalainen JM, Salonen JT. Systolic blood pressure response to exercise testing is related to the risk of acute myocardial infarction in middle-aged men. Eur J Cardiovasc Prev Rehabil 2006;13:421–8. [9] Erikssen G, Bodegard J, Bjornholt JV, Liestol K, Thelle DS, Erikssen J. Exercise testing of healthy men in a new perspective: from diagnosis to prognosis. Eur Heart J 2004;25:978–86. [10] Mundal R, Kjeldsen SE, Sandvik L, Erikssen G, Thaulow E, Erikssen J. Exercise blood pressure predicts cardiovascular mortality in middle-aged men. Hypertension 1994;24:56–62. [11] Lauer MS, Pashkow FJ, Harvey SA, Marwick TH, Thomas JD. Angiographic and prognostic implications of an exaggerated exercise systolic blood pressure response and rest systolic blood pressure in adults undergoing evaluation for suspected coronary artery disease. J Am Coll Cardiol 1995;26:1630–6. [12] Gupta MP, Polena S, Coplan N, Paganopoulos G, Dhingra C, Myers J, et al. Prognostic significance of systolic blood pressure increases in men during exercise stress testing. Am J Cardiol 2007;100:1609–13. [13] Kane GC, Askew JW, Chareonthaitawee P, Miller TD, Gibbons RJ. Hypertensive response with exercise does not increase the prevalence of abnormal Tc-99m SPECT stress perfusion images. Am Heart J 2008;155:930–7. [14] Bouzas-Mosquera A, Peteiro J, Broullon FJ, Alvarez-Garcia N, Garcia-Bueno L, Mosquera VX, et al. Prognostic value of an exaggerated exercise blood pressure response in patients with diabetes mellitus and known or suspected coronary artery disease. Am J Cardiol 2010;105:780–5. [15] Kurl S, Laukkanen JA, Rauramaa R, Lakka TA, Sivenius J, Salonen JT. Systolic blood pressure response to exercise stress test and risk of stroke. Stroke 2001;32:2036–41. [16] Kokkinos P, Chrysohoou C, Panagiotakos D, Narayan P, Greenberg M, Singh S. Betablockade mitigates exercise blood pressure in hypertensive male patients. J Am Coll Cardiol 2006;47:794–8. [17] Dubach P, Froelicher VF, Klein J, Oakes D, Grover-McKay M, Friis R. Exercise-induced hypotension in a male population. Criteria, causes, and prognosis. Circulation 1988;78:1380–7. [18] Diamond GA. A clinically relevant classification of chest discomfort. J Am Coll Cardiol 1983;1:574–5. [19] Bouzas-Mosquera A, Peteiro J, Álvarez-Garcia N, Broullon FJ, Mosquera VX, GarciaBueno L, et al. Prediction of mortality and major cardiac events by exercise echocardiography in patients with normal exercise electrocardiographic testing. J Am Coll Cardiol 2009;53:1981–90. [20] Peteiro J, Bouzas-Mosquera A, Estevez R, Pazos P, Pineiro M, Castro-Beiras A. Headto-head comparison of peak supine bicycle exercise echocardiography and treadmill exercise echocardiography at peak and at post-exercise for the detection of coronary artery disease. J Am Soc Echocardiogr 2012;25:319–26. [21] Irving JB, Bruce RA, DeRouen TA. Variations in and significance of systolic pressure during maximal exercise (treadmill) testing. Am J Cardiol 1977;39:841–8. [22] Morrow K, Morris CK, Froelicher VF, Hideg A, Hunter D, Johnson E, et al. Prediction of cardiovascular death in men undergoing noninvasive evaluation for coronary artery disease. Ann Intern Med 1993;118:689–95. [23] Schoenenberger AW, Schoenenberger-Berzins R, Suter PM, Erne P. Effects of weight on blood pressure at rest and during exercise. Hypertens Res 2013;36:1045–50. [24] Chmiel C, Wang M, Senn O, Del Prete V, Zoller M, Rosemann T, et al. Uncontrolled arterial hypertension in primary care–patient characteristics and associated factors. Swiss Med Wkly 2012;142:w13693. [25] Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Böhm M, et al. 2013 ESH/ESC guidelines for the management of arterial hypertension: the Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). Eur Heart J 2013;34:2159–219.
Contents lists available at ScienceDirect
European Journal of Internal Medicine journal homepage: www.elsevier.com/locate/ejim
Original Article
Exaggerated exercise blood pressure response and risk of stroke in patients referred for stress testing María del Carmen Bouzas-Mosquera a, Alberto Bouzas-Mosquera b,⁎, Jesús Peteiro b, Francisco J. Broullón c, Nemesio Álvarez-García b, Alfonso Castro-Beiras b a b c
Faculty of Health Sciences, European University, Madrid, Spain Department of Cardiology, Hospital Universitario A Coruña, A Coruña, Spain Department of Health Information Technology, Hospital Universitario A Coruña, A Coruña, Spain
a r t i c l e
i n f o
Article history: Received 2 March 2014 Received in revised form 22 May 2014 Accepted 27 May 2014 Available online 11 June 2014 Keywords: Exercise test Blood pressure Stroke Prognosis
a b s t r a c t Background/objectives: There is some evidence to suggest that exercise systolic blood pressure (SBP) may be associated with future risk of stroke in subjects without a history of coronary artery disease (CAD). However, the value of an exaggerated exercise SBP response (EESBPR) for predicting stroke in patients referred for stress testing for clinical reasons has not been investigated. Methods: We evaluated a community-based sample of 10,047 patients with known or suspected CAD who underwent treadmill exercise echocardiography. An EESBPR was defined as a peak exercise SBP of N 220 mmHg. The ratio of the increase in SBP during exercise to exercise workload (ΔSBPeEW) was also estimated. The endpoints were stroke of any type, ischemic stroke and hemorrhagic stroke. Median follow-up was 3.5 years. Results: Annualized rates of stroke of any type, ischemic stroke and hemorrhagic stroke were 0.6% (95% CI 0.53– 0.67), 0.49% (95% CI 0.42–0.56) and 0.12% (95% CI 0.09–0.15) in patients without EESBPR vs. 0.69% (95% CI 0.37– 1), 0.49% (95% CI 0.23–0.76) and 0.19% (95% CI 0.02–0.35) in those with EESBPR (p = 0.68, 0.90 and 0.39, respectively). Similarly, there was no significant univariate association between ΔSBPeEW and the occurrence of any endpoint. In multivariate analysis, hypertension, male sex, age, diabetes mellitus and resting SBP remained predictors of stroke of any type. EESBPR and ΔSBPeEW were not predictors of any of the endpoints evaluated. Conclusion: We did not observe any significant association between exercise SBP and the future occurrence of stroke in patients with known or suspected CAD referred for exercise echocardiography. © 2014 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
1. Introduction Although resting hypertension is a well-established risk factor for cardiovascular events, the prognostic value of exercise hypertension remains controversial. While some studies reported an association of an exaggerated exercise systolic blood pressure response (EESBPR) with future hypertension [1–7] and cardiovascular events [4,8–10] in healthy subjects, others evaluating patients with known or suspected coronary artery disease (CAD) disagree [11–14]. There is also some evidence to suggest that an EESBPR may be associated with future risk of stroke in middle-aged men without a history of CAD [15]. However, this issue has not been investigated in patients with known or suspected CAD, who represent the majority of subjects referred for noninvasive stress testing in clinical practice.
⁎ Corresponding author at: Department of Cardiology, Hospital Universitario A Coruña, As Xubias 84, 15006 A Coruña, Spain. Tel.: +34 981178184; fax: +34 981178258. E-mail address: [email protected] (A. Bouzas-Mosquera).
Our aim was to assess the value of an EESBPR for predicting stroke in patients with known or suspected CAD referred for exercise echocardiography.
2. Methods 2.1. Patients A total of 13,328 adult patients with known or suspected CAD who underwent treadmill exercise echocardiography at our hospital were initially considered for inclusion. To ensure comprehensive follow-up data, 1113 patients who did not belong to the geographic area of reference of our hospital were excluded. Given that beta-blocker therapy may blunt an EESBPR [16], 1218 patients who received betablockers within 48 h before the tests were also excluded; lastly, because of the well-known association of an exercise hypotensive or flat blood pressure response with poorer outcome [17], 950 patients in whom systolic blood pressure (SBP) failed to increase with exercise above the baseline values were also excluded. The remaining 10,047 patients constituted
http://dx.doi.org/10.1016/j.ejim.2014.05.013 0953-6205/© 2014 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
534
M.C. Bouzas-Mosquera et al. / European Journal of Internal Medicine 25 (2014) 533–537
Table 1 Baseline characteristics of the patients.
Male, n (%) Age, years Current smokers, n (%) Diabetes, n (%) Hypertension, n (%) Hypercholesterolemia, n (%) Family history of CAD, n (%) History of CAD, n (%) Prior myocardial infarction, n (%) ≤30 days before EE, n (%) N30 days before EE, n (%) Prior coronary revascularization, n (%) PCI, n (%) CABG, n (%) Chest pain, n (%) Typical angina, n (%) Atypical/probable angina, n (%) Nonanginal chest pain, n (%) Dyspnea, n (%) Uninterpretable resting ECG, n (%) Left bundle branch block, n (%) Atrial fibrillation, n (%) ACE-i/ARB, n (%) Nitrates, n (%) Calcium channel blockers, n (%)
All patients (n = 10047)
No EESBPR (n = 9645)
EESBPR (n = 402)
p
6105 (60.8) 62.0 ± 12.2 2504 (24.9) 1810 (18) 5227 (52) 4759 (47.4) 1305 (13) 2519 (25.1) 1900 (18.9) 850 (8.5) 1096 (10.9) 1534 (15.3) 1046 (10.4) 795 (7.9)
5861 (60.8) 62.0 ± 12.1 2408 (25) 1722 (17.9) 4935 (51.2) 4560 (47.3) 1237 (12.8) 2447 (25.4) 1846 (19.1) 829 (8.6) 1063 (11) 1486 (15.4) 1014 (10.5) 770 (8)
244 (60.7) 61.7 ± 10.3 96 (23.9) 88 (21.9) 291 (72.4) 199 (49.5) 68 (16.9) 72 (17.9) 54 (13.4) 21 (5.2) 33 (8.2) 48 (11.9) 32 (8) 25 (6.2)
0.97 0.51 0.62 0.04 b0.001 0.38 0.02 0.001 0.004 0.02 0.08 0.06 0.10 0.20
778 (7.7) 4251 (42.3) 2302 (22.9) 599 (6) 2354 (23.4) 669 (6.7) 453 (4.5) 3381 (33.7) 2207 (22) 900 (9)
758 (7.9) 4049 (42) 2229 (23.1) 583 (6) 2256 (23.4) 643 (6.7) 448 (4.6) 3235 (33.5) 2133 (22.1) 861 (8.9)
20 (5) 202 (50.2) 73 (18.2) 16 (4) 98 (24.4) 26 (6.5) 5 (1.2) 146 (36.3) 75 (18.7) 39 (9.7)
0.03 0.001 0.02 0.09 0.65 0.88 0.001 0.24 0.11 0.59
ACE-i denotes angiotensin converting enzyme inhibitors; ARB, angiotensin receptor blockers; CABG, coronary artery bypass grafting; CAD, coronary artery disease; ECG, electrocardiogram; EE, exercise echocardiography; EESBPR, exaggerated exercise systolic blood pressure response; PCI, percutaneous coronary intervention.
our sample. This study was approved by the Clinical Investigation Ethics Committee of Galicia. 2.2. Data collection Demographics, clinical data and stress testing results were entered in a dedicated database at the time of testing. Hypertension, hypercholesterolemia, and diabetes mellitus were defined on the basis of history and antihypertensive, lipid-lowering, or antidiabetic treatment, respectively. Patients referred for evaluation of chest pain were classified as having typical angina, atypical/probable angina or nonischemic chest pain [18]. A history of CAD was defined as previous myocardial
infarction, previous coronary revascularization or prior angiographic documentation of any ≥50% coronary stenosis. Resting ECG was considered uninterpretable in the presence of left bundle branch block, paced rhythm, left ventricular hypertrophy with strain, treatment with digoxin, preexcitation, or other repolarization abnormalities. All patients underwent symptom-limited treadmill exercise echocardiography. Protocols employed included the Bruce protocol in 8676 patients (86.4%), the modified Bruce protocol in 464 patients (4.6%), the Naughton protocol in 104 patients (1%), and other protocols in 803 patients (8%). Heart rate, blood pressure, and a 12-lead electrocardiogram were recorded at rest and at each stage of the protocol. Peak SBP was defined as the maximum value of SBP obtained during exercise.
Table 2 Data provided by exercise stress testing.
Systolic blood pressure, mm Hg Rest Peak Heart rate, beats/min Rest Peak RPP, ×103 mm Hg beats/min Rest Peak % of MAPHR Submaximal test, n (%) Exercise-induced chest pain, n (%) Ischemic ECG changes, n (%) Exercise workload, METs Left ventricular ejection fraction, % Rest Peak Resting wall motion abnormalities, n (%) Echocardiographic ischemia, n (%) Wall motion score index Rest Peak
All patients (n = 10047)
No EESBPR (n = 9645)
EESBPR (n = 402)
p
133 ± 19 171 ± 30
132 ± 18 168 ± 25
161 ± 23 243 ± 56
b0.001 b0.001
79 ± 15 149 ± 20
79 ± 14.8 149 ± 21
80 ± 14 148 ± 18
0.19 0.22
10.5 ± 2.6 25.6 ± 5.8 94.1 ± 11.5 1831 (18.2) 1368 (13.6) 1478 (14.7) 9.5 ± 3.1
10.4 ± 2.5 25.1 ± 5.3 94.1 ± 11.5 1747 (18.1) 1335 (13.8) 1413 (14.7) 9.5 ± 3.2
12.9 ± 3.1 34.7 ± 7.7 93.1 ± 10.3 84 (20.9) 33 (8.2) 65 (16.2) 8.3 ± 2.5
b0.001 b0.001 0.08 0.15 0.001 0.4 b0.001
58.4 ± 8.2 63.2 ± 12.3 2055 (20.5) 2744 (27.3)
58.3 ± 8.2 63.1 ± 12.3 1998 (20.7) 2653 (27.5)
60.3 ± 6.7 65.1 ± 10.5 57 (14.2) 91 (22.6)
b0.001 0.002 0.001 0.03
1.09 ± 0.24 1.18 ± 0.32
1.09 ± 0.23 1.19 ± 0.24
1.05 ± 0.18 1.14 ± 0.27
0.001 0.007
ECG denotes electrocardiographic; EESBPR, exaggerated exercise systolic blood pressure response; MAPHR, maximum age-predicted heart rate; METs, metabolic equivalents; RPP, ratepressure product.
M.C. Bouzas-Mosquera et al. / European Journal of Internal Medicine 25 (2014) 533–537 Table 3 Predictors of exaggerated exercise systolic blood pressure response in patients with known or suspected coronary artery disease.
Hypertension Typical angina Atrial fibrillation Resting systolic blood pressure Resting LVEF
Adjusted odds ratio (95% CI)
p
1.44 (1.13–1.84) 0.56 (0.35–0.92) 0.24 (0.09–0.61) 1.07 (1.06–1.08) 1.04 (1.02–1.05)
0.003 0.02 0.003 b0.001 b0.001
CI denotes confidence interval; LVEF, left ventricular ejection fraction.
An EESBPR was defined as a peak SBP N220 mm Hg [13,14], which is equivalent to the 95th percentile of peak SBP within the population under study. The increase in SBP during exercise (ΔSBPe) was calculated as the difference between peak and resting SBP. Since the degree of increase in SBP during exercise may depend on exercise capacity, we also estimated the ratio of the ΔSBPe to exercise workload in METs (ΔSBPeEW). Significant ischemic ST-segment changes during the tests were defined as the development of ST-segment deviation of ≥ 1 mm which was horizontal or sloping away from the isoelectric line 80 ms after the J point. A submaximal test was defined as failure to achieve 85% of maximum age-predicted heart rate. Echocardiographic images were acquired and analyzed as previously described [19]. Left ventricular ejection fraction and wall motion score index were estimated at rest and at peak exercise. Echocardiographic myocardial ischemia was defined as the appearance of new or worsening wall motion abnormalities with exercise, except worsening from akinesia to dyskinesia and isolated hypokinesia of the inferobasal segment. 2.3. Follow-up and end-points Follow-up data were retrieved from electronic medical records and discharge code databases through the Department of Health Information Technology of our institution. Data sources were linked using unique identifiers. End points were the occurrence of ischemic stroke (International Classification of Diseases, Ninth Revision [ICD-9] codes 433.x1 and 434.x1), hemorrhagic stroke (ICD-9 codes 430–432) or stroke of any type. Patients were censored at the last date in which vital status could be ascertained. Median follow-up was 3.5 years (interquartile range 0.2–8.0 years). 2.4. Statistical analysis Categorical variables were reported as percentages and comparison between groups based on the chi-square test. Continuous variables were reported as mean ± standard deviation and differences were assessed with the unpaired t-test or Mann–Whitney U test as
535
appropriate. Logistic regression analysis was performed to determine the predictors of EESBPR, and linear regression analysis was employed to assess the predictors of ΔSBPe and ΔSBPeEW. Annualized event rates were calculated by dividing the number of events by the total number of person-years at risk. Survival free of the end point of interest was estimated by the Kaplan–Meier method, and survival curves were compared with the log-rank test. Univariable and multivariable associations of clinical and exercise variables with stroke were assessed with Cox's proportional hazard models. Hazard ratios (HRs) with 95% confidence intervals (CI) were estimated. Different multivariate models were constructed forcing the inclusion of EESPBR, ΔSBPe and ΔSBPeEW. Interactions with sex, antihypertensive drug use, history of CAD or exercise echocardiography results were tested. A sensitivity analysis was also performed after exclusion of patients with either a history of CAD, history of atrial fibrillation, antihypertensive medication use, protocols other than the Bruce protocol, abnormal exercise echocardiography (i.e., resting or exercise-induced wall motion abnormalities) or reasons for stopping the tests other than physical exhaustion. Statistical analyses were performed using SPSS software, version 15.0 (SPSS, Chicago, IL). 3. Results Mean age was 62 ± 12 years, and 6105 patients (60.8%) were male. Demographic information, cardiac risk factors, history of cardiac events, chest pain symptoms, electrocardiographic features and medication use are listed in Table 1. A total of 402 patients (4%) developed an EESBPR during the tests. There were no significant differences in age or gender between patients with and without EESBPR, but the former had a higher prevalence of diabetes and hypertension, and a lower likelihood of prior history of CAD. There were no significant differences regarding antihypertensive medications between both groups (Table 1). Data obtained during the tests are shown in Table 2. Patients developing EESBPR had higher resting SBP, lower estimated functional capacity, and higher resting and peak left ventricular ejection fraction. These patients also developed less frequently exercise-induced chest pain. There were no significant differences between both groups regarding resting heart rate or the likelihood of submaximal tests (Table 2). Predictors of EESBPR are shown in Table 3. ΔSBPe was significantly correlated with age (R = − 0.138, p b 0.001), resting left ventricular ejection fraction (R = 0.119, p b 0.001), peak left ventricular ejection fraction (0.155, p b 0.001), the increase in left ventricular ejection fraction from rest to peak exercise (R = 0.120, p b 0.001), and estimated exercise capacity (R = 0.240, p b 0.001). In linear regression analysis, age (p b 0.001), sex (p b 0.001), hypertension (p b 0.001), estimated functional capacity (p b 0.001), resting left ventricular ejection fraction (p b 0.001) and the increase in left ventricular ejection fraction from rest to peak exercise (p b 0.001) were predictors of ΔSBPe, while age
Fig. 1. Event curves for (A) stroke of any type, (B) ischemic stroke and (C) hemorrhagic stroke in patients with and without exaggerated exercise systolic blood pressure response (EESBPR).
536
M.C. Bouzas-Mosquera et al. / European Journal of Internal Medicine 25 (2014) 533–537
Table 4 Multivariate predictors of stroke of any type, ischemic stroke and hemorrhagic stroke in patients with known or suspected coronary artery disease referred for stress testing. Adjusted hazard ratio (95% confidence interval) Any stroke Age Sex Hypertension Diabetes mellitus Hypercholesterolemia Smoking Family history of CAD Atrial fibrillation History of CAD Antihypertensive drugs Rest SBP (per 10 mm Hg) Rest LVEF EESBPR
1.04 (1.03–1.06), p 1.57 (1.19–2.06), p 1.56 (1.19–2.04), p 1.35 (1.02–1.78), p 0.99 (0.77–1.27), p 1.09 (0.80–1.48), p 0.88 (0.59–1.32), p 0.91 (0.53–1.54), p 1.10 (0.85–1.42), p 0.95 (0.74–1.23), p 1.07 (1.01–1.14), p 0.99 (0.98–1.01), p 0.88 (0.53–1.47), p
Ischemic stroke b 0.001 = 0.001 = 0.001 = 0.037 = 0.94 = 0.58 = 0.54 = 0.72 = 0.47 = 0.71 = 0.03 = 0.29 = 0.62
1.05 (1.03–1.06), p 1.68 (1.24–2.27), p 1.62 (1.20–2.18), p 1.49 (1.10–2.02), p 0.99 (0.75–1.30), p 1.13 (0.81–1.59), p 1.01 (0.66–1.55), p 0.85 (0.47–1.54), p 1.01 (0.76–1.34), p 1.00 (0.76–1.32), p 1.05 (0.98–1.13), p 0.99 (0.98–1.01), p 0.80 (0.44–1.45), p
Hemorrhagic stroke b 0.001 = 0.001 = 0.002 = 0.01 = 0.93 = 0.46 = 0.95 = 0.61 = 0.97 = 0.99 = 0.18 = 0.20 = 0.46
1.03 (1.00–1.06), p 1.13 (0.64–2.00), p 1.27 (0.71–2.26), p 1.02 (0.53–1.99), p 0.95 (0.55–1.65), p 0.91 (0.45–1.84), p 0.40 (0.12–1.29), p 1.31 (0.47–3.71), p 1.50 (0.87–2.60), p 0.86 (0.49–1.48), p 1.09 (0.95–1.24), p 1.00 (0.97–1.03), p 1.25 (0.46–3.44), p
= = = = = = = = = = = = =
0.05 0.68 0.42 0.94 0.87 0.79 0.12 0.61 0.14 0.58 0.24 0.92 0.66
CAD denotes coronary artery disease; EESBPR, exaggerated exercise systolic blood pressure response; ΔSBPe, increase in systolic blood pressure during exercise; ΔSBPeEW, ratio of increase in systolic blood pressure during exercise to exercise workload; LVEF, left ventricular ejection fraction; SBP, systolic blood pressure.
(p b 0.001), hypertension (p b 0.001), resting left ventricular ejection fraction (p = 0.026) and resting heart rate (p b 0.001) were predictors of ΔSBPeEW. During follow-up there were 232 ischemic strokes and 61 hemorrhagic strokes, and 285 patients had at least one stroke of any type. Annualized rates of stroke of any type, ischemic stroke and hemorrhagic stroke were 0.6% (95% CI 0.53–0.67), 0.49% (95% CI 0.42–0.56) and 0.12% (95% CI 0.09–0.15) in patients without EESBPR vs. 0.69% (95% CI 0.37–1), 0.49% (95% CI 0.23–0.76) and 0.19% (95% CI 0.02–0.35) in those with EESBPR (p = 0.68, 0.90 and 0.39, respectively) (Fig. 1). Similarly, there was no significant univariate association between ΔSBPeEW and the occurrence of any stroke (HR 1.02, 95% CI 0.99–1.05, p = 0.25), ischemic stroke (HR 1.00, 95% CI 0.97–1.05, p = 0.8) or hemorrhagic stroke (HR 1.05, 95% CI 0.99–1.10, p = 0.08). In Cox regression analysis, a history of hypertension at rest and resting SBP remained predictors of stroke of any type, along with gender, age and diabetes mellitus (Table 4). EESBPR, ΔSBPe or ΔSBPeEW were not predictors of any endpoint in their respective models after multivariate adjustment (Tables 4 and 5). There were no significant interactions with sex, antihypertensive drug use, history of atrial fibrillation or history of known CAD in any of these models. The lack of association between EESBPR, ΔSBPe or ΔSBPeEW with any of the endpoints in multivariate analyses remained consistent even after exclusion of patients with either a history of CAD, history of atrial fibrillation, antihypertensive medication use, protocols other than the Bruce protocol, abnormal exercise echocardiography results or reasons for stopping the tests other than physical exhaustion (Table 6). 4. Discussion In our study, we did not find any significant association between the degree of increase in SBP during exercise and the subsequent risk of stroke in a series of patients with known or suspected CAD. The prognostic value of a hypertensive response with exercise remains controversial. Some studies performed in asymptomatic healthy subjects found that exercise blood pressure response was associated with higher risk of cardiovascular events [4,8–10], while others evaluating patients with known or suspected CAD reported an association with improved outcome [11,12,14]. However, there is limited information regarding the value of an EESBPR for predicting stroke. Kurl and colleagues [15] found that a higher increase in SBP during exercise with bicycle ergometer was associated with higher risk of subsequent stroke in middle-aged men. Although this contrasts with our results, there are important differences in the characteristics of the patients, in the definition of the variables evaluated and in the technique employed between both studies. In this regard, patients with a history of CAD were excluded in Kurl and colleagues' study; in addition, there may be significant
differences in the magnitude of exercise-induced increase in SBP between bicycle ergometer and treadmill stress testing [20]. Exercise SBP depends on the balance between changes in cardiac output and systemic vascular resistance during exercise. In normal conditions, these two parameters change in opposite directions, with cardiac output increasing and systemic vascular resistance decreasing with exercise. An EESBPR may result from an increase in cardiac output and/or a blunted decrease in systemic vascular resistance due to pathological changes in arteries and arterioles. In this regard, our study showed a significant correlation between the increase in SBP during exercise and the increase in left ventricular ejection fraction from rest to peak exercise. While an exercise-induced increase in cardiac output may be favorable from a prognostic point of view, a blunted decrease in systemic vascular resistance is likely detrimental. These differences in the prognostic effect of both determinants of SBP response during exercise might explain, at least in part, the inconsistencies between different studies regarding the prognostic significance of an EESBPR. Other possible explanations include differences in the characteristics of the populations and in the definition of EESBPR. Furthermore, there is no consensus on normal values for exercise SBP response. The threshold employed in our study for defining an EESBPR was higher than those used in most previous reports; this threshold was selected because the prognostic value of moderate increases in blood pressure during exercise has been well established [21,22]. Also, in contrast with previous studies [11], we did not employ different cut-offs for defining EESBPR in men and women; nonetheless, the prevalence and prognostic value of EESBPR did not differ between both sexes using this definition. Finally, the degree of increase in SBP during exercise was correlated with estimated exercise capacity, which in turn may be a predictor of outcome, hence the rationale for controlling for exercise workload. Our study has the limitations inherent to an observational, retrospective, single-center study. These results only apply to patients undergoing treadmill exercise testing for clinical reasons; although they represent the vast majority of patients referred for noninvasive testing in clinical practice, these results cannot be extrapolated to other subgroups of subjects, such as healthy asymptomatic individuals. Important variables, such as body weight [23,24], anticoagulant therapy or left Table 5 Prognostic value of the increase in systolic blood pressure during exercise (ΔSBPe) and the ratio of ΔSBPe to exercise workload (ΔSBPeEW) for predicting stroke. Adjusted hazard ratio (95% confidence interval)a
Stroke of any type Ischemic stroke Hemorrhagic stroke a
ΔSBPe
ΔSBPeEW
1.00 (0.99–1.00), p = 0.25 1.00 (0.99–1.00), p = 0.18 1.00 (0.99–1.01), p = 0.90
1.01 (0.97–1.04), p = 0.73 0.99 (0.95–1.03), p = 0.62 1.04 (0.99–1.09), p = 0.15
Adjusted by all variables listed in Table 4 except EESBPR.
M.C. Bouzas-Mosquera et al. / European Journal of Internal Medicine 25 (2014) 533–537
537
Table 6 Prognostic value of exercise systolic blood pressure for predicting stroke after exclusion of patients with either a history of CAD, history of atrial fibrillation, antihypertensive medication use, protocols other than the Bruce protocol, abnormal exercise echocardiograms or reasons for stopping the tests other than physical exhaustion. Adjusted hazard ratio (95% confidence interval)a
Stroke of any type Ischemic stroke Hemorrhagic stroke
EESBPR
ΔSBPe
ΔSBPeEW
1.10 (0.37–3.23), p = 0.87 1.01 (0.29–3.48), p = 0.99 1.60 (0.17–15.2), p = 0.68
0.99 (0.98–1.00), p = 0.17 0.99 (0.98–1.01), p = 0.20 0.99 (0.96–1.02), p = 0.49
0.93 (0.83–1.05), p = 0.25 0.91 (0.80–1.05), p = 0.19 0.97 (0.75–1.26), p = 0.83
EESBPR denotes exaggerated exercise systolic blood pressure response; ΔSBPe, increase in systolic blood pressure during exercise; ΔSBPeEW, ratio of increase in systolic blood pressure during exercise to exercise workload. a Adjusted by age, sex, diabetes mellitus, hypertension, hypercholesterolemia, smoking, family history of coronary artery disease, resting systolic blood pressure and resting left ventricular ejection fraction.
atrial size, were not available for analysis. However, exclusion of patients with a history of atrial fibrillation did not change the results. Notwithstanding the exclusion of patients taking betablockers within 48 h before the tests, we cannot rule out residual beta-blocking effect in patients who had previously been prescribed betablockers and were instructed to withdraw them temporarily before testing. The number of hemorrhagic strokes during follow-up was relatively low; thus, the possibility of a type II error cannot be excluded and the results should be interpreted with caution; whether longer follow-up studies with higher statistical power might lead to different conclusions is unknown. Lastly, we did not evaluate the prognostic value of exercise diastolic blood pressure because only SBP can be measured reliably during exercise [25], but even the latter may be difficult to assess during treadmill stress testing. However, difficulties in real-world clinical practice should also be reflected in the prognostic value of testing. 5. Conclusions In our series of patients with known or suspected CAD referred for exercise echocardiography, we did not find any significant association between exercise SBP and stroke during follow-up. Thus, our study puts into question the clinical usefulness of blood pressure measurements during treadmill exercise testing for predicting the future occurrence of stroke in patients with known or suspected CAD, although further studies with higher statistical power are warranted. Conflict of interests The authors state that they have no conflicts of interest. References [1] Manolio TA, Burke GL, Savage PJ, Sidney S, Gardin JM, Oberman A. Exercise blood pressure response and 5-year risk of elevated blood pressure in a cohort of young adults: the CARDIA study. Am J Hypertens 1994;7:234–41. [2] Miyai N, Arita M, Morioka I, Miyashita K, Nishio I, Takeda S. Exercise BP response in subjects with high-normal BP: exaggerated blood pressure response to exercise and risk of future hypertension in subjects with high-normal blood pressure. J Am Coll Cardiol 2000;36:1626–31. [3] Singh JP, Larson MG, Manolio TA, O'Donnell CJ, Lauer M, Evans JC, et al. Blood pressure response during treadmill testing as a risk factor for new-onset hypertension. The Framingham heart study. Circulation 1999;99:1831–6. [4] Allison TG, Cordeiro MA, Miller TD, Daida H, Squires RW, Gau GT. Prognostic significance of exercise-induced systemic hypertension in healthy subjects. Am J Cardiol 1999;83:371–5. [5] Sharabi Y, Ben-Cnaan R, Hanin A, Martonovitch G, Grossman E. The significance of hypertensive response to exercise as a predictor of hypertension and cardiovascular disease. J Hum Hypertens 2001;15:353–6. [6] Majahalme S, Turjanmaa V, Tuomisto M, Lu H, Uusitalo A. Blood pressure responses to exercise as predictors of blood pressure level after 5 years. Am J Hypertens 1997;10:106–16.
[7] Dlin RA, Hanne N, Silverberg DS, Bar-Or O. Follow-up of normotensive men with exaggerated blood pressure response to exercise. Am Heart J 1983;106:316–20. [8] Laukkanen JA, Kurl S, Rauramaa R, Lakka TA, Venalainen JM, Salonen JT. Systolic blood pressure response to exercise testing is related to the risk of acute myocardial infarction in middle-aged men. Eur J Cardiovasc Prev Rehabil 2006;13:421–8. [9] Erikssen G, Bodegard J, Bjornholt JV, Liestol K, Thelle DS, Erikssen J. Exercise testing of healthy men in a new perspective: from diagnosis to prognosis. Eur Heart J 2004;25:978–86. [10] Mundal R, Kjeldsen SE, Sandvik L, Erikssen G, Thaulow E, Erikssen J. Exercise blood pressure predicts cardiovascular mortality in middle-aged men. Hypertension 1994;24:56–62. [11] Lauer MS, Pashkow FJ, Harvey SA, Marwick TH, Thomas JD. Angiographic and prognostic implications of an exaggerated exercise systolic blood pressure response and rest systolic blood pressure in adults undergoing evaluation for suspected coronary artery disease. J Am Coll Cardiol 1995;26:1630–6. [12] Gupta MP, Polena S, Coplan N, Paganopoulos G, Dhingra C, Myers J, et al. Prognostic significance of systolic blood pressure increases in men during exercise stress testing. Am J Cardiol 2007;100:1609–13. [13] Kane GC, Askew JW, Chareonthaitawee P, Miller TD, Gibbons RJ. Hypertensive response with exercise does not increase the prevalence of abnormal Tc-99m SPECT stress perfusion images. Am Heart J 2008;155:930–7. [14] Bouzas-Mosquera A, Peteiro J, Broullon FJ, Alvarez-Garcia N, Garcia-Bueno L, Mosquera VX, et al. Prognostic value of an exaggerated exercise blood pressure response in patients with diabetes mellitus and known or suspected coronary artery disease. Am J Cardiol 2010;105:780–5. [15] Kurl S, Laukkanen JA, Rauramaa R, Lakka TA, Sivenius J, Salonen JT. Systolic blood pressure response to exercise stress test and risk of stroke. Stroke 2001;32:2036–41. [16] Kokkinos P, Chrysohoou C, Panagiotakos D, Narayan P, Greenberg M, Singh S. Betablockade mitigates exercise blood pressure in hypertensive male patients. J Am Coll Cardiol 2006;47:794–8. [17] Dubach P, Froelicher VF, Klein J, Oakes D, Grover-McKay M, Friis R. Exercise-induced hypotension in a male population. Criteria, causes, and prognosis. Circulation 1988;78:1380–7. [18] Diamond GA. A clinically relevant classification of chest discomfort. J Am Coll Cardiol 1983;1:574–5. [19] Bouzas-Mosquera A, Peteiro J, Álvarez-Garcia N, Broullon FJ, Mosquera VX, GarciaBueno L, et al. Prediction of mortality and major cardiac events by exercise echocardiography in patients with normal exercise electrocardiographic testing. J Am Coll Cardiol 2009;53:1981–90. [20] Peteiro J, Bouzas-Mosquera A, Estevez R, Pazos P, Pineiro M, Castro-Beiras A. Headto-head comparison of peak supine bicycle exercise echocardiography and treadmill exercise echocardiography at peak and at post-exercise for the detection of coronary artery disease. J Am Soc Echocardiogr 2012;25:319–26. [21] Irving JB, Bruce RA, DeRouen TA. Variations in and significance of systolic pressure during maximal exercise (treadmill) testing. Am J Cardiol 1977;39:841–8. [22] Morrow K, Morris CK, Froelicher VF, Hideg A, Hunter D, Johnson E, et al. Prediction of cardiovascular death in men undergoing noninvasive evaluation for coronary artery disease. Ann Intern Med 1993;118:689–95. [23] Schoenenberger AW, Schoenenberger-Berzins R, Suter PM, Erne P. Effects of weight on blood pressure at rest and during exercise. Hypertens Res 2013;36:1045–50. [24] Chmiel C, Wang M, Senn O, Del Prete V, Zoller M, Rosemann T, et al. Uncontrolled arterial hypertension in primary care–patient characteristics and associated factors. Swiss Med Wkly 2012;142:w13693. [25] Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Böhm M, et al. 2013 ESH/ESC guidelines for the management of arterial hypertension: the Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). Eur Heart J 2013;34:2159–219.