Effect of Blood Pressure Variability on Cardiovascular Outcome in Diabetic and Nondiabetic Patients with Stroke Bu-Xing Chen, MD,*1 Jun-Ping Tian, MD,*1 Hong-Xia Wang, MD,† Jie Xu, MD,‡xk{ Feng-He Du, MD,* and Xing-Quan Zhao, MD‡xk{
Background: The association between blood pressure (BP) variability and stroke outcome is controversial, and there are few studies that have focused on the impact of BP variability in diabetic patients with stroke. Therefore, we aimed to examine the impact of BP variability on cardiovascular outcome in diabetic and nondiabetic patients with stroke. Methods: A total of 373 ischemic stroke patients with large artery atherosclerosis were recruited and followed up. Ambulatory BP monitoring was performed in all patients and divided according to the 25th and 75th percentiles interval of SD of daytime systolic BP (SBP). Kaplan-Meier analysis and Cox regression were used to assess the relationship between BP variability and cardiovascular outcomes including stroke recurrence, vascular events and cardiovascular death. Results: The 339 patients were included in the final analysis. During an average follow-up of 19.0 6 5.1 months (.6-26.8 months), 69 (20.4%) cardiovascular events occurred in all patients. Kaplan-Meier analysis found that there were no differences in cardiovascular events-free survival among the different BP variability groups in diabetic patients (P 5 .995); however, nondiabetic patients with greater BP variability showed a lesser cardiovascular events-free survival (P 5 .039). Through Cox regression we found the SD of daytime SBP (hazard ratio 1.103; 95% CI 1.0111.203) was associated with cardiovascular outcomes in nondiabetic patients with stroke. Conclusions: We show that SBP variability is associated with cardiovascular outcomes in stroke patients without diabetes, but we didn’t find a correlation between SBP variability and cardiovascular outcomes in stroke patients with diabetes. Key Words: Ischemic stroke—outcome—ambulatory blood pressure monitoring—blood pressure variability—diabetes mellitus. Ó 2014 by National Stroke Association
From the *Department of Cardiology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China; †Department of Cardiology, Beijing Shijingshan Hospital, Capital Medical University, Beijing, China; ‡Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China; xChina National Clinical Research Center for Neurological Diseases, Beijing, China; kCenter of Stroke, Beijing Institute for Brain Disorders, Beijing, China; and {Beijing Key Laboratory of Translational Medicines for Cerebrovascular Disease, Beijing, China. Received April 22, 2014; revision received May 21, 2014; accepted May 24, 2014. Supported by a grant from National Key Technology Research and Development Program of the Ministry of Science and Technology of
2450
China (2013BAI09B03), and a grant from Beijing Institute for Brain Disorders (BIBD-PXM2013_014226_07_000084). Address correspondence to Bu-Xing Chen, MD, Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, No. 6 Tian Tan Xi Li, Dongcheng District, Beijing 100050, P.R.China. E-mail: [email protected]; and Xing-Quan Zhao, MD, Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, No. 6 Tian Tan Xi Li, Dongcheng District, Beijing 100050, P.R.China. E-mail: [email protected]. 1 Both Bu-Xing Chen and Jun-Ping Tian are the first authors of this paper. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.05.030
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 9 (October), 2014: pp 2450-2457
BP VARIABILITY AND STROKE OUTCOME
Introduction Ambulatory blood pressure monitoring (ABPM) not only provides information on blood pressure (BP) levels but on the diurnal changes in BP as well.1 Previous studies have indicated that BP variability is an important trigger of vascular events2-4 and can provide a more precise estimate of individual cardiovascular risk.5 The association between BP variability and cardiovascular outcomes in patients with stroke has been investigated. Because of the methodologic differences, patient selection differences, and other differences, previous published studies were vague on the association of BP variability with stroke outcome. Some studies found that increased BP variability during an acute ischemic stroke (IS) was associated with worse clinical outcomes during a short period after an acute phase of stroke.6-11 On the contrary, some reports suggested that increased BP variability is of little prognostic value12-15 or may even indicate a good prognosis.16,17 Recently, diabetes mellitus has been shown to be an independent risk factor for IS.18 The prevalence of diabetes ranges from 21% to 44.4% among patients with acute IS.19 Jia et al19 suggested that the prevalence of abnormal glucose regulation was 68.7% among all the patients with acute stroke in China. In addition, some evidence indicates that the BP variability is increased in diabetic subjects20 and diabetes is associated with greater mortality from stroke.17,21 Up to now, there are few studies focusing on BP variability of diabetic patients with stroke. Therefore, we aimed to examine the impact of BP variability on cardiovascular outcomes in diabetic and nondiabetic patients with IS.
Figure 1. The patient selection flow in this study. Abbreviations: IS, ischemic stroke; SBP, systolic blood pressure.
2451
Patients and Methods Patients We performed an observational study of prevalent patients with IS who were admitted to the Cerebrovascular Center of Beijing Tiantan Hospital, People’s Republic of China, between May 2010 and August 2011. The patients were diagnosed with IS according to the World Health Organization criteria, which was confirmed by computed tomography (CT) or magnetic resonance imaging (MRI) of the brain. The enrolled IS patients, who were older than 18 years of age, were classified as having largeartery atherosclerosis according to the TOAST (ie, Trial of Org 10172 in Acute Stroke Treatment) criteria. None of patients suffered from infective endocarditis, atrial fibrillation, valvular heart disease, or malignant tumor that could lead to stroke. The patients who had any severe alimentary tract hemorrhage, infectious disease, diseases known to affect the autonomic nervous system, or severe heart failure (left ventricular ejection fraction ,45%) were excluded from the present study. In addition, patients with intracranial aneurysm who were confirmed by brain CT or MRI were not also included in our study. Therefore, 373 IS patients were recruited. The ethical committee of Beijing Tiantan Hospital approved the study protocol, and all patients or their designated relatives provided informed consent.
Baseline Characteristics The following baseline characteristics were investigated: age, gender, height, body weight, and duration of hypertension. Patients were defined as having
B.-X. CHEN ET AL.
2452
Table 1. Baseline characteristics of the study population and comparisons among the 3 different blood pressure variability groups Variables
Group 1 (#9.6)
Group 2 (.9.6 to #14.5)
Group 3 (.14.5)
Number Age, y Gender, male, % BMI, kg/m2 Diabetes mellitus, % Hypertension, % Previous stroke, % Coronary heart disease, % Dyslipidemia, % Current or previous smoker, % Moderate or heavy drinking, % Left ventricular ejection fraction, % Glycosylated hemoglobin, % Triglycerides, mmol/L Total cholesterol, mmol/L LDL cholesterol, mmol/L HDL cholesterol, mmol/L NIHSS scores at admission Total SBP, mm Hg Mean daytime SBP, mm Hg Mean nighttime SBP, mm Hg Total DBP, mm Hg Mean daytime DBP, mm Hg Mean nighttime DBP, mm Hg SD of total SBP, mm Hg SD of daytime SBP, mm Hg SD of nighttime SBP, mm Hg SD of total DBP, mm Hg SD of daytime DBP, mm Hg SD of nighttime DBP, mm Hg Dipper pattern Dipper Nondipper Cardiovascular events
86 55.7 6 9.5 79.1 24.9 6 3.1 47.4 75.6 24.4 14.0 70.9y 68.6 27.9 69.3 6 6.3 6.8 6 1.8 1.68 6 1.04 4.13 6 .94 2.43 6 .72 1.03 6 .24 4.8 6 4.8 130.4 6 16.9 131.8 6 17.0 125.8 6 17.9y,* 82.1 6 10.2 83.1 6 10.5 78.3 6 10.5 9.1 6 1.7 8.0 6 1.2 8.8 6 3.8*,y 8.3 6 1.9 7.6 6 1.9 8.3 6 2.9
170 59.3 6 11.8z 72.4 25.2 6 3.3 46.5 86.5 24.1 11.8 77.6 58.8 30.0 67.7 6 6.3 6.7 6 1.7 1.71 6 1.01 4.27 6 1.13 2.56 6 .96 1.05 6 .23 4.9 6 4.9 136.5 6 16.5* 137.6 6 16.9* 132.5 6 18.2 82.7 6 11.0 83.5 6 11.1 79.7 6 12.0 12.5 6 2.2* 11.9 6 1.4* 10.6 6 4.6 9.9 6 1.9* 9.6 6 2.0* 9.1 6 3.1
83 62.5 6 10.8y 68.7 25.3 6 3.4 36.4 89.2 28.9 7.2 88.0 56.6 15.7z 68.5 6 6.7 6.4 6 1.6 1.44 6 1.04 3.98 6 1.24 2.36 6 1.09 1.04 6 .22 5.8 6 4.9 142.6 6 16.8y 144.0 6 17.0y 137.8 6 19.9z 84.4 6 12.4 85.1 6 12.8 81.7 6 13.2 17.3 6 2.7y 17.3 6 2.7y 12.0 6 4.5z 11.9 6 2.3y 11.7 6 2.4y 10.2 6 3.5y,z
18.6 81.4 16 (18.6)
15.9 84.1 34 (20.0)
28.9 71.1 19 (22.9)
Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; NIHSS, National Institutes of Health Stroke Scale; SBP, systolic blood pressure. *P , .01, group 1 versus group 2, group 2 versus group 3. yP , .01, group 1 versus group 3. zP , .05, group 1 versus group 2, group 2 versus group 3.
hypertension if their BP was $140/90 mm Hg on repeated measurements during the hospitalization or if they were taking an antihypertensive medication. Other risk factors were defined as follows: history of stroke (defined as a medical chart–confirmed history of stroke, including IS, intracerebral hemorrhage, or subarachnoid hemorrhage), coronary heart disease (a reported history of myocardial infarction or cardiac surgery, or with a final diagnosis of myocardial infarction at discharge), diabetes mellitus (defined as a reported history of diabetes mellitus, or use of insulin or oral hypoglycemic agents),21,22 dyslipidemia (total cholesterol measurement $240 mg/ dL, high-density lipoprotein measurement ,35 mg/dL, or the use of lipid-lowering agents), current or previous smoking (defined as an individual who smoked at the time of stroke or had quit smoking within 1 year), and moderate or heavy drinking ($2 standard alcoholic beverages consumed per day). Other clinical features
Figure 2. The Kaplan-Meier of cardiovascular events-free survival according to BP variability in all patients. Abbreviation: BP, blood pressure.
BP VARIABILITY AND STROKE OUTCOME
2453
Table 2. Subgroup analysis in IS patients with diabetes mellitus Variables
Group 1 (#9.4)
Group 2 (.9.4 to #14.0)
Group 3 (.14.0)
Number Age, y Gender, male, % BMI, kg/m2 Hypertension, % Previous stroke, % Coronary heart disease, % Dyslipidemia, % Current or previous smoker, % Moderate or heavy drinking, % Left ventricular ejection fraction, % Glycosylated hemoglobin, % Triglycerides, mmol/L Total cholesterol, mmol/L LDL cholesterol, mmol/L HDL cholesterol, mmol/L NIHSS scores at admission Total SBP, mmHg Mean daytime SBP, mmHg Mean nighttime SBP, mmHg Total DBP, mmHg Mean daytime DBP, mmHg Mean nighttime DBP, mmHg SD of total SBP, mmHg SD of daytime SBP, mmHg SD of nighttime SBP, mmHg SD of total DBP, mmHg SD of daytime DBP, mmHg SD of nighttime DBP, mmHg Dipper pattern Dipper Nondipper Cardiovascular events
35 60.3 6 8.1 68.6 25.0 6 3.7 85.7 31.4 20.0 77.1 57.1 14.3 69.0 6 5.5 8.5 6 1.9 1.83 6 1.21 4.42 6 1.10 2.60 6 .82 1.09 6 .25 6.3 6 4.3 134.9 6 16.2 135.7 6 16.5 132.1 6 17.2 82.8 6 8.4 83.7 6 8.5 80.0 6 9.5 9.2 6 1.7 8.1 6 1.1 9.0 6 3.8y,z 7.9 6 1.4 7.4 6 1.4 7.9 6 2.7y
70 61.2 6 9.7 65.7 25.8 6 3.0 91.4 22.9 11.4 78.6 52.9 32.9 67.5 6 5.3 7.7 6 1.8 1.67 6 .77 4.21 6 1.14 2.49 6 .96 1.04 6 .23 4.9 6 5.2 140.2 6 15.5 141.3 6 15.9 136.7 6 16.9 83.8 6 10.0 84.6 6 10.0 80.8 6 11.4 12.3 6 1.9* 11.7 6 1.3* 11.0 6 3.5 9.7 6 2.0* 9.3 6 2.2* 8.8 6 2.6
34 62.1 6 10.5 58.8 25.6 6 2.5 91.2 35.3 5.9 79.4 50.0 20.6 67.1 6 7.5 7.8 6 2.0 1.49 6 .81 4.16 6 1.51 2.52 6 1.40 1.03 6 .18 6.2 6 5.6 145.4 6 14.0y 147.8 6 15.1y 137.6 6 14.8 84.0 6 9.5 85.2 6 10.0 79.3 6 9.3 17.1 6 2.9y 16.9 6 2.9y 11.7 6 4.1 11.1 6 2.0y 10.7 6 2.2y 9.8 6 2.5
17.1 82.9 12 (34.3)
12.9 87.1 19 (27.1)
32.4 67.6 10 (29.4)
Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HDL, high-density lipoprotein; IS, ischemic stroke; LDL, low-density lipoprotein; NIHSS, National Institutes of Health Stroke Scale; SBP, systolic blood pressure. *P , .01, group 1 versus group 2, group 2 versus group 3. yP , .01, group 1 versus group 3. zP , .05, group 1 versus group 2.
included severity of stroke on admission (National Institutes of Health Stroke Scale score). Electrocardiogram, echocardiography, carotid artery ultrasound, and brain CT or MRI were performed in IS patients. Left ventricular ejection fraction was determined by echocardiography. Biochemistry parameters such as glycosylated hemoglobin, triglycerides, total cholesterol, low-density lipoprotein cholesterol, and high-density lipoprotein cholesterol were examined.
24-Hour ABPM ABPM was performed with a portable lightweight device (Mobil-O-Graph; S/N: B19086, I.E.M., Stolberg, Germany). The accuracy of the ABPM was confirmed in each subject by simultaneous auscultation and sphygmo-
manometry. The patients underwent 24-hour ABPM after the acute stroke phase. Patients carried the device for 24 hours, and BP measurements were taken every 30 minutes during the day and evening (from 06:00 to 22:00) and every 60 minutes at night (from 22:00 to 06:00). The patients were instructed to keep their nondominant arm still and relaxed at the side during the measurements. Patients received verbal and written instructions on the monitors and completed a diary in which they recorded sleep medication, posture, and symptoms. An acceptable 24-hour ABPM recording for our study should have had $23 acceptable measurements during the 24 hours, $6 measurements at night, and effective reading records during the whole day .75%. We calculated mean 24-hour, day, and night systolic and diastolic BP. SD as an index of BP variability was calculated for
2454
24 hours and separately for day and night. Patients were classified according to the percent of nocturnal SBP dipping [100 3 (nocturnal SBP-daytime SBP)/daytime SBP]. The classifications were dipper, if the nocturnal fall in BP was $10%; nondipper, if the nocturnal fall in BP was ,10%.
Follow-Up and Events After an acute phase of stroke onset, the outcomes of all patients were assessed through telephone follow-up, including recurrence of stroke and time (rehospitalization with a diagnosis of new IS or intracerebral hemorrhage, diagnosed according to the Sacco criteria), combined vascular events (including stroke recurrence, myocardial infarction, angina pectoris, heart failure), rehospitalization (including stroke recurrence, cardiovascular disease, or other diseases), and death from any cause. The telephone follow-up was conducted centrally for all included patients and was based on a shared standardized interview protocol. If the patients answered they had experienced rehospitalization for any of the aforementioned events, we would contact the hospitals that admitted these patients to verify the diagnosis. All events were based on clear documentation in medical records.
Statistical Analysis Continuous variables were expressed as mean 6 SD or median (range), whereas categorical variables were expressed as ratio or percentage. The comparison of continuous variables among the 3 groups was analyzed by the use of analysis of variance (for post hoc analysis, Student-Newman-Keuls was performed if equal variance was assumed and Tamhane’s T2 test was performed if equal variance was not assumed). When the variables were not normally distributed, the nonparametric Kruskal-Wallis test was used. The comparison of categorical variables was performed using the c2test. The cumulative cardiovascular events-free survival was plotted as Kaplan-Meier curves among the 3 different groups, and the differences were assessed by the log-rank test. The associations between BP variability and cardiovascular outcomes were analyzed in multivariable Cox regression models, after we adjusted for potential confounders, including age, gender, dyslipidemia, hypertension, previous stroke, coronary heart disease, and stroke severity. Adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) were reported separately. A 2-tailed P-value less than .05 was considered to be statistically significant.
B.-X. CHEN ET AL.
to the TOASTcriteria. A total of 28 patients were excluded from analyses as having had noncardiovascular events during this period, and 339 patients were included in the final analysis (mean age 59.2 6 11.3, male 73.2%). During an average follow-up of 19.0 6 5.1 months (0.626.8 months), 69 (20.4%) cardiovascular events occurred in all enrolled patients. All patients were divided into the 3 groups according to the 25th and 75th percentiles interval of SD of daytime SBP. The division points were 9.6 (25%) and 14.5 (75%) respectively. Seen from Table 1, patients with greater BP variability were older and had greater BP. Kaplan-Meier curves are reported in Figure 2. There were no differences in cardiovascular events-free survival among the different BP variability groups in all patients (log-rank c2 5 1.974, P 5 .373). The IS patients were stratified according to diabetes mellitus or no diabetes, in which the information for 25 patients was missed. The subgroup analysis on the patients with diabetes mellitus (n 5 139) is shown in Table 2. The division points for the 25th and 75th percentile interval of SD of daytime SBP were 9.4 and 14.0, respectively. Patients with greater BP variability had greater BP compared with those in lower BP variability. Using Kaplan-Meier analysis (Fig 3), we found that there were no differences in cardiovascular events-free survival among the different BP variability subgroups with diabetes mellitus (log-rank c2 5 .010, P 5 .995). The 175 IS patients without diabetes mellitus were included in another subgroup analysis (mean age 58.2 6 12.2, male 80.0%; Table 3). The division points for the 25th and 75th percentile interval of SD of daytime SBP were 9.9 and 14.8, respectively. Patients with greater BP variability were older, had a greater proportion of
Results The flow diagram of the study is reported in Figure 1. Overall, 367 IS patients completed follow-up. All patients were diagnosed as large-artery atherosclerosis according
Figure 3. The Kaplan-Meier of cardiovascular events-free survival according to BP variability in IS patients with diabetes mellitus. Abbreviations: BP, blood pressure; IS, ischemic stroke.
BP VARIABILITY AND STROKE OUTCOME
2455
Table 3. Subgroup analysis in IS patients without diabetes mellitus Variables
Group 1 (#9.9)
Group 2 (.9.9 to #14.8)
Group 3 (.14.8)
Number Age, y Gender, male, % BMI, kg/m2 Hypertension, % Previous stroke, % Coronary heart disease, % Dyslipidemia, % Current or previous smoker, % Moderate or heavy drinking, % Left ventricular ejection fraction, % Glycosylated hemoglobin, % Triglycerides, mmol/L Total cholesterol, mmol/L LDL cholesterol, mmol/L HDL cholesterol, mmol/L NIHSS scores at admission Total SBP, mm Hg Mean daytime SBP, mm Hg ) Mean nighttime SBP, mm Hg Total DBP, mm Hg Mean daytime DBP, mm Hg Mean nighttime DBP, mm Hg SD of total SBP (mmHg) SD of daytime SBP, mm Hg SD of nighttime SBP, mm Hg SD of total DBP, mm Hg SD of daytime DBP, mm Hg SD of nighttime DBP, mm Hg Dipper pattern Dipper Nondipper Cardiovascular events
44 52.0 6 10.2*,y 90.9 24.8 6 2.6 63.6z 20.5 9.1 65.9 79.5 40.9 68.7 6 6.6 5.7 6 0.4 1.44 6 .66 4.06 6 .92 2.46 6 .84 .99 6 .20 3.8 6 5.3 126.0 6 16.9z 127.8 6 17.1 119.7 6 16.8z 80.8 6 11.7 82.2 6 12.1 76.3 6 11.2x 9.0 6 1.7 7.8 6 1.3 8.9 6 3.9 8.4 6 2.2 7.6 6 2.3 8.6 6 3.0
88 58.5 6 12.5 77.3 24.5 6 3.5 81.8 25.0 12.5 79.5 63.6 23.9 67.8 6 7.1 5.7 6 0.9 1.74 6 1.26 4.19 6 1.12 2.50 6 .91 1.05 6 .22 5.3 6 4.9 132.3 6 16.1 133.3 6 16.3 128.4 6 18.2 81.0 6 11.3 81.6 6 11.6 78.7 6 12.0 12.7 6 2.3* 12.2 6 1.4* 10.2 6 5.4 10.1 6 1.8* 9.9 6 1.9* 9.0 6 3.3
43 64.0 6 10.6z 74.4 24.9 6 3.6 90.7y 34.9 9.3 86.0 62.8 16.3x 69.5 6 6.0 5.7 6 0.5 1.43 6 1.22 4.00 6 1.02 2.31 6 .80 1.03 6 .24 5.7 6 4.4 142.3 6 16.8*,y 143.2 6 16.4*,y 139.1 6 21.9*,y 84.3 6 11.9 84.7 6 11.7 82.8 6 14.4 17.7 6 2.7y 17.8 6 2.6y 12.5 6 4.6y,z 12.3 6 2.5y 12.3 6 2.6y 10.5 6 4.0x,z
18.2 81.8 4 (9.1)
17.0 83.0 12 (13.6)
30.2 69.8 11 (25.6)
Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HDL, high-density lipoprotein; IS, ischemic stroke; LDL, low-density lipoprotein; NIHSS, National Institutes of Health Stroke Scale; SBP, systolic blood pressure. *P , .01, group 1 versus group 2, group 2 versus group 3. yP , .01, group 1 versus group 3. zP , .05, group 1 versus group 2, group 2 versus group 3. xP , .05, group 1 versus group 3.
hypertension and greater BP than those in lower BP variability. During an average follow-up of 18.8 6 4.9 months (5.5-26.7 months), 27 (15.4%) cardiovascular events occurred. The event rates in group 1, group 2, and group 3 were 9.1%, 13.6%, and 25.6%, respectively. KaplanMeier curves are reported in Figure 4. Nondiabetic patients who had a greater BP variability showed a lower cardiovascular events-free survival (log-rank c2 5 6.466, P 5 .039). Furthermore, we used multivariable Cox regression analyses to investigate the association between BP variability and risk of cardiovascular outcomes in 175 IS patients without diabetes mellitus. After we adjusted for potential confounders, including gender, dyslipidemia, hypertension, previous stroke, coronary heart disease, and SD of nighttime SBP, age (HR 1.040; 95% CI
1.002-1.079), functional status score (HR 1.089; 95% CI 1.024-1.158), and SD of daytime SBP (HR 1.103; 95% CI 1.011-1.203) were associated with cardiovascular outcomes in IS patients without diabetes mellitus (Table 4).
Discussion Our study demonstrated that SBP variability was associated with cardiovascular events in IS patients without diabetes mellitus. On the contrary, there was no correlation between BP variability and cardiovascular events in patients with diabetes mellitus and the whole study population. Using a Kaplan-Meier analysis, we found that nondiabetic patients with greater BP variability showed a significant lesser cardiovascular events-free survival. Thus, our study contradicts earlier data by Phillips
B.-X. CHEN ET AL.
2456
Figure 4. The Kaplan-Meier of cardiovascular events-free survival according to BP variability in IS patients without diabetes mellitus. Abbreviations: BP, blood pressure; IS, ischemic stroke.
et al,15 who analyzed 24-hour ABPM recordings of 110 control subjects and 91 stroke survivors and found that there was no significant difference in BP variability between stroke survivors and controls, a finding not explained by the presence of reverse dippers (raisers) in both groups. Phillips et al15 tested the prognostic value of BP variability, measured as cusums-derived circadian alteration magnitude. Cusums-derived circadian alteration magnitude of BP is a measure of an absolute variation of BP, regardless of BP profile. In contrast, the measurement of BP variability in our study was different from Phillips’ study. ABPM was used in our study and could reflect patient’s status more accurately. In contrast, our result was similar to those reported by Yong and Kaste,10 who reported that BP variability is increased after acute-phase outcomes after acute stroke. SBP variability is a strong and independent predictor of the 3-month outcome after acute stroke and intravenous thrombolysis. Table 4. Cox regression analysis of cardiovascular outcomes in IS patients without diabetes mellitus (n 5 175) Covariate
HR
95% CI
P value
Age, y SD of daytime SBP, mm Hg NIHSS scores at admission
1.040 1.103
1.002-1.079 1.011-1.203
.039 .027
1.089
1.024-1.158
.007
Abbreviations: CI, confidence interval; HR, hazard ratio; IS, ischemic stroke; NIHSS, National Institutes of Health Stroke Scale; SBP, systolic blood pressure. Adjusted for: gender (P 5 .661), dyslipidemia (P 5 .436), hypertension (P 5 .618), previous stroke (P 5 .276), coronary heart disease (P 5 .555), and SD of nighttime SBP (P 5 .423).
There are some differences between our study and the previous reports. First, most of previous studies investigated the relationship between BP variability during an acute IS and short-term clinical outcomes after an acute phase of stroke (from 10 days to 3 months after an acute stroke). In our study, we observed BP variability after an acute phase of stroke and longterm outcome in patients with stroke. Second, the use of different methods of BP variability assessment, for example, visit-to-visit BP variability recordings in the previous study versus ABPM readings in our study, and the use of difference methods of BP measurement, may explain some of the differences. BP variability pointed to short-term BP variation in our study. Kellertet al7 did not use 24-hour ABPM and included only patients who received thrombolysis. In addition, patients were stratified according to diabetes in our study, in which it was different from other studies. In our study, patients with greater BP variability were older and had a greater proportion of hypertension and greater BP compared with those in lower BP variability without diabetes mellitus. Furthermore, Cox analysis found the strong association of BP variability with cardiovascular outcomes after adjustment for gender, dyslipidemia, hypertension, previous stroke, coronary heart disease, and SD of nighttime SBP in IS patients without diabetes mellitus. In our study, we didn’t find a relationship between BP variability and cardiovascular outcomes in all patients and diabetes subgroup, perhaps because diabetic patients have a sign of unbalanced autonomic control of circulation and/or of an increased arterial stiffness, and an independent predictor of cardiovascular complications.20 Correspondingly, patients with diabetes mellitus were excluded in Dawson’s studies on BP and stroke outcome.8 In addition, our study showed a high rate of cardiovascular events (20.4%), possibly because cardiovascular outcomes included rehospitalization with stroke recurrence, myocardial infarction, angina pectoris, and heart failure. In contrast, our result was similar to that reported by Wang et al23 in the China National Stroke Registry. These authors reported that the cumulative stroke recurrence rate was 17.7% at 1 year after stroke onset. Some weaknesses of the present study should also be discussed. First, our data came from in-hospital patients, which could have patient selection bias. Second, the present study comprised a relatively small number of patients. A larger sample and long-term follow-up should be further investigated. Third, a comparison between patients with diabetes mellitus and without diabetes mellitus was not included in the present study. In fact, diabetes status also was associated with cardiovascular outcomes in IS patients according to our data. In addition, the patients with small-vessel disease or cardioembolism were not included in our study and need to be further investigated in the future.
BP VARIABILITY AND STROKE OUTCOME
In summary, we show that SBP variability is associated with cardiovascular outcomes in stroke patients without diabetes mellitus. We didn’t find a correlation between SBP variability and cardiovascular outcomes in stroke patients with diabetes mellitus, perhaps because there lies in an unbalanced autonomic dysfunction and/or an increased arterial stiffness in diabetic patients. Acknowledgments: The authors wish to express their great appreciation for the work performed by the staff of the ABPM Room at Beijing Tiantan Hospital, Capital Medical University.Appendix
Supplementary Data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.05. 030.
References 1. Harper G, Fotherby MD, Panayiotou BJ, et al. The changes in blood pressure after stroke: abolishing the ‘white coat effect’ with 24 h ambulatory monitoring. J Intern Med 1994;235:343-346. 2. Rothwell PM. Limitations of the usual blood-pressure hypothesis and importance of variability, instability, and episodic hypertension. Lancet 2010;375:938-948. 3. Rothwell PM, Howard SC, Dolan E, et al. Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet 2010; 375:895-905. 4. Poortvliet RK, Ford I, Lloyd SM, et al. Blood pressure variability and cardiovascular risk in the PROspective Study of Pravastatin in the Elderly at Risk (PROSPER). PLoS One 2012;7:e52438. 5. Floras JS. Blood pressure variability: a novel and important risk factor. Can J Cardiol 2013;29:557-563. 6. Geeganage C, Tracy M, England T, et al, for the TAIST Investigators. Relationship between baseline blood pressure parameters (including mean pressure, pulse pressure, and variability) and early outcome after stroke data from the Tinzaparin in acute ischaemic stroke trial (TAIST). Stroke 2011;42:491-493. 7. Kellert L, Sykora M, Gumbinger C, et al. Blood pressure variability after intravenous thrombolysis in acute stroke does not predict intracerebral hemorrhage but poor outcome. Cerebrovasc Dis 2012;33:135-140. 8. Dawson SL, Manktelow BN, Robinson TG, et al. Which parameters of beat-to-beat blood pressure and variability
2457
9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
19.
20.
21.
22.
23.
best predict early outcome after acute ischemic stroke? Stroke 2000;31:463-468. Endo K, Kario K, Koga M, et al. Impact of early blood pressure variability on stroke outcomes after thrombolysis: the SAMURAI rt-PA registry. Stroke 2013;44:816-818. Yong M, Kaste M. Association of characteristics of blood pressure profiles and stroke outcomes in the ECASS-II trial. Stroke 2008;39:366-372. Liu W, Liu R, Sun W, et al, CASISP Study Group. Different impacts of blood pressure variability on the progression of cerebral microbleeds and white matter lesions. Stroke 2012;43:2916-2922. Weiss A, Beloosesky Y, Kenett RS, et al. Systolic blood pressure during acute stroke is associated with functional status and long-term mortality in the elderly. Stroke 2013; 44:2434-2440. Carlberg B, Asplund K, H€ agg E. The prognostic value of admission blood pressure in patients with acute stroke. Stroke 1993;24:1372-1375. Fiorelli M, Alperovitch A, Argentino C, et al. Prediction of long-term outcome in the early hours following acute ischaemic stroke. Arch Neurol 1995;52:250-255. Phillips RA, Butkevich A, Sheinart KF, Tuhrim S. Dipping is superior to cusums analysis in assessment of the risk of stroke in a case–control study. Am J Hypertens 2001;14(7 Pt 1):649-652. Allen CM. Predicting the outcome of acute stroke: a prognostic score. J Neurol Neurosurg Psychiatry 1984;47:475-480. Jørgensen H, Nakayama H, Raaschou HO, Olsen TS. Effect of blood pressure and diabetes on stroke in progression. Lancet 1994;344:156-159. Goldstein LB, Adams R, Becker K, et al. Primary prevention of ischemic stroke: a statement for healthcare professionals from the Stroke Council of the American Heart Association. Circulation 2001;103:163-182. Jia Q, Zheng H, Zhao X, et al, Investigators for the Survey on Abnormal Glucose Regulation in Patients With Acute Stroke Across China (ACROSS-China). Abnormal glucose regulation in patients with acute stroke across China: prevalence and baseline patient characteristics. Stroke 2012;43:650-657. Parati G, Bilo G. Should 24-h ambulatory blood pressure monitoring be done in every patient with diabetes? Diabetes Care 2009;32(Suppl 2):S298-S304. Jia Q, Zhao X, Wang C, et al. Diabetes and poor outcomes within 6 months after acute ischemic stroke: the China National Stroke Registry. Stroke 2011;42:2758-2762. Xu J, Zhao X, Wang Y, et al. Impact of a better persistence with antihypertensive agents on ischemic stroke outcomes for secondary prevention. PLoS One 2013;8:e65233. Wang Y, Xu J, Zhao X, et al. Association of hypertension with stroke recurrence depends on ischemic stroke subtype. Stroke 2013;44:1232-1237.
Background: The association between blood pressure (BP) variability and stroke outcome is controversial, and there are few studies that have focused on the impact of BP variability in diabetic patients with stroke. Therefore, we aimed to examine the impact of BP variability on cardiovascular outcome in diabetic and nondiabetic patients with stroke. Methods: A total of 373 ischemic stroke patients with large artery atherosclerosis were recruited and followed up. Ambulatory BP monitoring was performed in all patients and divided according to the 25th and 75th percentiles interval of SD of daytime systolic BP (SBP). Kaplan-Meier analysis and Cox regression were used to assess the relationship between BP variability and cardiovascular outcomes including stroke recurrence, vascular events and cardiovascular death. Results: The 339 patients were included in the final analysis. During an average follow-up of 19.0 6 5.1 months (.6-26.8 months), 69 (20.4%) cardiovascular events occurred in all patients. Kaplan-Meier analysis found that there were no differences in cardiovascular events-free survival among the different BP variability groups in diabetic patients (P 5 .995); however, nondiabetic patients with greater BP variability showed a lesser cardiovascular events-free survival (P 5 .039). Through Cox regression we found the SD of daytime SBP (hazard ratio 1.103; 95% CI 1.0111.203) was associated with cardiovascular outcomes in nondiabetic patients with stroke. Conclusions: We show that SBP variability is associated with cardiovascular outcomes in stroke patients without diabetes, but we didn’t find a correlation between SBP variability and cardiovascular outcomes in stroke patients with diabetes. Key Words: Ischemic stroke—outcome—ambulatory blood pressure monitoring—blood pressure variability—diabetes mellitus. Ó 2014 by National Stroke Association
From the *Department of Cardiology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China; †Department of Cardiology, Beijing Shijingshan Hospital, Capital Medical University, Beijing, China; ‡Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China; xChina National Clinical Research Center for Neurological Diseases, Beijing, China; kCenter of Stroke, Beijing Institute for Brain Disorders, Beijing, China; and {Beijing Key Laboratory of Translational Medicines for Cerebrovascular Disease, Beijing, China. Received April 22, 2014; revision received May 21, 2014; accepted May 24, 2014. Supported by a grant from National Key Technology Research and Development Program of the Ministry of Science and Technology of
2450
China (2013BAI09B03), and a grant from Beijing Institute for Brain Disorders (BIBD-PXM2013_014226_07_000084). Address correspondence to Bu-Xing Chen, MD, Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, No. 6 Tian Tan Xi Li, Dongcheng District, Beijing 100050, P.R.China. E-mail: [email protected]; and Xing-Quan Zhao, MD, Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, No. 6 Tian Tan Xi Li, Dongcheng District, Beijing 100050, P.R.China. E-mail: [email protected]. 1 Both Bu-Xing Chen and Jun-Ping Tian are the first authors of this paper. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.05.030
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 9 (October), 2014: pp 2450-2457
BP VARIABILITY AND STROKE OUTCOME
Introduction Ambulatory blood pressure monitoring (ABPM) not only provides information on blood pressure (BP) levels but on the diurnal changes in BP as well.1 Previous studies have indicated that BP variability is an important trigger of vascular events2-4 and can provide a more precise estimate of individual cardiovascular risk.5 The association between BP variability and cardiovascular outcomes in patients with stroke has been investigated. Because of the methodologic differences, patient selection differences, and other differences, previous published studies were vague on the association of BP variability with stroke outcome. Some studies found that increased BP variability during an acute ischemic stroke (IS) was associated with worse clinical outcomes during a short period after an acute phase of stroke.6-11 On the contrary, some reports suggested that increased BP variability is of little prognostic value12-15 or may even indicate a good prognosis.16,17 Recently, diabetes mellitus has been shown to be an independent risk factor for IS.18 The prevalence of diabetes ranges from 21% to 44.4% among patients with acute IS.19 Jia et al19 suggested that the prevalence of abnormal glucose regulation was 68.7% among all the patients with acute stroke in China. In addition, some evidence indicates that the BP variability is increased in diabetic subjects20 and diabetes is associated with greater mortality from stroke.17,21 Up to now, there are few studies focusing on BP variability of diabetic patients with stroke. Therefore, we aimed to examine the impact of BP variability on cardiovascular outcomes in diabetic and nondiabetic patients with IS.
Figure 1. The patient selection flow in this study. Abbreviations: IS, ischemic stroke; SBP, systolic blood pressure.
2451
Patients and Methods Patients We performed an observational study of prevalent patients with IS who were admitted to the Cerebrovascular Center of Beijing Tiantan Hospital, People’s Republic of China, between May 2010 and August 2011. The patients were diagnosed with IS according to the World Health Organization criteria, which was confirmed by computed tomography (CT) or magnetic resonance imaging (MRI) of the brain. The enrolled IS patients, who were older than 18 years of age, were classified as having largeartery atherosclerosis according to the TOAST (ie, Trial of Org 10172 in Acute Stroke Treatment) criteria. None of patients suffered from infective endocarditis, atrial fibrillation, valvular heart disease, or malignant tumor that could lead to stroke. The patients who had any severe alimentary tract hemorrhage, infectious disease, diseases known to affect the autonomic nervous system, or severe heart failure (left ventricular ejection fraction ,45%) were excluded from the present study. In addition, patients with intracranial aneurysm who were confirmed by brain CT or MRI were not also included in our study. Therefore, 373 IS patients were recruited. The ethical committee of Beijing Tiantan Hospital approved the study protocol, and all patients or their designated relatives provided informed consent.
Baseline Characteristics The following baseline characteristics were investigated: age, gender, height, body weight, and duration of hypertension. Patients were defined as having
B.-X. CHEN ET AL.
2452
Table 1. Baseline characteristics of the study population and comparisons among the 3 different blood pressure variability groups Variables
Group 1 (#9.6)
Group 2 (.9.6 to #14.5)
Group 3 (.14.5)
Number Age, y Gender, male, % BMI, kg/m2 Diabetes mellitus, % Hypertension, % Previous stroke, % Coronary heart disease, % Dyslipidemia, % Current or previous smoker, % Moderate or heavy drinking, % Left ventricular ejection fraction, % Glycosylated hemoglobin, % Triglycerides, mmol/L Total cholesterol, mmol/L LDL cholesterol, mmol/L HDL cholesterol, mmol/L NIHSS scores at admission Total SBP, mm Hg Mean daytime SBP, mm Hg Mean nighttime SBP, mm Hg Total DBP, mm Hg Mean daytime DBP, mm Hg Mean nighttime DBP, mm Hg SD of total SBP, mm Hg SD of daytime SBP, mm Hg SD of nighttime SBP, mm Hg SD of total DBP, mm Hg SD of daytime DBP, mm Hg SD of nighttime DBP, mm Hg Dipper pattern Dipper Nondipper Cardiovascular events
86 55.7 6 9.5 79.1 24.9 6 3.1 47.4 75.6 24.4 14.0 70.9y 68.6 27.9 69.3 6 6.3 6.8 6 1.8 1.68 6 1.04 4.13 6 .94 2.43 6 .72 1.03 6 .24 4.8 6 4.8 130.4 6 16.9 131.8 6 17.0 125.8 6 17.9y,* 82.1 6 10.2 83.1 6 10.5 78.3 6 10.5 9.1 6 1.7 8.0 6 1.2 8.8 6 3.8*,y 8.3 6 1.9 7.6 6 1.9 8.3 6 2.9
170 59.3 6 11.8z 72.4 25.2 6 3.3 46.5 86.5 24.1 11.8 77.6 58.8 30.0 67.7 6 6.3 6.7 6 1.7 1.71 6 1.01 4.27 6 1.13 2.56 6 .96 1.05 6 .23 4.9 6 4.9 136.5 6 16.5* 137.6 6 16.9* 132.5 6 18.2 82.7 6 11.0 83.5 6 11.1 79.7 6 12.0 12.5 6 2.2* 11.9 6 1.4* 10.6 6 4.6 9.9 6 1.9* 9.6 6 2.0* 9.1 6 3.1
83 62.5 6 10.8y 68.7 25.3 6 3.4 36.4 89.2 28.9 7.2 88.0 56.6 15.7z 68.5 6 6.7 6.4 6 1.6 1.44 6 1.04 3.98 6 1.24 2.36 6 1.09 1.04 6 .22 5.8 6 4.9 142.6 6 16.8y 144.0 6 17.0y 137.8 6 19.9z 84.4 6 12.4 85.1 6 12.8 81.7 6 13.2 17.3 6 2.7y 17.3 6 2.7y 12.0 6 4.5z 11.9 6 2.3y 11.7 6 2.4y 10.2 6 3.5y,z
18.6 81.4 16 (18.6)
15.9 84.1 34 (20.0)
28.9 71.1 19 (22.9)
Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; NIHSS, National Institutes of Health Stroke Scale; SBP, systolic blood pressure. *P , .01, group 1 versus group 2, group 2 versus group 3. yP , .01, group 1 versus group 3. zP , .05, group 1 versus group 2, group 2 versus group 3.
hypertension if their BP was $140/90 mm Hg on repeated measurements during the hospitalization or if they were taking an antihypertensive medication. Other risk factors were defined as follows: history of stroke (defined as a medical chart–confirmed history of stroke, including IS, intracerebral hemorrhage, or subarachnoid hemorrhage), coronary heart disease (a reported history of myocardial infarction or cardiac surgery, or with a final diagnosis of myocardial infarction at discharge), diabetes mellitus (defined as a reported history of diabetes mellitus, or use of insulin or oral hypoglycemic agents),21,22 dyslipidemia (total cholesterol measurement $240 mg/ dL, high-density lipoprotein measurement ,35 mg/dL, or the use of lipid-lowering agents), current or previous smoking (defined as an individual who smoked at the time of stroke or had quit smoking within 1 year), and moderate or heavy drinking ($2 standard alcoholic beverages consumed per day). Other clinical features
Figure 2. The Kaplan-Meier of cardiovascular events-free survival according to BP variability in all patients. Abbreviation: BP, blood pressure.
BP VARIABILITY AND STROKE OUTCOME
2453
Table 2. Subgroup analysis in IS patients with diabetes mellitus Variables
Group 1 (#9.4)
Group 2 (.9.4 to #14.0)
Group 3 (.14.0)
Number Age, y Gender, male, % BMI, kg/m2 Hypertension, % Previous stroke, % Coronary heart disease, % Dyslipidemia, % Current or previous smoker, % Moderate or heavy drinking, % Left ventricular ejection fraction, % Glycosylated hemoglobin, % Triglycerides, mmol/L Total cholesterol, mmol/L LDL cholesterol, mmol/L HDL cholesterol, mmol/L NIHSS scores at admission Total SBP, mmHg Mean daytime SBP, mmHg Mean nighttime SBP, mmHg Total DBP, mmHg Mean daytime DBP, mmHg Mean nighttime DBP, mmHg SD of total SBP, mmHg SD of daytime SBP, mmHg SD of nighttime SBP, mmHg SD of total DBP, mmHg SD of daytime DBP, mmHg SD of nighttime DBP, mmHg Dipper pattern Dipper Nondipper Cardiovascular events
35 60.3 6 8.1 68.6 25.0 6 3.7 85.7 31.4 20.0 77.1 57.1 14.3 69.0 6 5.5 8.5 6 1.9 1.83 6 1.21 4.42 6 1.10 2.60 6 .82 1.09 6 .25 6.3 6 4.3 134.9 6 16.2 135.7 6 16.5 132.1 6 17.2 82.8 6 8.4 83.7 6 8.5 80.0 6 9.5 9.2 6 1.7 8.1 6 1.1 9.0 6 3.8y,z 7.9 6 1.4 7.4 6 1.4 7.9 6 2.7y
70 61.2 6 9.7 65.7 25.8 6 3.0 91.4 22.9 11.4 78.6 52.9 32.9 67.5 6 5.3 7.7 6 1.8 1.67 6 .77 4.21 6 1.14 2.49 6 .96 1.04 6 .23 4.9 6 5.2 140.2 6 15.5 141.3 6 15.9 136.7 6 16.9 83.8 6 10.0 84.6 6 10.0 80.8 6 11.4 12.3 6 1.9* 11.7 6 1.3* 11.0 6 3.5 9.7 6 2.0* 9.3 6 2.2* 8.8 6 2.6
34 62.1 6 10.5 58.8 25.6 6 2.5 91.2 35.3 5.9 79.4 50.0 20.6 67.1 6 7.5 7.8 6 2.0 1.49 6 .81 4.16 6 1.51 2.52 6 1.40 1.03 6 .18 6.2 6 5.6 145.4 6 14.0y 147.8 6 15.1y 137.6 6 14.8 84.0 6 9.5 85.2 6 10.0 79.3 6 9.3 17.1 6 2.9y 16.9 6 2.9y 11.7 6 4.1 11.1 6 2.0y 10.7 6 2.2y 9.8 6 2.5
17.1 82.9 12 (34.3)
12.9 87.1 19 (27.1)
32.4 67.6 10 (29.4)
Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HDL, high-density lipoprotein; IS, ischemic stroke; LDL, low-density lipoprotein; NIHSS, National Institutes of Health Stroke Scale; SBP, systolic blood pressure. *P , .01, group 1 versus group 2, group 2 versus group 3. yP , .01, group 1 versus group 3. zP , .05, group 1 versus group 2.
included severity of stroke on admission (National Institutes of Health Stroke Scale score). Electrocardiogram, echocardiography, carotid artery ultrasound, and brain CT or MRI were performed in IS patients. Left ventricular ejection fraction was determined by echocardiography. Biochemistry parameters such as glycosylated hemoglobin, triglycerides, total cholesterol, low-density lipoprotein cholesterol, and high-density lipoprotein cholesterol were examined.
24-Hour ABPM ABPM was performed with a portable lightweight device (Mobil-O-Graph; S/N: B19086, I.E.M., Stolberg, Germany). The accuracy of the ABPM was confirmed in each subject by simultaneous auscultation and sphygmo-
manometry. The patients underwent 24-hour ABPM after the acute stroke phase. Patients carried the device for 24 hours, and BP measurements were taken every 30 minutes during the day and evening (from 06:00 to 22:00) and every 60 minutes at night (from 22:00 to 06:00). The patients were instructed to keep their nondominant arm still and relaxed at the side during the measurements. Patients received verbal and written instructions on the monitors and completed a diary in which they recorded sleep medication, posture, and symptoms. An acceptable 24-hour ABPM recording for our study should have had $23 acceptable measurements during the 24 hours, $6 measurements at night, and effective reading records during the whole day .75%. We calculated mean 24-hour, day, and night systolic and diastolic BP. SD as an index of BP variability was calculated for
2454
24 hours and separately for day and night. Patients were classified according to the percent of nocturnal SBP dipping [100 3 (nocturnal SBP-daytime SBP)/daytime SBP]. The classifications were dipper, if the nocturnal fall in BP was $10%; nondipper, if the nocturnal fall in BP was ,10%.
Follow-Up and Events After an acute phase of stroke onset, the outcomes of all patients were assessed through telephone follow-up, including recurrence of stroke and time (rehospitalization with a diagnosis of new IS or intracerebral hemorrhage, diagnosed according to the Sacco criteria), combined vascular events (including stroke recurrence, myocardial infarction, angina pectoris, heart failure), rehospitalization (including stroke recurrence, cardiovascular disease, or other diseases), and death from any cause. The telephone follow-up was conducted centrally for all included patients and was based on a shared standardized interview protocol. If the patients answered they had experienced rehospitalization for any of the aforementioned events, we would contact the hospitals that admitted these patients to verify the diagnosis. All events were based on clear documentation in medical records.
Statistical Analysis Continuous variables were expressed as mean 6 SD or median (range), whereas categorical variables were expressed as ratio or percentage. The comparison of continuous variables among the 3 groups was analyzed by the use of analysis of variance (for post hoc analysis, Student-Newman-Keuls was performed if equal variance was assumed and Tamhane’s T2 test was performed if equal variance was not assumed). When the variables were not normally distributed, the nonparametric Kruskal-Wallis test was used. The comparison of categorical variables was performed using the c2test. The cumulative cardiovascular events-free survival was plotted as Kaplan-Meier curves among the 3 different groups, and the differences were assessed by the log-rank test. The associations between BP variability and cardiovascular outcomes were analyzed in multivariable Cox regression models, after we adjusted for potential confounders, including age, gender, dyslipidemia, hypertension, previous stroke, coronary heart disease, and stroke severity. Adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) were reported separately. A 2-tailed P-value less than .05 was considered to be statistically significant.
B.-X. CHEN ET AL.
to the TOASTcriteria. A total of 28 patients were excluded from analyses as having had noncardiovascular events during this period, and 339 patients were included in the final analysis (mean age 59.2 6 11.3, male 73.2%). During an average follow-up of 19.0 6 5.1 months (0.626.8 months), 69 (20.4%) cardiovascular events occurred in all enrolled patients. All patients were divided into the 3 groups according to the 25th and 75th percentiles interval of SD of daytime SBP. The division points were 9.6 (25%) and 14.5 (75%) respectively. Seen from Table 1, patients with greater BP variability were older and had greater BP. Kaplan-Meier curves are reported in Figure 2. There were no differences in cardiovascular events-free survival among the different BP variability groups in all patients (log-rank c2 5 1.974, P 5 .373). The IS patients were stratified according to diabetes mellitus or no diabetes, in which the information for 25 patients was missed. The subgroup analysis on the patients with diabetes mellitus (n 5 139) is shown in Table 2. The division points for the 25th and 75th percentile interval of SD of daytime SBP were 9.4 and 14.0, respectively. Patients with greater BP variability had greater BP compared with those in lower BP variability. Using Kaplan-Meier analysis (Fig 3), we found that there were no differences in cardiovascular events-free survival among the different BP variability subgroups with diabetes mellitus (log-rank c2 5 .010, P 5 .995). The 175 IS patients without diabetes mellitus were included in another subgroup analysis (mean age 58.2 6 12.2, male 80.0%; Table 3). The division points for the 25th and 75th percentile interval of SD of daytime SBP were 9.9 and 14.8, respectively. Patients with greater BP variability were older, had a greater proportion of
Results The flow diagram of the study is reported in Figure 1. Overall, 367 IS patients completed follow-up. All patients were diagnosed as large-artery atherosclerosis according
Figure 3. The Kaplan-Meier of cardiovascular events-free survival according to BP variability in IS patients with diabetes mellitus. Abbreviations: BP, blood pressure; IS, ischemic stroke.
BP VARIABILITY AND STROKE OUTCOME
2455
Table 3. Subgroup analysis in IS patients without diabetes mellitus Variables
Group 1 (#9.9)
Group 2 (.9.9 to #14.8)
Group 3 (.14.8)
Number Age, y Gender, male, % BMI, kg/m2 Hypertension, % Previous stroke, % Coronary heart disease, % Dyslipidemia, % Current or previous smoker, % Moderate or heavy drinking, % Left ventricular ejection fraction, % Glycosylated hemoglobin, % Triglycerides, mmol/L Total cholesterol, mmol/L LDL cholesterol, mmol/L HDL cholesterol, mmol/L NIHSS scores at admission Total SBP, mm Hg Mean daytime SBP, mm Hg ) Mean nighttime SBP, mm Hg Total DBP, mm Hg Mean daytime DBP, mm Hg Mean nighttime DBP, mm Hg SD of total SBP (mmHg) SD of daytime SBP, mm Hg SD of nighttime SBP, mm Hg SD of total DBP, mm Hg SD of daytime DBP, mm Hg SD of nighttime DBP, mm Hg Dipper pattern Dipper Nondipper Cardiovascular events
44 52.0 6 10.2*,y 90.9 24.8 6 2.6 63.6z 20.5 9.1 65.9 79.5 40.9 68.7 6 6.6 5.7 6 0.4 1.44 6 .66 4.06 6 .92 2.46 6 .84 .99 6 .20 3.8 6 5.3 126.0 6 16.9z 127.8 6 17.1 119.7 6 16.8z 80.8 6 11.7 82.2 6 12.1 76.3 6 11.2x 9.0 6 1.7 7.8 6 1.3 8.9 6 3.9 8.4 6 2.2 7.6 6 2.3 8.6 6 3.0
88 58.5 6 12.5 77.3 24.5 6 3.5 81.8 25.0 12.5 79.5 63.6 23.9 67.8 6 7.1 5.7 6 0.9 1.74 6 1.26 4.19 6 1.12 2.50 6 .91 1.05 6 .22 5.3 6 4.9 132.3 6 16.1 133.3 6 16.3 128.4 6 18.2 81.0 6 11.3 81.6 6 11.6 78.7 6 12.0 12.7 6 2.3* 12.2 6 1.4* 10.2 6 5.4 10.1 6 1.8* 9.9 6 1.9* 9.0 6 3.3
43 64.0 6 10.6z 74.4 24.9 6 3.6 90.7y 34.9 9.3 86.0 62.8 16.3x 69.5 6 6.0 5.7 6 0.5 1.43 6 1.22 4.00 6 1.02 2.31 6 .80 1.03 6 .24 5.7 6 4.4 142.3 6 16.8*,y 143.2 6 16.4*,y 139.1 6 21.9*,y 84.3 6 11.9 84.7 6 11.7 82.8 6 14.4 17.7 6 2.7y 17.8 6 2.6y 12.5 6 4.6y,z 12.3 6 2.5y 12.3 6 2.6y 10.5 6 4.0x,z
18.2 81.8 4 (9.1)
17.0 83.0 12 (13.6)
30.2 69.8 11 (25.6)
Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; HDL, high-density lipoprotein; IS, ischemic stroke; LDL, low-density lipoprotein; NIHSS, National Institutes of Health Stroke Scale; SBP, systolic blood pressure. *P , .01, group 1 versus group 2, group 2 versus group 3. yP , .01, group 1 versus group 3. zP , .05, group 1 versus group 2, group 2 versus group 3. xP , .05, group 1 versus group 3.
hypertension and greater BP than those in lower BP variability. During an average follow-up of 18.8 6 4.9 months (5.5-26.7 months), 27 (15.4%) cardiovascular events occurred. The event rates in group 1, group 2, and group 3 were 9.1%, 13.6%, and 25.6%, respectively. KaplanMeier curves are reported in Figure 4. Nondiabetic patients who had a greater BP variability showed a lower cardiovascular events-free survival (log-rank c2 5 6.466, P 5 .039). Furthermore, we used multivariable Cox regression analyses to investigate the association between BP variability and risk of cardiovascular outcomes in 175 IS patients without diabetes mellitus. After we adjusted for potential confounders, including gender, dyslipidemia, hypertension, previous stroke, coronary heart disease, and SD of nighttime SBP, age (HR 1.040; 95% CI
1.002-1.079), functional status score (HR 1.089; 95% CI 1.024-1.158), and SD of daytime SBP (HR 1.103; 95% CI 1.011-1.203) were associated with cardiovascular outcomes in IS patients without diabetes mellitus (Table 4).
Discussion Our study demonstrated that SBP variability was associated with cardiovascular events in IS patients without diabetes mellitus. On the contrary, there was no correlation between BP variability and cardiovascular events in patients with diabetes mellitus and the whole study population. Using a Kaplan-Meier analysis, we found that nondiabetic patients with greater BP variability showed a significant lesser cardiovascular events-free survival. Thus, our study contradicts earlier data by Phillips
B.-X. CHEN ET AL.
2456
Figure 4. The Kaplan-Meier of cardiovascular events-free survival according to BP variability in IS patients without diabetes mellitus. Abbreviations: BP, blood pressure; IS, ischemic stroke.
et al,15 who analyzed 24-hour ABPM recordings of 110 control subjects and 91 stroke survivors and found that there was no significant difference in BP variability between stroke survivors and controls, a finding not explained by the presence of reverse dippers (raisers) in both groups. Phillips et al15 tested the prognostic value of BP variability, measured as cusums-derived circadian alteration magnitude. Cusums-derived circadian alteration magnitude of BP is a measure of an absolute variation of BP, regardless of BP profile. In contrast, the measurement of BP variability in our study was different from Phillips’ study. ABPM was used in our study and could reflect patient’s status more accurately. In contrast, our result was similar to those reported by Yong and Kaste,10 who reported that BP variability is increased after acute-phase outcomes after acute stroke. SBP variability is a strong and independent predictor of the 3-month outcome after acute stroke and intravenous thrombolysis. Table 4. Cox regression analysis of cardiovascular outcomes in IS patients without diabetes mellitus (n 5 175) Covariate
HR
95% CI
P value
Age, y SD of daytime SBP, mm Hg NIHSS scores at admission
1.040 1.103
1.002-1.079 1.011-1.203
.039 .027
1.089
1.024-1.158
.007
Abbreviations: CI, confidence interval; HR, hazard ratio; IS, ischemic stroke; NIHSS, National Institutes of Health Stroke Scale; SBP, systolic blood pressure. Adjusted for: gender (P 5 .661), dyslipidemia (P 5 .436), hypertension (P 5 .618), previous stroke (P 5 .276), coronary heart disease (P 5 .555), and SD of nighttime SBP (P 5 .423).
There are some differences between our study and the previous reports. First, most of previous studies investigated the relationship between BP variability during an acute IS and short-term clinical outcomes after an acute phase of stroke (from 10 days to 3 months after an acute stroke). In our study, we observed BP variability after an acute phase of stroke and longterm outcome in patients with stroke. Second, the use of different methods of BP variability assessment, for example, visit-to-visit BP variability recordings in the previous study versus ABPM readings in our study, and the use of difference methods of BP measurement, may explain some of the differences. BP variability pointed to short-term BP variation in our study. Kellertet al7 did not use 24-hour ABPM and included only patients who received thrombolysis. In addition, patients were stratified according to diabetes in our study, in which it was different from other studies. In our study, patients with greater BP variability were older and had a greater proportion of hypertension and greater BP compared with those in lower BP variability without diabetes mellitus. Furthermore, Cox analysis found the strong association of BP variability with cardiovascular outcomes after adjustment for gender, dyslipidemia, hypertension, previous stroke, coronary heart disease, and SD of nighttime SBP in IS patients without diabetes mellitus. In our study, we didn’t find a relationship between BP variability and cardiovascular outcomes in all patients and diabetes subgroup, perhaps because diabetic patients have a sign of unbalanced autonomic control of circulation and/or of an increased arterial stiffness, and an independent predictor of cardiovascular complications.20 Correspondingly, patients with diabetes mellitus were excluded in Dawson’s studies on BP and stroke outcome.8 In addition, our study showed a high rate of cardiovascular events (20.4%), possibly because cardiovascular outcomes included rehospitalization with stroke recurrence, myocardial infarction, angina pectoris, and heart failure. In contrast, our result was similar to that reported by Wang et al23 in the China National Stroke Registry. These authors reported that the cumulative stroke recurrence rate was 17.7% at 1 year after stroke onset. Some weaknesses of the present study should also be discussed. First, our data came from in-hospital patients, which could have patient selection bias. Second, the present study comprised a relatively small number of patients. A larger sample and long-term follow-up should be further investigated. Third, a comparison between patients with diabetes mellitus and without diabetes mellitus was not included in the present study. In fact, diabetes status also was associated with cardiovascular outcomes in IS patients according to our data. In addition, the patients with small-vessel disease or cardioembolism were not included in our study and need to be further investigated in the future.
BP VARIABILITY AND STROKE OUTCOME
In summary, we show that SBP variability is associated with cardiovascular outcomes in stroke patients without diabetes mellitus. We didn’t find a correlation between SBP variability and cardiovascular outcomes in stroke patients with diabetes mellitus, perhaps because there lies in an unbalanced autonomic dysfunction and/or an increased arterial stiffness in diabetic patients. Acknowledgments: The authors wish to express their great appreciation for the work performed by the staff of the ABPM Room at Beijing Tiantan Hospital, Capital Medical University.Appendix
Supplementary Data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.05. 030.
References 1. Harper G, Fotherby MD, Panayiotou BJ, et al. The changes in blood pressure after stroke: abolishing the ‘white coat effect’ with 24 h ambulatory monitoring. J Intern Med 1994;235:343-346. 2. Rothwell PM. Limitations of the usual blood-pressure hypothesis and importance of variability, instability, and episodic hypertension. Lancet 2010;375:938-948. 3. Rothwell PM, Howard SC, Dolan E, et al. Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet 2010; 375:895-905. 4. Poortvliet RK, Ford I, Lloyd SM, et al. Blood pressure variability and cardiovascular risk in the PROspective Study of Pravastatin in the Elderly at Risk (PROSPER). PLoS One 2012;7:e52438. 5. Floras JS. Blood pressure variability: a novel and important risk factor. Can J Cardiol 2013;29:557-563. 6. Geeganage C, Tracy M, England T, et al, for the TAIST Investigators. Relationship between baseline blood pressure parameters (including mean pressure, pulse pressure, and variability) and early outcome after stroke data from the Tinzaparin in acute ischaemic stroke trial (TAIST). Stroke 2011;42:491-493. 7. Kellert L, Sykora M, Gumbinger C, et al. Blood pressure variability after intravenous thrombolysis in acute stroke does not predict intracerebral hemorrhage but poor outcome. Cerebrovasc Dis 2012;33:135-140. 8. Dawson SL, Manktelow BN, Robinson TG, et al. Which parameters of beat-to-beat blood pressure and variability
2457
9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
19.
20.
21.
22.
23.
best predict early outcome after acute ischemic stroke? Stroke 2000;31:463-468. Endo K, Kario K, Koga M, et al. Impact of early blood pressure variability on stroke outcomes after thrombolysis: the SAMURAI rt-PA registry. Stroke 2013;44:816-818. Yong M, Kaste M. Association of characteristics of blood pressure profiles and stroke outcomes in the ECASS-II trial. Stroke 2008;39:366-372. Liu W, Liu R, Sun W, et al, CASISP Study Group. Different impacts of blood pressure variability on the progression of cerebral microbleeds and white matter lesions. Stroke 2012;43:2916-2922. Weiss A, Beloosesky Y, Kenett RS, et al. Systolic blood pressure during acute stroke is associated with functional status and long-term mortality in the elderly. Stroke 2013; 44:2434-2440. Carlberg B, Asplund K, H€ agg E. The prognostic value of admission blood pressure in patients with acute stroke. Stroke 1993;24:1372-1375. Fiorelli M, Alperovitch A, Argentino C, et al. Prediction of long-term outcome in the early hours following acute ischaemic stroke. Arch Neurol 1995;52:250-255. Phillips RA, Butkevich A, Sheinart KF, Tuhrim S. Dipping is superior to cusums analysis in assessment of the risk of stroke in a case–control study. Am J Hypertens 2001;14(7 Pt 1):649-652. Allen CM. Predicting the outcome of acute stroke: a prognostic score. J Neurol Neurosurg Psychiatry 1984;47:475-480. Jørgensen H, Nakayama H, Raaschou HO, Olsen TS. Effect of blood pressure and diabetes on stroke in progression. Lancet 1994;344:156-159. Goldstein LB, Adams R, Becker K, et al. Primary prevention of ischemic stroke: a statement for healthcare professionals from the Stroke Council of the American Heart Association. Circulation 2001;103:163-182. Jia Q, Zheng H, Zhao X, et al, Investigators for the Survey on Abnormal Glucose Regulation in Patients With Acute Stroke Across China (ACROSS-China). Abnormal glucose regulation in patients with acute stroke across China: prevalence and baseline patient characteristics. Stroke 2012;43:650-657. Parati G, Bilo G. Should 24-h ambulatory blood pressure monitoring be done in every patient with diabetes? Diabetes Care 2009;32(Suppl 2):S298-S304. Jia Q, Zhao X, Wang C, et al. Diabetes and poor outcomes within 6 months after acute ischemic stroke: the China National Stroke Registry. Stroke 2011;42:2758-2762. Xu J, Zhao X, Wang Y, et al. Impact of a better persistence with antihypertensive agents on ischemic stroke outcomes for secondary prevention. PLoS One 2013;8:e65233. Wang Y, Xu J, Zhao X, et al. Association of hypertension with stroke recurrence depends on ischemic stroke subtype. Stroke 2013;44:1232-1237.
