SEARCH

SEARCH BY CITATION

Keywords:

  • brain circulation and metabolism;
  • cerebrovascular disease/stroke;
  • Doppler ultrasound;
  • sleep apnea;
  • transcranial Doppler

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Acknowledgements
  10. References

Background and purpose:  Obstructive sleep apnea syndrome (OSAS) is an independent risk factor for stroke. Impairment of cerebral autoregulation may play a potential role in the pre-disposition to stroke of OSAS patients. In this study, we aimed to assess dynamic cerebral autoregulation (DCA) during wakefulness in OSAS patients and a group of matched controls.

Methods:  Patients and controls were examined in the morning after an overnight complete polysomnography. Mean cerebral blood flow velocity (CBFV) in the middle cerebral artery and mean arterial blood pressure (ABP) were continuously recorded using transcranial Doppler and Finapres. DCA was assessed using the Mx autoregulatory index. Mx is a moving correlation coefficient between mean CBFV and mean ABP. More positive value of Mx indicates worse autoregulation.

Results:  Eleven OSAS patients (mean age ± SD; 52.6 ± 7.9) and 9 controls (mean age ± SD; 49.1 ± 5.3) were enrolled. The mean apnea–hypopnea index (AHI) in the OSAS group was of 22.7 ± 11.6. No significant difference was found between the two groups as for age, body mass index, mean ABP and endtidal CO2 pressure. Cerebral autoregulation was impaired in OSAS patients compared with controls (Mx index: 0.414 ± 0.138 vs. 0.233 ± 0.100; P = 0.009). The severity of autoregulation impairment correlated to the severity of the sleep respiratory disturbance measured by the AHI (P = 0.003).

Conclusion:  Cerebral autoregulation is impaired in patients with OSAS during wakefulness. Impairment of cerebral autoregulation is correlated with the severity of OSAS.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Acknowledgements
  10. References

Obstructive sleep apnea syndrome (OSAS) is a risk factor for stroke. Stroke incidence in OSAS patients is increased independently from other classical risk factors including hypertension [1–3]. The mechanisms underlying the pre-disposition of OSAS patients to stroke are not well known. Potential mechanisms include hypoxemia during apneas and the resulting increase in sympathetic activity [4,5], hypertension [6–8], cardiac arrhythmia [9,10], increased platelet aggregation [11], diminished vasodilating endothelial capacity [12], and altered cerebral hemodynamics [13,14].

Descriptive studies of arterial blood pressure (ABP) and cerebral blood flow (CBF) as assessed by trancranial Doppler (TCD) performed during obstructive sleep apneas have shown parallel variations of mean ABP and cerebral blood flow velocity (CBFV) during and immediately after the occurrence of obstructive apnea [15–18]. It has then been hypothesized that changes in CBFV might have resulted from cerebral autoregulation failure [13,17]. Cerebral autoregulation, however, was not formally assessed in these studies. In addition, several other potential mechanisms may contribute to changes in CBFV during obstructive sleep apneas. These include retention of CO2 which profoundly modifies cerebral autoregulation [19,20] and negative intrathoracic pressure which results in increased central venous pressure and elevated intracranial pressure during apneas [13]. There are only a few studies of cerebral hemodynamics in patients with OSAS during wakefulness. These studies focused on chemoregulation of CBF (i.e. cerebrovascular reactivity to hypercapnia or hypoxia) which was found to be altered in patients with OSAS [14,21,22]. In this study we aimed to assess cerebral autoregulation (i.e. cerebral pressure autoregulation also termed mechanoregulation [23]) in OSAS patients compared with controls. Cerebral autoregulation was assessed in daytime, during wakefulness.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Acknowledgements
  10. References

Study population

Consecutive patients addressed to a Sleep Unit for investigation of excessive diurnal sleepiness, sleep disturbance or unexplained chronic asthenia were considered for possible inclusion. To be included patients had to be aged 18–75 years. Exclusion criteria were: a history of cerebrovascular disease, extracranial carotid stenosis (> 50% reduction of the arterial lumen) or occlusion, intracranial artery stenosis (> 50% reduction of the arterial lumen) as detected by duplex ultrasonography and TCD performed before inclusion, lack of a temporal bone acoustic window on either side, asymmetrical blood flow velocities between middle cerebral arteries (left-right asymetry of more than 20%), hypercapnia, respiratory failure requiring oxygen supplementation, heart failure, cardiac dysrrhythmia, other clinical conditions that may influence cerebral arterial velocities such as a febrile illness or overt anemia, narcolepsy, treated sleep apnea syndrome, sleep apnea syndrome secondary to neuromuscular or endocrin disase, and central or mixed sleep apnea syndrome. Patients signed informed consent. The research project was approved by the local institutional ethics committee.

Polysomnography

Patients from the OSAS and the control groups were enrolled after an overnight complete polysomnography during which the following parameters had been monitored: electroencephalogram, electrooculogram, anterior tibial and chin electromyogram, measurements of oro-nasal airflow pressure, chest and abdominal excursions, snoring, oxyhemogmobin saturation (finger pulse oxymetry) and three-lead electrocardiogram.

Polysomnograms were scored by a sleep physician who was blind to the results of DCA assessment. Scoring was made following the standard sleep staging criteria. Respiratory events were scored manually according to the recommendations of the American Academy of Sleep Medicine Task Force [24]. Apnea was defined as the cessation of oro-basal airflow of a minimal duration of 10 s. Persistent thoraco-abdominal excursions during apneas qualified them as obstructive. Hypopneas were defined using the following criteria: substantial reduction in airflow (> 50%); moderate reduction in airflow (< 50%) associated with desaturation (> 3%); or moderate reduction in airflow (< 50%) associated with electroencephalographic evidence of arousal. The minimal duration required to define a hypopnea was of 10 s. Persistent thoraco-abdominal excursions defined an obstructive hypopnea. A minimum of 50% obstructive respiratory events were required to characterize sleep apnea syndrome as obstructive. Sleep apnea syndrome was diagnosed using the threshold of 10 apneas or hypopneas per hour. Control subjects had no sleep apnea as defined by these criteria.

Screening for carotid or intracranial artery stenosis

All patients were screened for carotid and intracranial artery stenosis using duplex ultrasonography (ATL Ultramark 9 HDI; ATL/Philips Ultrasound, Bothell, WA, USA) and TCD (Multidop X2 DWL, Singen, Germany). Ultrasound examinations were performed in the morning after polysomnography, prior to the assessment of cerebral autoregulation.

Assessment of dynamic cerebral autoregulation

Measurements were performed with patients supine, within 2–3 h after awaking. During DCA assessment, the state of persistent wakefulness was clinically monitored. Mean CBFV in the middle cerebral artery (MCA), mean ABP and endtidal CO2 pressure (ETPCO2) were recorded over 15 min. The MCA was insonated unilaterally using DWL Multidop device with a 2 MHz probe through the temporal bone at a depth of 50–55 mm. Then the probe was fixed using a rigid headframe (Lam rack, DWL, Germany). This allowed to keep the angle of insonation constant during the continuous monitoring of MCA spectral outline. DCA assessment was performed on the right side, except for patients who had a more accessible left temporal acoustic window. Continuous, non-invasive ABP recording was achieved using a servo-controlled finger plethysmograph (Finapres, Ohmeda, CO, USA) with the patients’ left hand positioned at heart level. Finapres ABP values were controlled using an automated Dinamap monitoring system. ETPCO2 was measured with a nasal capnometer using infrared spectrophotometry (POET TE Plus, Criticare Systems, Waukesha, WI, USA).

Analog outputs from the pressure monitor, the TCD unit (maximal frequency outline), and the capnometer were connected to an analog-to-digital convertor fitted into a computer supporting the Biosan software (Biological Signals Analyser) version 2.2 developed by P. Smielewski and M. Czonyka for data collection and assessment of DCA [25]. Data were analysed using its newer version, ICM+ [26]. The algorithm for DCA assessment uses the correlation coefficient Mx [25,27].

The Mx autoregulatory index is a moving correlation coefficient between mean CBFV and mean ABP. It represents a mathematical approach to quantifying the relationship between spontaneous fluctuations of ABP and CBFV. Using this approach, altered DCA will manifest as an increase in Mx values. The steps of calculation of Mx were as follows: (i) mean ABP and mean CBFV were calculated after spectral filtration to reduce the influence of noise and averaged over 5 s periods; (ii) 36 consecutive 5-s periods were used to calculate the Pearson’s correlation coefficient between mean ABP and mean CBFV over 3-min periods of the recording time series and, (iii) the resulting sets of 3-min correlation coefficients were then averaged yielding the autoregulatory index Mx.

Mx close to +1 denotes that slow fluctuations in ABP produce synchronized slow changes in CBFV, indicating defective DCA. Mx around 0 indicates that variations in ABP are not associated with fluctuations in CBFV, indicating that DCA is preserved. Mx values close to zero indicate intact autoregulation. Mx values above 0.3 indicate disturbed autoregulation. The magnitude of the increase of Mx reflects the severity of autoregulation impairment [27,28].

Statistical analysis

Non-parametric statistical methods were used, because variables did not have normal distributions. Mx values were compared in the OSAS and control groups using the Mann–Whitney U-test. We tested the correlation between Mx and the number of apneas and hypopneas per hour (AHI) amongst OSAS patients and controls using the Spearman rank correlation test. We then used linear and non-linear regression to determine the model best fitting the distribution of Mx according to AHI. Age, mean ABP, body mass index (BMI), ETPCO2 and the prevalence of risk factors for vascular disease were compared in both groups using the Mann–Whitney U-test or the Pearson chi-squared test. Data were given as mean ± SD. Results were considered to be significant for a P-value below 0.05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Acknowledgements
  10. References

Twenty patients were enrolled: 11 OSAS patients and 9 controls. The mean AHI in the OSAS group was 22.7 ± 11.6. The diagnoses in controls were: snoring without OSAS in seven patients, restless legs syndrome in one patient, and no sleep disturbance on polysomnography in one patient. The AHI in control patients was 3.8 ± 2.4. Characteristics of OSAS patients and controls are described in Table 1.

Table 1.   Clinical characteristics of OSAS patients and controls. Data are given as mean ± SD
 OSAS (n = 11)Controls (n = 9)P-value
  1. *Number of patients.

Age52.6 ± 7.9 49.1 ± 5.30.569
Male/female ratio 9/2  4/50.160
Hypertension* 0  10.45
Diabetes* 0  0NA
Hyperlipemia* 6  30.406
Smoking* 4  50.653
Waist circumference (cm)97.5 ± 7.5101.8 ± 9.20.323
Body mass index26.7 ± 3.5 26.8 ± 2.70.849

No significant difference was found between the two groups as for age, hypertension, cigarette smoking, diabetes, dyslipidemia, BMI, and waist circumference (Table 1). There were nine men in the OSAS group and four in the control group (P = 0.16). No significant differences were found between the two groups as for mean ABP, mean CBFV, and ETPCO2 (Table 2).

Table 2.   Mean ABP, mean CBFV and ETPCO2 recordings during dynamic cerebral autoregulation assessment. Data are given as mean ± SD
 OSASControlsP-value
  1. ABP, arterial blood pressure; CBFV, cerebral blood flow velocity; MCA, middle cerebral artery; ETPCO2, endtidal CO2 pressure.

Mean ABP (mmHg)93.0 ± 10.291.1 ± 15.40.569
Mean CBFV in the right MCA (cm/s)52.9 ± 10.459.1 ± 10.20.403
Mean CBFV in the left MCA (cm/s)54.5 ± 10.758.6 ± 9.20.595
ETPCO2 (mmHg)39.6 +/−2.540.1 +/−2.80.704

Mx was higher in OSAS patients (mean Mx ± SD; 0.414 ± 0.138) compared with controls (mean Mx ± SD; 0.233 ± 0.100); P = 0.009; Mann–Whitney U-test (Fig. 1). There was a strong correlation between Mx and AHI amongst patients and controls (Rho = 0.679; P = 0.003; Spearman rank correlation test.). The scatterplot of Mx values according to AHI in OSAS patients and controls is shown in Fig. 2. The model best fitting the data was a logarithmic relationship (R2 = 0.556, P = 0.0002). No correlation was found between Mx and age (P = 0.796), mean ABP (P = 0.272), and BMI (P = 0.618). Mx values did not significantly differ between males and females (P = 0.234).

image

Figure 1.  Mean ± SD of Mx in OSAS patients (0.414 ± 0.138) and controls (0.233 ± 0.100); P = 0.009; Mann–Whitney U-test. Vertical bars represent 95% confidence intervals.

Download figure to PowerPoint

image

Figure 2.  Scatter plot of Mx values according to the apnea–hypopnea index (AHI) in OSAS patients and controls. There was a significant correlation between Mx and AHI (Rho = 0.679; P = 0.003; Spearman rank correlation test). The graph represents the regression model best fitting the data (Mx = 0.134 + 0.093 lnAHI).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Acknowledgements
  10. References

This study demonstrates that cerebral autoregulation is impaired in patients with OSAS during wakefulness. There was a strong relationship between cerebral autoregulation impairment and the number of apneas and hypopneas during sleep, suggesting that patients with the most severe OSAS had the most severe alteration of cerebral autoregulation.

Cerebral autoregulation was assessed through the measurement of the transient changes in CBF that appear within a few seconds of spontaneous variations in ABP. We used the Mx method, a cerebral autoregulation algorithm based on the continuous analysis of spontaneous fluctuations of ABP and CBFV [25]. The Mx method has shown good agreement with other methods that are in use for autoregulation assessment [27,29]. The mean Mx value in the OSAS group (0.414) indicated worse autoregulation as compared with controls (0.233). This finding is consistent with previous observations that showed normal Mx values to be beneath 0.30 [27,28].

Precautions were taken in this study to avoid known or potential factors that could flaw DCA assessment. Patients with carotid stenosis or occlusion as well as patients with a history of cerebrovascular disease were excluded. DCA measurements were performed in awake patients, during daytime, remote from CO2 retention episodes that occurred during sleep apneas. ETPCO2, basal CBFV and ABP values were similar in the OSAS and control groups. Cerebral autoregulation assessment was performed at the same time of the day in both groups.

To our knowledge, this is the first study to report impairment of cerebral autoregulation in OSAS patients, this impairment of cerebral autoregulation being observed during wakefulness. Previous reports have described the occurrence of parallel variations of ABP and CBFV which were observed during and immediately after obstructive sleep apneas [15–18]. A likely explanation to these changes was the concomitant hypercapnia which is known to deeply alter cerebral autoregulation [19,20]. Other studies in awake patients during daytime have shown an impairment of cerebral vasoreactivity to CO2 [14,22]. This study performed during wakefulness in patients with normal endtidal PCO2 indicates an intrinsic impairment of cerebral-pressure autoregulation. It appears thus that both cerebrovascular chemoregulation (i.e. cerebrovascular reactivity to hypoxia and hypercapnia) and mechanoregulation (i.e. the arterial blood pressure–cerebral blood flow relationship) are altered in OSAS patients.

Impaired cerebral autoregulation may contribute to the increased risk of stroke in OSAS patients through two physiopathological pathways: (i) increased vulnerability to drops in ABP leading to cerebral ischaemia, and (ii) excess of flow in cerebral vessels during surges in ABP leading to capillary damage [30]. In OSAS, the cerebral impact of autoregulation failure is probably potentiated by the abrupt changes of ABP that occur during apneic episodes [13]. In addition, impaired cerebral autoregulation might contribute to the poorer neurological outcome that has been reported in stroke patients presenting with sleep apneic episodes [31–33].

Limitations

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Acknowledgements
  10. References

Our study is based on a small sample of controls and patients. Nevertheless, because of relatively strong significance of presented results and the positive correlation of Mx with the AHI, we believe that the findings are generally sound. A second potential limitation is that controls defined with an AHI beneath 10 included snorers. The AHI in controls was 3.8 ± 2.4. This may have lessened the difference in autoregulation impairent between OSAS patients (AHI > 10) and controls (AHI < 10) which however remained significant.

We used arterial pressure [34], instead of cerebral perfusion pressure [25], for assessment of autoregulation. This might have produced a difference if OSAS patients had increased intracranial pressure. Elevated intracranial pressure was, however, very unlikely in our patients because we assessed cerebral autoregulation during wakefulness in daytime and the continuously monitored ETPCO2 was normal in all patients.

The Mx method has been validated in several scenarios against other methods for autoregulation assessment that also use TCD [27,29,34,35] but it has not been validated against positron emission tomography which is currently considered the gold standard for the assessment of impaired autoregulation. The use of TCD for cerebral blood flow measurement requires that the diameter of the insonated MCA is constant during the autoregulatory test to interpret relative changes in CBFV as relative changes in CBF. There is, however, considerable evidence from previous studies that the diameter of MCA does not change during autoregulatory testing [36–38].

Finally, the cross-sectional nature of the study and the limited number of patients do not allow us to establish a causative link between SAS and impaired autoregulation. An interventional study showing the improvement of initially impaired cerebral autoregulation after treatment with nocturnal ventilation would be necessary to demonstrate causality.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Acknowledgements
  10. References

Our findings indicate an impairment of cerebral autoregulation in OSAS patients during wakefulness. This impairment was positively correlated to the severity of the sleep respiratory disturbance as measured by the AHI.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Acknowledgements
  10. References

The study was supported by les Hopitaux de Toulouse and was funded by a grant from the ‘Association pour la Lutte contre les Maladies Cérébro-vasculaires’. We thank Dr Brigitte Guidolin and Dr Thierry Montemayor for their participation in data collection and Dr Gérard Tap for statistical analysis of our data.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Limitations
  8. Conclusion
  9. Acknowledgements
  10. References