Effect of Weight Loss on Upper Airway Collapsibility in Obstructive Sleep Apnea 1- 3
ALAN R. SCHWARTZ,4 AVRAM R. GOLD, NORMAN SCHUBERT, ALEXANDRA STRYZAK, ROBERT A. WISE, SOlBERT PERMUTT, and PHILIP L. SMITH
Introduction
Obstructive sleep apnea is a common disorder in middle-aged, obese men (1). In recent studies, we and others have demonstrated that significant improvements in obstructive sleep apnea accompany reductions in body weight (2-4), suggesting that obesity plays a major role in the pathogenesis of this disorder. Nevertheless, the mechanism for decreases in apnea severity with weight loss is unclear. It is clear from many studies that obstructive sleep apnea is caused by recurrent occlusion of the pharynx during sleep (5, 6). Additional work has shown that airway occlusion in these patients is due to the increase in upper airway collapsibility that occurs during sleep (7-11). Although the mechanism for the improvement in obstructive sleep apnea with weight loss has not been delineated, it is possible that reductions in apnea severity are due to decreases in upper airway collapsibility. In previous studies, differences in upper airway collapsibility have been related to alterations in the upper airway critical pressure (Pcrit), which has been defined by the nasal pressure (PN) below which the airway occludes (7-9). Specifically, these studies have demonstrated elevations in Pcrit in apneic patients and lower levels of Pcrit in association with greater degrees of upper airway patency. Therefore, the purpose of the present study was to test the hypothesis that weight loss in apneic patients results in decreases in upper airway Pcrit, and that these decreases are associated with reductions in apnea severity. Methods Subject Selection Patients with obstructive apneas and hypopneas referred to the Johns Hopkins Sleep Disorder Center were recruited for this study. Eligible patients demonstrated a disordered breathing rate (DBR) of greater than 10 episodes/h in nonrapid eye movement (non494
SUMMARY Previous investigators have demonstrated in patients with obstructive sleep apnea that weight reduction results in a decrease in apnea severity. Although the mechanism for this decrease is not clear, we hypothesize that decreases In upper airway collapsibility account for decreases in apnea severity with weight loss. To determine whether weight loss causes decreases in collapsibility, we measured the upper airway critical pressure (Perit) before and after a 17.4 ± 3.4% (mean ± SO) reduction in body mass index in 13 patients with obstructive sleep apnea. Thirteen weightstable control subjects matched for age, body mass index, gender (all men), and non-REM disordered breathing rate (DBR) also were studied before and after usual care intervention. During nonREM sleep, maximal inspiratory airflow was measured by varying the level of nasal pressure and Pcrit was determined by the level of nasal pressure below which maximal inspiratory airflow ceased. In the weight loss group, a significant decrease in DBR from 83.3 ± 31.0 to 32.5 ± 35.9 episodeslh and in Perit from 3.1 ± 4.2 to - 2.4 ± 4.4 cm H2 0 (p < 0.00001) was demonstrated. Moreover,decreases in Perit were associated with nearly complete elimination of apnea in each patient whose Perit fell below -4 cm H20 . In contrast, no significant change in DBR and a minimal reduction in Pcrit from 5.2 ± 2.3 to 4.2 ± 1.8 cm H2 0 (p = 0.031) was observed in the "usual care" group. We conclude that (1) weight loss is associated with decreases in upper airway collapsibility In obstructive sleep apnea, and that (2) the resolution of sleep apnea depends on the absolute level to which Perit falls. AM REV RESPIR DIS 1991; 144:494-498
REM) sleep on an overnight screening sleep study. Patients with concomitant medical illness on screening history, phyical examination, or pulmonary function testing were excluded. Written informed consent was obtained from each patient for this study, which was approved by the Johns Hopkins Medical Institution Human Investigations Review Board.
Study Design To examine the effect of weight loss on apnea severity and upper airway collapsibility, all patients referred to the Johns Hopkins Sleep Disorder Center in whom obstructive sleep apnea was newly diagnosed were invited to join either a weight loss group or a "usual care" control group. The weight loss group was enrolled in a program of intensive dietary counseling and behavior modification designed to promote weight reduction, and an age-, weight-, and sex-matched usual care group received usual follow-up care from physicians at the sleep center. This usual care group did not receive any special weight loss counseling. Intervention Weight loss. The patients in the weight loss group attended weekly hour-long sessions with a dietician who offered dietary counseling, reviewed food logs, and charted body weight, as previously described (2). A goal
was set for each patient to lose approximately 15% of body weight, a target previously associated with significant reductions in apnea severity (2). Sleep studies were repeated in patients in the weight loss group when they achieved their goal weight. Those patients who did not achieve their goal weight were restudied at the termination of the weight loss program 25 months later if they lost at least 5OJo of their initial body weight. Patients who did not lose at least 5OJo of body weight were not restudied and were not included in the final weight loss group. All weight loss patients also received usual follow-up care as described below.
(Received in original form November 9, ]990 and in revised form April ]7, ]99]) 1 From the Division of Pulmonary and Critical Care Medicine, Department of Medicine, Johns Hopkins Asthma and Allergy Center at Francis Scott Key Medical Center, Baltimore, Maryland. 2 Supported by Grant HL-37379-02from the National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to Alan R. Schwartz, M.D., Division of Pulmonary/Critical Care Medicine, Johns Hopkins Asthma and Allergy Center, Hopkins Bayview Research Campus, 301 BayviewBoulevard, Baltimore, MD 21224. 4 Recipient of Clinical Investigator Award HL02031-01Al from the National Institutes of Health.
495
WEIGHT LOSS AND UPPER AIRWAY COLLAPSIBILITY IN OBSTRUCTIVE SLEEP APNEA
Of the 23 patients who initially enrolled in the weight loss program, 13 lost the minimum 5OJo of body weight required for inclusion in the weight loss group. Review of food logs from the patients who did not lose weight suggests that their failure to lose weight was due to dietary noncompliance. When those in the weight loss program who lost weight were compared with those who failed to lose weight, no significant differences were found in initial entry characteristics including sex (those who lost weight versus those who did not, all men versus eight men and one woman), body mass index (42.0 ± 7.1 versus 38.31 ± 4.3 kg/m', mean ± SD) or non-REM DBR (83.3 ± 31.0 versus 90.9 ± 46.7 episodes/h). Usual care. Patients in both the weight loss and the usual care groups received general advice regarding the importance of losing weight, avoiding alcohol and sedatives, and specific treatment recommendations, including nocturnal nasal continuous positive airway pressure (CPAP), from a staff physician at the Johns Hopkins Sleep Disorder Center. All patients in both groups received treatment with CPAP and returned for routine followup care at the Johns Hopkins Sleep Disorder Center every 3 to 6 months. The intervention period in the usual care group was matched to that in the weight loss group.
Experimental Techniques Polysomnography. Standard polysomnographic techniques (12) were employed to define the respiratory pattern during sleep, and they are outlined briefly here. Surface electroencephalographic electrodes placed at C 3 A 2 and C 3-O" a submental electrode, and left and right electrooculograms were used to stage sleep. An ear oximeter (Biox III; Bioximetry Technology, Boulder, CO) recorded oxyhemoglobin saturation (Sa02), and heart rate and rhythm were monitored. A polygraph (No. 780; Grass Instruments, Quincy, MA) ran continuously at 10 mm/s to simultaneously record all physiologic data throughout the night. Nocturnal sleep studies were performed between the hours of 11:00 P.M. and 8:00 A.M. Polysomnograms were scored for disordered breathing episodes (either obstructive apneas or hypopneas) and for changes in Saor The baseline and nadir Sa02 were measured for non-REM disordered breathing episodes to calculate the average baseline and average low Sao 2, as previously described (12).
a pneumotachograph, which was inserted in the breathing circuit. A thermistor was placed at the mouth to assure that oral breathing did not occur. All pressures, flows, and polysomnographic parameters were recorded continuously on the polygraph recorder. Patients were allowed to initiate sleep with nasal pressure (PN) maintained at approximately 1 em H 20 . After the initiation of stable non-REM sleep, PN was varied stepwise every 15 min over a range that included the PN at which maximal inspiratory airflow (VImax) became zero, Le., the upper airway completely occluded, as previously described (7-9).
Data Analysis Upper airway Peril. During spontaneous inspirations during non-REM sleep, VImax was measured at each level of PN. Pcrit was determined from the relationship between VImax and PN, and it was given by the level of PN below which VImax became zero, as previously described (7-9). In figure 1, this relationship is illustrated before and after weight loss in one patient who demonstrated a fall in the Pcrit from +0.4 to -6.1 em H 20 . Vtmax at atmospheric PN. The, effect of intervention on upper airway airflow was examined by determining Vrmax at atmospheric PN from the least squares linear regression of VImax versus PN. In the patient illustrated in figure 1, VImax at atmospheric PN was found to increase from zero to 118.3mlls during sleep after weight loss. Statistics. To determine whether Pcrit changed in the weight loss and usual care groups, paired two-tailed t tests were performed. The response of the weight loss group was compared with that of the usual care group using Student's two-tailed t test. A p < 0.05 significance level was accepted to confirm or reject the major hypothesis that weight loss decreased upper airway collapsibility. Least squares linear regression was used to examine the relationship between VImax and PN, and between Pcrit and body mass index (Minitab, Inc., State College, PA).
were matched initially for gender (all men), age, weight, and body mass index. Spirometry, lung volumes, diffusion capacity, arterial blood gases, and the prescribed level of nasal CPAP were also similar between these groups (table 1). After intervention, body mass index decreased in the weight loss group by 17.4 ± 3.4070 (mean ± SD) and increased minimally by 0.1 ± 0.3 % in the usual care group (table 2). Intervention was maintained for 16.9 ± 10 months in the weight loss group and for 18.4 ± 9.5 months in the usual care group.
Sleep-disordered Breathing Before intervention, no significant differences in non-REM DBR, baseline Sao2 or average low Sao, were found between the weight loss and usual care groups (table 2). Weight loss intervention resulted in a significant decrease in DBR without significant changes in baseline and average low Sao.. In contrast, DBR did not change, and baseline and average low Sao, increased significantly in the usual care group.
Upper Airway Collapsibility Before intervention, no significant difference in Pcrit was detected between the weight loss and the usual care groups (table 2). With intervention, Pcrit decreased from 3.1 ± 4.2 to - 2.4 ± 4.4 em H 20 (p < 0.00(01) in the weight loss group and from 5.2 ± 2.3 to 4.2 ± 1.8 em H 20 (p = 0.031) in the usual care group (table 2). The decrease in Pcrit was greater in the weight loss group than in the usual care group (p = 0.00(2). Moreover, the change in Pcrit in the weight loss group tended to correlate with the change in body mass index (p .= 0.056) (figure 2).
Results
Patient Characteristics
Relationship between Pcrit and Sleep-disordered Breathing
The weight loss and usual care groups
In the weight loss group, reductions in
400
Upper airway pressure-flow relationships. In a second overnight sleep study, upper airway pressure-flow relationships were recorded, as previously described (7-9). Briefly, each subject was studied in the supine position and breathed through a tightly fitting nasal mask that was connected to a variable pressure source at the inflow port and a variable resistor at the outflow port. Pressure in the nasal mask was measured with a catheter connected to a port in the mask. Esophageal pressure was measured with a latex balloon catheter, which was passed perinasally and positioned 10 em above the gastroesophageal junction. Nasal airflow was measured with
"I
Fig. 1. max versus PN is illustrated for one patient before and after a 24·kg weight loss. Pre-weight loss (solid line, solid circles) and post-weight loss (dashed line, open circles) pressure-flow relationships are shown. Perit is represented by the PN below which max became zero. In this patient, Perit decreased from 0.4 cm H2 0 before weight loss to -6.1 cm H 20 after weight loss.
300
1
V1max (mils)
200
"I
100
O..--........_..._.......-....--...--r-__- - r - _ -8
-6
-4
-2
o
2
PN (em
4 H20)
6
8
10
__ 12
14
496
SCHWARTZ, GOLD, SCHUBERT, STRYZAK, WISE, PERMUTT. AND SMITH
DBR and Pcrit was observed in the usual care group (bottom panel). After weight loss, DBR fell below 20 episodes/h in each patient in whom Pcrit fell below -4 ern H 20, and obstructive hypopneas were observed in the only one of these patients in whom some disordered breathing episodes remained. In six of the seven remaining weight loss patients (figure 4) and in every usual care patient, however, disordered breathing rates above 20 episodes/h were observed after intervention in association with levels of Pcrit above -4 em H 20. These findings suggest the presence of a minimally negative threshold (rv - 4 em H 20) below which decreases in Pcrit result in nearly complete resolution of sleep apnea.
TABLE 1 INITIAL PATIENT CHARACTERISTICS* Weight Loss Group
Usual Care Group
13:0 46.9 ± 8.9 129.1 ± 19.6 42.0 ± 7.1 84.0 ± 4.2 92.8 ± 16.4 65.3 ± 20.9 84.6 ± 19.4 99.1 ± 18.6 7.39 ± 0.03 40.9 ± 7.3 81.3 ± 12.9 12.5 ± 3.0
13:0 43.9 ± 10.5 121.7 ± 16.3 38.2 ± 5.0 79.2 ± 6.0 100.9 ± 13.6 73.8 ± 18.2 104.0 ± 27.2 107.3 ± 20.5 7.39 ± 0.03 43.7 ± 3.8 81.6 ± 9.6 13.7 ± 1.9
Sex, M:F Age, yr Weight, kg BMI, kg/m 2 FEV 1 , % TLC, % predicted FRC, % predicted RV, % predicted DLco, % predicted pH PAC02, mm Hgt Pao2, mm Hgt PN prescribed, cm H2 0
=
Definition of abbreviations: BMI • Values are mean ± SO. t Arterial partial pressures.
body mass index; PN
=
nasal pressure; RV
=
residual volume .
TABLE 2 EFFECT OF INTERVENTION
Before BMI, kg/m 2 Non-REM DBR, episodes/h Baseline SA0 2, % Average low SA0 2 , % Pcrit, cm H2O Vlmax, mlls+
42.0 83.3 92.5 84.0 3.1 13.5
After
7.1 31.0 3.9 11.8 4.2 25.6
± ± ± ± ± ±
34.7 32.5 93.9 89.0 -2.4 102.7
Definition of abbreviations: BMI = body mass index; OBR Vlmax = maximal inspiratory airflow; NS = not significant. • p = 0.0004 versus after usual care. t p = 0.0002 versus after usual care. t At atmospheric nasal pressure.
2
Change in Pcrf t (em H 2O)
-2
0 0
-4 -6
0
OJ 0
-12 -15
ALAN R. SCHWARTZ,4 AVRAM R. GOLD, NORMAN SCHUBERT, ALEXANDRA STRYZAK, ROBERT A. WISE, SOlBERT PERMUTT, and PHILIP L. SMITH
Introduction
Obstructive sleep apnea is a common disorder in middle-aged, obese men (1). In recent studies, we and others have demonstrated that significant improvements in obstructive sleep apnea accompany reductions in body weight (2-4), suggesting that obesity plays a major role in the pathogenesis of this disorder. Nevertheless, the mechanism for decreases in apnea severity with weight loss is unclear. It is clear from many studies that obstructive sleep apnea is caused by recurrent occlusion of the pharynx during sleep (5, 6). Additional work has shown that airway occlusion in these patients is due to the increase in upper airway collapsibility that occurs during sleep (7-11). Although the mechanism for the improvement in obstructive sleep apnea with weight loss has not been delineated, it is possible that reductions in apnea severity are due to decreases in upper airway collapsibility. In previous studies, differences in upper airway collapsibility have been related to alterations in the upper airway critical pressure (Pcrit), which has been defined by the nasal pressure (PN) below which the airway occludes (7-9). Specifically, these studies have demonstrated elevations in Pcrit in apneic patients and lower levels of Pcrit in association with greater degrees of upper airway patency. Therefore, the purpose of the present study was to test the hypothesis that weight loss in apneic patients results in decreases in upper airway Pcrit, and that these decreases are associated with reductions in apnea severity. Methods Subject Selection Patients with obstructive apneas and hypopneas referred to the Johns Hopkins Sleep Disorder Center were recruited for this study. Eligible patients demonstrated a disordered breathing rate (DBR) of greater than 10 episodes/h in nonrapid eye movement (non494
SUMMARY Previous investigators have demonstrated in patients with obstructive sleep apnea that weight reduction results in a decrease in apnea severity. Although the mechanism for this decrease is not clear, we hypothesize that decreases In upper airway collapsibility account for decreases in apnea severity with weight loss. To determine whether weight loss causes decreases in collapsibility, we measured the upper airway critical pressure (Perit) before and after a 17.4 ± 3.4% (mean ± SO) reduction in body mass index in 13 patients with obstructive sleep apnea. Thirteen weightstable control subjects matched for age, body mass index, gender (all men), and non-REM disordered breathing rate (DBR) also were studied before and after usual care intervention. During nonREM sleep, maximal inspiratory airflow was measured by varying the level of nasal pressure and Pcrit was determined by the level of nasal pressure below which maximal inspiratory airflow ceased. In the weight loss group, a significant decrease in DBR from 83.3 ± 31.0 to 32.5 ± 35.9 episodeslh and in Perit from 3.1 ± 4.2 to - 2.4 ± 4.4 cm H2 0 (p < 0.00001) was demonstrated. Moreover,decreases in Perit were associated with nearly complete elimination of apnea in each patient whose Perit fell below -4 cm H20 . In contrast, no significant change in DBR and a minimal reduction in Pcrit from 5.2 ± 2.3 to 4.2 ± 1.8 cm H2 0 (p = 0.031) was observed in the "usual care" group. We conclude that (1) weight loss is associated with decreases in upper airway collapsibility In obstructive sleep apnea, and that (2) the resolution of sleep apnea depends on the absolute level to which Perit falls. AM REV RESPIR DIS 1991; 144:494-498
REM) sleep on an overnight screening sleep study. Patients with concomitant medical illness on screening history, phyical examination, or pulmonary function testing were excluded. Written informed consent was obtained from each patient for this study, which was approved by the Johns Hopkins Medical Institution Human Investigations Review Board.
Study Design To examine the effect of weight loss on apnea severity and upper airway collapsibility, all patients referred to the Johns Hopkins Sleep Disorder Center in whom obstructive sleep apnea was newly diagnosed were invited to join either a weight loss group or a "usual care" control group. The weight loss group was enrolled in a program of intensive dietary counseling and behavior modification designed to promote weight reduction, and an age-, weight-, and sex-matched usual care group received usual follow-up care from physicians at the sleep center. This usual care group did not receive any special weight loss counseling. Intervention Weight loss. The patients in the weight loss group attended weekly hour-long sessions with a dietician who offered dietary counseling, reviewed food logs, and charted body weight, as previously described (2). A goal
was set for each patient to lose approximately 15% of body weight, a target previously associated with significant reductions in apnea severity (2). Sleep studies were repeated in patients in the weight loss group when they achieved their goal weight. Those patients who did not achieve their goal weight were restudied at the termination of the weight loss program 25 months later if they lost at least 5OJo of their initial body weight. Patients who did not lose at least 5OJo of body weight were not restudied and were not included in the final weight loss group. All weight loss patients also received usual follow-up care as described below.
(Received in original form November 9, ]990 and in revised form April ]7, ]99]) 1 From the Division of Pulmonary and Critical Care Medicine, Department of Medicine, Johns Hopkins Asthma and Allergy Center at Francis Scott Key Medical Center, Baltimore, Maryland. 2 Supported by Grant HL-37379-02from the National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to Alan R. Schwartz, M.D., Division of Pulmonary/Critical Care Medicine, Johns Hopkins Asthma and Allergy Center, Hopkins Bayview Research Campus, 301 BayviewBoulevard, Baltimore, MD 21224. 4 Recipient of Clinical Investigator Award HL02031-01Al from the National Institutes of Health.
495
WEIGHT LOSS AND UPPER AIRWAY COLLAPSIBILITY IN OBSTRUCTIVE SLEEP APNEA
Of the 23 patients who initially enrolled in the weight loss program, 13 lost the minimum 5OJo of body weight required for inclusion in the weight loss group. Review of food logs from the patients who did not lose weight suggests that their failure to lose weight was due to dietary noncompliance. When those in the weight loss program who lost weight were compared with those who failed to lose weight, no significant differences were found in initial entry characteristics including sex (those who lost weight versus those who did not, all men versus eight men and one woman), body mass index (42.0 ± 7.1 versus 38.31 ± 4.3 kg/m', mean ± SD) or non-REM DBR (83.3 ± 31.0 versus 90.9 ± 46.7 episodes/h). Usual care. Patients in both the weight loss and the usual care groups received general advice regarding the importance of losing weight, avoiding alcohol and sedatives, and specific treatment recommendations, including nocturnal nasal continuous positive airway pressure (CPAP), from a staff physician at the Johns Hopkins Sleep Disorder Center. All patients in both groups received treatment with CPAP and returned for routine followup care at the Johns Hopkins Sleep Disorder Center every 3 to 6 months. The intervention period in the usual care group was matched to that in the weight loss group.
Experimental Techniques Polysomnography. Standard polysomnographic techniques (12) were employed to define the respiratory pattern during sleep, and they are outlined briefly here. Surface electroencephalographic electrodes placed at C 3 A 2 and C 3-O" a submental electrode, and left and right electrooculograms were used to stage sleep. An ear oximeter (Biox III; Bioximetry Technology, Boulder, CO) recorded oxyhemoglobin saturation (Sa02), and heart rate and rhythm were monitored. A polygraph (No. 780; Grass Instruments, Quincy, MA) ran continuously at 10 mm/s to simultaneously record all physiologic data throughout the night. Nocturnal sleep studies were performed between the hours of 11:00 P.M. and 8:00 A.M. Polysomnograms were scored for disordered breathing episodes (either obstructive apneas or hypopneas) and for changes in Saor The baseline and nadir Sa02 were measured for non-REM disordered breathing episodes to calculate the average baseline and average low Sao 2, as previously described (12).
a pneumotachograph, which was inserted in the breathing circuit. A thermistor was placed at the mouth to assure that oral breathing did not occur. All pressures, flows, and polysomnographic parameters were recorded continuously on the polygraph recorder. Patients were allowed to initiate sleep with nasal pressure (PN) maintained at approximately 1 em H 20 . After the initiation of stable non-REM sleep, PN was varied stepwise every 15 min over a range that included the PN at which maximal inspiratory airflow (VImax) became zero, Le., the upper airway completely occluded, as previously described (7-9).
Data Analysis Upper airway Peril. During spontaneous inspirations during non-REM sleep, VImax was measured at each level of PN. Pcrit was determined from the relationship between VImax and PN, and it was given by the level of PN below which VImax became zero, as previously described (7-9). In figure 1, this relationship is illustrated before and after weight loss in one patient who demonstrated a fall in the Pcrit from +0.4 to -6.1 em H 20 . Vtmax at atmospheric PN. The, effect of intervention on upper airway airflow was examined by determining Vrmax at atmospheric PN from the least squares linear regression of VImax versus PN. In the patient illustrated in figure 1, VImax at atmospheric PN was found to increase from zero to 118.3mlls during sleep after weight loss. Statistics. To determine whether Pcrit changed in the weight loss and usual care groups, paired two-tailed t tests were performed. The response of the weight loss group was compared with that of the usual care group using Student's two-tailed t test. A p < 0.05 significance level was accepted to confirm or reject the major hypothesis that weight loss decreased upper airway collapsibility. Least squares linear regression was used to examine the relationship between VImax and PN, and between Pcrit and body mass index (Minitab, Inc., State College, PA).
were matched initially for gender (all men), age, weight, and body mass index. Spirometry, lung volumes, diffusion capacity, arterial blood gases, and the prescribed level of nasal CPAP were also similar between these groups (table 1). After intervention, body mass index decreased in the weight loss group by 17.4 ± 3.4070 (mean ± SD) and increased minimally by 0.1 ± 0.3 % in the usual care group (table 2). Intervention was maintained for 16.9 ± 10 months in the weight loss group and for 18.4 ± 9.5 months in the usual care group.
Sleep-disordered Breathing Before intervention, no significant differences in non-REM DBR, baseline Sao2 or average low Sao, were found between the weight loss and usual care groups (table 2). Weight loss intervention resulted in a significant decrease in DBR without significant changes in baseline and average low Sao.. In contrast, DBR did not change, and baseline and average low Sao, increased significantly in the usual care group.
Upper Airway Collapsibility Before intervention, no significant difference in Pcrit was detected between the weight loss and the usual care groups (table 2). With intervention, Pcrit decreased from 3.1 ± 4.2 to - 2.4 ± 4.4 em H 20 (p < 0.00(01) in the weight loss group and from 5.2 ± 2.3 to 4.2 ± 1.8 em H 20 (p = 0.031) in the usual care group (table 2). The decrease in Pcrit was greater in the weight loss group than in the usual care group (p = 0.00(2). Moreover, the change in Pcrit in the weight loss group tended to correlate with the change in body mass index (p .= 0.056) (figure 2).
Results
Patient Characteristics
Relationship between Pcrit and Sleep-disordered Breathing
The weight loss and usual care groups
In the weight loss group, reductions in
400
Upper airway pressure-flow relationships. In a second overnight sleep study, upper airway pressure-flow relationships were recorded, as previously described (7-9). Briefly, each subject was studied in the supine position and breathed through a tightly fitting nasal mask that was connected to a variable pressure source at the inflow port and a variable resistor at the outflow port. Pressure in the nasal mask was measured with a catheter connected to a port in the mask. Esophageal pressure was measured with a latex balloon catheter, which was passed perinasally and positioned 10 em above the gastroesophageal junction. Nasal airflow was measured with
"I
Fig. 1. max versus PN is illustrated for one patient before and after a 24·kg weight loss. Pre-weight loss (solid line, solid circles) and post-weight loss (dashed line, open circles) pressure-flow relationships are shown. Perit is represented by the PN below which max became zero. In this patient, Perit decreased from 0.4 cm H2 0 before weight loss to -6.1 cm H 20 after weight loss.
300
1
V1max (mils)
200
"I
100
O..--........_..._.......-....--...--r-__- - r - _ -8
-6
-4
-2
o
2
PN (em
4 H20)
6
8
10
__ 12
14
496
SCHWARTZ, GOLD, SCHUBERT, STRYZAK, WISE, PERMUTT. AND SMITH
DBR and Pcrit was observed in the usual care group (bottom panel). After weight loss, DBR fell below 20 episodes/h in each patient in whom Pcrit fell below -4 ern H 20, and obstructive hypopneas were observed in the only one of these patients in whom some disordered breathing episodes remained. In six of the seven remaining weight loss patients (figure 4) and in every usual care patient, however, disordered breathing rates above 20 episodes/h were observed after intervention in association with levels of Pcrit above -4 em H 20. These findings suggest the presence of a minimally negative threshold (rv - 4 em H 20) below which decreases in Pcrit result in nearly complete resolution of sleep apnea.
TABLE 1 INITIAL PATIENT CHARACTERISTICS* Weight Loss Group
Usual Care Group
13:0 46.9 ± 8.9 129.1 ± 19.6 42.0 ± 7.1 84.0 ± 4.2 92.8 ± 16.4 65.3 ± 20.9 84.6 ± 19.4 99.1 ± 18.6 7.39 ± 0.03 40.9 ± 7.3 81.3 ± 12.9 12.5 ± 3.0
13:0 43.9 ± 10.5 121.7 ± 16.3 38.2 ± 5.0 79.2 ± 6.0 100.9 ± 13.6 73.8 ± 18.2 104.0 ± 27.2 107.3 ± 20.5 7.39 ± 0.03 43.7 ± 3.8 81.6 ± 9.6 13.7 ± 1.9
Sex, M:F Age, yr Weight, kg BMI, kg/m 2 FEV 1 , % TLC, % predicted FRC, % predicted RV, % predicted DLco, % predicted pH PAC02, mm Hgt Pao2, mm Hgt PN prescribed, cm H2 0
=
Definition of abbreviations: BMI • Values are mean ± SO. t Arterial partial pressures.
body mass index; PN
=
nasal pressure; RV
=
residual volume .
TABLE 2 EFFECT OF INTERVENTION
Before BMI, kg/m 2 Non-REM DBR, episodes/h Baseline SA0 2, % Average low SA0 2 , % Pcrit, cm H2O Vlmax, mlls+
42.0 83.3 92.5 84.0 3.1 13.5
After
7.1 31.0 3.9 11.8 4.2 25.6
± ± ± ± ± ±
34.7 32.5 93.9 89.0 -2.4 102.7
Definition of abbreviations: BMI = body mass index; OBR Vlmax = maximal inspiratory airflow; NS = not significant. • p = 0.0004 versus after usual care. t p = 0.0002 versus after usual care. t At atmospheric nasal pressure.
2
Change in Pcrf t (em H 2O)
-2
0 0
-4 -6
0
OJ 0
-12 -15
