Long-term Effects of Treatment with Nasal Continuous Positive Airway Pressure on Daytime Lung Function and Pulmonary Hemodynamics in Patients with Obstructive Sleep Apnea 1 , 2
EMILIA SFORZA, JEAN KRIEGER, EMMANUEL WEITZENBWM, MICHEL APPRILL, ELIANE LAMPERT, and JULIA RATAMAHARO
Introduction
Continuous positive airway pressure (CPAP) applied through the nares (1) is generally acknowledged to be an effective treatment for patients with obstructive sleep apnea (OSA) syndrome (2-5). Its short-term effects have been well described. It eliminates apneas and snoring during sleep (6) and normalizes the sleep pattern (7); it also stabilizes the heart rate and eliminates pulmonary hypertension peaks (8). Within a fewnights the patient's daytime somnolence disappears (9). However, the long-term effects of CPAP on lung function and pulmonary hemodynamics have not been systematically investigated. A few studies have reported occasional patients (10, 11) whose hypoxemia and hypercapnia were reversed after long-term treatment. Issa and coworkers (12) reported that "spirometry, lung volumes, and lung compliance did (not) change in a group of patients after 3 to 4 years of continuous use of nCPAP therapy" and that "the improvement in cardiovascular systemis evident; ... this improvement is partly due to a reduction in pulmonary artery pressure followingthe relief of nocturnal hypoxemia." However, no data were given in support of these statements. We report here the results of a prospective follow-up evaluation of lung function and pulmonary hemodynamics after at least 1 yr of home treatment with nasal CPAP in a group of 54 unselected patients with OSA. Methods Patients Fifty-four consecutive patients (48 men and six women) were studied. They all gave informed consent. They had a sleep apnea syndrome evidenced by heavy snoring, varying degrees of daytime somnolence, and the poly866
SUMMARY Fifty-four patients with obstructive sleep apnea (OSA) syndrome received long-term treatment with nasal continuous positive airway pressure (CPAP).The effects on daytime lung function and pUlmonary hemodynamics were prospectively evaluated by repeating pulmonary function tests, including right heart catheterization after a follow-up period of at least 1 yr (554 ± 28 days, mean ± SEM). Pa02 increased in the patient group as a whole from 69.9 ± 1.4 to 72.8 ± 1.4 mm Hg (p < 0.02). The increase in Pao2 was greater (from 60.4 ± 1.0 to 66.4 ± 2.1, P < 0.01) in those patients who were hypoxemic prior to treatment. PaC02 decreased significantly only In the SUbgroup of patients with significant hypercapnia prior to treatment (n = 7), from 48.5 ± 1.3 to 44.5 ± 1.5 mm Hg (p < 0.01). The improvement in daytime blood gases seemed to be related to an increase In the alveolar ventilation, from 5.2 ± 0.2 to 5.9 ± 0.3 Llmin without a change in the alveolar-arterial p0 2 difference, as calculated In 25 patients. Both the red blood cell count and the hematocrit decreased significantly, from 5,347 ± 90 to 5,024 ± 61 103 /mm3 and from 49.4 ± 0.9 to 47.1 ± 0.6%, P < 0.001 and p < 0.02, respectively. No change was observed in the resting pulmonary arterial pressure. We conclude that nasal CPAP is effective in improving daytime blood gases in patients with OSA. AM REV RESPIR DIS 1990; 141:866-870
graphic demonstration of 10or more hypopneas plus apneas per hour of sleep. None of the patients had a history of overt chronic lung disease such as asthma, pulmonary fibrosis, sarcoidosis, or emphysema, but patients with a history of tobacco smoking, whether associated or not with chronic cough and sputum, were not excluded. Among the 54 patients, 17 had never smoked, 17 had quit smoking more than 1 yr prior to the study, three wereoccasional smokers (lessthan 1cigarette/day), and 17wereregular smokers (seven between 1 and 15 cigarettes/day, 10 more than 16cigarettes/day).None of them received long-term oxygen therapy. All patients had a full evaluation, including polysomnography and respiratory function tests, to evaluate the severity of the syndrome prior to initiating treatment. The evaluation was performed at least 6 wk after a possible episode of acute cardiorespiratory failure. Twenty-one of them also had an evaluation of the ventilatory responses to hypoxia and to hypercapnia, which was available to us only later in the course of the study. After at least I yr (554 ± 28 days, mean ± SEM) of home treatment with CPAP, follow-up investigations wereproposed to all
patients. Some patients refused part of the follow-up investigations; polysomnography and spirometry were obtained in all patients, blood gas analysis in 52, right heart catheterization in 47, and blood cellcounts in 51.Ventilatory responses to hypoxia and hypercapnia were repeated in all 21 patients. Protocol Polysomnography was performed during two consecutive nights, without and with CPAP, as part of the baseline evaluation, and during one night with CPAP as part of the followup reevaluation. The method has been described elsewhere (13). Briefly, it included an
(Received in original form May 18, 1989 and in revised form September 7, 1989) 1 From the Service d'Explorations Fonctionnelles du Systerne Nerveux, Explorations Fonctionnelles Respiratoires PavilIon Laennec, and Explorations Fonctionnelles Respiratoires PavilIon Poincare. 2 Correspondence and requests for reprints should be addressed to Jean Krieger, M.D., Service d'Explorations Fonctionnelles du Systeme Nerveux, 67091 Strasbourg, Cedex, France.
867
EFFECTS OF NASAL CPAP ON DAYTIME WNG FUNCTION IN OSA
electroencephalogram, an electrooculogram, and an electromyogram of chin muscles. Breathing was analyzed with a Fleisch no. 2 pneumotachograph and electronic integrator (Godart Statham, Bilthoven,the Netherlands) attached to a face mask and either thoracic and abdominal strain gauges (n = 24) or an esophageal balloon (n = 30) connected to a pressure transducer (Validyne MP 45; Validyne, Northridge, CAl. Oxygen saturation was measured continuously with an ear oximeter (BioxIII; Ohmeda, Boulder, CO). During the CPAP night, the level of CPAP, applied via a nasal mask with a commercial device (Pression Plus; SEFAM, Vandoeuvre-Ies-Nancy, France), was rapidly increased above the starting level of 2 em H,O until apneas and snoring were abolished. The pressure reached ranged from 4 to 14 em H,O. Sleep was scored using the method of Rechtschaffen and Kales (14). The number of arousals, the sleep efficiency, and the time awake after sleep onset were determined. Central, mixed, and obstructive apneas were defined according to standard criteria (15). Hypopneas weredefined as a 50010 drop in tidal volume from its value in quiet wakefulness prior to sleep onset. The apnea index was the mean number of apneas per hour of sleep, and the apnea plus hypepnea index was the mean number of apneas plus hypopneas per hour of sleep. The mean durations of apneas and hypopneas werealso computed. The following Sao, indices were measured: mean Sao, during 10 min of wakefulness prior to sleep onset, mean of the minimal Sao, after each apnea, minimal Sao" mean Sao, during sleep, and the percentage of sleep time spent below 90 and 80% Saor Oximeter values below 40% wereconsidered to be equal to 40% because of the alinearity of the instrument in this range. Conventional spirography was performed with a IO-L closed spirograph. Static lung volumes were measured by the closed circuit helium dilution method. The reference values used were those of the European Community (16). Right heart catheterization was performed as previously described (17). 'The hemodynamic measurements werealways done while the patient was awake. For the purposes of this study, we used small-diameter floated catheters, either Grandjean flexopulmocaths (French size no. 4) or Swan-Ganz catheters (French size no. 5). The catheters were introduced percutaneously into an antecubital vein. Almost all the patients (42 of 47) underwent an 8-min steady-state exercise on a bicycle ergometer in the supine position. The load was one of 40 watts. Hemodynamic measurements were performed during the last minute of exercise. The ventilatory responses to hypoxia and hypercapnia were measured using the rebreathing method of Read (18)and of Rebuck and Campbell (19),adapted as previously described (20).Airflow, Sao" and end-tidal CO, were monitored using a pneumotachograph,
an ear oximeter, and a rapid infrared analyzer, respectively. The ventilatory responses to hypoxia and hypercapnia were expressed as the slope of the regression of ventilation against Sao, (VE/Sao,) and end-tidal CO, (VE/PETCO,), respectively. The position of the regression line was expressed as VE at PETco, = 60 mm Hg (V 60), rather than the PETco, extrapolated to zero ventilation, which would have resulted in large errors for small errors in slope (21). The overall ventilation was measured, and the alveolar ventilation was computed (22) for a subgroup of 25 patients. The alveolararterial Po, difference was calculated, PAo, being obtained from the ideal alveolar air equation (23). These 25 patients did not differ from the entire patient group for any of the parameters measured. During the home treatment period, the rate of use was measured using the time counter built into the CPAP device. The difference in the time counter readings divided by the duration of treatment was used as an evaluation of the mean daily use of CPAP.
Statistical Analysis Baseline and follow-up values were compared using Student's paired t test. This was done for the entire patient group and for subgroups of hypoxemic (Pao, ~ 65 mm Hg), hypercapnic (Paeo, ~ 45 mm Hg), and pulmonary hypertensive (Ppa ~ 20 mm Hg) patients. In order to investigate the determinants of the changes observed, correlations were calculated between the difference before and after treatment and various parameters, using least squares regression and Pearson's correlation coefficient. Rejection of the null hypothesis required p < 0.05. Results are given as mean ± SEM.
Results
Baseline Evaluation Polysomnographic data. All patients had poor sleep ,quality, as evidenced by the high number of arousals, the low total sleep time and sleep efficiency, and the high time awake after sleeponset (table I). The respiratory sleep disturbances are givenin table l. All subjects had predominantly obstructive and mixed apneas, with sporadic central events in some patients. The total frequency of respiratory events and the sleep hypoxemia covered a wide range of severity (figure I). In most patients the lowest Sao, and the longest apneas were recorded during REM sleep. Daytime data. The anthropometric parameters during the baseline evaluation are shown in table 2. The mean age for the group was 53 ± 1yr with a duration of symptoms of 4 ± 1 yr. Nineteen subjects had systemic arterial hypertension defined as a systolic
TABLE 1 BASELINE POLYSOMNOGRAPHIC DATA Mean
SEM
270 59
18 3 10 6 3 3
Total sleep time, min Sleep efficiency, % Wake after sleep onset, min Apnea index Obstructive apneas, % Mixed apneas, % Central apneas, % Apnea + hypopnea index, nih Mean apnea duration, s Mean hypopnea duration, s Mean Sao" % Mean lowest Sao" % Time spent below Sao, = 90%, % of total sleep time Time spent below Sao, = 80%, % of total sleep time
177 72
83 14 3 91
24
1 4 1
15
o
92 86
1 1
24
3
6
2
pressure ~ 160 and/or a diastolic pressure ~ 100 mm Hg. Twenty-one patients had a hematocrit greater than 50070. The spirometric, blood gas, and right heart hemodynamic data are given in table 3. Twenty-one patients had resting hypoxemia defined as Pao, ~ 65 mm Hg and seven had hypercapnia defined as Paco, ~ 45 mm Hg. Five of the seven hypercapnic patients also had a Pao, ~ 65 mm Hg.
ilJlJJLlJ IIJllll[ o
25
1 2 3
35
4
45
5
0 1 2 3 4 5 6
55
0
PaC0 2
10
20
30
40
R PAP
40
a
60
20
80
40
60
100
80
Ex PAP
:l6JlJIl o
60
120 180
A+H Index
0
40
80
% Time
Sa02
EMILIA SFORZA, JEAN KRIEGER, EMMANUEL WEITZENBWM, MICHEL APPRILL, ELIANE LAMPERT, and JULIA RATAMAHARO
Introduction
Continuous positive airway pressure (CPAP) applied through the nares (1) is generally acknowledged to be an effective treatment for patients with obstructive sleep apnea (OSA) syndrome (2-5). Its short-term effects have been well described. It eliminates apneas and snoring during sleep (6) and normalizes the sleep pattern (7); it also stabilizes the heart rate and eliminates pulmonary hypertension peaks (8). Within a fewnights the patient's daytime somnolence disappears (9). However, the long-term effects of CPAP on lung function and pulmonary hemodynamics have not been systematically investigated. A few studies have reported occasional patients (10, 11) whose hypoxemia and hypercapnia were reversed after long-term treatment. Issa and coworkers (12) reported that "spirometry, lung volumes, and lung compliance did (not) change in a group of patients after 3 to 4 years of continuous use of nCPAP therapy" and that "the improvement in cardiovascular systemis evident; ... this improvement is partly due to a reduction in pulmonary artery pressure followingthe relief of nocturnal hypoxemia." However, no data were given in support of these statements. We report here the results of a prospective follow-up evaluation of lung function and pulmonary hemodynamics after at least 1 yr of home treatment with nasal CPAP in a group of 54 unselected patients with OSA. Methods Patients Fifty-four consecutive patients (48 men and six women) were studied. They all gave informed consent. They had a sleep apnea syndrome evidenced by heavy snoring, varying degrees of daytime somnolence, and the poly866
SUMMARY Fifty-four patients with obstructive sleep apnea (OSA) syndrome received long-term treatment with nasal continuous positive airway pressure (CPAP).The effects on daytime lung function and pUlmonary hemodynamics were prospectively evaluated by repeating pulmonary function tests, including right heart catheterization after a follow-up period of at least 1 yr (554 ± 28 days, mean ± SEM). Pa02 increased in the patient group as a whole from 69.9 ± 1.4 to 72.8 ± 1.4 mm Hg (p < 0.02). The increase in Pao2 was greater (from 60.4 ± 1.0 to 66.4 ± 2.1, P < 0.01) in those patients who were hypoxemic prior to treatment. PaC02 decreased significantly only In the SUbgroup of patients with significant hypercapnia prior to treatment (n = 7), from 48.5 ± 1.3 to 44.5 ± 1.5 mm Hg (p < 0.01). The improvement in daytime blood gases seemed to be related to an increase In the alveolar ventilation, from 5.2 ± 0.2 to 5.9 ± 0.3 Llmin without a change in the alveolar-arterial p0 2 difference, as calculated In 25 patients. Both the red blood cell count and the hematocrit decreased significantly, from 5,347 ± 90 to 5,024 ± 61 103 /mm3 and from 49.4 ± 0.9 to 47.1 ± 0.6%, P < 0.001 and p < 0.02, respectively. No change was observed in the resting pulmonary arterial pressure. We conclude that nasal CPAP is effective in improving daytime blood gases in patients with OSA. AM REV RESPIR DIS 1990; 141:866-870
graphic demonstration of 10or more hypopneas plus apneas per hour of sleep. None of the patients had a history of overt chronic lung disease such as asthma, pulmonary fibrosis, sarcoidosis, or emphysema, but patients with a history of tobacco smoking, whether associated or not with chronic cough and sputum, were not excluded. Among the 54 patients, 17 had never smoked, 17 had quit smoking more than 1 yr prior to the study, three wereoccasional smokers (lessthan 1cigarette/day), and 17wereregular smokers (seven between 1 and 15 cigarettes/day, 10 more than 16cigarettes/day).None of them received long-term oxygen therapy. All patients had a full evaluation, including polysomnography and respiratory function tests, to evaluate the severity of the syndrome prior to initiating treatment. The evaluation was performed at least 6 wk after a possible episode of acute cardiorespiratory failure. Twenty-one of them also had an evaluation of the ventilatory responses to hypoxia and to hypercapnia, which was available to us only later in the course of the study. After at least I yr (554 ± 28 days, mean ± SEM) of home treatment with CPAP, follow-up investigations wereproposed to all
patients. Some patients refused part of the follow-up investigations; polysomnography and spirometry were obtained in all patients, blood gas analysis in 52, right heart catheterization in 47, and blood cellcounts in 51.Ventilatory responses to hypoxia and hypercapnia were repeated in all 21 patients. Protocol Polysomnography was performed during two consecutive nights, without and with CPAP, as part of the baseline evaluation, and during one night with CPAP as part of the followup reevaluation. The method has been described elsewhere (13). Briefly, it included an
(Received in original form May 18, 1989 and in revised form September 7, 1989) 1 From the Service d'Explorations Fonctionnelles du Systerne Nerveux, Explorations Fonctionnelles Respiratoires PavilIon Laennec, and Explorations Fonctionnelles Respiratoires PavilIon Poincare. 2 Correspondence and requests for reprints should be addressed to Jean Krieger, M.D., Service d'Explorations Fonctionnelles du Systeme Nerveux, 67091 Strasbourg, Cedex, France.
867
EFFECTS OF NASAL CPAP ON DAYTIME WNG FUNCTION IN OSA
electroencephalogram, an electrooculogram, and an electromyogram of chin muscles. Breathing was analyzed with a Fleisch no. 2 pneumotachograph and electronic integrator (Godart Statham, Bilthoven,the Netherlands) attached to a face mask and either thoracic and abdominal strain gauges (n = 24) or an esophageal balloon (n = 30) connected to a pressure transducer (Validyne MP 45; Validyne, Northridge, CAl. Oxygen saturation was measured continuously with an ear oximeter (BioxIII; Ohmeda, Boulder, CO). During the CPAP night, the level of CPAP, applied via a nasal mask with a commercial device (Pression Plus; SEFAM, Vandoeuvre-Ies-Nancy, France), was rapidly increased above the starting level of 2 em H,O until apneas and snoring were abolished. The pressure reached ranged from 4 to 14 em H,O. Sleep was scored using the method of Rechtschaffen and Kales (14). The number of arousals, the sleep efficiency, and the time awake after sleep onset were determined. Central, mixed, and obstructive apneas were defined according to standard criteria (15). Hypopneas weredefined as a 50010 drop in tidal volume from its value in quiet wakefulness prior to sleep onset. The apnea index was the mean number of apneas per hour of sleep, and the apnea plus hypepnea index was the mean number of apneas plus hypopneas per hour of sleep. The mean durations of apneas and hypopneas werealso computed. The following Sao, indices were measured: mean Sao, during 10 min of wakefulness prior to sleep onset, mean of the minimal Sao, after each apnea, minimal Sao" mean Sao, during sleep, and the percentage of sleep time spent below 90 and 80% Saor Oximeter values below 40% wereconsidered to be equal to 40% because of the alinearity of the instrument in this range. Conventional spirography was performed with a IO-L closed spirograph. Static lung volumes were measured by the closed circuit helium dilution method. The reference values used were those of the European Community (16). Right heart catheterization was performed as previously described (17). 'The hemodynamic measurements werealways done while the patient was awake. For the purposes of this study, we used small-diameter floated catheters, either Grandjean flexopulmocaths (French size no. 4) or Swan-Ganz catheters (French size no. 5). The catheters were introduced percutaneously into an antecubital vein. Almost all the patients (42 of 47) underwent an 8-min steady-state exercise on a bicycle ergometer in the supine position. The load was one of 40 watts. Hemodynamic measurements were performed during the last minute of exercise. The ventilatory responses to hypoxia and hypercapnia were measured using the rebreathing method of Read (18)and of Rebuck and Campbell (19),adapted as previously described (20).Airflow, Sao" and end-tidal CO, were monitored using a pneumotachograph,
an ear oximeter, and a rapid infrared analyzer, respectively. The ventilatory responses to hypoxia and hypercapnia were expressed as the slope of the regression of ventilation against Sao, (VE/Sao,) and end-tidal CO, (VE/PETCO,), respectively. The position of the regression line was expressed as VE at PETco, = 60 mm Hg (V 60), rather than the PETco, extrapolated to zero ventilation, which would have resulted in large errors for small errors in slope (21). The overall ventilation was measured, and the alveolar ventilation was computed (22) for a subgroup of 25 patients. The alveolararterial Po, difference was calculated, PAo, being obtained from the ideal alveolar air equation (23). These 25 patients did not differ from the entire patient group for any of the parameters measured. During the home treatment period, the rate of use was measured using the time counter built into the CPAP device. The difference in the time counter readings divided by the duration of treatment was used as an evaluation of the mean daily use of CPAP.
Statistical Analysis Baseline and follow-up values were compared using Student's paired t test. This was done for the entire patient group and for subgroups of hypoxemic (Pao, ~ 65 mm Hg), hypercapnic (Paeo, ~ 45 mm Hg), and pulmonary hypertensive (Ppa ~ 20 mm Hg) patients. In order to investigate the determinants of the changes observed, correlations were calculated between the difference before and after treatment and various parameters, using least squares regression and Pearson's correlation coefficient. Rejection of the null hypothesis required p < 0.05. Results are given as mean ± SEM.
Results
Baseline Evaluation Polysomnographic data. All patients had poor sleep ,quality, as evidenced by the high number of arousals, the low total sleep time and sleep efficiency, and the high time awake after sleeponset (table I). The respiratory sleep disturbances are givenin table l. All subjects had predominantly obstructive and mixed apneas, with sporadic central events in some patients. The total frequency of respiratory events and the sleep hypoxemia covered a wide range of severity (figure I). In most patients the lowest Sao, and the longest apneas were recorded during REM sleep. Daytime data. The anthropometric parameters during the baseline evaluation are shown in table 2. The mean age for the group was 53 ± 1yr with a duration of symptoms of 4 ± 1 yr. Nineteen subjects had systemic arterial hypertension defined as a systolic
TABLE 1 BASELINE POLYSOMNOGRAPHIC DATA Mean
SEM
270 59
18 3 10 6 3 3
Total sleep time, min Sleep efficiency, % Wake after sleep onset, min Apnea index Obstructive apneas, % Mixed apneas, % Central apneas, % Apnea + hypopnea index, nih Mean apnea duration, s Mean hypopnea duration, s Mean Sao" % Mean lowest Sao" % Time spent below Sao, = 90%, % of total sleep time Time spent below Sao, = 80%, % of total sleep time
177 72
83 14 3 91
24
1 4 1
15
o
92 86
1 1
24
3
6
2
pressure ~ 160 and/or a diastolic pressure ~ 100 mm Hg. Twenty-one patients had a hematocrit greater than 50070. The spirometric, blood gas, and right heart hemodynamic data are given in table 3. Twenty-one patients had resting hypoxemia defined as Pao, ~ 65 mm Hg and seven had hypercapnia defined as Paco, ~ 45 mm Hg. Five of the seven hypercapnic patients also had a Pao, ~ 65 mm Hg.
ilJlJJLlJ IIJllll[ o
25
1 2 3
35
4
45
5
0 1 2 3 4 5 6
55
0
PaC0 2
10
20
30
40
R PAP
40
a
60
20
80
40
60
100
80
Ex PAP
:l6JlJIl o
60
120 180
A+H Index
0
40
80
% Time
Sa02
