INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: obstructive sleep apnea; pulmonary hypertension; CPAP; hypertension; hemodynamics

Mild daytime pulmonary hypertension (PH) is a common complication in patients with moderate to severe obstructive sleep apnea (OSA) (1) and can occur in the absence of lung or heart disease (5, 8, 9). The potential mechanisms of daytime PH in OSA are elevated daytime pulmonary vascular tone secondary to hypoxic pulmonary vasoconstriction (HPV), hypoxia-induced endothelial dysfunction (10), and pulmonary vascular remodeling (PRm) (9).

Whether PH and PRm in OSA are reversible with treatment is uncertain. Experiments in animal models of intermittent hypoxia-related PRm indicate that morphologic changes are potentially reversible with abolition of hypoxia (11). However, there is a paucity of human studies on the subject. OSA is a reversible disorder and therefore is a model of episodic (nocturnal) repetitive hypoxia in humans, suitable to investigate the development and reversal of PH and hypoxic PRm. If PH were reversible by effective treatment of OSA, it would strengthen the case for a casual link between OSA and PH.

The effects of treatment with nocturnal nasal continuous positive airway pressure (CPAP) ventilation on pulmonary hemodynamics and gas exchange in OSA are controversial. Although some studies reported a decrease in pulmonary artery pressure (Ppa) with treatment (2), others showed no change (12, 13). These earlier studies included patients with very mild sleep apnea and coexisting chronic airflow limitation. The studies were also weakened by a lack of objective monitoring of CPAP compliance.

The purpose of the present study was, first, to determine whether PH in patients with OSA without lung or cardiac disease is reversible with treatment. Second, if reversal of PH occurred, we aimed to establish whether this was associated with: (1) changes in gas exchange consistent with reversal of ventilation-perfusion inequality, (2) altered responsiveness to hypoxia, or (3) improved pulmonary vascular compliance, which would imply reversal of PRm.

In this study the results are presented of 4 mo of nasal CPAP treatment on pulmonary hemodynamics, lung function, and gas exchange in patients with moderate to severe OSA without lung or cardiac disease. Objective compliance with CPAP was monitored by an inbuilt digital pressure-sensitive compliance meter.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Selection

Twenty-two patients with OSA without lung or heart disease or systemic hypertension (21 male, age 49.9 ± 2.5 yr, mean ± SEM) from the 32 patients included in our previous study (9) were followed after commencing nasal CPAP treatment. None of the patients had been treated previously and none were taking antihypertensive medication or drugs affecting cardiac or pulmonary function. All subjects gave their written informed consent before commencement of the study. The protocol was approved by the Research and Ethics Committee of the Repatriation General Hospital-Daw Park.

Study Protocol

Resting arterial blood gas analysis and Doppler echocardiography, including dobutamine stress-Doppler echocardiography were conducted at baseline and repeated at 1 and 4 mo after initiation of nasal CPAP treatment. Detailed lung function measurements were performed at baseline and after 4 mo of CPAP treatment.

Polysomnography

Standard polysomnographic recordings were made before CPAP treatment using a computerized data acquisition system (Sleepwatch, Compumedics, Melbourne, Australia) as previously described (9). Sleep was scored manually in 30-s epochs using the system of Rechtschaffen and Kales (14). Hypopnea was defined as a 50% or greater decrease from baseline values in abdominal and thoracic excursions or airflow lasting 10 s or more. Apnea was defined as cessation of airflow for 10 s or more and characterized as obstructive, mixed, or central. Sleep-disordered breathing was quantitated by calculating the apnea-hypopnea index (AHI; apneas plus hypopneas per hour of sleep) and various indices of arterial O₂ saturation (Sa_O2) during sleep (lowest Sa_O2 and percent time Sa_O2 < 90%).

Lung Function

Detailed lung function tests, including forced expiratory lung volumes, static lung volumes, single-breath carbon monoxide diffusing capacity (DL_CO), closing volume (CV) by single-breath nitrogen testing, and resting arterial blood gas analysis were performed with patients in the sitting position as described previously (9).

Pulmonary Hemodynamics and Cardiac Output

Pulsed Doppler echocardiography (Acuson Computed Sonography System; Acuson, Mountain View, CA) was used to estimate systolic and mean Ppa () and cardiac output () by the methods of Morera and colleagues (15) and Spodick and Koito (16) respectively. Detailed descriptions of these techniques are given in our earlier reports (5, 9, 17). The Doppler methods provide reliable estimates of catheter-derived Ppa and thermodilution , with high reproducibility (17). PH was defined as Doppler-estimated of  20 mm Hg.

Doppler Assessment of the Left Ventricular Diastolic Function

Transmitral flow measurements of integrated early passive (E) and late atrial (A) filling velocities were recorded during different experimental conditions. The ratio of integrated velocity-time intervals of A and E waves (A/E ratio) was used as an index of left ventricular (LV) diastolic function (18, 19).

Response of the Pulmonary Circulation to Changes in Inspired Oxygen Concentration and Pulmonary Blood Flow

As described previously (9), resting Doppler hemodynamic measurements were obtained during room air breathing, sustained isocapnic hypoxia (fraction of inspired oxygen [FI_O2] = 0.11, Sa_O2 ~ 80%), and hyperoxia (FI_O2 = 0.50, Sa_O2 ~ 99%). Hemodynamic measurements were also made during an incremental dobutamine infusion with patients breathing 50% O₂ (9). Briefly, the dobutamine infusion rate was increased at 8-min intervals to 10, 20, and 30 µg/kg/min and measurements were made at steady-state at the end of each infusion period. All physiologic signals were recorded using a computerized data acquisition system (WINDAQ, version 1.15; Dataq Instruments Inc., Akron, OH) with digital sampling speed of 50 Hz.

CPAP Treatment and Monitoring of Effective Compliance

The therapeutic pressure required to abolish obstructive apneas and hypopneas was manually titrated in the sleep laboratory in all patients. For the purpose of this study, the patients were given a CPAP pump with an inbuilt pressure-sensitive compliance meter (Sullivan III; ResMed, Sydney, Australia). The effective compliance (time spent at the effective pressure) with nasal CPAP was digitally recorded by a microprocessor in the CPAP pump and downloaded to a personal computer (PC) at 1-mo and 4-mo follow-up visits.

Data Analysis and Statistics

Because the purpose of the study was to measure the effects of CPAP treatment on pulmonary hemodynamics and gas exchange, it was decided to remove noncompliant patients from further analysis. As the minimal effective treatment adherence in terms of PH and pulmonary vascular resistance is not known, it was decided to use a conservative threshold of an average of 2 h per night as the cutoff. Baseline hemodynamic, lung function, and gas exchange parameters were compared with corresponding values after 1 and 4 mo of CPAP treatment, using analysis of variance (ANOVA) for repeated measures. Where the F statistic indicated statistical significance, post hoc comparisons were performed using the Neumann-Keuls method. Student's t test was used to compare numerical variables (e.g., age, body mass index) between patients with and without PH. The chi-square test was used to compare categorical variables (e.g., smoking history). Results are presented as mean ± SEM. A p value of < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline Anthropomorphic, Lung Function, Sleep-Disordered Breathing, and Pulmonary Hemodynamic Measurements

By design, all routine lung function tests of patients were within the normal range: FEV₁ (100.9 ± 2.1% predicted, range: 84 to 117% predicted); FVC (95.1 ± 1.7% predicted, range: 81 to 114% predicted); DL_CO (100.9 ± 2.8% predicted, range 76 to 122% predicted); TLC (95.0 ± 2.1% predicted, range 81 to 116% predicted). Patients had a wide range of sleep apnea severity (apnea index [AI]: 20.0 ± 4.4/h, range 0 to 58/h; AHI: 48.6 ± 5.2/h, range 16 to 102/h; minimum Sa_O2: 79.2 ± 2.3%, range 51 to 91%; time spent with Sa_O2 < 90%: 33.8 ± 11.9 min, range 0 to 218 min; average desaturation per event: 6.1 ± 0.8%, range 2 to 15%). Doppler echocardiography and 12-lead electrocardiogram (ECG) measurements during the experiments did not reveal any unsuspected LV or valvular heart disease.

Five patients (all males) were pulmonary hypertensive (PH) and 15 patients (14 males) were pulmonary normotensive (non-PH). PH and non-PH groups were not different with respect to age, body mass index, or smoking history. Ppa elevations in PH patients were mild ( ranged from 20 to 31 mm Hg, and systolic Ppa from 26 to 43 mm Hg).

CPAP Compliance

Two patients failed to meet the minimum requirement of compliance, and their results were excluded from analysis. For the remaining 20 patients, the CPAP prescribed and delivered through the mask ranged from 7 to 13 cm H₂O (mean CPAP = 10.4 ± 0.46 cm H₂O). Average CPAP compliance for the periods 0 to 1 mo and 1 to 4 mo was 4.9 ± 0.3 and 5.2 ± 0.4 h per night respectively. Average effective compliance for the whole 4-mo period was 5.1 ± 0.3 h per night. Seventeen patients used CPAP effectively for an average of more than 4 h per night.

Effect of CPAP on Pulmonary Hemodynamics

During CPAP treatment there was a fall of daytime (room air breathing) in the 20 compliant patients with OSA, which reached statistical significance after 4 mo of treatment (Table 1, Figure 1). A statistically significant decline in was also present when all 22 patients were included ( = 17.0 ± 1.2 mm Hg and 14.5 ± 0.8 mm Hg at baseline and 4 mo of CPAP treatment respectively; p < 0.05). In the one noncompliant patient with daytime PH there was no decline in (Figure 1). CPAP treatment did not significantly change the measured at each FI_O2 and each rate of dobutamine infusion. Total pulmonary vascular resistance (TPVR) fell significantly during CPAP treatment.

View larger version (21K): [in this window] [in a new window]   Figure 1.   Daytime Ppa during room air breathing in 20 patients with OSA over 4 mo of CPAP treatment. The biggest drop in Ppa was observed in pulmonary hypertensive patients with Ppa  20 mm Hg.

The effects of different levels of FI_O2 on at various stages of CPAP treatment are shown in Figure 2. There was a significant decrease in with CPAP treatment at all levels of FI_O2. Sa_O2 was the same for each stage of treatment at each FI_O2 (Sa_O2 at baseline, 1 mo, and 4 mo: FI_O2 0.11: 81.1 ± 0.4%, 80.4 ± 0.3%, 80.9 ± 0.3%; FI_O2 0.21: 96.5 ± 0.2%, 96.2 ± 0.1%, 96.6 ± 0.2%; FI_O2 0.5: 98.9 ± 0.1%, 99.0 ± 0.1%, 99.0 ± 0.1%). The hypoxic constrictor response (measure of pulmonary vascular reactivity) measured as the difference in Ppa between FI_O2 of 0.5 (no hypoxic pulmonary vasoconstrictor stimulus) and FI_O2 of 0.11 (marked hypoxic pulmonary vasoconstrictor stimulus) was significantly lower after 4 mo of CPAP treatment (Ppa/ Sa_O2 before CPAP, 10.0 ± 1.6 mm Hg; after 4 mo CPAP, 6.3 ± 0.8 mm Hg, p < 0.05).

View larger version (16K): [in this window] [in a new window]   Figure 2.   Ppa breathing varying inspired O2 concentrations at baseline (BL), 1 mo (1/12), and 4 mo (4/ 12) of CPAP treatment (n = 20). SaO2 values at each level of inspired O2 concentration remained the same throughout treatment. Ppa was lower after 4/12 of CPAP treatment at all levels of inspired oxygen. p < 0.05 versus baseline, p < 0.05 versus 1 mo of CPAP treatment.

Pulmonary hemodynamic responses to increased pulmonary blood flow during dobutamine infusion are shown in Table 1 and Figure 3. There was a significant stepwise increase in (p < 0.01) at each level of dobutamine infusion under hyperoxic breathing, which was accompanied by a fall in TPVR. The significant finding is that the pressure/flow (Ppa/) curves seem to move downward in a parallel fashion after CPAP treatment (Figure 3). In the five pulmonary hypertensive patients there was a trend for the pressure-flow curve to change its slope after 4-mo CPAP treatment (Figure 4); however, the sample size is insufficient for meaningful statistical comparisons.

View larger version (14K): [in this window] [in a new window]   Figure 3.   Pulmonary hemodynamic responses to increasing dobutamine infusion rates at baseline (BL), 1 mo (1/12), and 4 mo (4/ 12) of CPAP treatment in 20 compliant patients with OSA. Points on the curves correspond to different dobutamine infusion rates: from left to right, 0, 10, 20, and 30 µg/kg/min, respectively. Patients breathed 50% oxygen to abolish HPV. One patient did not consent to follow-up dobutamine infusion; two additional patients developed intolerable side effects at the highest dobutamine infusion rate, and data were not obtained at that infusion rate. All Ppa and values obtained are shown, but statistical comparisons were performed only in patients with complete data sets (n = 17). Ppa was significantly higher at all levels of dobutamine infusion (p < 0.05) than Ppa at 0 µg/kg/min dobutamine at BL, 1/12, and 4/12.

View larger version (12K): [in this window] [in a new window]   Figure 4.   Pulmonary hemodynamic responses to increasing dobutamine infusion rates at baseline (BL) and 4 mo (4/12) of CPAP treatment in five compliant pulmonary hypertensive patients with OSA.

Effect of CPAP on Lung Function and Gas Exchange

Lung function and gas exchange measurements before and after 4 mo of CPAP treatment were performed in all subjects (see Table E1 in the online data supplement). A complete set of single-breath nitrogen measurements was obtained in 14 of 20 patients with OSA. There was a small, but statistically significant decline in forced expiratory lung volumes (FEV₁, FVC) and Pa_CO2, and a small, but statistically significant increase in total lung capacity (TLC).

Body mass index for the group was 32.0 ± 0.8 (range: 24.0 to 36.7). Body weight was stable throughout the treatment: 96.3 ± 3.0 kg, 96.6 ± 2.9 kg, and 96.7 ± 3.1 kg before, and after 1 mo and 4 mo treatment with CPAP respectively (not significant [NS]).

Effect of CPAP on LV Function, Heart Rate, and Systemic Blood Pressure

A complete set of integrated velocity-time interval recordings (i.e., A/E ratios) was obtained in 20 patients under various oxygen breathing conditions. Suitable measurements were obtained in 19 patients at dobutamine infusion rates of 10 and 20 µg/kg/min and in 11 patients at dobutamine infusion rate of 30 µg/kg/min. One of the 20 compliant patients refused follow-up dobutamine stress-Doppler echocardiography. At the highest dobutamine flow rate, another two patients developed intolerable side effects, and six patients had A/E Doppler velocity tracings unsuitable for analysis (fusion of A and E envelopes).

The A/E ratio while breathing various inspired oxygen concentrations was unchanged after CPAP treatment (Table 2). However, A/E ratio during dobutamine infusion decreased after CPAP treatment, reaching statistical significance at the dobutamine flow rate of 20 µg/kg/min, suggesting improvement of diastolic function during conditions of increased after CPAP treatment.

Changes in heart rate and systemic blood pressure (BP) during various inspired oxygen concentrations and dobutamine flow rates throughout the CPAP treatment are shown in Table 3. The heart rate while breathing various inspired oxygen concentrations was unchanged after CPAP treatment, but heart rate during dobutamine infusion decreased after CPAP treatment. There was no change in systolic BP with CPAP treatment. However, diastolic BP fell significantly during room air and 50% oxygen breathing, as well as during 20 µg/kg/min dobutamine infusion flow rate after 4 mo of CPAP treatment.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The main finding of this study is that CPAP treatment in patients with OSA, who had no lung or cardiac disease, improved daytime pulmonary hemodynamics. The greatest pulmonary hemodynamic improvement (i.e., decrease in the daytime Ppa and TPVR) was observed in patients with OSA who had sustained daytime PH. The next most important finding is that CPAP treatment appeared to reduce hypoxic pulmonary vascular reactivity, i.e., the propensity of the pulmonary circulation to constrict in response to a hypoxic stimulus. These findings are consistent with the hypothesis that intermittent nocturnal hypoxia, such as occurs in OSA, causes pulmonary endothelial dysfunction, which is reversible with abolition of OSA with CPAP.

Improvement in Pulmonary Hemodynamics with CPAP

Previously published results of CPAP treatment on daytime Ppa and right ventricular function in patients with OSA are controversial. One of the early studies reported a substantial decrease in Ppa and pulmonary vascular resistance, and significant improvement in right ventricular ejection fraction in patients with OSA treated with tracheostomy (2). Another study (20) reported an increase in right ventricular ejection fraction from 30 to 39% in seven obese patients with OSA treated with CPAP for 6 to 24 mo. However, no measurements of and were performed in that study. A Polish group (21) studied pulmonary hemodynamics in 19 patients with OSA and reported a normal average for the group (16.6 ± 5.7 mm Hg) which did not change significantly after 1 yr of CPAP treatment (15.8 ± 4.0 mm Hg). However, in two patients with resting daytime PH, dropped from 33 to 25 mm Hg and 28 to 18 mm Hg respectively. This is in concordance with the results of the present study, which showed that the largest decrease in occurred in patients with daytime PH.

Another European group (12) studied the effects of more than 1-yr CPAP treatment on daytime lung function and hemodynamics in patients with OSA, and found a slight but significant improvement in daytime blood gases with no change in the resting daytime Ppa. A more recent study from the same group extended the CPAP follow-up to 5 yr in 65 unselected patients with severe OSA (13). A significant improvement in daytime Pa_O2 was reported in a subgroup of hypoxemic patients with OSA, with a slight increase in Pa_CO2 over the 5-yr period. In 44 patients with OSA who underwent right heart catheterization at 5-yr follow-up, Chaouat and colleagues (13) reported no significant change in resting daytime Ppa, which on average remained within normal limits throughout the treatment (16 ± 5 mm Hg versus 17 ± 5 mm Hg; NS). The main limitations of the latter study are the inclusion of patients with airflow limitation and daytime hypoxemia, and the lack of an objective measurement of patient's compliance with CPAP. However, similar to the study of Hawrylkiewicz and colleagues (21), this group showed a trend toward a fall in in 11 patients with daytime PH (24 ± 5 mm Hg before CPAP versus 20 ± 7 mm Hg after 5 yr of CPAP, p = 0.14 [NS]). It is possible that CPAP reversed the component of PH secondary to OSA, whereas the chronic obstructive pulmonary disease (COPD) component persisted or increased with time.

In the current study CPAP significantly improved pulmonary hemodynamics by lowering both Ppa and TPVR, without significant change in resting LV function and despite the small decline of forced expiratory lung volumes. The strength of our study is that the CPAP compliance was monitored and that patients with clinical lung or cardiac disease were excluded, so any hemodynamic improvement can be solely attributed to abolition of apnea/nocturnal hypoxemia by CPAP.

Potential Reasons for Decrease in Ppa with CPAP Treatment

Decrease in pulmonary vascular reactivity to hypoxia. The results from this study indicate decreased hypoxic pulmonary vascular reactivity after CPAP treatment. One possible explanation for this is that CPAP treatment over 4 mo improved pulmonary endothelial function, which is known to control the pulmonary vascular tone (10, 22, 23). It can be speculated that intermittent hypoxemia (10) together with the increased shear stress (23, 24) due to increased flow rates after sleep apneas causes pulmonary endothelial dysfunction and alters the balance between vasoconstricting and vasodilating endothelial modulators, thereby causing an exaggerated hypoxic pulmonary vasoconstrictor response. Endothelial dysfunction has been postulated as one of the causes for systemic hypertension associated with OSA (25).

Reversal of PRm. Another possible explanation for the decrease in Ppa and pulmonary vascular reactivity to hypoxia is reversal of the structural changes of PRm (i.e., increased pulmonary arteriolar muscularization and medial thickening). Different pressure/flow (Ppa/) slopes between PH and non-PH patients with OSA, as shown previously (9), suggest the presence of stiffer, remodeled pulmonary vasculature in pulmonary hypertensive patients with OSA. In this study the Ppa/ curves seemed to move downward in a parallel fashion after treatment (Figure 3). This suggests vasodilatation due to improved endothelial function rather than reversal of PRm. However, in the five pulmonary hypertensive patients of the present study there was a trend for the pressure-flow curve to change its slope after 4-mo CPAP treatment, implying reversal of PRm (Figure 4), which might have contributed to the CPAP-induced decrease in Ppa.

Change in LV Diastolic Function

It was recently suggested by Sanner and colleagues (8) that, in the absence of lung disease, PH in OSA is associated with LV dysfunction and is of postcapillary type. The same group also suggested that in the absence of LV dysfunction the level of nocturnal hypoxemia (expressed as time spent with Sa_O2 < 90%) is an independent predictor of daytime Ppa, which was not confirmed in our previous studies (5, 9). In Sanner and colleagues' study (8) LV dysfunction, assessed by catheter measurement of pulmonary capillary wedge pressure (Ppcw), was strongly associated with the presence of systemic arterial hypertension (all eight patients with elevated Ppcw were hypertensive). Thus, at least some of the PH in their patients with OSA could have been a result of hypertensive LV hypertrophy with diastolic dysfunction. In the current study, however, patients with known systemic hypertension were eliminated and no single BP measurement during room air breathing exceeded 150/90 mm Hg. Also, as shown previously (9), daytime PH in patients with OSA was not associated with LV dysfunction as assessed by A/E velocity time intervals, which are strongly correlated with left atrial filling pressures and Ppcw (19). A/E integral ratios in subjects studied before and after CPAP treatment were within the range reported for age-matched normal subjects (26).

An interesting finding of the current study is the improvement after CPAP treatment of LV diastolic function during increased flow rates with dobutamine infusion, which was paralleled by a decrease in the heart rate at higher flow rates. A possible explanation is that patients with OSA in this study had some degree of LV diastolic dysfunction, which was uncovered at higher flow rates or higher heart rate. However, LV function (and filling) is heart rate dependent (27) and it is therefore not surprising that the reduction of heart rate during dobutamine infusion after 4 mo of CPAP treatment was paralleled by an improvement of LV diastolic function. It is therefore unlikely that patients in this study had any LV dysfunction or that changes in their LV function had a significant role in their pulmonary hemodynamic improvement with CPAP.

One possible explanation for the decline in heart rate during dobutamine infusion after CPAP treatment is that CPAP treatment restored the sympathovagal balance by increasing the vagal and decreasing the sympathetic nervous activity (28). It has been previously shown that sympathetic nerve activity, as well as circulating plasma norepinephrine levels, is increased in patients with OSA (29, 30) and this is potentially reversible with treatment (30, 31). Long-term CPAP studies confirmed the decrement of awake muscle sympathetic activity in compliant individuals with OSA (32). A similar mechanism might operate to cause a reduction of diastolic BP with CPAP, as observed in this study population.

Change in Lung Function and Pulmonary Gas Exchange

Very small, but statistically significant changes in forced expiratory volumes (FEV₁ and FVC), TLC, and Pa_CO2 were recorded in patients with OSA over a 4-mo treatment period with CPAP (see Table E1). Although pulmonary vascular resistance is related to lung volume (33), these very small changes are unlikely to result in significant hemodynamic consequences. The functional residual capacity (FRC; i.e., the resting lung volume) at which all hemodynamic measurements were made was unchanged throughout the study period. Indices of arterial oxygenation (Pa_O2 and Sa_O2) were also unchanged. Therefore, the observed improvement in pulmonary hemodynamics cannot be attributed to improved oxygenation and reversal of daytime HPV. The small decline in Pa_CO2 of 1 mm Hg over the 4-mo treatment period is also unlikely to have affected pulmonary hemodynamics.

It can only be speculated as to the mechanisms of the small decline in forced expiratory volumes. There was no change in body weight to explain these changes. One possible explanation is that small seasonal variations in lung function might have existed in some of the patients, particularly in current smokers or ex-smokers (n = 9) (34). Another possibility is a small, but real decline in lung function over a 4-mo treatment period because nine of 20 patients were current smokers or ex-smokers. Both explanations are, however, unlikely because FEV₁/FVC ratio and maximal midexpiratory flow (MMEF) remained unchanged. Other possible explanations for the observed small change in lung volumes are the increased variability of lung function in the morning hours (35), when most of the lung function measurements in this study were made, or diurnal changes in pulmonary function caused by endogenous circadian rhythms (36).

Previous studies have reported conflicting results of CPAP treatment on lung function in patients with OSA. One study showed a small decline in FEV₁ after 5 yr of CPAP treatment (13), whereas another reported improvement in airflow obstruction (37).

Systemic Blood Pressure in Patients with OSA during CPAP Treatment

A small, but significant fall in diastolic BP was observed in this group of normotensive patients with OSA. This supports the hypothesis that a direct link may exist between OSA and systemic BP control. Patients with OSA have a blunted venodilatory responsiveness to bradykinin, which contributes to development of systemic hypertension, and appears to be reversible with nasal CPAP treatment (38). It has also been demonstrated that treatment of OSA with nasal CPAP normalizes elevated atrial natriuretic peptide levels (39) and reverses a depressed baroreflex (40). Several studies have reported an immediate decrease in nocturnal and morning awake intra-arterial BP after CPAP therapy (41, 42), as well as a modest decline in mean daytime systolic and diastolic BP (43). Suzuki and colleagues (44) found that BP reduction depended on the level of pretreatment BP, i.e., systolic and diastolic BP decreased only in hypertensive patients with OSA, whereas in normotensive patients there was no significant change. In contrast, our study found a significant fall in diastolic BP despite excluding known hypertensive patients. Most earlier studies suffer from the common weakness that the effective compliance with CPAP was not monitored. In the current study effective compliance with CPAP was monitored and confirmed. Therefore, it seems likely that the reduction of diastolic BP with CPAP is a genuine effect of the abolition of apneas.

Methodological Considerations

As mentioned earlier, the major limitation of the current study is the lack of a placebo control arm. Also of concern is whether the small changes observed in Ppa after CPAP treatment could be expected to be reliably detected by Doppler methods. In our previous study (17) we showed that under repeatability conditions (four independent, equally spaced replicate Doppler tracings over 1 h) the coefficient of variation in was 8.9%. For a baseline of approximately 16 mm Hg this variability would be equivalent to approximately ± 1.45 mm Hg. Thus, the observed group mean difference of 2.9 mm Hg after CPAP treatment is twice that expected from variability ("error") of the measurement technique in a typical patient. Other factors that convince us that the changes we observed after treatment were real treatment effects rather than experimental errors are that: (1) changes in Ppa of similar magnitude were found using independent Doppler measurements in the same subjects under a range of different conditions (e.g., hypoxia, hyperoxia, dobutamine infusion at various flow rates); and (2) there was a progressive decrease in group through the treatment period (i.e., from 1 mo to 4 mo).

Another limitation of this study was the small number of pulmonary hypertensive patients included in the CPAP follow-up (n = 5), which made testing the hypothesis of reversal of PRm statistically difficult. It is anticipated that a much larger sample size of pulmonary hypertensive patients with OSA would be required to show a difference in pressure/flow slope with CPAP treatment and thereby infer a reversal of PRm.

Finally, we did not confirm by follow-up polysomnography ongoing effectiveness of CPAP treatment in abolishing apneas and hypopneas. However, we are confident that CPAP pressure requirements did not change over the 4-mo period, because body weight in our subjects did not change.

Conclusion

CPAP treatment in patients with OSA was associated with beneficial pulmonary and systemic cardiovascular effects. It reduced Ppa and TPVR as well as systemic diastolic blood pressure. CPAP also reduced hypoxic pulmonary vascular reactivity. These hemodynamic improvements in patients with OSA are consistent with the hypothesis of reversal or improvement in endothelial dysfunction after CPAP treatment. Ppa was decreased by CPAP treatment without any concomitant change in alveolar arterial oxygen pressure (AaPO₂) or Pa_O2. The results are inconclusive regarding the effects of CPAP treatment on PRm in OSA.