|Home | About | Journals | Submit | Contact Us | Français|
Elevated systolic blood pressure (SBP) recorded by 24-hour blood pressure monitoring (24H BPM) on the first day of acute stroke is associated in elderly patients, with an unfavorable outcome. Herein, we assessed, by 24H BPM, the impact of the change in SBP levels during the first week of stroke on short-term functional status and long-term mortality in elderly patients.
One hundred and fifty acute stroke patients (69 males), mean age at admission 83.6±5.5 years, 82% with ischemic stroke, were investigated. 24H BPM was recorded within 24 hours of admission and 1 week later. After 7 days, patients were assessed for functional status according to the modified Rankin scale (mRS) and were subsequently followed for mortality up to 7.5 years (mean 3.16±2.29).
After 7 days, SBP decreased from 147±21 to 140±20mm Hg (P < 0.001). Functional status improved and mRS decreased from 4.2 to 3.7. During follow-up, 58 patients (17 males and 41 females) had died. Mortality rate was higher in females (69% vs. 45%; P < 0.01) and in patients with a history of congestive heart failure. The average admission SBP predicted short-term functional status and long-term mortality. However, the change in SBP corrected for admission levels, gender, age and other variables was not associated with short-term functional status and long-term mortality.
There is no evidence of association between change in SBP during the first week of stroke and short-term functional status and long-term mortality in this group of stroke patients.
Elevated blood pressure (BP) is commonly observed during an acute stroke and usually returns to normal within a few days.1–4 Elevated BP during an acute ischemic stroke might be advantageous by improving cerebral perfusion to the ischemic tissue or detrimental by exacerbating edema and hemorrhagic transformation of the ischemic tissue.1 It is unclear whether high BP should be treated in acute ischemic stroke.5
Most intervention studies have failed to discover an association between the fall in BP and outcome.6 Castillo et al. found that a fall in systolic BP (SBP) of more than 20mm Hg was associated with a poor outcome.7 Leira et al. observed that in elderly patients (>70 years) reductions in SBP was associated with a worse prognosis.8 Giantin et al., in a small study using 24H BP monitoring (BPM), found an inverse correlation between the decrease of 24H BP within the first week and functional status in elderly patients with acute stroke.9
It should be noted that most of the studies were limited by their small size.5 Recently, a meta-regression analysis showed that modest BP reductions may reduce death and dependency, whereas large falls or increases in BP are associated with a worse outcome. However, these results were not adjusted for baseline BP values or for other factors such as previous hypertensive disease and previous antihypertensive treatment.10,11
Several randomized controlled trials evaluated the effect of therapeutic BP lowering during the acute phase of stroke on outcome.7–9,12 Most data regarding the association between the change in BP during the acute phase of stroke and outcome were derived from interventional studies that included different agents, evaluated short-term outcome, used casual BP levels and included only very few elderly patients.
We recently evaluated the effect of admission SBP on short- and long-term outcome in very elderly stroke patients.13 In the present study, we extended the follow-up and assessed the relationship between a spontaneous change in BP, assessed by 24H BPM, and short- and long-term outcomes.
All study patients diagnosed with acute stroke were consecutively admitted and evaluated on admission to an acute geriatric ward between 15 May 2007 and 15 March 2011. Subjects were included if they had experienced a documented acute stroke within 24 hours preadmission. Ischemic stroke and intracranial hemorrhage were diagnosed by clinical findings and head computerized tomography. Patients younger than 70 years and patients with transient ischemic attacks, or other causes of acute neurological deficit, were excluded. Further exclusions included patients unable to continue taking antihypertensive medications, patients with atrial fibrillation, patients with inaccurate 24H BPM, and those with a hypertensive crisis (>220/120mm Hg) who had received immediate antihypertensive treatment. Patients, who did not have a second 24H BPM 1 week after admission, were also excluded. The study was approved by the local ethics committee and all patients or their next of kin signed an informed consent.
BP and heart rate were measured on admission and the reported values were the average of 2 measurements taken 1 minute apart. However, when the 2 systolic measurements differed by more than 5mm Hg, additional measurements were taken until 2 similar values were obtained, which were then included in the analysis. 24H BPM was recorded within 24 hours (3−24 hours) of admission and 1 week later. Antihypertensive treatment continued unchanged during hospitalization. Patients who were taking antihypertensive medications prior to admission continued the same medications at the same dose. However, in a few patients with low BP levels, some antihypertensive drugs were withheld.
Data were obtained from the patients’ medical records and included clinical evaluation findings performed on each subject upon hospital admission, standard physical examination, and detailed medical history. Functional status was determined on admission and 1 week after hospitalization according to the modified Rankin scale (mRS).14 The following variables were recorded: age, height, weight, BP, heart rate, ethnic origin, and prescribed medications. Comorbid conditions were also identified from the medical records.
Hypertension was defined as the presence of a prescription for antihypertensive medications or a BP recording >140/90mm Hg on 2 or more repeated measurements prior to hospitalization. Diabetes mellitus was defined as the presence of hypoglycemic agents or a recording of fasting blood glucose of ≥126mg/dl on at least 2 measurements prior to hospitalization. Renal failure was defined when serum creatinine of ≥1.5mg/dl was found on at least 2 measurements.
Mortality data until 1 December 2013 were extracted from the Ministry of Interior Affairs registry.
24H BPM was performed by the Oscar2 24-HR ABP (SunTech Medical, Morrisville, NC). The monitor was mounted on the left arm or the paralyzed arm of right hemiparesis or hemiplegia patients. A mercury sphygmomanometer was initially attached to the monitor through a Y connector to ensure agreement between the 2 modes of measurements. BP was measured every 20 minutes during the day (from 06:00 to 22:00) and every 30 minutes at night (from 22:00 to 06:00). An acceptable recording for our study was a 24H BPM with at least 50 acceptable measurements. The average SBP and diastolic BP (DBP) were calculated for 24 hours.
Data are presented as mean ± SDs. Patients were categorized into 2 groups according to their change in average SBP assessed by 24H BPM during the first week. Group 1 included patients with a decrease of >10% in the average SBP and group 2 included patients with a decrease of ≤10% in the average SBP. Paired t-test was used to analyze the differences between parameters recorded on admission and 1 week later. Unpaired t-test was used to determine significant differences between the sexes and between those with a decrease of >10% and those with a decrease of ≤10% in average SBP (groups 1 and 2). A proportional hazard Cox regression was used to analyze the impact of variables on survival. A nonparametric Kaplan–Meier analysis was conducted and presented using survival curves accounting for right censoring of survival data post-admission. A binary logistic regression was used to identify factors affecting a reduction of mRS post-hospitalization. All calculations were performed with MINITAB™ Version 16.2.
Of 177 patients who participated in the original study, 150 patients with acute stroke (69 males), mean age 83.6±5.5 years (range 70–99), underwent a 24H BPM within the first 24 hours of hospitalization and 1 week later and were included in the present study. Those who were excluded from the original study had the same baseline characteristics as those who were included in the study. Most patients had experienced an ischemic stroke, were severely handicapped on admission, and suffered from hypertension, dyslipidemia, diabetes mellitus, and ischemic heart disease (Table 1). Females were older, had a faster heart rate, were more handicapped on admission than males, and were less prone to have a history of smoking and ischemic heart disease (Table 1). Most patients received beta blockers, renin angiotensin system blockers, and diuretics (Table 1).
Admission casual BP before beginning 24H BPM was 151±24 (range 113–208)/78±15 (range 45–125) mm Hg. The average 24H BP was 147±21/74±11mm Hg. Average SBP was 148±20 in patients with a history of hypertension and 144±23 in patients without a history of hypertension (P = 0.36). In 40 patients (26.6%), the average 24H SBP was ≥160mm Hg; 44 patients (36.6%) received 1 antihypertensive drug, 47 (39.2%) received 2 antihypertensive drugs, and 29 (24.2%) received 3 or more antihypertensive drugs. Antihypertensive treatment remained unchanged during the study.
One week after stroke, 24H SBP decreased by 7mm Hg (P < 0.01) and DBP decreased by 3mm Hg (P < 0.01). Heart rate did not change (P = 0.767). SBP increased or was unchanged in 49 patients and DBP increased or remained unchanged in 57 patients. SBP decreased by >10% in 33 patients. Patients who exhibited a decrease of >10% in SBP had higher admission BP and heart rate levels, had worse functional status on admission, and were more likely to have congestive heart failure, diabetes mellitus, and chronic renal failure than those who exhibited a decrease of ≤10% (Table 2). The change in 24H SBP after 1 week correlated to the 24H admission SBP (r = −0.428; P < 0.0001).
After a week of hospitalization, the mRS decreased from 4.2±1.1 to 3.7±1.6 (P < 0.01). Functional status as determined by the mRS improved in 71 patients, deteriorated in 12, and remained unchanged in the rest. Admission SBP as measured by 24H BPM was inversely associated with a 1-week favorable functional outcome. The decrease in mRS was the same in those who exhibited >10% and those who exhibited ≤10% decrease in 24H SBP (−0.30 and −0.53 respectively; P = 0.162).
The change in SBP as measured by 24H BPM had no effect on short-term improvement in functional outcome.
During the follow-up period of 7.5 years (mean 3.16±2.29), 58 patients died. Patients who died were more likely to be female, had a higher rate of congestive heart failure, and a higher admission 24H SBP and DBP than those who survived (Table 3). The change in SBP after 1 week as measured by 24H BPM was −9.6±20 in those who died and -4.8±12.5 in those who survived (P = 0.107, Table 3). The rate of mortality was higher in those who exhibited >10% decrease in 24H SBP [19 of 33 (57.6%)) than in those who exhibited ≤10% decrease in 24H SBP [39 of 117 (P = 0.02). However, after adjustment for age, gender, congestive heart failure, smoking, history of hypertension history of previous stroke, and admission 24H SBP, the risk of mortality was the same in those who exhibited >10% and ≤10% decrease in 24H SBP (Table 4).
In the present study, we demonstrated that in elderly patients with acute stroke, the spontaneous change in 24H SBP during the first week was not associated with short-term functional status and long-term mortality. It is well known that BP levels are commonly elevated during acute stroke and fall spontaneously during subsequent days.2–4
The association between the change in BP after acute stroke and outcome has been studied by several investigators. Most of the studies used antihypertensive agents to lower BP; hence it is difficult to isolate the effect of the drug from the effect of spontaneous BP change.
Elevated admission SBP is associated with more extensive cerebral damage.15 The Safe Implementation of Thrombolysis in Stroke-International Stroke Thrombolysis Register showed a strong association of high SBP after thrombolysis with a poor outcome.16
In the present study, we confirm the findings that elevated admission SBP, as assessed by 24H BPM, are associated with poor short- and long-term outcomes.
The effect of BP decrease on outcome in patients with acute stroke is not clear. Several studies have demonstrated that a decrease in BP after stroke is associated with poor short-term outcome.7,9,17 Castillo et al. showed that reductions in SBP and DBP by >20mm Hg during the first day after stroke were associated with short-term poor outcome.7 However, they studied the BP change only on the first day after stroke and about a third of their patients had received new BP lowering agents.
Data from the Scandinavian Candesartan Acute Stroke Trial showed that patients with a large decrease in SBP had poor outcome, but the change in BP was assessed 2 days after the stroke and follow-up was maintained for only 6 months.17
Giantin et al. evaluated a small group of elderly patients with acute stroke using 24H BPM, showing that BP fall during the first week after stroke is associated with poor short-term outcome.9 Other studies have demonstrated that a decrease in BP after stroke is associated with a favorable outcome.11,18,19
It is noteworthy that most studies were based on casual measurements, included very few elderly patients2–4,7 and patient follow-up was short-term. The advantage of 24H BPM over casual BP values is that more measurements are yielded over time and the “white coat effect” and observer bias is eliminated. 24H BPM facilitates using a smaller sample and receiving accurate information.
Bhalla et al. evaluated patients with acute stroke with 24H BPM. However, their group was small and they did not report the effect of change in BP on outcome.20 Giantin et al. evaluated a small group of elderly patients with acute stroke by using 24H BPM, showing that BP fall during the first week after stroke is associated with poor short-term outcome.9 However, their study group was very small, treatment had been changed during the first week after the stroke and patients were not followed up long-term. Rossi et al. evaluated a small group of patients with acute stroke by using 24H ABPM showing that a reduction of BP was associated with early neurological improvement.21 However, the study group was small and included only a few elderly patients.
Our study is unique since we included only elderly patients, used 24H BPM twice, did not change antihypertensive treatment during the week, and engaged in a short- and long-term follow-up. The fact that we did not interfere pharmacologically to lower BP enabled us to assess the association between a spontaneous change in BP after stroke and outcome. Our results showed that a spontaneous fall in SBP had no effect on short-term outcome. During long-term follow-up, the crude data showed that a fall in BP was associated with increased mortality, but after adjustment for other risk factors and admission SBP, the association between the change in BP and outcome disappeared. Admission SBP, as assessed by 24H ABPM, was associated with poor outcome confirming our previous analysis.13
Similar to our previous observation, the decrease in SBP after 1 week was related to admission BP.22 Since admission SBP was a major determinant of poor outcome, the effect of change in BP disappeared after adjustment to admission SBP. Thus, admission elevated SBP rather than the fall in SBP after stroke, is associated with poor short-term outcome and long-term mortality. To the best of our knowledge, our study is the only one using 2 24H BPM to assess the change in SBP in very elderly patients followed for up to 7.5 years. Elevated admission SBP levels could reflect a response to partial loss of brain auto-regulation and dependence of brain perfusion on systemic arterial pressure. The magnitude of SBP rise on admission might be related to severity of brain injury and probably SBP levels on admission and change during the first week are consequences of the same phenomena; the magnitude of stroke.
Since in acute stroke elderly patients, a spontaneous fall in BP is not associated with improved outcome, it seems reasonable not to lower SBP in acute stroke in these patients.
Our study has several limitations. Firstly, the sample size is small, and the statistical power is low and therefore the evidence in support of a “null finding” is weak, however; only a few small studies have employed 24H BPM in elderly patients. Secondly, the vast majority of our patients have had an ischemic stroke and therefore, our results may apply only to patients with ischemic stroke. Thirdly, data relating to drug treatment, BP control, and BP variability after discharge were unavailable. It is well known that these parameters can predict long-term outcome in patients with ischemic stroke,23,24 and therefore we were unable to comment on the effect of treatment on long-term mortality.
The present study suggests that in very elderly patients increasing 24-hour SBP levels following acute stroke predict poor outcome. However, the fall in SBP during the first week after ischemic stroke has no effect on outcome.
The authors declared no conflict of interest.
We thank Mrs Phyllis Curchack Kornspan for her editorial services.