Abstract
This study discusses the association between blood pressure (BP) variability at different time periods within first 24 hours after admission and the functional outcome in acute ischemic stroke (AIS). We observed BP variability within first 24 hours after admission and evaluated the association between BP variability at different time periods (4 am‐8 am, 10 am‐2 pm, 4 pm‐8 pm, 10 pm‐2 am) and the functional outcome in AIS. National Institute of Health Stroke Scale (NIHSS) and modified Rankin Scale (mRS) were applied to evaluate short‐ (7 days) and long‐term functional outcome. The 24 hours after admission and early morning (4 am‐8 am) systolic blood pressure (SBP) variability were associated with poor outcome at day 7 (adjusted OR = 1.567, 95% CI = 1.076‐2.282; adjusted OR = 1.507, 95% CI = 1.028‐2.209, respectively). Compared with the impact of the 24‐hour BP variability on long‐term functional outcome, the early morning SBP was proved to be a strongly independent predictor for functional outcome at 3 months (adjusted OR = 1.505, 95% CI = 1.053‐2.152), 6 months (adjusted OR = 1.560, 95% CI = 1.048‐2.226), and 12 months (adjusted OR = 1.689, 95% CI = 1.104‐2.584). The BP variability in other time period groups was shown to have no influence on functional outcome. In addition, attempts to explain early morning BP variability with baseline characteristic factors at admission found that baseline SBP is the most influential (2.71%) factor. About 95.87% of the SBP variability in early morning was unexplained. In our study, early morning SBP variability is the strongest independent predictor for functional outcome in (AIS) patients, and baseline SBP after admission should be monitored as a control indicator of early morning SBP variability in the treatment of AIS patients.
Keywords: 24‐hour blood pressure variability, acute ischemic stroke, coefficient of variation, early morning, outcome
1. INTRODUCTION
Stroke is the second most common cause of death in the world and also a complication of hypertension.1 Acute ischemic stroke (AIS) is one of the common types, with characterizations such as high disability, high mortality, and high recurrence rate.2 The poor prognosis after stroke brings a heavy burden to both families and society, and hence, it has become the most concerned issue by clinicians and families.
The systematic analysis of the global burden of disease in 2013, in which stroke burden and risk factors of stroke in 188 countries were analyzed, showed that hypertension ranks the first in the risk factors of stroke.2 High blood pressure (BP) level occurs in up to 80% of all patients with acute stroke.3 A meta‐analysis of 32 studies (10 892 patients were involved) summarized that high BP after admission in AIS is independently associated with poor outcomes.4 Nevertheless, in the acute phase of stroke, BP at admission can fluctuate due to factors such as level of consciousness, physiological, psychological, and pathological ones.5 BP variability refers to the degree of fluctuation of BP within a specified period, and it is a complex phenomenon that includes short‐term variability occurring within 24 hours after admission and variability over more‐prolonged periods.6 Previous studies7, 8, 9 also have proved that BP variability obtained by using 24‐hour ambulatory blood pressure monitoring (ABPM), especially systolic blood pressure (SBP) variability, was a risk factor for poor outcome in AIS patients.
At present, variant in daytime and nighttime of BP within 24 hours after admission and stroke risk are widely concerned, and daytime BP better related to stroke risk than did nighttime BP.10 A prospective cohort study of 7458 people recruited from Denmark, Belgium, Japan, Sweden, Uruguay, and China11 showed that daytime BP and nocturnal BP monitoring within 24 hours can provide valuable prognostic information. But, the method of dividing different time periods was not carefully designed. Previous studies divided 24‐hour recording periods into daytime of 16 hours (6 am‐10 pm) and nighttime of 8 hours (10 pm‐6 am), and found that mean daytime and nighttime BP was associated with the functional outcome after AIS.12 However, the relationship between BP variability at different time periods within the first 24 hours after admission and the functional outcome of AIS has not yet been investigated. Here, we hypothesize that BP variability at different time periods within the first 24 hours after admission may be actively indicative for functional outcome prediction, especially the early morning time periods. Previous studies performed in Japan have reported morning BP fluctuation to be related to the incidence of stroke supported our hypothesis.13, 14 However, another study performed in a white population shows no independent association of the early morning BP morning surge with cardiovascular outcomes.15 Here, in order to find out BP variability which time periods is more significant for poor prognosis, in this study, we used cohort study and aimed to investigate the influence of BP variability at different time periods within the first 24 hours after admission on the functional outcome at short term (7 days) and long term (3, 6, and 12 months) in patients with AIS.
2. METHODS
2.1. Study design and participants
We performed a single‐center, prospective cohort study, and AIS patients were recruited from April 1, 2016, to March 31, 2018. All patients were hospitalized within 24 hours after symptom onset at the Stroke ward of the Department of Neurology, the First Affiliated Hospital of Harbin Medical University, Harbin, China. Our diagnostic criteria for AIS patients were based on the Guidelines for Diagnosis and Treatment for Acute Ischemic Stroke.16 The diagnoses were also further confirmed by computed tomography (CT) and (or) magnetic resonance imaging (MRI) scanning with standard imaging techniques. Patients were excluded with any one of the following reasons: (a) lack of complete vascular imaging and complete ambulatory BP recordings; (b) modified Rankin Scale (mRS) score ≥3 before stroke onset and severe disturbance of consciousness (Glasgow coma scale score ≤8); (c) other severe systemic diseases other than AIS, such as intracerebral hemorrhage, hypertension encephalopathy, liver failure, renal failure, respiratory failure, heart failure, hypothyroidism, malignant tumors, and other disease which could influence the daily life ability of the patients; (d) lack of detailed value of BP or refused to receive continuous BP monitoring; (e) lack of informed consent or refused to cooperate; (f) non‐classified stroke according to Trial of Org 10172 in Acute Stroke Treatment; and (g) deceased or leaves during the hospitalization process within 7 days.
2.2. Ethics
This study was approved by the Ethics Committee of the First Affiliated Hospital of Harbin Medical University. Informed consent statement was obtained from each patient or family member.
2.3. Questionnaire and data collection
We adopted a self‐designed, unified format questionnaire in this study. This questionnaire has included information such as general demographic information (age, gender, average monthly income, educational level, and physical activity level); behavioral factors (drinking and smoking status); medical history (hypertension, diabetes mellitus, stroke, coronary heart disease and atrial fibrillation); and anthropometric parameters (height and weight were measured by World Health Organization regulations17). Furthermore, we recorded the National Institutes of Health Stroke Scale (NIHSS) score at admission. Clinical information, laboratory parameters (total cholesterol, triglyceride, high, and low‐density lipoprotein, fasting blood glucose, and homocysteine), and imaging data (computed tomography, magnetic resonance imaging, and ultrasound) were also included in the questionnaire.
Patients were assessed with the National Institutes of Health Stroke Scale (NIHSS; with higher scores indicating a greater deficit) at baseline, 24 hours after admission, and on day 7. The mRS, a measure of disability, was used to assess patients at month 3, 6, and 12 by professional neurologists.
2.4. Definition of baseline factors
Current smokers among the patients were defined as person who smokes at least one cigarette per day for 1 year or more and still smokes within 6 months before admission. Former smokers were defined as stopped smoking for at least 6 months before admission.18 Drinkers were defined as people who consumed ≥50 g alcoholic per week continuously for one more year and still drinking before admission. Body mass index (BMI) was calculated as dividing the weight of kilograms by the square of the height of the meter, underweight, normal weight, and overweight, and obese was defined as BMI < 18.5 kg/m2, 18.5 kg/m2 ≤ BMI < 24 kg/m2, 24 kg/m2 ≤ BMI < 28 kg/m2, ≥28 kg/m2, respectively.19 Abnormal blood liquid includes a large of lipid abnormalities and involve total cholesterol (TC) ≥6.20 mmol/L, high‐density lipoprotein cholesterol (HDL‐C) <1.03 mmol/L, low‐density lipoprotein cholesterol (LDL‐C) ≥4.13 mmol/L, and triglyceride (TG) ≥2.25 mmol/L.20
2.5. BP monitoring and analysis of BP variability
Baseline BP was measured using standard mercury sphygmomanometer when the patients reached the emergency room. We verified the accuracy of blood monitor monthly by standard mercury sphygmomanometer to maintain quality control. During the first 24 hours in stroke ward, SBP and diastolic blood pressure (DBP) values were recorded every 2 hours using a non‐invasive automatic bedside monitoring instrument (Japan Photoelectric Industrial Co. Ltd, model BSM‐6701C, Registration License SFDA (I) 20113212379).
The coefficient of variation (CV; 100 × standard deviation/mean) of SBP and DBP value was calculated to represent BP variability during the first 24 hours. The CV statistics of BP including the mean, median, standard deviation (SD), P25, P75, maximum and minimum were also calculated. As BP values in this study were recorded every 2 hours, we creatively self‐designed divided BP values of the first 24 hours after admission into four groups, 4 am‐8 am, 10 am‐2 pm, 4 pm‐8 pm, and 10 pm‐2 am These BP values were recorded at three time integer points in each time period. We calculated BP variability of each period using these three successive BP measurements.
2.6. Outcome definitions
Poor outcome at day 7 (short term) was defined by NIHSS score ≥5 in hospital.21 We estimated the functional outcome after discharge using the ordinal score on the modified Rankin scale (mRS; range, 0‐6, with higher scores indicating greater disability) at 3, 6, and 12 months (long term) visit. The mRS score was used as a clinician‐reported measure of global disability and assessed by professional neurologists, and functional independence (defined as a score on the mRS of 0‐2) represented a good outcome, whereas functional dependence (defined as a score on the mRS of ≥3) represented a poor one.22
2.7. Statistical analysis
Two trained investigators typed all data repeatedly into Epidata software with the blind method. Professional staffs initiated logic check in order to ensure accuracy. Patients' baseline characteristics were presented as mean ± SD for continuous variables and as frequency (%) for categorical variables. Shapiro‐Wilk test was used for examine the normality of BP variability. Pearson's chi‐square test was applied to examine the categorical variables. Multiple logistic regression models were used to examine the association between BP variability and functional outcome. To further analyze the impact of baseline factors on BP variability, multivariable linear regression model was applied. Values of P < .05 were considered to indicate statistically significant in all tests (two‐tailed). All statistical analyses were performed using the statistical analysis software (SAS version 9.4).
3. RESULTS
3.1. Patients characteristics
A total of 704 eligible patients (mean age: 63.30 ± 12.00 years, 34.38% females) with AIS were enrolled in the data analysis process. Table 1 shows the baseline characteristics and demographics of these enrolled patients. The mean ± SD of NIHSS scores at baseline were 9.42 ± 6.43. Among all cases, 549 cases (77.98%) had history of hypertension, and there were 520 cases (73.86%) with admission SBP ≥140 mm Hg; 388 (55.11%) cases had history of smoking. Characteristics of the CV of BP are shown in Table 2. Test of normality indicated the skewed distribution of CV of BP at any different time periods (all P < .0001). Compared with other time groups, the median value of CV of the first early morning (4 am‐8 am) BP was significantly higher. Patients lacking mRS data at 3 months (n = 47), 6 months (n = 53), and 12 month (n = 27), were additionally excluded in the analysis for functional outcome.
Table 1.
Demographics and baseline characteristics of 704 enrolled patients with acute ischemic stroke
| Indexes | Number (%) or mean ± SD |
|---|---|
| Age (y) | |
| <50 | 90 (12.78) |
| 50‐60 | 175 (24.86) |
| 60‐70 | 233 (33.10) |
| ≥70 | 206 (29.26) |
| Gender | |
| Male | 462 (65.62) |
| Female | 242 (34.38) |
| Average monthly income (RMB) | |
| <1000 | 148 (21.02) |
| 1000‐2000 | 190 (26.99) |
| 2000‐3000 | 247 (35.09) |
| ≥3000 | 119 (16.90) |
| Degree of education | |
| ≤Primary school | 190 (26.99) |
| Middle school or technical secondary | 294 (41.76) |
| High school or junior college | 188 (26.70) |
| ≥University | 32 (4.55) |
| BMI (kg/m2) | |
| <18.5 | 30 (4.26) |
| 18.5‐24 | 350 (49.72) |
| 24‐28 | 239 (33.95) |
| ≥28 | 85 (12.07) |
| Smoking status | |
| Never smoker | 316 (44.89) |
| Current smoker | 294 (41.76) |
| Former smoker | 94 (13.35) |
| Drinking | 297 (42.19) |
| Degree of physical activity | |
| Sedentary | 112 (15.91) |
| Light | 497 (70.60) |
| Moderate to high | 95 (13.49) |
| History of hypertension | 549 (78.00) |
| History of diabetes | 181 (25.71) |
| History of stroke | 305 (43.32) |
| History of coronary heart disease | 181 (25.71) |
| History of atrial fibrillation | 75 (10.65) |
| FBG (mmol/L) | 6.66 ± 2.78 |
| TC (mmol/L) | 4.12 ± 1.18 |
| TG (mmol/L) | 1.83 ± 1.13 |
| HDL (mmol/L) | 1.29 ± 0.32 |
| LDL (mmol/L) | 3.16 ± 0.81 |
| Hcy (umol/L) | 18.05 ± 14.25 |
| Admission NIHSS scores | 9.42 ± 6.43 |
| Admission SBP (mm Hg) | |
| <120 | 39 (5.54) |
| 120‐139 | 145 (20.60) |
| 140‐159 | 252 (35.80) |
| 160‐179 | 162 (23.01) |
| ≥180 | 106 (15.05) |
| TOAST classification | |
| Large artery atherosclerosis | 390 (55.40) |
| Small vessel occlusion | 233 (33.10) |
| Cardioembolism | 81 (11.50) |
| Delay (from symptom onset to blood pressure data collection) | |
| <12 h | 530 (75.28) |
| ≥12 h | 174 (24.72) |
All values are expressed with mean ± standard deviation, or frequency (%).
Abbreviations: FBG, fasting blood glucose; Hcy, homocysteine; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; NIHSS, National Institutes of Health Stroke Scale; TC, total cholesterol; TG, triglycerides; TOAST, Trial of Org 10172 in acute stroke trial.
Table 2.
Characteristics of coefficient of variation in blood pressure
| CV | Mean | SD | Median | P25‐P75 | Max | Min | W (P) |
|---|---|---|---|---|---|---|---|
| 24 h | |||||||
| SBP CV | 10.97 | 3.68 | 10.31 | 8.31‐12.98 | 28.93 | 1.98 | 0.9599 (<.00010) |
| DBP CV | 14.39 | 5.65 | 13.30 | 10.58‐17.70 | 41.95 | 3.97 | 0.9303 (<.00010) |
| 4 am‐8 am | |||||||
| SBP CV | 9.57 | 6.00 | 8.36 | 5.09‐12.55 | 37.19 | 0.35 | 0.9283 (<.00010) |
| DBP CV | 12.34 | 8.15 | 10.63 | 6.35‐17.20 | 47.50 | 0.00 | 0.9224 (<.00010) |
| 10 am‐2 pm | |||||||
| SBP CV | 8.48 | 5.63 | 7.41 | 4.41‐11.12 | 37.82 | 0.00 | 0.9114 (<.00010) |
| DBP CV | 11.36 | 7.85 | 9.89 | 5.69‐15.41 | 67.83 | 0.00 | 0.8987 (<.00010) |
| 4 pm‐8 pm | |||||||
| SBP CV | 8.74 | 5.46 | 7.85 | 4.91‐11.41 | 32.56 | 0.00 | 0.9220 (<.00010) |
| DBP CV | 11.38 | 8.18 | 9.23 | 5.52‐15.85 | 56.19 | 0.00 | 0.8980 (<.00010) |
| 10 pm‐2 am | |||||||
| SBP CV | 9.07 | 5.44 | 8.15 | 5.09‐12.44 | 32.33 | 0.00 | 0.9461 (<.00010) |
| DBP CV | 12.02 | 7.76 | 10.75 | 6.09‐16.22 | 49.69 | 0.00 | 0.9272 (<.00010) |
Abbreviations: CV, coefficient of variation; DBP, diastolic blood pressure; SBP, systolic blood pressure; SD, standard deviation.
3.2. Effects of the CV of BP on short‐term functional outcome
Table 3 indicates the association between the CV of BP at different time periods and functional outcome at day 7. Since the CV of SBP and DBP showed a skewed distribution during the first 24 hours after admission and other time periods, we further divided each CV of SBP and DBP into low variability (≤median) and high variability (>median) layers using the median value of BP variability parameters. As shown in Table 3, 405 patients (57.53%) had good functional outcome and 299 patients (42.47%) had poor functional outcome at day 7. The CV of SBP during the first 24 hours after admission was associated with poor functional outcome at day 7: The CV ≥10 or 11 of SBP was a determinant of poor outcome (odds ratio [OR] = 1.363, 95% confidence interval [CI] = 1.007‐1.846; OR = 1.494, 95% CI = 1.104‐2.023), and CV ≥11 of SBP was significantly associated with poor outcome even after adjusting for possible confounding factors (adjusted OR = 1.567, 95% CI = 1.076‐2.282). However, the CV of DBP had no relationship with poor functional outcome at day 7.
Table 3.
The association of the CV of blood pressure at different time periods with the functional outcome at day 7
| Groups | NIHSS < 5 (n = 405) | NIHSS ≥ 5 (n = 299) | Unadjusted OR (95% CI) | Adjusted OR (95% CI) |
|---|---|---|---|---|
| 24 h | ||||
| SBP CV < 10 | 192 (47.41) | 119 (39.80) | 1.363 (1.007‐1.846)* | 1.262 (0.874‐1.824) |
| SBP CV ≥ 10 | 213 (52.59) | 180 (60.20) | ||
| SBP CV < 11 | 251 (61.98) | 156 (52.17) | 1.494 (1.104‐2.023)* | 1.567 (1.076‐2.282)* |
| SBP CV ≥ 11 | 154 (38.02) | 143 (47.83) | ||
| DBP CV < 13 | 194 (47.90) | 138 (46.15) | 1.073 (0.795‐1.447) | 0.805 (0.557‐1.165) |
| DBP CV ≥ 13 | 211 (52.10) | 161 (53.85) | ||
| DBP CV < 14 | 236 (58.27) | 162 (54.18) | 1.181 (0.874‐1.596) | 0.866 (0.598‐1.254) |
| DBP CV ≥ 14 | 169 (41.73) | 137 (45.82) | ||
| 4 am‐8 am | ||||
| SBP CV < 8 | 211 (52.10) | 126 (42.14) | 1.493 (1.105‐2.018)* | 1.528 (1.024‐2.237)* |
| SBP CV ≥ 8 | 194 (47.90) | 173 (57.86) | ||
| SBP CV < 9 | 232 (57.28) | 145 (48.49) | 1.424 (1.055‐1.923)* | 1.507 (1.028‐2.209)* |
| SBP CV ≥ 9 | 173 (42.72) | 154 (51.51) | ||
| DBP CV < 10 | 204 (50.37) | 131 (43.81) | 1.302 (0.964‐1.757) | 1.030 (0.702‐1.511) |
| DBP CV ≥ 10 | 201 (49.63) | 168 (56.19) | ||
| DBP CV < 11 | 224 (55.31) | 146 (48.83) | 1.297 (0.961‐1.750) | 0.975 (0.664‐1.433) |
| DBP CV ≥ 11 | 181 (44.69) | 153 (51.17) | ||
| 10 am‐2 pm | ||||
| SBP CV < 7 | 189 (46.67) | 136 (45.48) | 1.049 (0.777‐1.415) | 1.025 (0.713‐1.473) |
| SBP CV ≥ 7 | 216 (53.33) | 163 (54.52) | ||
| SBP CV < 8 | 225 (55.56) | 158 (52.84) | 1.116 (0.826‐1.506) | 1.187 (0.825‐1.707) |
| SBP CV ≥ 8 | 180 (44.44) | 141 (47.16) | ||
| DBP CV < 9 | 189 (46.67) | 137 (45.82) | 1.035 (0.767‐1.396) | 1.060 (0.734‐1.530) |
| DBP CV ≥ 9 | 216 (53.33) | 162 (54.18) | ||
| DBP CV < 10 | 208 (51.36) | 149 (49.83) | 1.063 (0.788‐1.433) | 1.108 (0.769‐1.597) |
| DBP CV ≥ 10 | 197 (48.64) | 150 (50.17) | ||
| 4 pm‐8 pm | ||||
| SBP CV < 7 | 178 (43.95) | 129 (43.14) | 1.033 (0.764‐1.397) | 1.074 (0.745‐1.548) |
| SBP CV ≥ 7 | 227 (56.05) | 170 (56.86) | ||
| SBP CV < 8 | 218 (53.83) | 150 (50.17) | 1.158 (0.859‐1.562) | 1.171 (0.813‐1.686) |
| SBP CV ≥ 8 | 187 (46.17) | 149 (49.83) | ||
| DBP CV < 9 | 210 (51.85) | 136 (45.48) | 1.291 (0.957‐1.742) | 1.186 (0.824‐1.707) |
| DBP CV ≥ 9 | 195 (48.15) | 163 (54.52) | ||
| DBP CV < 10 | 220 (54.32) | 153 (51.17) | 1.135 (0.841‐1.531) | 1.068 (0.742‐1.539) |
| DBP CV ≥ 10 | 185 (45.68) | 146 (48.83) | ||
| 10 pm‐2 am | ||||
| SBP CV < 8 | 207 (51.11) | 136 (45.48) | 1.253 (0.929‐1.691) | 1.166 (0.812‐1.676) |
| SBP CV ≥ 8 | 198 (48.89) | 163 (54.52) | ||
| SBP CV < 9 | 241 (59.51) | 163 (54.52) | 1.226 (0.907‐1.658) | 1.132 (0.786‐1.630) |
| SBP CV ≥ 9 | 164 (40.49) | 136 (45.48) | ||
| DBP CV < 10 | 185 (45.68) | 138 (46.15) | 0.981 (0.727‐1.324) | 0.837 (0.583‐1.202) |
| DBP CV ≥ 10 | 220 (54.32) | 161 (53.85) | ||
| DBP CV < 11 | 205 (50.62) | 155 (51.84) | 0.952 (0.706‐1.284) | 0.879 (0.613‐1.261) |
| DBP CV ≥ 11 | 200 (49.38) | 144 (48.16) | ||
Abbreviations: 95% CI, 95% confidence interval; CV, coefficient of variation; DBP, diastolic blood pressure; n, number; NIHSS, National Institutes of Health Stroke Scale (poor discharge outcome was defined as NIHSS ≥ 5); OR, odds ratio; SBP, systolic blood pressure; adjusted factors for age, gender, average monthly income, degree of education, BMI, smoking status, drinking status, degree of physical activity, history of hypertension, history of diabetes, history of stroke, history of coronary heart disease, history of atrial fibrillation, fasting blood glucose, abnormal blood lipid, Hcy, NIHSS scores for admission, delay time.
Statistically significant difference (P < .05).
Among four different time periods (4 am‐8 am, 10 am‐2 pm, 4 pm‐8 pm, and 10 pm‐2 am), only the CV of SBP during 4 am‐8 am was significantly correlated with poor functional outcome at day 7: The CV ≥8 of SBP during 4 am‐8 am had poor outcome (OR = 1.493, 95% CI = 1.105‐2.018; adjusted OR = 1.528, 95% CI = 1.024‐2.237); the CV ≥9 of SBP during 4 am‐8 am had poor outcome (OR = 1.424, 95% CI = 1.055‐1.923; adjusted OR = 1.507, 95% CI = 1.028‐2.209, respectively). There was no relationship between CV of DBP during the first 4 am‐8 am and outcome.
3.3. Effects of the CV of BP on long‐term functional outcome
Of the 704 cases included in this study, 657 (93.32%), 604 (86.00%) and 577 (81.96%) cases had completed mRS scores at 3, 6, and 12 months, respectively. We performed the multivariable logistic regression analysis to see if the association between the CV of BP during the first 4 am‐8 am and poor outcome still exist on long‐term prognosis. The CV of SBP during the first 4 am‐8 am was found to be significantly related to poor functional outcome at 3, 6, and 12 months (Table 4): The CV ≥8 or 9 of SBP was associated with poor outcome at 3 months (adjusted OR = 1.573, 95% CI = 1.098‐2.254; adjusted OR = 1.505, 95% CI = 1.053‐2.152, respectively); the CV ≥8 or 9 of SBP was associated with poor outcome at 6 months (adjusted OR = 1.678, 95% CI = 1.123‐2.507; adjusted OR = 1.560, 95% CI = 1.048‐2.322, respectively); the CV ≥8 of SBP was associated with poor outcome at 12 months (OR = 1.648, 95% CI = 1.107‐2.453; adjusted OR = 1.689, 95% CI = 1.104‐2.584). However, there was no relationship between early morning (4 am‐8 am) DBP variability and functional outcome. In addition, compared with the effect of 24‐hour BP variability on long‐term functional outcome (Table 5), early morning BP variability (4 am‐8 am) was found to be a more significant predictive indication in AIS for long‐term functional outcome.
Table 4.
The relationship between the coefficient of variation of blood pressure at the 4 am‐8 am and functional outcome at 3, 6, and 12 mo
| Index | 3‐mo outcome | 6‐mo outcome | 12‐mo outcome | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| mRS < 3 (n = 441) | mRS ≥ 3 (n = 216) | Unadjusted OR (95% CI) | Adjusted OR (95% CI) | mRS < 3 (n = 448) | mRS ≥ 3 (n = 156) | Unadjusted OR (95% CI) | Adjusted OR (95% CI) | mRS < 3 (n = 448) | mRS ≥ 3 (n = 129） | Unadjusted OR (95% CI) | Adjusted OR (95% CI) | |
| SBPCV < 8 | 229 (51.93) | 88 (40.74) | 234 (52.23) | 63 (40.38) | 236 (52.68) | 52 (40.31) | ||||||
| SBP CV ≥ 8 | 212 (48.07) | 128 (59.26) | 1.571 (1.130‐2.184)* | 1.573 (1.098‐2.254)* | 214 (47.77) | 93 (59.62) | 1.614 (1.115‐2.336)* | 1.678 (1.123‐2.507)* | 212 (47.32) | 77 (59.69) | 1.648 (1.107‐2.453)* | 1.689 (1.104‐2.584)* |
| SBP CV < 9 | 253 (57.37) | 100 (46.30) | 258 (57.59) | 73 (46.79) | 257 (57.37) | 62 (48.06) | ||||||
| SBP CV ≥ 9 | 188 (42.63) | 116 (53.70) | 1.561 (1.125‐2.166)* | 1.505 (1.053‐2.152)* | 190 (42.41) | 83 (53.21) | 1.544 (1.071‐2.226)* | 1.560 (1.048‐2.322)* | 191 (42.63) | 67 (51.94) | 1.454 (0.982‐2.154) | 1.421 (0.935‐2.161) |
| DBPCV < 10 | 215 (48.75) | 97 (44.91) | 219 (48.88) | 70 (44.87) | 211 (47.10) | 63 (48.84) | ||||||
| DBPCV ≥ 10 | 226 (51.25) | 119 (55.09) | 1.167 (0.842‐1.618) | 1.005 (0.701‐1.441) | 229 (51.12) | 86 (55.13) | 1.175 (0.815‐1.694) | 1.073 (0.718‐1.602) | 237 (52.90) | 66 (51.16) | 0.933 (0.630‐1.380) | 0.852 (0.557‐1.303) |
| DBPCV < 11 | 238 (53.97) | 107 (49.54) | 242 (54.02) | 78 (50.00) | 235 (52.46) | 69 (3.49) | ||||||
| DBPCV ≥ 11 | 203 (46.03) | 109 (50.46) | 1.194 (0.862‐1.654) | 1.021 (0.712‐1.464) | 206 (45.98) | 78 (50.00) | 1.175 (0.816‐1.692) | 1.047 (0.701‐1.563) | 213 (47.54) | 60 (46.51) | 0.959 (0.648‐1.421) | 0.855 (0.558‐1.311) |
Abbreviations: 95% CI, 95% confidence interval; CV, coefficient of variation; DBP, diastolic blood pressure; mRS, modified Rankin Scale (poor outcome was defined as mRS ≥ 3); n, number; OR, odds ratio; SBP, systolic blood pressure; adjusted factors for age, gender, average monthly income, degree of education, BMI, smoking status, drinking status, degree of physical activity, history of hypertension, history of diabetes, history of stroke, history of coronary heart disease, history of atrial fibrillation, fasting blood glucose, abnormal blood lipid, Hcy, NIHSS scores for admission, delay time.
Statistically significant difference (P < .05).
Table 5.
The relationship between the coefficient of variation of blood pressure within 24 h after admission and functional outcome at 3, 6, and 12 mo
| Index | 3‐mo outcome | 6‐mo outcome | 12‐mo outcome | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| mRS < 3 (n = 441) | mRS ≥ 3 (n = 216) | Unadjusted OR (95% CI) | Adjusted OR (95% CI) | mRS < 3 (n = 448) | mRS ≥ 3 (n = 156) | Unadjusted OR (95% CI) | Adjusted OR (95% CI) | mRS < 3 (n = 448) | mRS ≥ 3 (n = 129） |
Unadjusted OR (95% CI) |
Adjusted OR (95% CI) |
|
| SBP CV < 10 | 205 (46.49) | 89 (41.20) | 211 (47.10) | 59 (37.82) | 208 (46.43) | 51 (39.53) | ||||||
| SBP CV ≥ 10 | 236 (53.51) | 127 (58.80) | 1.240 (0.892‐1.723) | 1.112 (0.771‐1.603) | 237 (52.90) | 97 (62.18) | 1.464 (1.008‐2.125)* | 1.374 (0.910‐2.074) | 240 (53.57) | 78 (60.47) | 1.325 (0.890‐1.975) | 1.223 (0.788‐1.899) |
| SBP CV < 11 | 267 (60.54) | 116 (53.70) | 276 (61.61) | 77 (49.36) | 276 (61.61) | 65 (50.39) | ||||||
| SBP CV ≥ 11 | 174 (39.46) | 100 (46.30) | 1.323 (0.952‐1.837) | 1.211 (0.839‐1.747) | 172 (38.39) | 79 (50.64) | 1.646 (1.140‐2.377)* | 1.565 (1.037‐2.361)* | 172 (38.39) | 64 (49.61) | 1.580 (1.065‐2.343)* | 1.512 (0.975‐2.347) |
| DBP CV < 13 | 220 (49.89) | 96 (44.44) | 225 (50.22) | 63 (40.38) | 225 (50.22) | 54 (41.86) | ||||||
| DBP CV ≥ 13 | 221 (50.11) | 120 (55.56) | 1.244 (0.897‐1.726) | 1.064 (0.739‐1.534) | 223 (49.78) | 93 (59.62) | 1.489 (1.029‐2.155)* | 1.239 (0.820‐1.871) | 223 (49.78) | 75 (58.14) | 1.401 (0.943‐2.082) | 1.184 (0.758‐1.848) |
| DBP CV < 14 | 264 (59.86) | 113 (52.31) | 267 (59.60) | 80 (51.28) | 267 (59.60) | 68 (52.71) | ||||||
| DBP CV ≥ 14 | 177 (40.14) | 103 (47.69) | 1.360 (0.979‐1.887) | 1.127 (0.782‐1.625) | 181 (40.40) | 76 (48.72) | 1.401 (0.972‐2.021) | 1.124 (0.747‐1.691) | 181 (40.40) | 61 (47.29) | 1.323 (0.893‐1.962) | 1.099 (0.708‐1.707) |
Abbreviations: 95% CI, 95% confidence interval; CV, coefficient of variation; DBP, diastolic blood pressure; mRS, modified Rankin Scale (poor outcome was defined as mRS ≥ 3); n, number; OR, odds ratio; SBP, systolic blood pressure; adjusted factors for age, gender, average monthly income, degree of education, BMI, smoking status, drinking status, degree of physical activity, history of hypertension, history of diabetes, history of stroke, history of coronary heart disease, history of atrial fibrillation, fasting blood glucose, abnormal blood lipid, Hcy, NIHSS scores for admission, delay time.
Statistically significant difference (P < .05).
To further find out which admission baseline factor would explain most of the early morning SBP variability in AIS patients (Figure 1), linear multivariable regression analysis model manifested baseline SBP accounts for the largest proportion (2.71%) of explained variance. Surprisingly, all other factors also explained less than one percent of variance (history of smoking and BMI) and 95.87% of the SBP variability in early morning unexplained.
Figure 1.
Baseline factors on admission explaining early morning blood pressure. BMI, body mass index; NIHSS, National Institutes of Health Stroke Scale
4. DISCUSSION
To the best of our knowledge, this is the first study which discusses the significance of different time period BP variability within first 24 hours after admission in AIS for short‐ and long‐term outcomes. Our study's main highlights indicated that when data were unadjusted and adjusted for confounding factor, SBP variability not only within the first 24 hours after admission but also within the first early morning time period (4 am‐8 am within the first 24 hours after admission) was positive to predict the short‐term functional outcome (7 days). Other time periods (10 am‐2 pm, 4 pm‐8 pm, 10 pm‐2 am) within the first 24 hours had no association with functional outcome. More importantly, compared with the impact of 24‐hour BP variability on long‐term functional outcome, the first early morning (4 am‐8 am) SBP variability was a more reasonable indicator in AIS for predicting long‐term functional outcomes (at 3, 6, and 12 months). Moreover, when we attempted to explain early morning BP variability in the linear multivariable regression analysis by baseline data, the baseline SBP was the strongest predictor for the first early morning (4 am‐8 am) SBP variability, followed by history of smoking and BMI. However, there was still 95.87% variability in early morning that remained unexplained.
Our study surprisingly found that, compared to the first 24‐hour BP variability after admission, the first early morning BP variability is a more valuable predictor for short‐ and long‐term functional outcomes in AIS. The conclusion is in line with the observations of an independent relation between the morning BP variability and risk outcome.13, 14 However, no association of the morning BP variability with poor outcome has recently been observed in another study.15 The first early morning BP variability, which is a short‐term (within the 24 hours) BP variability, indicated the 4 am‐8 am BP variability within the first 24 hours after admission in our study. For the short‐term BP variability, higher BP variability was shown to be associated with an increased risk of death and disability after stroke.23, 24 Here, our study shows the significant inverse association between 4 am‐8 am BP variability and on short‐ and long‐term functional outcomes in AIS patients. Moreover, it was hypothesized that short‐term BP variability could induce endothelial cell damage and inflammation, and this might be one of the potential mechanisms of target organ damage.25 This hypothesis was supported by our conclusion above.
The observed relationship between early morning BP variability and outcomes can also be explained from several different aspects. It has long been known that BP is characterized by spontaneous variations. Over the 24 hours, BP undergoes variations in response to physical activity, sleep, and stimuli of various physiological or pathological nature and duration. Stroke‐specific mechanisms play a role in acute hypertensive response in patients with AIS. Because of the widespread distribution of the BP control areas in brain, most stroke lesion could involve these areas to some extent. Increased intracranial pressure, stress response to sudden illness, headache, or urinary retention may also lead to this hypertensive response. Thus, dynamic BP fluctuation becomes greater in AIS. Meanwhile, cerebral auto‐regulation is impaired, and cerebral circulation become an increasingly pressure‐dependant after stroke. The greater BP fluctuations may exacerbate the deleterious effects of ischemic or reperfusion injury on salvageable ischemic tissue. Our breakthrough discovery is the predictive of the 4 am‐8 am BP variability, that is, the first early morning SBP variability. Compared with the first 24‐hour BP variability, morning blood pressure surge (MBPS) in the acute phase of AIS appears to be the key factor underlying the first early morning BP variability. MBPS means an excessive increase in BP in the morning from the lowest BP at night, which is also characterized by important neurohormone changes during morning hours.26, 27 MBPS has been reported as the strongest independent risk factor for stroke in elderly hypertensive patients,13 and a 10 mm Hg morning SBP (the 2 hours after waking up) surge was associated with a 1.28% increase (95% CI: 1.10‐1.48) in the risk of stroke event complications. The International Data Base, which analyzed 5645 patients randomly from eight countries, revealed that MBPS could increase the risk of coronary events by 45%.26 The above may reveal the potential mechanism of the first early morning (4 am‐8 am) SBP variability from a pathophysiological point of view.
Interestingly, SBP variability was the strongest predictor compared with DBP variability in our study. Most studies, reporting the effect of BP variability on short‐ and long‐term functional outcome after AIS, have found SBP variability to be independently associated with poor outcome.7, 8, 9, 28 SBP variability is also recognized as a risk factor of stroke and cardiovascular events, independent from absolute BP level.8, 29 As for the value of early morning SBP or DBP variability in AIS, our findings recommend that early morning SBP variability is a better risk indicator in AIS. In Framingham Heart Study, systolic rather than diastolic BP was demonstrated to be the predominant risk factor for coronary heart disease.30 According to our data, increased morning SBP variability was significantly associated with hazard of poor outcome in followed up periods. Meanwhile, we did not discover any association between morning DBP variability and clinical outcome.
In summary, early morning SBP variability is a complex phenomenon affected by multiple factors. It is generally considered under the result of interactions between intrinsic (hypertension, stroke, and autonomic regulation) and extrinsic (history of hypertensive drugs, smoking, drinking, emotional stressors, and body position change) factors.6 We analyzed the influence of all the general demographics information, medical history, and baseline clinical information on early morning SBP variability by multivariable linear regression. Even by doing so, 95.87% the first early morning SBP variability was unexplained in our study. Higher baseline SBP, history of smoking, and BMI have been found to be correlated with increased morning SBP variability. Pre‐treatment SBP was reported to be the most influential baseline characteristic (5.55% explained variance) for 24‐hour BP variability in patients with AIS treated with intravenous thrombolysis.31 Interestingly, Wang et al32 also confirmed the baseline SBP was positively related to the CV of BP. A controlled trial study33 showed that the BP variability was increased in the smoking group compared with non‐smokers and related to increased sympathetic activity in the vascular system of smokers in the morning. In a cross‐sectional study of 32 482 cases, Chen et al19 found that every one‐unit increase in BMI was related with an increase in SBP variability of 0.077. BP management is necessary for AIS patients in clinical practice, but the individual strategies have not yet been established. Therefore, physicians can actively monitors BP, advocate cessation of smoking, recommend obese patients to lose weight, and factors intervened that may be associated with early morning SBP variability, in order to reduce morning SBP variability and improve the prognosis of AIS.
This study, besides the main limitations of uncontrolled data, still has several additional limitations. First, the selection criteria we used, including the inclusion and exclusion criteria, might be biased. Second, since this study was a single‐center study and was conducted in a single university hospital, and patients were comprised of the elderly Chinese population, the result might also be biased and need to be tested in other population. Third, the follow‐up of some cases was completed through phone calls, and some patients were lost to follow‐up or passed away at 3, 6, and 12 months, indicating that those follow‐up data may be underpowered. Fourth, the early morning BP variability data of the second day, the third day, and even longer time after AIS were not available after in our study, which should display the natural course of early morning BP variability after stroke. These uncertainties should be addressed in future prospective studies. Fourth, we creatively self‐designed divided BP values of the first 24 hours after admission into four groups may lead to selective bias. Last but not least, we demonstrated no influence of other baseline factors, such as stroke subtypes, on the first early morning SBP variability by multivariable linear regression, but more factors still should be explored in future studies. Further randomized, controlled clinical trials should be conducted to confirm whether the treatment of reducing early morning SBP variability after stroke could improve the prognosis in AIS. Despite these limitations, we interpret our study as novel and meaningful for patients care and future study.
5. CONCLUSION
In conclusion, this study illustrates the significance of early morning (4 am‐8 am) BP variability within the first 24 hours of admission in AIS patients and the influencing factors for early morning BP variability. The main strength of our findings indicated that the first early morning SBP variability, not the DBP variability, is more effective in predicting the short‐ and long‐term functional outcomes after AIS. Our study also suggests that we should focus on not only standard BP parameters but also early morning BP variability after AIS. Early morning BP variability after onset may represent a modifiable therapeutic target in AIS.
CONFLICT OF INTEREST
No potential conflict of interest and source of funding was declared by the authors.
AUTHOR CONTRIBUTIONS
Study design: Ying Tang, Jingbo Zhao. Data collection: Xuyang Geng, Xinyao Liu, Jiamin Wang, Hongwei Sun, Anqi Feng, Yanyan Sun, Hongwei Sun. Data analysis and interpretation of the results: Xuyang Geng, Jingbo Zhao. Manuscript drafting: Xuyang Geng, Xinyao Liu. Manuscript edition: Xuyang Geng. Administrative and material support: Ying Tang, Fang Li, Jingbo Zhao. Study supervision: Ying Tang, Jingbo Zhao.
ACKNOWLEDGMENTS
The authors acknowledge and thank all participants and medical staff for their support.
Geng X, Liu X, Li F, et al. Blood pressure variability at different time periods within first 24 hours after admission and outcomes of acute ischemic stroke. J Clin Hypertens. 2020;22:194–204. 10.1111/jch.13785
Funding information
This study was supported by National Natural Science Foundation of China (Grant number 81771508); Department of Education, Heilongjiang Province, China (Grant number 12521291); and Natural Science Foundation of Heilongjiang Province of China (Grant number D201235).
Contributor Information
Jingbo Zhao, Email: zhaojb168@sina.com.
Ying Tang, Email: hydtangying@hotmail.com.
REFERENCES
- 1. Tikhonoff V, Zhang H, Richart T, Staessen JA. Blood pressure as a prognostic factor after acute stroke. Lancet Neurol. 2009;8:938‐948. [DOI] [PubMed] [Google Scholar]
- 2. Feigin VL, Roth GA, Naghavi M, et al. Global burden of stroke and risk factors in 188 countries, during 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet Neurol. 2016;15:913‐924. [DOI] [PubMed] [Google Scholar]
- 3. Qureshi AI. Acute hypertensive response in patients with stroke: pathophysiology and management. Circulation. 2008;118:176‐187. [DOI] [PubMed] [Google Scholar]
- 4. Willmot M, Leonardi‐Bee J, Bath PM. High blood pressure in acute stroke and subsequent outcome: a systematic review. Hypertension. 2004;43:18‐24. [DOI] [PubMed] [Google Scholar]
- 5. Carlberg B, Asplund K, Hagg E. Factors influencing admission blood pressure levels in patients with acute stroke. Stroke. 1991;22:527‐530. [DOI] [PubMed] [Google Scholar]
- 6. Parati G, Ochoa JE, Lombardi C, Bilo G. Assessment and management of blood‐pressure variability. Nat Rev Cardiol. 2013;10:143‐155. [DOI] [PubMed] [Google Scholar]
- 7. Manning LS, Rothwell PM, Potter JF, Robinson TG. Prognostic significance of short‐term blood pressure variability in acute stroke: systematic review. Stroke. 2015;46:2482‐2490. [DOI] [PubMed] [Google Scholar]
- 8. 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. [DOI] [PubMed] [Google Scholar]
- 9. Yong M, Kaste M. Association of characteristics of blood pressure profiles and stroke outcomes in the ECASS‐II trial. Stroke. 2008;39:366‐372. [DOI] [PubMed] [Google Scholar]
- 10. Ohkubo T, Hozawa A, Nagai K, et al. Prediction of stroke by ambulatory blood pressure monitoring versus screening blood pressure measurements in a general population: the Ohasama study. J Hypertens. 2000;18:847‐854. [DOI] [PubMed] [Google Scholar]
- 11. Boggia J, Li Y, Thijs L, et al. Prognostic accuracy of day versus night ambulatory blood pressure: a cohort study. Lancet. 2007;370:1219‐1229. [DOI] [PubMed] [Google Scholar]
- 12. Tomii Y, Toyoda K, Suzuki R, et al. Effects of 24‐hour blood pressure and heart rate recorded with ambulatory blood pressure monitoring on recovery from acute ischemic stroke. Stroke. 2011;42:3511‐3517. [DOI] [PubMed] [Google Scholar]
- 13. Kario K, Ishikawa J, Pickering TG, et al. Morning hypertension: the strongest independent risk factor for stroke in elderly hypertensive patients. Hypertens Res. 2006;29:581‐587. [DOI] [PubMed] [Google Scholar]
- 14. Kario K, Pickering TG, Umeda Y, et al. Morning surge in blood pressure as a predictor of silent and clinical cerebrovascular disease in elderly hypertensives: a prospective study. Circulation. 2003;107:1401‐1406. [DOI] [PubMed] [Google Scholar]
- 15. Verdecchia P, Angeli F, Mazzotta G, et al. Day‐night dip and early‐morning surge in blood pressure in hypertension: prognostic implications. Hypertension. 2012;60:34‐42. [DOI] [PubMed] [Google Scholar]
- 16. Sun W, Ou Q, Zhang Z, Qu J, Huang Y. Chinese acute ischemic stroke treatment outcome registry (CASTOR): protocol for a prospective registry study on patterns of real‐world treatment of acute ischemic stroke in China. BMC Complement Altern Med. 2017;17:357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Physical status: the use and interpretation of anthropometry. Report of a WHO Expert Committee. World Health Organ Tech Rep Ser. 1995;854:1‐452. [PubMed] [Google Scholar]
- 18. Peng H, Han SH, Liu HY, Chandni V, Cai XQ, Zhang YH. Relationship of inflammation and endothelial dysfunction with risks to cardiovascular disease among people in Inner Mongolia of China. Biomed Environ Sci. 2013;26:792‐800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Chen H, Zhang R, Zheng Q, Yan X, Wu S, Chen Y. Impact of body mass index on long‐term blood pressure variability: a cross‐sectional study in a cohort of Chinese adults. BMC Public Health. 2018;18:1193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Kopin L, Dyslipidemia LC. Dyslipidemia. Ann Intern Med. 2017;167(11):ITC81‐ITC96. [DOI] [PubMed] [Google Scholar]
- 21. Ruuskanen EI, Laihosalo M, Kettunen J, et al. Predictors of discharge to home after thrombolytic treatment in right hemisphere infarct patients. J Cent Nerv Syst Dis. 2010;2:73‐79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Banks JL, Marotta CA. Outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials: a literature review and synthesis. Stroke. 2007;38:1091‐1096. [DOI] [PubMed] [Google Scholar]
- 23. Buratti L, Cagnetti C, Balucani C, et al. Blood pressure variability and stroke outcome in patients with internal carotid artery occlusion. J Neurol Sci. 2014;339:164‐168. [DOI] [PubMed] [Google Scholar]
- 24. Kang J, Ko Y, Park JH, et al. Effect of blood pressure on 3‐month functional outcome in the subacute stage of ischemic stroke. Neurology. 2012;79:2018‐2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Mancia G. Short‐ and long‐term blood pressure variability: present and future. Hypertension. 2012;60:512‐517. [DOI] [PubMed] [Google Scholar]
- 26. Li Y, Thijs L, Hansen TW, et al. Prognostic value of the morning blood pressure surge in 5645 subjects from 8 populations. Hypertension. 2010;55:1040‐1048. [DOI] [PubMed] [Google Scholar]
- 27. Kario K. Morning surge and variability in blood pressure: a new therapeutic target? Hypertension. 2005;45:485‐486. [DOI] [PubMed] [Google Scholar]
- 28. Geeganage C, Tracy M, England T, et al. 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. [DOI] [PubMed] [Google Scholar]
- 29. Rothwell PM. Limitations of the usual blood‐pressure hypothesis and importance of variability, instability, and episodic hypertension. Lancet. 2010;375:938‐948. [DOI] [PubMed] [Google Scholar]
- 30. Franklin SS, Larson MG, Khan SA, et al. Does the relation of blood pressure to coronary heart disease risk change with aging? The Framingham Heart Study. Circulation. 2001;103:1245‐1249. [DOI] [PubMed] [Google Scholar]
- 31. Kellert L, Hametner C, Ahmed N, et al. Reciprocal interaction of 24‐hour blood pressure variability and systolic blood pressure on outcome in stroke thrombolysis. Stroke. 2017;48:1827‐1834. [DOI] [PubMed] [Google Scholar]
- 32. Wang A, Li Z, Yang Y, et al. Impact of baseline systolic blood pressure on visit‐to‐visit blood pressure variability: the Kailuan study. Ther Clin Risk Manag. 2016;12:1191‐1196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Ragueneau I, Michaud P, Demolis JL, Moryusef A, Jaillon P, Funck‐Brentano C. Effects of cigarette smoking on short‐term variability of blood pressure in smoking and non smoking healthy volunteers. Fundam Clin Pharmacol. 1999;13:501‐507. [DOI] [PubMed] [Google Scholar]