| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2005;36:2614.)
© 2005 American Heart Association, Inc.
Original Contributions |
From the Neurology Department, University Hospital of Zurich, Zurich, Switzerland.
Correspondence to Claudio L. Bassetti, Neurologische Poliklinik, UniversitätsSpital Zürich, Frauenklinikstrasse 26, CH-8091 Zürich, Switzerland. E-mail claudio.bassetti{at}usz.ch
| Abstract |
|---|
|
|
|---|
Methods We studied 41 consecutive patients admitted within 96 hours after stroke onset. Stroke severity on admission (National Institutes of Health Stroke Scale [NIHSS]) and stroke outcome at discharge (modified Rankin Disability Scale [mRS]) were assessed. Nocturnal breathing was assessed with an ambulatory device the first night after admission. SA was defined by an apnea-hypopnea-index (AHI)
10/hour, and moderate-severe SA (MSSA) was defined by an AHI >30/hour. BP monitoring was performed during the first 36 hours after admission. A nondipping status (NDS) was defined by a ratio >0.9 of mean systolic BP during nights 1 to 2/mean systolic BP during day 2.
Results SA was found in 28 (68%) and MSSA in 11 (27%) of 41 patients. A correlation was found between AHI and both NIHSS (r=0.331; P=0.035) and mRS (r=0.341; P=0.031). Patients with MSSA had higher systolic and diastolic BP values during night 1 (P=0.003), day 2 (P=0.004), and night 2 (P=0.03). NDS was found in 26 (63%) patients. Nondippers had a similar AHI but higher NIHSS (P=0.004) and mRS (P=0.005) than dippers. AHI and NDS were confirmed to be independent predictors for both stroke severity and stroke outcome in a multiple stepwise linear regression model.
Conclusions SA severity is associated with high 24-hour BP values but only weakly with stroke severity and outcome. Conversely, NDS is linked with a more severe stroke and a poorer evolution but not with SA severity. These data suggest different, although overlapping, pathophysiological and clinical implications of circadian and nocturnal BP values in acute stroke.
Key Words: blood pressure dipping/nondipping ischemic stroke sleep apnea stroke outcome
| Introduction |
|---|
|
|
|---|
Evolution of BP in acute stroke and its impact on outcome has been addressed in several studies. Eighty percent of patients with acute stroke are hypertensive on admission, and elevated BP spontaneously declines over the following days.14,15 Patients with both increased and decreased BP may have an unfavorable prognosis, suggesting a bell-shaped relationship between BP and stroke outcome.16,17 In addition, an impairment of circadian BP modulation with loss of physiological nocturnal BP dipping has been documented in the first days after stroke.1821
Considering the strong association between SA and both hypertension and stroke and the impact of elevated BP levels on stroke outcome,19,22,23 it is surprising that the relationship among BP, SA, and stroke was studied only once in the literature.24 To test the hypothesis of a link between SA and high BP in acute stroke, we correlated BP values with SA severity and short-term outcome in 41 patients with acute ischemic stroke.
| Methods |
|---|
|
|
|---|
Clinical Evaluation
Cardiovascular risk factors including family history, arterial hypertension (BP >140/90 mm Hg measured
3 times before stroke), diabetes (fasting glucose level >140 mg/dL), smoking status, hypercholesterinemia (cholesterol level >250 mg/dL), adipositas (body mass index [BMI]
25 kg/m2), and previous history of coronary heart disease were recorded.
Stroke assessment included NIHSS and Scandinavian Stroke Scale (SSS).26 Short-term stroke outcome was estimated by the modified Rankin Disability Scale (mRS)27 at hospital discharge.
Nocturnal Breathing
Overnight respirography was performed with an ambulatory device (Autoset Embletta PDS; ResMed) during the first night after admission (night 1). This respirography has been validated before in patients with stroke (C.L.B., et al, unpublished data, 2005). The analysis was performed automatically and corrected manually. Apnea was defined by a cessation of oronasal airflow
10 seconds and hypopnea by a reduction of oronasal airflow
10 seconds by
50% or
30% when associated with an oxygen desaturation
4%. Apnea-hypopnea index (AHI) was defined by the mean number of apneas and hypopneas per hour and apnea index (AI) by the mean number of apneas per hour between lights off and on. Moderate-severe SA (MSSA) was defined by AHI >30/hour, mild SA by AHI 10 to 30/hour, and no SA by AHI <10/hour. A differentiation between obstructive (OAI) and central (CAI) AI was undertaken according to standard criteria. The percentage of time with oxygen saturation <90%, as well as the oxygen desaturation-index (ODI) were calculated. The magnitude of the oxygen desaturation considered for ODI was
4%. Body position was not monitored.
BP
BP and pulse rate were monitored with an ambulatory device (bp one, Cardiette) over 36 hours in intervals of 30 minutes beginning from night 1. Mean systolic and diastolic BP values and pulse rates were calculated in all of the patients for each period of time (night 1, day 2, and night 2, defined individually). NDS was defined as a ratio >0.9 of mean systolic BP during nights 1 and 2/mean systolic BP during day 2.18,28,29 The number of antihypertensive drugs used on admission were recorded.
Statistics
Statistical analysis was performed with the SPSS software package. Continuous data were presented as mean/SD and categorial variables as numbers/percentage. Student unpaired t tests were used for comparison of patients groups and for testing of continuous variables. Not normally distributed data were compared with the MannWhitney test. Correlations were calculated using the Pearson coefficient. A P<0.05 was considered to be statistically significant. Multiple stepwise regression models (with entry value set to 0.05 and removal value set to 0.1 in the univariate analysis) were tested using stroke severity (SSS at day 1) and stroke outcome (mRS) as dependent variables. Age, gender, number of cardiovascular risk factors, BMI, history of arterial hypertension, diabetes, smoking status, hypercholesterinemia, coronary heart diseases, adipositas, mean systolic BP night 1/day/night 2, mean diastolic BP night 1/day/night 2, mean pulse night 1/day/night 2, number of antihypertensive drugs, NDS, and AHI were candidate predictors.
| Results |
|---|
|
|
|---|
|
Sleep Breathing
The mean AHI was 23±22/hour (0 to 101), AI was 11±16/hour (0 to 70), OAI was 2±3/hour (0 to 15), CAI was 8±15/hour (0 to 61), ODI was 14±15/hour (0 to 64), and percentage of time with oxygen saturation <90% was 10±20% (0 to 95). SA was found in 28 of 41 patients (68%). Eleven patients (27%) had a MSSA (mean AHI, 50±21/hour [31 to 101]; OAI, 5±5/hour [0 to 15]; and CAI, 24±23/hour [1 to 61]). Seventeen patients had a mild SA (mean AHI, 20±7/hour [10 to 29]; OAI, 3±3/hour [0 to 9]; and CAI, 6±7/hour [0 to 20]). In 13 patients (32%), no SA was found (mean AHI, 4±3/hour [0 to 8]; OAI, 1±1/hour [0 to 2]; and CAI, 1±1/hour [0 to 3]).
Considering the entire group of patients (n=41), a weak but statistically significant correlation between SA severity (AHI) and stroke severity on admission (NIHSS day 1 [r=0.331; P=0.035], SSS day 1 [r=0.407; P=0.008]) and short-term outcome (mRS at discharge [r=0.341; P=0.031]) was observed. AHI remained significantly associated with both stroke severity (P=0.015) and stroke outcome (P=0.031) in a multiple stepwise linear regression model (Table 2).
|
When compared with patients without SA (n=13), patients with MSSA (n=11) had a higher BMI (P=0.001), a lower SSS on day 1 (P=0.046), and a longer duration of hospitalization (P=0.008). These and other results are shown in Table 3.
|
BP
History of arterial hypertension was present in 26 of 41 patients (63%). Twenty-five (61%) patients were treated with antihypertensive drugs. The mean BP during night 1 was 140±25 (100 to 239)/87±16 (63 to 147) mm Hg, during day 2 it was 153±24 (117 to 220)/97±15 (72 to 138) mm Hg, and during night 2 it was 144±31 (86 to 209)/90±20 (51 to 129) mm Hg. Patients with MSSA had higher systolic and diastolic BP values during night 1, day 2, and night 2 than patients without SA (Table 3; Figure).
|
Twenty-six patients (63%) were nondippers. Nondippers had more severe strokes (NIHSS day 1, P=0.004; SSS day 1, P=0.003), worse outcome (mRS, P=0.005), and also a lower number of antihypertensive drugs on admission (P=0.009). NDS remained significantly associated with both stroke severity (P=0.001) and stroke outcome (P=0.016) in a multiple-stepwise linear regression model (Table 2).
Conversely, there were no significant differences between dippers and nondippers regarding history of hypertension and respiratory parameters. A reverse dipping (BP night>BP day) was found in 7 patients (17%), 3 of whom had MSSA. These and other results are shown in Table 4.
|
| Discussion |
|---|
|
|
|---|
SA, BP, and Stroke Outcome
The high frequency of SA (69%) and MSSA (27%) in stroke victims is in line with previous reports.810 Our study also confirms the high proportion of hypopneas and central apneas in the first few nights after stroke (mean latency to sleep recording, 37±25 h),10 which contrasts with the predominance of obstructive and apneic events thereafter.8,9,30 The pathophysiological mechanisms involved therein remain unclear at this point.
In patients with SA, we found significantly higher systolic and diastolic BP values than in patients without SA. Furthermore, a linear, dose-response relationship between SA severity (AHI) and BP levels was observed, similar to what has been reported in the general SA population.3 These findings have, to our best knowledge, never been reported before in stroke victims. Turkington et al24 found an increased BP variability (expressed as 10 and 15 mm Hg dips in BP) in 7 patients with SA (AHI>10/h), as compared with 5 patients without SA but did not provide data on absolute BP values.
Several mechanisms may explain higher BP values in stroke patients with SA including oscillations in cardiac output and heart rate, increased sympathetic activity, and impaired baroreceptor control.31,32 Considering the observation of a similar BP evolution in patients with and without SA, sleep disordered breathing appears to be a factor contributing to the overall poststroke increase of BP.
We observed only a weak relationship between SA severity and both stroke severity and short-term stroke outcome. Several previous studies also failed to find any significant link between stroke severity and SA.810,12 Data on SA and short-term stroke outcome are scarce and contradictory. In 1 study, worsening of neurological deficits in the acute phase was observed in patients with SA in the absence, however, of any detrimental effect on short-term stroke outcome.12 Conversely, Kaneko et al33 reported, as we did in the present study, an association between SA severity and duration of hospitalization.
The only weak association found between SA severity (as estimated by the AHI) and short-term stroke outcome could be related not only to the size of the present study, but also to the high frequency of central apneas recorded in our patients in the very acute phase of stroke. In fact, central respiratory events are known to have less detrimental effects (eg, on cerebral blood flow) than obstructive respiratory events.34 In addition, a spontaneous improvement of central respiratory events can often be observed in the subacute phase of stroke.10 Additional studies are, however, needed to test the hypothesis that obstructive SA persisting beyond the very acute phase after stroke may have a more significant, detrimental effect on its short-term outcome.
Dipping Status, SA, and Stroke Outcome
We confirm that NDS, observed in 63% of patients, is a frequent phenomenon in acute ischemic stroke. This observation was made first by Sander and Klingelhofer.35 In a series of 86 patients with acute stroke, Lip et al18 found mean day-night differences in systolic and diastolic BP <10%, demonstrating that NDS was a common finding in this stage. Jain et al21 reported a NDS in 88% of 50 patients examined in the first 120 hours after stroke. In 58%, the mean nocturnal BP was even higher than daytime BP (reverse dipping).
Our data also suggest that loss of physiological nocturnal BP dipping is a stronger predictor for a worse stroke outcome than increased circadian BP values. Similar to us, Bhalla et al19 found that a decrease in day-night BP difference was associated with poor outcome in 72 patients 1 week after stroke. In hypertensive patients, Verdecchia et al36 reported that the number of combined fatal and nonfatal cardiovascular events per 100 patient-years was higher in patients with reduced nocturnal BP variation (4.99 in nondippers; 1.79 in dippers). Size and design of our study do not allow us to test the hypothesis of a link between stroke outcome and increased BP variability,24 BP values at both extremes of its spectrum,17 or hypotensive BP values after the first 24 hours.16
NDS is thought to arise from an imbalance between sympathetic and parasympathetic tone. This may be because of an increased, stroke-related catecholamine and cortisol release, possibly reflecting the cerebral response to a decreased perfusion in the ischemic penumbra but also changes in the sleep-wake cycle and the acute stress of hospitalization. Based on our data, stroke severity and use of antihypertensive drugs at stroke onset, but not SA, influence the dipping status of acute stroke patients. Other studies have suggested a role also of stroke topography (cortical versus noncortical) and stroke etiology (hemorrhagic versus ischemic).1821 Obviously, we cannot exclude, based on our data, the possibility that, in larger size studies including more patients, an association between NDS and SA of obstructive type could be found.
| Conclusions |
|---|
|
|
|---|
| Acknowledgments |
|---|
| Footnotes |
|---|
Received March 15, 2005; revision received August 16, 2005; accepted September 12, 2005.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
W. T. McNicholas, M. R. Bonsignore, and the Management Committee of EU COST ACTION B26 Sleep apnoea as an independent risk factor for cardiovascular disease: current evidence, basic mechanisms and research priorities Eur. Respir. J., January 1, 2007; 29(1): 156 - 178. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2005 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |