Blood Pressure Changes in Acute Cerebral Infarction and Hemorrhage
Background and Purpose We sought to investigate the changes in blood pressure (BP) that occur after hospitalization of patients with different types of acute stroke.
Methods Twenty-four–hour ambulatory BP monitoring was performed on days 1 and 7 after admission to the hospital in 72 patients with acute stroke (44 thromboembolic strokes, 18 lacunar infarcts, and 10 intracerebral hemorrhages) and in 22 control patients. Stroke was categorized clinically into the above stroke subtypes with radiological confirmation. The controls were patients admitted with a range of acute medical problems other than stroke who were not severely ill or in significant pain. Left ventricular hypertrophy was assessed with echocardiography. Multiple linear regression was used to determine the effect of stroke category on BP after adjustment for the effects of potential confounders.
Results Patients with thromboembolic and lacunar strokes had significantly higher systolic BP (SBP) on day 1 than control subjects (mean, 8.6% and 13.2%, respectively). Diastolic BP (DBP) was also significantly higher for patients with thromboembolic and lacunar strokes on day 1 (mean, 11.7% and 14.6%, respectively). Patients with intracerebral hemorrhage had SBP 9.7% and DBP 6.3% higher than control subjects on day 1, but the results did not achieve statistical significance. By day 7 there was no significant difference in SBP or DBP between the stroke subgroups and control subjects.
Conclusions BP is elevated after stroke but resolves spontaneously after 7 days. This transient elevation in BP does not appear to result solely from the stress of hospitalization.
The treatment of hypertension in the acute phase of stroke is still a controversial issue.1 Few studies have provided accurate data concerning BP changes in the acute phase of various subtypes of stroke. Hence, uncertainty surrounds the problem of the management of elevated BP after acute stroke.2
An elevation of BP at the time of presentation to the hospital has been previously described in stroke patients.3 4 The exact mechanism for this transient elevation and subsequent decline in BP accompanying acute stroke has not been elucidated. It has been suggested that the initial BP rise was a physiological response to brain ischemia and that BP fell as recovery of cerebral function occurred and normal vascular autoregulation was restored.4 Carlberg et al5 suggested that this phenomenon was due to the acute mental stress of hospitalization. However, their study lacked a control group to support their conclusion. Furthermore, few data are available to determine whether strokes of different etiologies are associated with different patterns of BP change at the time of and after presentation to the hospital. The present study investigated 24-hour BP recordings in patients with strokes of various types and in hospital control subjects immediately after presentation to the hospital and 7 days after admission.
Subjects and Methods
Ninety-four patients, consisting of 72 acute stroke patients and 22 control subjects who presented to St George Hospital, a 600-bed teaching hospital, were studied during a 14-month period. Stroke patients were included in the study if they were older than 40 years and presented to the Emergency Department within 24 hours of the onset of an acute focal neurological deficit. Exclusion criteria were subarachnoid hemorrhage on CT scanning, previous cerebrovascular event leading to a persistent neurological deficit, the presence of an autonomic neuropathy, or a diagnosis of Parkinson’s disease. The control patients were selected randomly from acute medical admissions seen within the first 24 hours of admission to the Emergency Department and were older than 40 years, not severely ill, not dehydrated, hemodynamically stable with no evidence of heart failure, and not in any pain. A history and examination, including an assessment of stroke severity by the Canadian Neurological Score,6 were obtained at the time of admission. The information recorded included any history of treatment for hypertension, diabetes, atrial fibrillation, ischemic heart disease, smoking, and the time of onset of symptoms. If the stroke occurred during the night, the time of onset was considered to be the time the patient awoke and noticed the neurological deficit. The control patients underwent a medical history, examination, and assessment similar to those of the stroke patients.
Stroke was categorized clinically into the following groups: lacunar infarction, thromboembolic infarction, and intracerebral hemorrhage.7 This classification was chosen to reflect the probable underlying vascular mechanism of the stroke. Lacunar infarction was defined as occlusion of a single basal perforating artery, presenting as either a pure motor stroke, pure sensory stroke, sensory motor stroke, or ataxic hemiparesis without clinical signs of cortical involvement. Thromboembolic infarction (or hemispheric infarction) was defined as an anterior circulation infarct (involving ischemia of the territories of the middle cerebral artery) or posterior circulation infarct (involving the vertebrobasilar circulation). A cerebral CT scan was performed within 48 hours of admission. If the initial CT scan showed no acute changes, it was repeated after 1 week with the use of contrast. The clinical classification of each stroke was confirmed by a panel of radiologists and neurologists.
BPs were recorded with the use of a SpaceLabs 90207 ambulatory BP monitor for the first 24 hours of admission and on day 7 of hospitalization. The monitor was programmed to record BP at half-hourly intervals from 8 am to 9 pm and at hourly intervals overnight. No new antihypertensive medications were begun during the 7-day study period. Patients who were receiving antihypertensive therapy at the time of admission to hospital were maintained on the same medication for the duration of the study. However, patients who could not swallow because of their stroke received no antihypertensive medication during the 7 days of the study.
In addition to ambulatory BP monitoring, echocardiography was performed to determine the presence or absence of LVH. Standard M-mode echocardiography was performed to define LVH.8 Measurements of the thickness of the interventricular septum in diastole were compared with measurements of posterior wall thickness in diastole. Concentric LVH was defined as septal and posterior wall thickness of greater than 1.2 cm. The severity of LVH was graded as mild (1.2 to 1.4 cm), moderate (1.5 to 1.9 cm), or severe (>2.0 cm).
Informed consent was obtained from patients or relatives. Ethical approval was obtained from the South Eastern Sydney Area Health Service Ethics Committee.
ANOVA or the χ2 test was used to test for significant differences between groups in baseline characteristics. Multiple linear regression with generalized estimating equations was used to determine the effect of diagnostic category on BP after adjustment for the effects of potential confounders, as determined with the aid of the Statistical Package for Interactive Data Analysis.9
The daytime BP measurements used in analysis were determined from the mean of the two BP measurements made per hour. In some cases, single missing BP measurements occurred. In these cases, a BP measurement was estimated by averaging the measurement before and the measurement after the missing value.
To explore the effect of diagnostic category on BP, the technique of multiple linear regression was used with hourly BP measurement as the outcome variable after adjustment for potential confounders (history of hypertension, LVH, age, sex, diabetes, atrial fibrillation, and smoking). Since the BP distribution was skewed, the distribution was normalized by log transformation. Adjustment was made for possible correlations between repeated measures on the same subject with the use of generalized estimating equations.10
The 72 acute stroke patients consisted of 18 patients with lacunar infarctions, 44 with thromboembolic infarctions, and 10 with intracerebral hemorrhages. The principal diagnoses of the control patients were pneumonia (n=2), chronic airflow limitation (n=5), delirium (n=1), urinary tract infection (n=1), unstable diabetes (n=1), cellulitis (n=3), thrombophlebitis (n=1), deep venous thrombosis (n=2), pulmonary emboli (n=3), failure to cope (n=2), and senile gait disorder (n=1). The mean ages for the stroke subgroups and control subjects were similar: control, 77±10 years; lacunar infarction, 73±10 years; thromboembolic infarction, 74±10 years; and intracerebral hemorrhage, 67±14 years (P=NS). Fifty-seven percent of the stroke group were female, compared with 45% of the control group (P=NS). A previous history of hypertension was present in 41% of control subjects, 67% of those with lacunar infarctions, 57% of those with thromboembolic infarctions, and 60% of those with intracerebral hemorrhages (P=NS). Overall, 60% (43/72) of stroke patients had a history of hypertension compared with 41% of control subjects (P=NS). Thirty-five of the 43 stroke subjects with a history of hypertension were receiving antihypertensive medication before admission. Ten of the 35 subjects who were receiving antihypertensive therapy before admission could not receive oral therapy because of their stroke and did not receive their usual antihypertensive medications for the duration of the study (6 with thromboembolic infarctions, 2 with lacunar strokes, and 2 with intracerebral hemorrhages).
The mean time for admission to the study since the stroke was 11.8 hours, and the mean time of admission to the study since the stroke event was similar for each stroke subgroup (P=.94) (Table 1⇓).
Atrial fibrillation was present in 23% of control subjects, 30% of thromboembolic strokes, and 20% of all strokes (P=NS). The same proportion of the patients with stroke and control subjects smoked (14%, P=NS). More of the control subjects suffered from diabetes mellitus than the patients with stroke (32% versus 21%), but the difference was not statistically significant.
Mean 24-hour BPs for days 1 and 7 by diagnostic group are shown in Table 2⇓. Unadjusted mean 24-hour SBPs by diagnostic group for days 1 and 7 are plotted in Fig 1⇓. The differences in SBP profiles can be seen on day 1, with a fall in the 24-hour BP profiles of all stroke subgroups toward control values on day 7. The plots of unadjusted mean 24-hour DBPs reveal a pattern of change similar to the SBP pattern from day 1 to day 7, with the BP profiles of all stroke groups converging toward the level of control on day 7 (Fig 2⇓). The SBP profiles for lacunar and thromboembolic strokes on day 1 show normal diurnal rhythm, similar to that of control subjects (Fig 1⇓, top panel). However, the SBP profile for intracerebral hemorrhage shows large fluctuations and variability with a loss of normal diurnal rhythm. On day 7, the diurnal pattern for all groups was less marked than on day 1.
The results of the multiple regression model fitted by the generalized estimating equations method for SBP are given in Table 3⇓. A history of hypertension, the presence of LVH, or age greater than 70 years was associated with a significant increase in SBP of 6.7%, 7.1%, and 5.8%, respectively. Patients with thromboembolic and lacunar strokes had significantly higher SBP on day 1 than control subjects by 8.6% and 13.2%, respectively (mean values). For patients with intracerebral hemorrhage on day 1, SBP was 9.7% higher than that of control subjects, but this did not achieve statistical significance, largely because of the small sample size of the intracerebral hemorrhage group. In addition, the intracerebral hemorrhage group demonstrated wide fluctuations in the 24-hour SBP profile (Fig 1⇑, top panel). There was no significant difference in SBP between stroke groups and control subjects on day 7. The fall in SBP occurring between days 1 and 7 was significant for both thromboembolic and lacunar stroke groups but not significant for the intracerebral hemorrhage stroke group. There was no significant difference in SBP between days 1 and 7 for the control group.
Table 4⇓ details the results of multiple regression for DBP. The effect of history of hypertension, LVH, or age was not apparent for DBP. Patients with thromboembolic and lacunar strokes had significantly higher DBP than control subjects on day 1, by 11.7% and 14.6%, respectively; however, the difference was not significant on day 7. Intracerebral hemorrhage patients had DBP 6.3% higher than control subjects on day 1, but this difference was not statistically significant. The fall in DBP occurring between days 1 and 7 was significant for both thromboembolic and lacunar stroke groups but not significant for the intracerebral hemorrhage group. There was no significant difference in DBP between days 1 and 7 for the control group.
This study used 24-hour ambulatory BP monitoring to distinguish between the BP profiles of lacunar infarction, thromboembolic infarction, and intracerebral hemorrhage in their acute phases. The study also compared the time course of BP in patients with these strokes with a control group. The results indicate that BP was elevated in patients with lacunar and thromboembolic strokes at the time of admission to the hospital, but BP levels fell to control values after 7 days. The initial BPs of patients with intracerebral hemorrhage appeared higher than control values, but the difference did not achieve statistical significance. Lacunar infarction appeared to be particularly associated with elevated BP at the time of presentation. This transient elevation in BP did not appear to be due to the stress of hospitalization, since it was not present in patients admitted to the hospital with a range of medical conditions other than stroke. This conclusion is in contrast to that of Carlberg et al,5 who proposed that the rise in BP in acute stroke was due to the stress of hospitalization.
Wallace and Levy4 first documented the spontaneous changes in BP associated with acute stroke and demonstrated that BP declined to levels significantly lower than on admission within the first 10 days of hospitalization. Except for intracerebral hemorrhage, the average BP and BP variability declined to be significantly lower than on admission as early as day 4. However, this study did not use 24-hour ambulatory BP monitoring and used control patients who were awaiting elective cataract surgery or herniorrhaphy, who may not have undergone the same stress as patients admitted with an acute illness.
Britton et al3 evaluated BP changes in stroke patients and patients admitted with acute medical illnesses using two hourly BP readings during the awake hours. They found that BP declined spontaneously in the first few days after hospitalization in both stroke patients and control subjects. These authors suggested that the spontaneous BP decline was due to a combination of factors, including adaptation to the environment and, in stroke patients, a reduction in brain perfusion. The fall in BP in both stroke patients and control subjects seen in their study, but not seen in our study, may have been due to the fact that we used 24-hour ambulatory BP monitoring, which measures nocturnal and daytime BPs and reduces the likelihood of a “white coat” hypertension effect. Alternately, differences in the BP responses between the two studies may have reflected differences in the control patients chosen. The relevance of the transient increase in BP in acute stroke remains to be determined.
This study confirmed a high prevalence of LVH in both stroke patients and elderly hospitalized control subjects. A history of hypertension was also common in all stroke groups and in control subjects. However, the relatively small sample size of each of the groups probably precluded the possibility of detecting a relationship between stroke subtypes and the presence of LVH.
The BPs in intracerebral hemorrhage patients in this study appeared higher than those of control subjects on both study days, but the difference was not significantly significant, most likely because of the small sample size of the intracerebral hemorrhage group and the large fluctuations in BP in this stroke subtype. A larger intracerebral hemorrhage group would have been needed to detect a statistically significant difference between this group and the control group.
Lip et al11 recently described the use of ambulatory BP recordings for the assessment of BP after acute stroke. Their study differed from ours in that they only measured BP on one occasion immediately after admission to the hospital, they did not distinguish between thromboembolic infarction and lacunar infarction, and they did not study a control group. They found no difference between mean nighttime and mean daytime BPs in the stroke patients as a whole and concluded that acute stroke is associated with a failure of BP to dip at night. In contrast, we found a normal diurnal variation in BP immediately after admission to the hospital in both control and stroke patients, with the exception of intracerebral hemorrhage. Our results suggest that the failure of BP to dip at night in stroke patients observed by Lip et al11 may have been due to characteristics of their study population rather than a feature of acute stroke.
In conclusion, elevation of BP in stroke patients at the time of hospitalization appears to be a transient phenomenon that resolves spontaneously. Whether this transient rise in BP is a cause or an effect of stroke is unknown. However, the BP response does not appear to be explained by the stress of hospitalization. The observed magnitude of the acute rise in BP accompanying stroke appears to vary depending on the type of stroke and appears greatest for lacunar infarction. Further studies are needed to establish whether lowering of BP at the time of presentation will be of benefit for each or any of the stroke subtypes.
Selected Abbreviations and Acronyms
|DBP||=||diastolic blood pressure|
|LVH||=||left ventricular hypertrophy|
|SBP||=||systolic blood pressure|
The authors would like to acknowledge the excellent assistance of Robyn Parsons, Denise Abbott, and Dr Dennis Cordato in the preparation of this manuscript.
- Received December 5, 1996.
- Revision received April 23, 1997.
- Accepted April 23, 1997.
- Copyright © 1997 by American Heart Association
Powers WJ. Acute hypertension after stroke: the scientific basis for treatment decisions. Neurology. 1993;43:461-467.
Phillips SJ. Pathophysiology and management of hypertension in acute ischemic stroke. Hypertension. 1994;23:131-136.
Britton M, Carlsson A, De Faire U. Blood pressure course in patients with acute stroke and matched controls. Stroke. 1986;17:861-864.
Carlberg C, Asplund K, Hagg E. Factors influencing admission blood pressure levels in patients with acute stroke. Stroke. 1991;22:527-530.
Cote R, Battista RN, Wolfson C, Boucher J, Adam J, Hachinski V. The Canadian Neurological Scale: validation and reliability assessment. Neurology. 1989;39:638-643.
Weyman AE. Principles and Practice of Echocardiography. 2nd ed. Philadelphia, Pa: Lea & Febiger; 1994.
Gebski V, Leung O, McNeil D, Lunn D. Statistical Package for Interactive Data Analysis, Version 6. Sydney, Australia: The Statistical Computing Laboratory Pty Ltd; 1992.
Lip GYH, Zarifis J, Farooqi S, Page A, Sagar G, Beevers G. Ambulatory blood pressure monitoring in acute stroke. Stroke. 1997;28:31-35.