| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Pneumology (Drs, Martínez-García and Soler-Cataluña) and Neurology Units (Dr. Galiano-Blancart), Service of Internal Medicine (Drs. Cabero-Salt and Román-Sánchez), Requena General Hospital, Valencia, Spain.
Correspondence to: Miguel Angel Martínez-García, MD, Hospital General de Requena, Unidad de Neumología (Servicio de Medicina Interna), Paraje Casa Blanca s/n, 46320-Requena, Valencia, Spain; e-mail: med013413{at}nacom.es
Abstract
Study objective: The aim of this study was to analyze the role of dysphagia as a model of pharyngeal muscle dysfunction in the time course of nocturnal disordered breathing (NDB) in patients who experienced a first-ever ischemic stroke.
Design: Prospective study.
Patients and interventions: Fifty-nine consecutive patients (mean age, 73.2 years; SD, 12.8 years) were studied. Clinical sleep and neurologic data and vascular risk factors were recorded. Two nocturnal studies using a portable autotitration device (AutoSet Portable Plus II system; ResMed; Sydney, NSW, Australia) were performed in both the acute phase (mean duration, 1.23 days; SD, 0.7 day) and the stable phase (mean duration 65.9 days; SD, 12.5 days) of the neurologic event in all patients.
Results: The mean total apnea-hypopnea index (AHI) measured with the autotitration device in the acute phase was 34.9 (SD, 25.2) vs 20.1 (SD, 21.7) in the stable phase, both with predominance of obstructive apnea. Patients with dysphagia (n = 30) showed the largest number of obstructive apneic episodes (OAIs) in the acute phase (AHI, 40 episodes; OAI, 30.4 episodes), with a significant reduction in this type of apnea during the stable phase of stroke (AHI, 24.7 episodes; OAI, 17.7 episodes), coinciding with the recovery of pharyngeal muscle function. In contrast, nondysphagic patients (n = 29) showed no significant changes in NDB from the acute to the stable phase of stroke. Logistic regression analysis found dysphagia to be the best independent predictor of AHI reduction of > 50% from baseline (odds ratio, 13.4; 95% confidence interval, 3.3 to 39.6; p = 0.001).
Conclusion: The present study shows significant improvement in the number obstructive apneic events occurring in the stable phase of a first-ever ischemic stroke in patients with transient pharyngeal muscle alterations secondary to the neurologic lesion.
Key Words: AutoSet • dysphagia • ischemic stroke • obstructive apnea • sleep apnea
In the past few decades, sufficient evidence has been accumulated to regard obstructive sleep apnea-hypopnea syndrome (OSAS) as an independent risk factor for the development of systemic arterial hypertension123 and probably also for cardiovascular diseases.456 However, growing controversy exists over the relation between OSAS and cerebrovascular disease.78910111213141516 In this context, a cross-sectional study, the Sleep Heart Health Study,10 which was conducted with a population base of > 6,000 individuals, revealed a positive linear relation between the apnea-hypopnea index (AHI) and the probability of stroke, adjusting for possible confounding variables.
It seems well established that a greater than expected prevalence of sleep or nocturnal disordered breathing (NDB) of an obstructive nature exists in the acute phase in patients who have experienced ischemic stroke, regardless of the severity and location of the neurologic lesion.8914 On the other hand, some studies7813 have reported a significant decrease in sleep-disordered breathing or NDB several weeks after the neurologic event. The physiopathologic mechanism underlying this relation is not clear. Obstructive events could be explained both by the presence of OSAS prior to stroke and, probably, by muscular dysfunction as a consequence of the neurologic event proper.16
Neurologic dysphagia has been demonstrated in 30 to 40% of patients with hemispheric or brainstem stroke.1718 It can be speculated that dysphagia, as a model of subjects with pharyngeal muscles affected by stroke, would facilitate the development of obstructive events or worsen existing events during the acute phase of stroke, followed by improvement as the patient recovers from the neurologic process. However, none of the previous studies have analyzed the course of NDB in patients with dysphagia. The aim of the present study thus was to analyze the time course of NDB from the acute to the stable phase of a first-ever ischemic stroke, particularly in the subgroup of patients with dysphagia, which was assessed as swallowing difficulty under certain circumstances in our patients.
Materials and Methods
Consecutive patients admitted to our hospital during the year 2002 with a first-ever ischemic stroke were initially screened for inclusion in the study. Patients with an important reduction in consciousness (Glasgow coma scale score, < 8), short-term terminal background disease, daytime heart or respiratory failure, or using psychoactive drugs (fundamentally benzodiazepines) were excluded from the study. We also excluded patients who refused to give informed consent to participation in the study, or who had been referred for an onset of stroke symptoms > 72 h before arriving in our center. The study was approved by the local human ethics committee, and informed consent was obtained from all patients or their families.
Stroke and Vascular Risk Factors Assessment
The diagnosis and location of ischemic stroke were determined by a neurologist, based on the evaluation of the existing neurologic defects and of cerebral CT scans conducted in the first few hours after patient admission to the emergency department, and again several days later. Stroke subtypes were classified as hemispheric, brainstem, or cerebellar depending on the location, and also according to the presence or absence of a lacunar syndrome. Functional disability and neurologic impairment at hospital admission were evaluated based on the following widely used neurologic scales: the Barthel index,19 which assesses daily activity on a scale of 0 to 100 (a score of 100 corresponding to full patient autonomy); and the Canadian scale,20 which quantifies stroke-related symptom severity on a scale from 0 to 10 (with decreasing scores indicating greater severity). Dysphagia was considered in the occurrence of any of the following situations after stroke: manifest incapacity to swallow solids or liquids; oral dribbling after administering 10 mL of water; or repetitive cough or an absence of laryngeal movement on swallowing. Laryngeal movement was assessed by direct visual inspection.21 Pharyngeal function was assessed by examining the presence or absence of palatal elevation (motor function) and the nausea reflex (pharyngeal sensation). An examination was performed in all patients by an experienced neurologist on hospital admission and prior to the stable-phase study performed using an autotitration device (AutoSet Portable Plus II system; ResMed; Sydney, NSW, Australia), without knowledge of the autotitration study results. Dysphagic patients prior to the neurologic event and patients with nonneurologic causes of dysphagia or pharyngeal muscle dysfunction were excluded.
Data were collected on the existence of the following vascular risk factors: internal carotid stenosis (> 50% of the vascular lumen), assessed by continuous Doppler flowmetry, transcranial Doppler flowmetry and magnetic resonance angiographic confirmation where appropriate; current smoking (> 10 cigarettes per day); arterial hypertension (defined according to World Health Organization criteria22 or by the current use of antihypertensive drugs); hypercholesterolemia (concentration of > 250 mg/dL in peripheral blood); diabetes mellitus; atrial fibrillation; ischemic heart disease; and fibrinogen concentration in peripheral blood.
Nocturnal Study Assessment
OSAS-related clinical manifestations prior to ischemic stroke were recorded as follows: a patient was considered to have a significant snoring disorder if snoring occurred every night or almost every night. Significant witnessed apneic episodes were those occurring every night or almost every night, or repeatedly on the same night, and daytime hypersomnia was assessed by the validated Spanish version of the Epworth sleepiness scale that was completed by a relative of the patient.23 Demographic and anthropometric variables (ie, age, sex, body mass index [BMI]), and neck circumference) were also recorded.
All patients finally included in the study underwent two nonattended nocturnal evaluations using the portable autotitration system (AutoSet Portable Plus II; ResMed), one during the acute phase (ie, within the first 72 h after the onset of symptoms) and the other during the stable phase (ie, between 60 and 90 days after the acute episode). Patients in whom both autotitration studies could not be made were excluded from the study. The autotitration device used (AutoSet Portable Plus II; ResMed) is an auto-continuous positive airway pressure model in which the diagnostic mode allows the recording of different respiratory variables and patient heart rate. The system has been described and validated by several studies.242526 Nasal ventilation was measured using a nasal cannula in the anterior of the nares. Nasal cannulas are connected to a pressure transducer in the autotitration device. This provided a semiquantitative reading of nasal ventilation. A respiratory event was considered to constitute apnea when the nasal flow dropped to < 25% of the recent average for at least 10 consecutive seconds, while hypopnea was considered when the nasal flow was between 25% and 50% of the recent average for at least 10 consecutive seconds. Respiratory effort was measured using a respiratory band. The raw data from the respiratory band (a signal that varies with band tension) was sampled every second, and the amplitude was averaged. The resultant trace was then mirror-imaged below the baseline and shaded to yield the respiratory effort trace. All apneas were classified by manual analysis following the subdivision given by the equipment. The presence of some respiratory effort (measured by the amplitude of the respiratory effort trace) during an apnea was indicative of an obstructive event, while if the respiratory effort trace showed zero or near-zero amplitude during an apnea, the respiratory event was considered to be central. The autotitration device automatically and independently calculates an approximate AHI and an approximate apnea index, since the total sleep time was unknown, with derivation from both indexes of the corresponding hypopnea index calculated by the autotitration device. The AHI was defined as the number of respiratory events (ie, apneas or hypopneas) documented per hour of recording. Cheyne-Stokes respiration was defined as central apnea alternating with hyperpnea in a crescendo-decrescendo pattern in the respiratory effort trace in > 10% of the recording. Nocturnal desaturation was also evaluated as the nocturnal counting time with an oxygen saturation of < 90% (CT90). All data were calculated as a function of the total recording time. All tests were performed in our hospital, in rooms conveniently prepared to the effect by trained personnel. On the morning after the autotitration device study, the patient (with the help of a relative if necessary) completed a form indicating a subjective appraisal of the amount (in hours) and quality (good, regular or poor) of sleep. Those tests in which the patient claimed to have slept at least 3 h with a minimum sleep quality rated as "regular" were considered to be valid. Tests involving some technical malfunction or patient-caused disconnection resulting in < 3 h of valid recording were regarded as nonvalid tests. In this latter case, the autotitration device study was repeated. No polysomnographic studies were performed in any patient.
Statistical Analysis
A statistical software package (SPSS for Windows, version 9.0; SPSS; Chicago, IL) was used throughout. The quantitative variables were tabulated as the mean, SD, mean differences, and 95% confidence interval (CI), and the qualitative variables as the absolute value, percentage, odds ratio (OR), and 95% CI. Baseline data of dysphagic and nondysphagic patients were compared by the Student t test for independent samples in those cases in which a normal distribution was observed (as determined by the Kolmogorov-Smirnov test); alternatively, the nonparametric Mann-Whitney U test or Wilcoxon test was used in the absence of a normal distribution. Qualitative variables were compared using the 
2 test for independent samples, or the McNemar test for paired samples. Time courses from baseline data were compared within groups (intragroup differences) and between groups (intergroup differences) using a repeated-measures analysis of variance test with Bonferroni correction.
Stepwise logistic regression analysis was used to evaluate the relationship between dysphagia and AHI changes calculated by the autotitration device (as a dichotomous dependent variable) from the acute phase to the stable phase of stroke adjusted for other confusion variables. The following four independent variables were initially introduced in the equation: presence/absence of dysphagia; AHI in the autotitration device study in the acute phase of stroke; and neurologic data (ie, Barthel index and Canadian scale). The results were expressed as the OR and 95% CI for statistically significant relationships. Since the study included 59 patients, this implied that approximately 15 patients per variable included in the equation. The cutoff point in the percentage change of AHI calculated by the autotitration device from the acute phase to the stable phase of stroke was taken to be the point providing the greatest dysphagia-diagnosing capacity in our patients, in those situations in which the cutoff point was considered to be clinically significant by the investigators. A receiver operating characteristic (ROC) curve was plotted to select this cutoff point. The curve plotted the relationship between the sensitivity and specificity of different cutoff points in the changes of AHI calculated by the autotitration device (from 5 to 100%) for identifying the presence of dysphagia. The greatest diagnostic value corresponds to the cutoff point that generates the largest area under the curve (AUC). In all cases, statistical significance was accepted for p < 0.05.
Results
Of the 127 patients who were initially screened for inclusion in the study, an acute-phase autotitration study could be performed in 107. The second autotitration study, performed in the stable phase of stroke, could not be conducted in 48 patients. The most common reason for withdrawal from the study was refusal to continue participating in the study (34 patients), either expressly or because the patients failed to present for the stable-phase autotitration device study (Fig 1 ). The patients in whom the stable-phase autotitration device study could not be performed showed increased functional repercussions as a result of stroke (mean Barthel index score, 74 [SD, 39] vs 61 [SD, 38], respectively; p = 0.02; mean Canadian scale score, 7.8 [SD, 2.9] vs 7.4 [SD, 2.8]; p = 0.02). No statistically significant differences in vascular risk factors, AHI calculated by autotitration device, or retrospective OSAS-related clinical manifestations collected from family members were found (Table 1 ).
|
|
Stroke and Vascular Risk Factors Assessment
Table 1 shows the stroke characteristics and prevalence of vascular risk factors in our study population. Fifty-one percent of the patients experienced acute hemiplegia. Fifty percent of these individuals showed significant recovery in the stable phase of stroke, while 36.7% developed permanent monoplegia, and the rest had permanent hemiplegia. On the other hand, 34% of the patients showed acute monoplegia that totally resolved in 65% of cases during the stable phase of stroke.
Nocturnal Study Assessment
The autotitration study in the acute phase of stroke was carried out after a mean duration of 1.23 days (SD, 0.7 days), while the study was conducted in stable phase of stroke after 65.9 days (SD, 12.5 days). The comparative analysis of the results obtained in both autotitration studies is summarized in Table 2 . Significant improvement was observed in AHI calculated by autotitration device (p = 0.001), which was fundamentally in relation to obstructive apnea index (OAI) [p = 0.008] and nocturnal oxygen desaturation (ie, CT90) [p = 0.006]. There were no significant differences between acute-phase and stable-phase autotitration study parameters related to BMI and the body position of the patients during the night.
|
|
|
|
|
Our series showed greater than expected prevalence of NDB compared to reported sleep and NDB in patients in the acute phase of a first-ever ischemic stroke. This may be related to the fact that we used a portable autotitration device (AutoSet Portable Plus II; ResMed) and that sleep was never measured. However, a significant decrease in obstructive-type NDB calculated by the autotitration device was observed following stabilization of the neurologic event only in patients with dysphagia (defined as swallowing difficulties under certain circumstances in our patients).
Agreement is lacking regarding the mechanisms by which stroke can generate obstructive NDB.47916 It is known that stroke affecting both the vertebrobasilar and hemispheric regions can cause transient or permanent alterations in pharyngeal muscle tone,18 and this may in turn facilitate the development of obstructive events or worsen preexisting events. In this sense, the analysis of obstructive NDB among patients with dysphagia as a model of subjects with pharyngeal muscle or sensory alterations could offer a satisfactory explanation. In our study, patients with dysphagia showed a greater number of obstructive apneas in the acute-phase autotitration study and were the individuals that showed the greatest decrease in such obstructive events in the stable phase of stroke, coinciding with the recovery of pharyngeal motor or sensory dysfunction, regardless of the location and clinical picture associated with OSAS. In contrast, neither obstructive nor central events were significantly improved in patients without dysphagia. In addition, there was no difference between dysphagic and nondysphagic patients in the number of obstructive events occurring in the stable phase of stroke. This observation would support the hypothesis that the main mechanism by which stroke generates obstructive events comprises the transient loss of pharyngeal muscle tone following the neurologic event, since these are the type of respiratory events that improve significantly with stabilization of the neurologic process. In our study, only five of the patients subjected to the autotitration study in the stable phase of stroke failed to have their dysphagia improve. Moreover, dysphagia was the best independent predictor of AHI improvement calculated by autotitration device of > 50% from baseline in patients with ischemic stroke. Although we found more severe neurologic impairment in dysphagic patients at hospital discharge, as evaluated by the Barthel index score, this variable was not found to be an independent predictor of improvement in AHI calculated by autotitration device. The choice of a cutoff point of 50% in improvement of AHI calculated by autotitration device from the acute phase to the stable phase of stroke was not entirely arbitrary. This percentage corresponds to the cutoff point providing the greatest predictive capacity of dysphagia for important changes in AHI calculated by autotitration device based on the calculated AUC-ROC. Since a search of the literature yielded no definition of the minimum improvement in AHI calculated by autotitration device that can be considered clinically significant, we consider that 50% improvements in AHI calculated by autotitration device among our patients would be clinically significant in most cases. This cutoff point was therefore used for posterior calculations.
Turkington et al,11 in a study of 120 patients (mean age, 79 years), showed that BMI (23.6 kg/m2) and neck circumference (38.3 cm), but not pharyngeal dysfunction or dysphagia, were the best independent predictors of the development of upper airway obstruction in the first 24 h after acute stroke. The reasons for the difference between the results of that study and our own findings regarding the relation between obstructive events and the presence of dysphagia in the acute-phase autotitration study are not easily explained. Probably, several factors could contribute to this discrepancy, including the possible error introduced by using the autotitration device to calculate the number of NDB episodes; the different timing of the nocturnal study (ie, first 24 h after ischemic stroke in the study by Turkington et al11 vs the first 72 h in our study), the characteristics of the two series (with fewer obese individuals in the study by Turkington et al11 [BMI, 23.6 vs 28.8 kg/m2, respectively]), and, fundamentally, the different principal objectives of the studies. In this sense, we were interested mainly in analyzing which factors could independently contribute to the time course of obstructive events from the acute phase to the stable phase of stroke, and the role of dysphagia. In comparison, the study by Turkington et al11 was designed to analyze the best predictors of upper airway obstruction in the first 24 h after stroke. As a result, the dependent variables in the logistic regression analyses of the two studies were different.
On the other hand, the presence of selection bias was implied by the greater clinical severity of stroke in the patients excluded from the study because of major disease severity or nonperformance of the two autotitration studies. Such bias would make our patients nonrepresentative of the general stroke population, due to the selection of the less serious cases. The patients in whom the two autotitration studies could not be carried out did not present significant differences in terms of the autotitration parameters or clinical manifestations of OSAS vs the patients included in the study. Finally, a possible limitation in our study is represented by the definition of dysphagia as swallowing difficulty under certain circumstances in our patients, since no dynamic or imaging studies were performed to assess pharyngeal muscle function.
In conclusion, our results support the hypothesis that stroke might generate an increase in the presence of obstructive apneas in the acute phase, mainly due to the transient loss of pharyngeal muscle tone, since such events decreased significantly once muscular function or sensory dysfunction recovered. Since the study design only allowed the definition of an association between dysphagia and obstructive NDB, we think that further studies are needed to determine whether a true cause-effect relation exists between the two conditions, and to determine the effectiveness of continuous positive airway pressure treatment in the acute phase of neurologic events in patients with dysphagia, despite the known low adherence to this treatment among acute stroke patients.8
Footnotes
Abbreviations: AHI = apnea-hypopnea index; AUC = area under the curve; BMI = body mass index; CI = confidence interval; CT90 = nocturnal counting time with an oxygen saturation of < 90%; NDB = nocturnal disordered breathing; OAI = obstructive apnea index; OR = odds ratio; OSAS = obstructive sleep apnea-hypopnea syndrome; ROC = receiver operating characteristic
Received for publication January 11, 2005. Accepted for publication July 10, 2005.
References
Read all eLetters
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
