Abstract
Study Objectives:
Because air leakage from masks is known as a common cause of low adherence to continuous positive airway pressure therapy, we analyzed the risk factors for air leakage related to parameters associated with auto-titrating positive airway pressure, polysomnography, InBody Test, and rhinomanometry.
Methods:
Usage data and medical records of 120 auto-titrating positive airway pressure users were reviewed retrospectively. All patients used a nasal or pillow mask and were carefully monitored at scheduled follow-ups.
Results:
Use of a pillow mask, sex (male), age, and abdominal fat percentage were significantly associated with high average air leakage. The higher the auto-titrating positive airway pressure average and mean pressure, the more likely patients exhibited high rates of air leakage. The percentage of patients with high average air leakage increased over time (up to 6 months of follow-up).
Conclusions:
Older male patients using a pillow mask and those with a high abdominal fat percentage and high auto-titrating positive airway pressure may require close follow-up and continuous monitoring for air leakage. Because air leakage from a mask can change over time, mask-sealing capacity should be reassessed and masks should be changed regularly.
Citation:
Kang YJ, Cho J-H, Park C-S. Analysis of risk factors for air leakage in auto-titrating positive airway pressure users: a single-center study. J Clin Sleep Med. 2022;18(1):75–88.
Keywords: OSA, APAP, mask type, air leakage
BRIEF SUMMARY
Current Knowledge/Study Rationale: Many previous studies have attempted to determine the effect of masks or device type on air leakage and adherence. However, the effects of nasal cavity resistance and body composition on air leakage and other causes air leakage have yet to be fully evaluated.
Study Impact: This study showed that age, sex, mask type, obesity, and auto-titrating positive airway pressure may be related to high average air leakage and that average air leakage from a mask may change over time. Continuous monitoring for air leakage and scheduled changes of mask may be important in patients with risk factors.
INTRODUCTION
The prevalence of sleep apnea syndrome, as defined by an apnea-hypopnea index (AHI) score of 5 events/h or higher and daytime somnolence, is estimated at 4% in men and 2% in women and the prevalence of sleep-disordered breathing occurs in 4% of men and 9% of women, by the definition of AHI of 5 events/h or higher.1 Obstructive sleep apnea (OSA) is also associated with an increased risk of cardiovascular or cerebrovascular disease and reduced quality of life.2–4 In addition, high morbidity and mortality rates are associated with OSA.5
Since the positive airway pressure (PAP) device was introduced by Sullivan in Australia in 1981, continuous positive airway pressure (CPAP) has been widely used for OSA treatment and is recommended as first-line treatments for moderate to severe OSA.6
To date, 3 main devices have been developed for OSA: CPAP, bilevel PAP, and auto-titrating PAP (APAP). Pressure-relief modes such as expiratory pressure relief by ResMed and C-flex by Respironics have been developed to temporarily reduce pressure during exhalation and ease patient discomfort. APAP devices were developed to reduce mean CPAP pressure and pressure-related side effects of CPAP devices, which are controlled by an auto-adjusting pressure algorithm. Either a CPAP or APAP device is now recommended for ongoing treatment of OSA in adults.7,8
There was a meaningful study that compared 3 different methods of CPAP delivery: (1) autotitration, (2) autotitration for 1 week followed by fixed pressure (95th centile), and (3) fixed pressure by algorithm based on neck size and dip rate. Although average pressures were reportedly lowest in a 6-month autotitration group, no significant difference was reported to clinical outcomes among those 3 groups.9
Low adherence to PAP treatment is a challenge for many patients with OSA patients. Kohler et al10 reported that the severity of sleep-disordered breathing, age, and CPAP device pressure were associated with long-term adherence to CPAP.
Van Ryswyk et al11 reported that approximately one-third of patients using CPAP complain of nasal symptoms and dry mouth and that mask-related problems such as fitting and leakage are reported in 12% of cases at 3 months and 7% of cases at 24 months. Among the side effects of CPAP, the most frequently reported were unintentional air leakage and its related symptoms.12
To date, a large number of trials have been conducted to make patients comfortable during PAP therapy and to increase adherence to PAP therapy. Various mask types, primarily nasal pillow, nasal, and full face, may be associated with unintentional air leakage, patient discomfort, and variable adherence. Previous studies have demonstrated that nasal pillows were effective alternative interfaces that achieved adherence rates equal to those of nasal masks.13,14 However, oronasal masks were reportedly associated with a higher risk of CPAP nonadherence compared with nasal or nasal pillow masks.15
With regard to the relationship between the type of CPAP device and air leakage, Lebret et al16 showed that the type of CPAP device did not influence the amount of air leakage.
Given these findings, air leakage can be considered one of the most troublesome side effects of PAP therapy and a possible cause of low adherence, which is one of the weakest aspects of PAP therapy.
A number of studies have examined the factors that influence air leakage, but none have determined whether the effects on air leakage are temporary or continuous. In addition, the effects of nasal cavity resistance and body composition on air leakage have not been demonstrated in detail.
The objectives of this study were to identify risk factors for air leakage in APAP users with respect to patient demographics, polysomnography (PSG), rhinomanometry, and body composition data and to determine whether such effects change over time.
METHODS
Data collection
From January 2017 to May 2020, APAP usage data and medical records from 122 patients with mild to severe OSA were reviewed retrospectively. All patients completed the Berlin questionnaire and were evaluated using the Epworth Sleepiness Scale, an otolaryngology examination, InBody Test (InBody 770, InBody Co., Ltd., Seoul, Korea), rhinomanometry, and full-night in-lab PSG. Patients with severe respiratory disease, severe cardiovascular disease, and cognitive impairment were excluded, and patients who used APAP manufactured by companies other than Respironics before January 2017 were also excluded to maintain data uniformity. Of the 122 patients included in this study, 1 patient using a full-face mask and 1 patient with follow-up loss were excluded. Finally, data of 120 patients were included in the analysis. Follow-ups were scheduled at first (2 weeks after using APAP), second (1 month from first follow-up, from 2 weeks to 1.5 months after using APAP), third (3 months from second follow-up, from 1.5 to 4.5 months after using APAP), and fourth (3 months from third follow-up, from 4.5 to 10.5 months after using APAP). The number of patients at each follow-up were 120, 119, 100, and 60 respectively (Figure 1).
Figure 1. Study flowchart: changes of enrolled patients with low and high h-average air leakage over time.
HAL = high h-average air leakage, LAL = low h-average air leakage, OSA = obstructive sleep apnea.
All patients were provided with an APAP device and mask by the same manufacturer (Philips Respironics, Murrysville, PA)—either a nasal mask (DreamWear or Wisp Nasal mask; Philips Respironics) or a pillow mask (DreamWear Gel Nasal Pillow mask; Philips Respironics). When a doctor prescribed an APAP device, the patient was given masks and cushions of different sizes and allowed to choose a mask that fit best at hospital. If a patient complained of mask leakage or felt uncomfortable, a different size of mask or cushion with same interface initially supplied could be used within the first 2 weeks (at the first follow-up).
Although many mask manufacturers suggest replacing mask cushions once or twice a month, Korean national health care insurance only covers the costs of mask replacement once a year. In this study no mask cushions were replaced over the course of the 6 months.
Although APAP pressures ranged from 4 cm H2O to 16 cm H2O, low-span APAP settings with the window of pressure variation less than 5 cm H2O were used and the upper pressure of low-span APAP settings was less than 12 cm H2O in all but 6 patients. A previous paper used low-span APAP setting with pressure range from 8 cm H2O to 12 cm H2O, which meant the window of low-span APAP pressure was 4 cm H2O.17
All patients used heated humidification and a pressure-relief mode (A-flex 2 or 3), but 6 patients did not temporarily use heated humidification. When a patient visited the clinic, detailed data for the 7 days prior to the follow-up day were extracted from the APAP device with Philips EncorePro2 (version 2.23.03) software as follows: average device pressure under 90% of time (90% ADP; cm H2O), mean pressure (cm H2O), large air leakage (%time), AHI (events/h), rate of device use (greater than 70% or not, greater than 50% or not), average usage (all days), average usage (days used). H-average air leakage in this study was defined as the highest value of daily average air leakages in liters per minute (LPM), which was checked after reviewing daily average air leakages in between follow-ups.
The protocol for this study was approved by the Institutional Review Committee, St. Vincent Hospital, The Catholic University, Suwon, Republic of Korea (VC20RISI0194).
Rhinomanometry for nasal air flow and resistance
Our clinic used active anterior rhinomanometry, which is the most commonly used rhinomanometric tool because of its superior performance in measuring nasal cavity resistance and patency compared with acoustic rhinomanometry. A technician skilled in the use of a rhinorrheograph MRP-3100 (Nihon Kohden Co., Tokyo, Japan) performed all measurements. All patients were examined in a temperature-controlled and humidity-controlled room. One nasal cavity was blocked, and the opposite side was opened. The technician measured the low rate of the open nasal cavity, with a nasal adapter positioned in front of one nostril. Because resistance in the nasal cavity can change due to the nasal cycle, transnasal pressure was changed from 75 to 150 Pa to measure basal resistance. Nasal cavity resistance was measured by dividing transnasal pressure by natural airflow (V) at each nostril. The value of transnasal resistance (Pa/cm3/second) was calculated using Standardization Committee on Objective Assessment of the Nasal Airway recommendations.
Polysomnography
Full-night in-laboratory PSG was performed at the sleep center of St. Vincent Hospital, Suwon, Gyeonggi Province, Republic of Korea. During PSG tests, sensors used to record various biological signals were as follows: 6 electroencephalography (F3, F4, C3, C4, O1, O2); 2 electrooculography; 1 electrocardiogram lead; and 3 electromyography leads at the submentum, anterior tibial muscle, left and right sides; and a body-position sensor. A nasal pressure transducer, oral thermistor, thoracic and abdominal respiratory plethysmography belts, a pulse oximeter, and a snoring sensor were used to monitor respiratory events during sleep.
The full-night in-laboratory PSG data were recorded Embla N7000/S7000 hardware (Embla Systems, Inc., Broomfield, CO). All PSG data were manually scored by 2 sleep technicians and reviewed by a sleep specialist according to criteria of the American Academy of Sleep Medicine Scoring Manual version 2.4.18
Statistical analysis
Data are expressed as means ± standard deviation or standard error, including the interquartile range for continuous variables and the number or percentage for categorical variables. Baseline continuous or categorical variables were analyzed by independent t-tests or Wilcoxon ranking sum tests based on normality, chi-square tests, or Fisher’s exact tests. For longitudinal data, a generalized linear mixed model was used to estimate the association between maximal air leakage and independent variables. The generalized linear mixed model protocol uses a fixed effect for the independent variable and a random effect over time (2 weeks, 1 month, 3 months, or 6 months) and a binomial distribution with a logit link function model. We used a variance component structure for random effects. In addition, we analyzed the association of h-average air leakage with 90% ADP, mean pressure, device-estimated AHI, large leaks (%time) and average usage time (average usage [all days], average usage [days used]). Multivariate analysis was not performed to prevent overfitting according to the rule of experiencing at least 10 events per variable. Results were not corrected for multiple tests as each of the independent variables was considered to be of interest. A P value < 0.05 was considered statistically significant. All data analyses were performed using SAS 9.4 (SAS Institute, Inc., Cary, NC).
RESULTS
Data of 120 patients were reviewed retrospectively and the demographic and clinical characteristics of all patients are provided in Table 1.
Table 1.
Demographic and clinical characteristics of the 120 enrolled patients.
| All Patients (n = 120) | Patients with LAL (n = 103) | Patients with HAL (n = 17) | |
|---|---|---|---|
| Anthropometric data | |||
| Sex (male n, %) | 99 (82.5) | 82 (79.6) | 17 (100.0) |
| Age (years) | 51.9 ± 10.5 | 52.1 ± 10.3 | 51.0 ± 11.3 |
| Nasal stuffiness (yes, n %) | 38 (31.7) | 34 (33.0) | 4 (23.5) |
| Mouth breathing (yes, n %) | 64 (53.3) | 56 (54.4) | 8 (47.1) |
| Physical examination | |||
| Tonsil size (1–4) | 1.8 ± 0.8 | 1.87 ± 0.08 | 1.53 ± 0.17 |
| Palatal position (1–3) | 2.5 ± 0.7 | 2.52 ± 0.07 | 2.06 ± 0.18 |
| Muller maneuver (0–3) | 0.9 ± 0.8 | 0.91 ± 0.08 | 0.76 ± 0.14 |
| Comorbidities | |||
| Treated for hypertension (%) | 35.0 | 32.7 | 50.0 |
| Treated for diabetes (%) | 12.5 | 12.5 | 12.5 |
| Treated for dyslipidemia (%) | 8.3 | 5.8 | 25.0 |
| Treated for COPD or asthma (%) | 1.7 | 1.9 | 0.0 |
| Treated for cardiovascular disease (%) | 5.0 | 4.8 | 6.3 |
| Treated for cerebellar vascular disease (%) | 3.3 | 2.9 | 6.3 |
| Treated for other disease (Parkinson, thyroid) (%) | 9.2 | 7.7 | 6.3 |
| Body composition | |||
| BMI (kg/m2) | 28.1 ± 3.9 | 27.88 ± 0.38 | 29.35 ± 0.97 |
| Body fat (kg) | 24.6 ± 7.7 | 24.32 ± 0.76 | 26.43 ± 2.22 |
| Body fat percentage (%) | 29.8 ± 6.5 | 29.99 ± 0.66 | 28.78 ± 1.73 |
| Visceral fat area (cm2) | 109.7 ± 37.2 | 108.39 ± 3.69 | 117.57 ± 10.82 |
| Abdominal fat percentage (%) | 0.9 ± 0.1 | 0.92 ± 0.01 | 0.95 ± 0.02 |
| Neck circumference (cm) | 39.7 ± 4.0 | 39.41 ± 0.42 | 41.35 ± 0.78 |
| Waist circumference (cm) | 95.1 ± 13.0 | 95.22 ± 1.07 | 94.22 ± 5.90 |
| Hip circumference (cm) | 103.6 ± 7.8 | 103.47 ± 0.81 | 104.41 ± 1.77 |
| Neck-hip ratio (NHR) | 0.4 ± 0.0 | 0.38 ± 0.00 | 0.40 ± 0.01 |
| Waist-hip ratio (WHR) | 0.9 ± 0.1 | 0.92 ± 0.01 | 0.90 ± 0.05 |
| Severity of OSA at diagnosis | |||
| AHI (events/h) | 50.2 ± 23.5 | 48.46 ± 2.25 | 60.64 ± 6.19 |
| Obstructive type | 22.1 ± 22.2 | 20.24 ± 2.02 | 33.55 ± 6.99 |
| Hypopnea type | 23.8 ± 13.8 | 24.36 ± 1.36 | 20.24 ± 3.29 |
| REM AHI | 45.7 ± 23.8 | 44.51 ± 2.31 | 53.07 ± 6.16 |
| Supine AHI | 53.8 ± 40.2 | 52.49 ± 4.21 | 61.69 ± 5.03 |
| Minimal saturation (%) | 74.2 ± 16.0 | 75.25 ± 1.60 | 67.88 ± 3.21 |
| Periodic limb movements in sleep index | 9.7 ± 20.1 | 10.65 ± 2.09 | 4.02 ± 2.62 |
| Epworth Sleepiness Scale (0–24) | 10.6 ± 5.2 | 10.45 ± 0.52 | 11.47 ± 1.39 |
| Berlin questionnaire (0–22) | 7.7 ± 3.5 | 7.68 ± 0.38 | 7.59 ± 0.39 |
| Rhinomanometry | |||
| Total resistance less than 75 Pa (before constriction) | 0.14 ± 0.08 | 0.14 ± 0.01 | 0.12 ± 0.01 |
| Total resistance less than 150 Pa (before constriction) | 0.20 ± 0.12 | 0.21 ± 0.01 | 0.16 ± 0.02 |
| Nasal flow of dominant side less than 75 Pa (before constriction) | 284.8 ± 182.5 | 282.33 ± 18.52 | 299.87 ± 58.31 |
| Nasal flow of dominant side less than 150 Pa (before constriction) | 381.0 ± 212.6 | 371.24 ± 22.03 | 439.20 ± 59.91 |
| Nasal flow of dominant side less than 300 Pa (before constriction) | 360.2 ± 290.4 | 330.54 ± 29.80 | 538.40 ± 72.43 |
| Resistance of dominant side less than 75 Pa (before constriction) | 0.42 ± 0.66 | 0.45 ± 0.07 | 0.29 ± 0.05 |
| Resistance of dominant side less than 150 Pa (before constriction) | 0.50 ± 0.50 | 0.53 ± 0.06 | 0.36 ± 0.06 |
| Resistance of dominant side less than 300 Pa (before constriction) | 0.63 ± 0.81 | 0.65 ± 0.09 | 0.54 ± 0.12 |
| Nasal flow of dominant side less than 75 Pa (after constriction) | 352.6 ± 204.4 | 341.99 ± 21.86 | 414.57 ± 63.54 |
| Nasal flow of dominant side less than 150 Pa (after constriction) | 431.9 ± 242.8 | 417.85 ± 25.91 | 514.00 ± 75.28 |
| Nasal flow of dominant side less than 300 Pa (after constriction) | 374.3 ± 342.2 | 370.93 ± 35.82 | 393.71 ± 119.38 |
| Resistance of dominant side less than 75 Pa (after constriction) | 0.28 ± 0.28 | 0.29 ± 0.03 | 0.22 ± 0.02 |
| Resistance of dominant side less than 150 Pa (after constriction) | 0.36 ± 0.32 | 0.37 ± 0.04 | 0.29 ± 0.04 |
| Resistance of dominant side less than 300 Pa (after constriction) | 0.39 ± 0.42 | 0.42 ± 0.05 | 0.21 ± 0.07 |
Data are reported as median values or percentages. AHI = apnea-hypopnea index, BMI = body mass index, COPD = chronic obstruction pulmonary disease, HAL = high h-average air leakage, LAL = low h-average air leakage, REM = rapid eye movement.
Overall changes in patients with low and high average air leakage during APAP use
In this study, the threshold for division into low h-average air leakage (LAL) and high h-average air leakage (HAL) subgroups was defined as 40 LPM. Philips Respironics devices provide total leakage data (unintentional leakage plus intentional air leakage) and tolerate leakage of up to twice an intentional leak. Intentional air leakage varies by mask type and PAP pressure. In this study, given that the upper pressure of low-span APAP treatment ranged from 9 to 16 cmH2O, the calculated intentional air leakage ranged from approximately 26 to 35 LPM for both mask types (DreamWear/Wisp nasal and DreamWear pillow mask). The definition of a large leak threshold is total leakage of more than twice the intentional leak at each pressure for 1.5 min, which ranges from 52 to 70 LPM in this study. A large leak (%time) was defined as the percentage of time spent with the large leak. Philips Respironics machines tolerate up to 2 times the intentional leaks. If the patient has a perfect mask seal, the device will notice a total leak value that is equal to the intentional leak from the exhalation port. Any leaks between the cushion and the patient’s face will be detected by device and considered to be unintentional leaks.19,20
Because the mean of 90% ADP of all patients was 9.4 cm H2O, we divided patients into 2 subgroups, with a median value between intentional and large-leak thresholds of 40 LPM at this pressure. The proportion of patients with a HAL greater than 40 LPM was between 14.2 and 25% during 4 follow-ups. This proportion increased continuously when follow-up assessments were repeated (Figure 1).
Estimated risk factors on HAL
The characteristics of patients with LAL and HAL were compared at each follow-up (Table 2).
Table 2.
Characteristics of patients with high and low h-average air leakage at 4 follow-ups.
| h-Average Air Leakage (LPM) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| First APAP Follow-up (2 Weeks) | Second APAP Follow-up (1 Month) | Third APAP Follow-up (3 Months) | Fourth APAP Follow-up (6 Months) | |||||||||
| < 40 (n = 103) | ≥ 40 (n = 17) | P | < 40 (n = 97) | ≥ 40 (n = 22) | P | < 40 (n = 79) | ≥ 40 (n = 21) | P | <40 (n = 45) | ≥40 (n = 15) | P | |
| Mask type | .699 | .173 | .097 | >.999 | ||||||||
| Nasal (n (%)) | 90 (87.4) | 14 (82.4) | 86 (88.7) | 17 (77.3) | 69 (87.3) | 15 (71.4) | 34 (75.6) | 11 (73.3) | ||||
| Pillow (n (%)) | 13 (12.6) | 3 (17.6) | 11 (11.3) | 5 (22.7) | 10 (12.7) | 6 (28.6) | 11 (24.4) | 4 (26.7) | ||||
| Sex (male, n (%)) | 82 (79.6) | 17 (100.0) | .041 | 77 (79.4) | 21 (95.5) | .118 | 61 (77.2) | 20 (95.2) | .068 | 37 (82.2) | 13 (86.7) | >.999 |
| Age (years) | >.999 | .030 | .032 | >.999 | ||||||||
| <65 (n (%)) | 93 (90.3) | 16 (94.1) | 91 (93.8) | 17 (77.3) | 74 (93.7) | 16 (76.2) | 39 (86.7) | 13 (86.7) | ||||
| ≥65 (n (%)) | 10 (9.7) | 1 (5.9) | 6 (6.2) | 5 (22.7) | 5 (6.3) | 5 (23.8) | 6 (13.3) | 2 (13.3) | ||||
| Nasal stuffiness (yes, n (%)) | 34 (33.0) | 4 (23.5) | .436 | 28 (28.9) | 10 (45.5) | .132 | 25 (31.6) | 7 (33.3) | .883 | 11 (24.4) | 4 (26.7) | >.999 |
| Mouth breathing (yes, n (%)) | 56 (54.4) | 8 (47.1) | .576 | 49 (50.5) | 15 (68.2) | .134 | 41 (51.9) | 13 (61.9) | .414 | 24 (53.3) | 12 (80.0) | .068 |
| Rhinomanometry | ||||||||||||
| Total resistance less than 75 Pa (before constriction) | 0.14 ± 0.01 | 0.12 ± 0.01 | .591 | 0.14 ± 0.01 | 0.12 ± 0.01 | .909 | 0.14 ± 0.01 | 0.14 ± 0.02 | .966 | 0.13 ± 0.01 | 0.13 ± 0.02 | .879 |
| Total resistance less than 150 Pa (before constriction) | 0.21 ± 0.01 | 0.16 ± 0.02 | .149 | 0.21 ± 0.01 | 0.16 ± 0.02 | .210 | 0.21 ± 0.02 | 0.19 ± 0.02 | .861 | 0.19 ± 0.02 | 0.19 ± 0.02 | .571 |
| Nasal flow of dominant side less than 75 Pa (before constriction) | 282.33 ± 18.52 | 299.87 ± 58.31 | .993 | 279.47 ± 19.10 | 299.65 ± 50.13 | .765 | 263.91 ± 22.19 | 340.56 ± 49.06 | .103 | 320.05 ± 34.92 | 290.17 ± 50.75 | .939 |
| Nasal flow of dominant side less than 150 Pa (before constriction) | 371.24 ± 22.03 | 439.20 ± 59.91 | .220 | 369.01 ± 21.59 | 424.24 ± 62.93 | .401 | 337.76 ± 23.81 | 449.33 ± 54.47 | .039 | 378.30 ± 37.90 | 452.67 ± 62.79 | .232 |
| Nasal flow of dominant side less than 300 Pa (before constriction) | 330.54 ± 29.80 | 538.40 ± 72.43 | .011 | 348.43 ± 29.25 | 441.88 ± 88.12 | .293 | 330.18 ± 30.12 | 490.28 ± 80.88 | .029 | 308.53 ± 42.01 | 454.00 ± 112.06 | .218 |
| Resistance of dominant side less than 75 Pa (before constriction) | 0.45 ± 0.07 | 0.29 ± 0.05 | .178 | 0.44 ± 0.08 | 0.37 ± 0.08 | .439 | 0.49 ± 0.10 | 0.30 ± 0.06 | .049 | 0.53 ± 0.16 | 0.34 ± 0.07 | .704 |
| Resistance of dominant side less than 150 Pa (before constriction) | 0.53 ± 0.06 | 0.36 ± 0.06 | .142 | 0.50 ± 0.05 | 0.51 ± 0.12 | .367 | 0.57 ± 0.07 | 0.40 ± 0.08 | .048 | 0.55 ± 0.11 | 0.50 ± 0.13 | .728 |
| Resistance of dominant side less than 300 Pa (before constriction) | 0.65 ± 0.09 | 0.54 ± 0.12 | .778 | 0.62 ± 0.09 | 0.72 ± 0.21 | .713 | 0.73 ± 0.11 | 0.58 ± 0.15 | .276 | 0.74 ± 0.18 | 0.59 ± 0.21 | .575 |
| Nasal flow of dominant side less than 75 Pa (after constriction) | 341.99 ± 21.86 | 414.57 ± 63.54 | .329 | 339.84 ± 22.39 | 400.27 ± 56.25 | .364 | 317.08 ± 19.63 | 325.25 ± 37.09 | .911 | 320.23 ± 26.21 | 297.60 ± 39.28 | .652 |
| Nasal flow of dominant side less than 150 Pa (after constriction) | 417.85 ± 25.91 | 514.00 ± 75.28 | .299 | 409.45 ± 25.16 | 516.27 ± 74.77 | .196 | 393.16 ± 25.80 | 431.00 ± 52.92 | .706 | 400.17 ± 38.53 | 444.40 ± 63.04 | .816 |
| Nasal flow of dominant side less than 300 Pa (after constriction) | 370.93 ± 35.82 | 393.71 ± 119.38 | .949 | 375.40 ± 36.20 | 324.80 ± 106.52 | .486 | 352.19 ± 39.28 | 469.25 ± 85.56 | .236 | 346.40 ± 54.65 | 502.40 ± 117.23 | .260 |
| Resistance of dominant side less than 75 Pa (after constriction) | 0.29 ± 0.03 | 0.22 ± 0.02 | .477 | 0.29 ± 0.03 | 0.24 ± 0.03 | .507 | 0.31 ± 0.04 | 0.27 ± 0.03 | .757 | 0.32 ± 0.07 | 0.30 ± 0.05 | .461 |
| Resistance of dominant side less than 150 Pa (after constriction) | 0.37 ± 0.04 | 0.29 ± 0.04 | .467 | 0.37 ± 0.04 | 0.32 ± 0.05 | .374 | 0.39 ± 0.05 | 0.35 ± 0.05 | .965 | 0.40 ± 0.08 | 0.41 ± 0.06 | .420 |
| Resistance of dominant side less than 300 Pa (after constriction) | 0.42 ± 0.05 | 0.21 ± 0.07 | .068 | 0.42 ± 0.05 | 0.26 ± 0.10 | .083 | 0.42 ± 0.06 | 0.41 ± 0.08 | .848 | 0.44 ± 0.09 | 0.49 ± 0.13 | .616 |
| Physical examination | ||||||||||||
| Tonsil size (1–4) | 1.87 ± 0.08 | 1.53 ± 0.17 | .037 | 1.86 ± 0.09 | 1.68 ± 0.17 | .261 | 1.76 ± 0.07 | 1.62 ± 0.15 | .279 | 1.64 ± 0.10 | 1.53 ± 0.19 | .482 |
| Palatal position (1–3) | 2.52 ± 0.07 | 2.06 ± 0.18 | .011 | 2.52 ± 0.07 | 2.18 ± 0.17 | .061 | 2.52 ± 0.08 | 2.33 ± 0.16 | .252 | 2.47 ± 0.11 | 2.13 ± 0.22 | .149 |
| Muller maneuver (0–3) | 0.91 ± 0.08 | 0.76 ± 0.14 | .610 | 0.90 ± 0.08 | 0.86 ± 0.19 | .732 | 0.84 ± 0.09 | 0.90 ± 0.15 | .590 | 0.89 ± 0.10 | 0.73 ± 0.21 | .419 |
| Body composition | ||||||||||||
| BMI (kg/m2) | 27.88 ± 0.38 | 29.35 ± 0.97 | .075 | 27.59 ± 0.35 | 30.33 ± 1.04 | .016 | 27.44 ± 0.38 | 29.82 ± 0.90 | .020 | 27.55 ± 0.56 | 31.16 ± 1.02 | .005 |
| Body fat (kg) | 24.32 ± 0.76 | 26.43 ± 2.22 | .239 | 23.84 ± 0.75 | 28.19 ± 2.12 | .106 | 23.77 ± 0.82 | 26.87 ± 1.75 | .175 | 23.74 ± 1.15 | 30.14 ± 2.02 | .008 |
| Body fat percentage (%) | 29.99 ± 0.66 | 28.78 ± 1.73 | .865 | 29.75 ± 0.71 | 30.20 ± 1.32 | .637 | 29.83 ± 0.81 | 30.16 ± 1.21 | .581 | 29.59 ± 1.06 | 31.82 ± 1.73 | .153 |
| Visceral fat area (cm2) | 108.39 ± 3.69 | 117.57 ± 10.82 | .289 | 106.60 ± 3.72 | 123.70 ± 9.58 | .136 | 106.93 ± 4.13 | 118.43 ± 8.08 | .223 | 104.66 ± 5.56 | 137.91 ± 9.04 | .005 |
| Abdominal fat percentage (%) | 0.92 ± 0.01 | 0.95 ± 0.02 | .089 | 0.92 ± 0.01 | 0.94 ± 0.01 | .085 | 0.91 ± 0.01 | 0.93 ± 0.01 | .169 | 0.91 ± 0.01 | 0.96 ± 0.01 | .003 |
| Neck circumference (cm) | 39.41 ± 0.42 | 41.35 ± 0.78 | .029 | 39.21 ± 0.43 | 41.88 ± 0.65 | .002 | 39.11 ± 0.51 | 41.93 ± 0.59 | .001 | 39.59 ± 0.74 | 42.14 ± 0.99 | .046 |
| Waist circumference (cm) | 95.22 ± 1.07 | 94.22 ± 5.90 | .078 | 94.02 ± 1.27 | 99.97 ± 3.70 | .007 | 93.71 ± 1.16 | 96.83 ± 4.88 | .002 | 94.23 ± 1.70 | 95.46 ± 7.60 | .076 |
| Hip circumference (cm) | 103.47 ± 0.81 | 104.41 ± 1.77 | .539 | 102.78 ± 0.79 | 107.66 ± 1.80 | .010 | 102.74 ± 0.90 | 106.68 ± 1.31 | .015 | 104.37 ± 1.17 | 109.46 ± 2.00 | .052 |
| Neck-hip ratio (NHR) | 0.38 ± 0.00 | 0.40 ± 0.01 | .120 | 0.38 ± 0.00 | 0.39 ± 0.01 | .294 | 0.38 ± 0.00 | 0.39 ± 0.00 | .125 | 0.38 ± 0.01 | 0.38 ± 0.01 | .867 |
| Waist-hip ratio (WHR) | 0.92 ± 0.01 | 0.90 ± 0.05 | .039 | 0.92 ± 0.01 | 0.93 ± 0.03 | .047 | 0.91 ± 0.01 | 0.91 ± 0.04 | .023 | 0.90 ± 0.01 | 0.87 ± 0.07 | .419 |
| Severity of OSA at diagnosis | ||||||||||||
| AHI (events/h) | 48.46 ± 2.25 | 60.64 ± 6.19 | .056 | 48.82 ± 2.35 | 55.64 ± 5.39 | .266 | 48.32 ± 2.48 | 51.15 ± 5.07 | .688 | 44.36 ± 3.10 | 61.67 ± 5.34 | .016 |
| Obstructive type | 20.24 ± 2.02 | 33.55 ± 6.99 | .055 | 20.15 ± 2.16 | 29.65 ± 5.32 | .025 | 19.65 ± 2.25 | 24.09 ± 5.13 | .283 | 16.38 ± 2.49 | 32.66 ± 5.87 | .016 |
| Hypopnea type | 24.36 ± 1.36 | 20.24 ± 3.29 | .341 | 24.64 ± 1.40 | 20.37 ± 2.95 | .210 | 24.21 ± 1.52 | 22.21 ± 3.38 | .446 | 23.08 ± 1.84 | 23.19 ± 3.84 | .878 |
| REM AHI | 44.51 ± 2.31 | 53.07 ± 6.16 | .111 | 44.10 ± 2.36 | 53.01 ± 5.53 | .121 | 44.92 ± 2.42 | 42.46 ± 6.10 | .793 | 43.28 ± 3.05 | 54.73 ± 5.17 | .050 |
| Supine AHI | 52.49 ± 4.21 | 61.69 ± 5.03 | .040 | 53.30 ± 4.45 | 56.56 ± 4.43 | .215 | 53.08 ± 5.36 | 57.20 ± 3.80 | .039 | 45.26 ± 3.19 | 60.25 ± 3.97 | .005 |
| Minimal saturation (%) | 75.25 ± 1.60 | 67.88 ± 3.21 | .009 | 74.89 ± 1.74 | 71.27 ± 2.17 | .019 | 75.87 ± 1.78 | 71.57 ± 2.55 | .045 | 78.62 ± 1.32 | 70.00 ± 2.99 | .011 |
| Periodic limb movements in sleep index | 10.65 ± 2.09 | 4.02 ± 2.62 | .198 | 9.86 ± 2.13 | 8.69 ± 3.51 | .863 | 8.97 ± 2.23 | 10.18 ± 3.31 | .361 | 11.22 ± 3.03 | 5.27 ± 3.53 | .322 |
| Epworth Sleepiness Scale (0–24) | 10.45 ± 0.52 | 11.47 ± 1.39 | .653 | 10.27 ± 0.55 | 12.10 ± 1.05 | .124 | 9.89 ± 0.59 | 11.95 ± 1.12 | .115 | 10.44 ± 0.90 | 10.79 ± 1.46 | .860 |
| Berlin questionnaire (0–22) | 7.68 ± 0.38 | 7.59 ± 0.39 | .395 | 7.73 ± 0.39 | 7.57 ± 0.43 | .425 | 7.81 ± 0.43 | 7.62 ± 0.76 | .967 | 6.93 ± 0.29 | 7.50 ± 0.51 | .370 |
| Average usage (all days) (minutes) | 346.8 ± 10.4 | 320.2 ± 16.5 | .195 | 336.9 ± 6.3 | 297.9 ± 18.5 | .058 | 317.1 ± 6.9 | 274.3 ± 17.9 | .140 | 295.7 ± 11.8 | 309.0 ± 15.4 | .202 |
| Average usage (days used) (minutes) | 353.1 ± 7.9 | 343.5 ± 14.8 | .317 | 358.3 ± 5.3 | 318.2 ± 14.5 | .009 | 347.9 ± 5.6 | 326.1 ± 12.3 | .247 | 342.3 ± 8.7 | 336.3 ± 11.0 | .647 |
Data are reported as medians or percentages. AHI = apnea-hypopnea index, APAP = auto-titrating positive airway pressure, BMI = body mass index, LPM = liters per minute, OSA = obstructive sleep apnea, REM = rapid eye movement.
Although previous reports have demonstrated that nasal pillows are effective alternative interfaces and achieve adherence rates comparable to a nasal mask, data of 120 patients were further analyzed after division into 2 subgroups according to mask type to identify the relationship between mask type and air leakage not from the point of view of adherence.13,14
On the first APAP follow-up, significant differences between the LAL and HAL groups were found in sex, nasal flow of the dominant side less than 300 Pa, tonsil size, palatal position, neck circumference, waist/hip ratio, supine AHI, and minimal saturation. However, because the period of the first follow-up was only 2 weeks after APAP use and was thought to be the initial adaptation period to the APAP and mask, these data were considered less important than the second, third, and fourth follow-up points and were not used in further evaluation.
On the second and third APAP follow-ups, age, body mass index (BMI), neck/waist/hip circumference, neck-hip ratio, and minimal saturation were significantly different between LAL and HAL groups.
On the fourth APAP follow-up (more than 6 months of APAP use), BMI, body fat, visceral fat area (cm2), abdominal fat percentage, neck circumference, and some PSG data (AHI, supine AHI, minimal saturation) were significantly different between LAL and HAL groups. From the second to the fourth APAP follow-up (most of the APAP use period), BMI, neck circumference, and minimal saturation were significantly different between LAL and HAL groups (Table 2).
Nasal flow and resistance of the dominant side before nasal cavity constriction, tonsil size, and palatal position were significantly different between LAL and HAL groups only at the first or second follow-ups (Table 2), but the odds ratios (ORs) of nasal flow and resistance parameters were approximately 1.0 or not significant (Table 3).
Table 3.
The estimated risk factors for high h-average air leakage.
| h-Average Air Leakage (> 40, LPM) | |||
|---|---|---|---|
| Beta (SE) | Crude OR (95% CI) | P | |
| Pillow mask (%) | 0.672 (0.315) | 1.96 (1.05–3.66) | .035 |
| Sex (female, %) | −1.532 (0.533) | 0.22 (0.08–0.62) | .005 |
| Age ≥ 65 (%) | 0.836 (0.366) | 2.31 (1.12–4.76) | .024 |
| Nasal stuffiness (yes, %) | 0.142 (0.274) | 1.15 (0.67–1.98) | .604 |
| Mouth breathing (yes, %) | 0.477 (0.266) | 1.61 (0.95–2.73) | .076 |
| Rhinomanometry | |||
| Total resistance less than 75 Pa (before constriction) | −1.833 (1.907) | 0.16 (0.00–7.03) | .339 |
| Total resistance less than 150 Pa (before constriction) | −2.436 (1.395) | 0.09 (0.01–1.39) | .084 |
| Nasal flow of dominant side less than 75 Pa (before constriction) | 0.001 (0.001) | 1.00 (1.00–1.00) | .304 |
| Nasal flow of dominant side less than 150 Pa (before constriction) | 0.002 (0.001) | 1.00 (1.00–1.00) | .013 |
| Nasal flow of dominant side less than 300 Pa (before constriction) | 0.002 (0.001) | 1.00 (1.00–1.00) | .001 |
| Resistance of dominant side less than 75 Pa (before constriction) | −0.692 (0.517) | 0.50 (0.18–1.40) | .184 |
| Resistance of dominant side less than 150 Pa (before constriction) | −0.431 (0.358) | 0.65 (0.32–1.32) | .230 |
| Resistance of dominant side less than 300 Pa (before constriction) | −0.091 (0.179) | 0.91 (0.64–1.30) | .611 |
| Nasal flow of dominant side less than 75 Pa (after constriction) | 0.001 (0.001) | 1.00 (1.00–1.00) | .264 |
| Nasal flow of dominant side less than 150 Pa (after constriction) | 0.001 (0.001) | 1.00 (1.00–1.00) | .041 |
| Nasal flow of dominant side less than 300 Pa (after constriction) | 0.000 (0.000) | 1.00 (1.00–1.00) | .296 |
| Resistance of dominant side less than 75 Pa (after constriction) | −0.877 (0.895) | 0.42 (0.07–2.46) | .330 |
| Resistance of dominant side less than 150 Pa (after constriction) | −0.454 (0.571) | 0.63 (0.20–1.97) | .428 |
| Resistance of dominant side less than 300 Pa (after constriction) | −0.546 (0.401) | 0.58 (0.26–1.28) | .177 |
| Physical examination | |||
| Tonsil size (1–4) | −0.480 (0.218) | 0.62 (0.40–0.95) | .030 |
| Palatal position (1–3) | −0.593 (0.171) | 0.55 (0.39–0.77) | .001 |
| Muller maneuver (0–3) | −0.101 (0.169) | 0.90 (0.65–1.26) | .551 |
| Body composition | |||
| BMI (kg/m2) | 0.162 (0.034) | 1.18 (1.10–1.26) | < .001 |
| Body fat (kg) | 0.062 (0.017) | 1.06 (1.03–1.10) | < .001 |
| Body fat percentage (%) | 0.008 (0.020) | 1.01 (0.97–1.05) | .679 |
| Visceral fat area (cm2) | 0.012 (0.004) | 1.01 (1.00–1.02) | .001 |
| Abdominal fat percentage (%) | 10.590 (2.590) | 2.88 (1.73–4.82) | < .001 |
| Neck circumference (cm) | 0.217 (0.045) | 1.24 (1.14–1.36) | < .001 |
| Waist circumference (cm) | 0.016 (0.012) | 1.02 (0.99–1.04) | .187 |
| Hip circumference (cm) | 0.059 (0.017) | 1.06 (1.03–1.10) | .001 |
| Neck-hip ratio (NHR) | 13.707 (5.629) | 1.15 (1.03–1.28) | .017 |
| Waist-hip ratio (WHR) | −0.787 (1.103) | 0.92 (0.74–1.15) | .477 |
| Severity of OSA at diagnosis | |||
| AHI (events/h) | 0.017 (0.006) | 1.02 (1.01–1.03) | .004 |
| Obstructive type | 0.020 (0.006) | 1.02 (1.01–1.03) | .001 |
| Hypopnea type | −0.016 (0.010) | 0.98 (0.96–1.00) | .114 |
| REM AHI | −0.020 (0.007) | 0.98 (0.97–0.99) | .008 |
| Supine AHI | −0.008 (0.007) | 0.99 (0.98–1.01) | .285 |
| Minimal saturation (%) | 0.012 (0.006) | 1.01 (1.00–1.02) | .044 |
| Periodic limb movements in sleep index | 0.004 (0.003) | 1.00 (1.00–1.00) | .194 |
| Epworth Sleepiness Scale (0–24) | 0.050 (0.024) | 1.05 (1.00–1.10) | .045 |
| Berlin questionnaire (0–22) | −0.004 (0.039) | 1.00 (0.92 BMI: body mass index, 1.08) | .913 |
AHI = apnea-hypopnea index, BMI = body mass index, CI = confidence interval, LPM = liters per minute, OR = odds ratio, REM = rapid eye movement.
A few PSG parameters such as supine AHI and minimal saturation were significantly different between LAL and HAL groups (Table 2) but the estimated ORs of all PSG parameters for HAL were small (approximately 1.00) (Table 3).
Although tonsil size and palatal position were significantly different only at the first follow-up, their ORs were 0.62 (confidence interval [CI] 0.40–0.95) and 0.55 (CI 0.39–0.77), respectively, which were contrary to our expectation (Table 2 and Table 3).
As shown in Table 3, pillow mask (OR 1.96, 95% CI 1.05–3.66), male sex, and age (more than 65 years, OR 2.31, 95% CI 1.12–4.76) were closely related to HAL with large ORs. In particular, abdominal fat percentage (OR 2.88) was more closely associated with HAL than with BMI (OR 1.18), body fat (OR 1.06), visceral fat area (OR 1.01), hip circumference (OR 1.06), and neck-hip ratio (OR 1.15).
For the average usage (all days) and average usage (days used), there was no significant difference between HAL and LAL groups except for the second follow-up. Only at the second follow-up, the average usage time of the LAL group was significantly longer than that of the HAL group (Table 2).
Association of APAP pressure and device-estimated AHI with risk of high average air leakage
In each subgroup, LAL and HAL, 90% ADP, mean pressure, and device-estimate AHI among second, third, and fourth follow-ups showed a decreasing trend (negative beta values) but were not statistically significant (Table 4). However, large leaks (%time) significantly increased in LAL but not in HAL groups (Table 4).
Table 4.
Device-estimated pressure, AHI, and large leakage in HAL and LAL groups.
| LAL (h-Average Air Leakage < 40 LPM) | HAL (h-Average Air Leakage ≥ 40 LPM) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Second APAP Follow-up (n = 97) | Third APAP Follow-up (n = 79) | Fourth APAP Follow-up (n = 45) | Beta (SE) | P | Second APAP Follow-up (n = 22) | Third APAP Follow-up (n = 21) | Fourth APAP Follow-up (n = 15) | Beta (SE) | P | |
| 90% Average device pressure | 9.1 ± 0.1 | 9.1 ± 0.2 | 9.0 ± 0.2 | −0.08 (0.12) | .513 | 10.7 ± 0.4 | 10.4 ± 0.4 | 10.4 ± 0.4 | −0.12 (0.28) | .664 |
| Mean pressure | 7.9 ± 0.1 | 7.8 ± 0.2 | 7.7 ± 0.2 | −0.08 (0.12) | .497 | 9.5 ± 0.4 | 9.1 ± 0.4 | 9.2 ± 0.4 | −0.19 (0.29) | .503 |
| AHI | 3.4 ± 0.2 | 3.4 ± 0.3 | 3.2 ± 0.2 | −0.11 (0.17) | .535 | 6.7 ± 1.3 | 5.8 ± 0.9 | 3.9 ± 0.3 | −1.38 (0.74) | .066 |
| Large leakage (%time) | 0.4 ± 0.1 | 0.4 ± 0.1 | 0.9 ± 0.3 | 0.26 (0.09) | .002 | 3.0 ± 0.6 | 3.3 ± 0.7 | 3.3 ± 0.7 | 0.15 (0.50) | .764 |
| Average usage (all days) (minutes) | 336.9 ± 6.3 | 317.1 ± 6.9 | 295.7 ± 11.8 | −0.005 (0.002) | .004 | 297.9 ± 18.5 | 274.3 ± 17.9 | 309.0 ± 15.4 | −0.006 (0.002) | .007 |
| Average usage (days used) (minutes) | 358.3 ± 5.3 | 347.9 ± 5.6 | 342.3 ± 8.7 | −0.006 (0.002) | .007 | 318.2 ± 14.5 | 326.1 ± 12.3 | 336.3 ± 11.0 | −0.009 (0.003) | .003 |
AHI = apnea-hypopnea index, APAP = auto-titrating positive airway pressure, HAL = high h-average air leakage group, LAL = low h-average air leakage group, LPM = liters per minute.
The average usage (all days) and average usage (days used) significantly decreased in LAL group but average usage for used days significantly increased in HAL group (Table 4). In addition, our results showed that AHI (OR 1.38, 95% CI 1.17–1.62, P < .001), 90% ADP (OR 1.93, 95% CI 1.52–2.45, P < 0.001), and mean pressure (OR 1.89, 95% CI 1.52–2.35, P < 0.001), and large leaks (OR 2.3, CI 1.66–3.20, P < 0.001) were risk factors for HAL (Table 5).
Table 5.
Association of device-estimated APAP parameters with risk of high h-average air leakage.
| h-Average Air Leakage (LPM) | Crude OR (95% CI) | P | ||||||
|---|---|---|---|---|---|---|---|---|
| Second APAP Follow-up (1 Month) | Third APAP Follow-up (3 Months) | Fourth APAP Follow-up (6 Months) | ||||||
| < 40 (n = 97) | ≥ 40 (n = 22) | < 40 (n = 79) | ≥ 40 (n = 21) | < 40 (n = 45) | ≥ 40 (n = 15) | |||
| 90% Average device pressure | 9.1 ± 0.1 | 10.7 ± 0.4 | 9.1 ± 0.2 | 10.4 ± 0.4 | 9.0 ± 0.2 | 10.4 ± 0.4 | 1.93 (1.52–2.45) |
<.001 |
| Mean pressure | 7.9 ± 0.1 | 9.5 ± 0.4 | 7.8 ± 0.2 | 9.1 ± 0.4 | 7.7 ± 0.2 | 9.2 ± 0.4 | 1.89 (1.52–2.35) |
< .001 |
| AHI | 3.4 ± 0.2 | 6.7 ± 1.3 | 3.4 ± 0.3 | 5.8 ± 0.9 | 3.2 ± 0.2 | 3.9 ± 0.3 | 1.38 (1.17–1.62) |
< .001 |
| Large leak (%time) | 0.4 ± 0.1 | 3.0 ± 0.6 | 0.4 ± 0.1 | 3.3 ± 0.7 | 0.9 ± 0.3 | 3.3 ± 0.7 | 2.30 (1.66–3.20) |
< .001 |
AHI = apnea-hypopnea index, APAP = auto-titrating positive airway pressure, CI = confidence interval, LPM = liters per minute, OR = odds ratio.
DISCUSSION
CPAP is recommended as a first-line treatment for moderate to severe OSA and low adherence remains a major problem in long-term PAP treatments.7,8
CPAP adherence is influenced by a variety of factors, such as knowledge of OSA/perception of CPAP treatment, mask interfaces, drying of upper airway mucosa, claustrophobia, early follow-up within the first week after CPAP treatment.21
The appropriate selection of mask interfaces, typically a nasal, nasal pillow, and oronasal mask, is thought to be the first step in effective CPAP adherence. Several studies of the relationship between mask type and CPAP adherence have been conducted. Borel et al15 used multivariate analysis to show that oronasal masks were associated with a higher risk of CPAP nonadherence compared with nasal or nasal pillow masks.
However, it was also reported that CPAP adherence did not differ between the 3 different mask interfaces (ie, nasal mask with or without chin strap and an oronasal mask) but fewer air leaks and mask-fitting problems were associated with nasal mask compared with nasal mask chin-straps and oronasal masks.22
Previous studies, including the 2 cited above, showed that nasal masks achieve superior adherence compared with pillow-type and oronasal-type masks.23–26
Air leakage and its side effects are prevalent in long-term CPAP-treated patients and Valentin et al reported that greater air leakage was associated with poor adherence to APAP therapy.27,28 Therefore, among a variety of factors affecting CPAP adherence, considerable attention has been paid to air leakage and mask-related factors.
Although Lebret et al12 reported that nasal obstruction, higher BMI, older age, and male sex are related to air leakage, our results suggest additional risk factors, such as pillow mask, male sex, and older age, are related to air leakage.
The pillow mask is well-tolerated, an effective interface, and smaller than the nasal-type and oronasal-type masks, which explains why it is frequently chosen by patients. However, our results showed that use of pillow masks may be a risk factor for air leakage and a negative factor for PAP adherence.
Aging is a well-known risk factor that can increase upper airway collapse due to low neuromuscular function, a relationship that was confirmed by our results.29
The unintentional leaks during APAP treatment were reported to be related to mouth opening, CPAP level, sleep position, and rapid eye movement sleep, and oronasal masks can reduce the risk of unintentional air leakage in the case of mouth breathing and rapid eye movement sleep.30
Our results showed that higher APAP pressure (90th percentile, mean pressure) may be related to HAL. Patients with a high 90% ADP or mean pressure should therefore be closely monitored for air leakage regardless of initial air leakage values.
A study with patients with extreme obesity (a median BMI of at least 43 kg/m2) reported that BMI, neck circumference, and waist circumference are not predictive of APAP pressure.31
In our study, only abdominal fat percentage was more closely associated with air leakage than with BMI, body fat, visceral fat area, hip circumference, and neck/hip ratio. This suggests that regional obesity be more closely related to elevated abdominal pressure and collapsibility of the upper airway and therefore to air leakage than to BMI.
Even after the selection of an appropriate mask type, multiple mask-related problems can be expected, including air leakage, stuffy or dry nose, tube problems, and claustrophobia, among others. Dry or stuffy nose can often be reduced with heated humidification, and claustrophobia can be reduced by desensitization to the mask, change of mask type, and sleep-onset detection.32
Contrary to our expectation that higher nasal resistance increases air leakage, rhinomanometry results failed to demonstrate that nasal resistance (obstruction) was related to air leakage, which may be attributable to the lack of a significant difference between LAL and HAL groups in total nasal resistance at all follow-ups.
However, previous studies have shown that surgery to decrease nasal resistance is associated with improved PAP compliance, lower PAP pressure, and improved quality of life in patients with OSA.33–35 A few studies have also reported that nasal airway resistance accounts for up to 60% of total airway resistance, and total nasal resistance does not change significantly between wakefulness and sleep.36,37 Further study of nasal resistance and air leakage is needed.
The relationship between air leakage and CPAP type may be an issue worth exploring. The type of PAP (fixed or auto) was reportedly not associated with a risk of leaks, but CPAP pressure level (mean) and oronasal masks were associated with a risk of leaks.16
Another potential issue involves the effect of air leakage on the function and capacity of PAP devices. Although an automatic analysis from APAP was reported to accurately detect residual sleep apnea, Coller et al reported that, in the presence of air leakage, APAP underestimated the pressure required to treat OSA and overestimated the pressure delivered at the upper airway.38,39 Baek et al40 found that higher air leakage was an independent risk factor for greater differences between auto-AHI and manual-AHI scoring. In summary of the abovementioned 2 papers, data estimated by PAP device may not be accurate in the presence of a large air leak. Therefore, effective PAP therapy may also not be possible in the presence of large air leakage.
In our study, the average usage (all days) and average usage (days used) included the time with large air leakage. The proportion of time with large leak was relatively small and may not have affected our results. However, in the presence of large air leak, the device may not properly sense air flow or events or respond or deliver pressures as intended by the auto-algorithms.
Generally speaking, the longer the duration of APAP use, the better patients adapt to APAP; however, our study showed that the longer the duration of APAP use, the higher the h-average air leakage, which may indicate why PAP compliance decreases over time in initially well-adapted APAP users (Figure 1).
According to a study by Valentin et al28, air leakage may be associated with poor adherence to APAP therapy and the proportion of time spent at large leak levels may be associated with nonadherence. Our results showed that a large leak (%time) was a risk factor for air leakage and that it increased even in the LAL group over time (Table 4 and Table 5). These results may be a product of weakened mask-sealing capacity because the silicone portion of the mask tends to stiffen over time, suggesting that the mask should be replaced before 6 months.
Several limitations of this study should be noted. First, each APAP device uses its own algorithm for air leakage, pressure change, and AHI change, which means our results may have been different if we used APAP devices from other manufacturers.41
Second, other conditions that affect air leakage, such as upper airway diseases, were not considered.
Third, further research with a larger population should be conducted to assess the effects of other parameters such as mask type and manufacturers on air leakage and the associations between air leakage and PAP treatment effectiveness.
Fourth, each patient’s orofacial anatomy and contours can affect mask fit and sealing effects differently; further evaluation of the relationship between orofacial anatomy and mask leakage is therefore needed.
Fifth, the results of age and nasal masks should be interpreted with caution, as our study did not have the necessary power to robustly evaluate those variables.
Finally, the use of single-type cushions could be the limitation of this study, because, if the previously used cushion had some problems such as air leak, it was replaced with extra cushion of same type but different size within 1 month. It might be possible that different type cushions reduced air leakage.
CONCLUSIONS
Our study showed that air leakage increased over time, particularly by 6 months after APAP treatment, which may be due to reduced sealing effects of the masks, suggesting a need for regular mask management. Among the factors related to air leakage, the most consequential appear to be age, sex, mask type, regional obesity, and PAP pressure. Close follow-up and continuous monitoring are recommended for such patients.
ACKNOWLEDGMENTS
Statistical consultation and analysis were performed by staff at the Department of Biostatistics of the Catholic Research Coordinating Center.
ABBREVIATIONS
- AHI,
apnea-hypopnea index
- APAP,
auto-titrating positive airway pressure
- BMI,
body mass index
- CI,
confidence interval
- CPAP,
continuous positive airway pressure
- HAL,
high h-average air leakage
- LAL,
low h-average air leakage
- LPM,
liters per minute
- OR,
odds ratio
- OSA,
obstructive sleep apnea
- PAP,
positive airway pressure
- PSG,
polysomnography
- 90% ADP,
average device pressure under 90% of time
DISCLOSURE STATEMENT
All authors have seen and approved this manuscript. Work for this study was performed at the College of Medicine, The Catholic University of Korea. The authors report no conflicts of interest.
REFERENCES
- 1. Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med. 1993; 328( 17): 1230– 1235. [DOI] [PubMed] [Google Scholar]
- 2. Marin JM, Carrizo SJ, Vicente E, Agusti AG. Long-term cardiovascular outcomes in men with obstructive sleep apnoea-hypopnoea with or without treatment with continuous positive airway pressure: an observational study. Lancet. 2005; 365 ( 9464): 1046– 1053. [DOI] [PubMed] [Google Scholar]
- 3. Johnson KG, Johnson DC. Frequency of sleep apnea in stroke and TIA patients: a meta-analysis. J Clin Sleep Med. 2010; 6( 2): 131– 137. [PMC free article] [PubMed] [Google Scholar]
- 4. D’Ambrosio C, Bowman T, Mohsenin V. Quality of life in patients with obstructive sleep apnea: effect of nasal continuous positive airway pressure—a prospective study. Chest. 1999; 115( 1): 123– 129. [DOI] [PubMed] [Google Scholar]
- 5. Marshall NS, Wong KK, Cullen SR, Knuiman MW, Grunstein RR. Sleep apnea and 20-year follow-up for all-cause mortality, stroke, and cancer incidence and mortality in the Busselton Health Study cohort. J Clin Sleep Med. 2014; 10( 4): 355– 362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Sullivan CE, Issa FG, Berthon-Jones M, Eves L. Reversal of obstructive sleep apnoea by continuous positive airway pressure applied through the nares. Lancet. 1981; 1( 8225): 862– 865. [DOI] [PubMed] [Google Scholar]
- 7. Kushida CA, Littner MR, Hirshkowitz M, et al. ; American Academy of Sleep Medicine. Practice parameters for the use of continuous and bilevel positive airway pressure devices to treat adult patients with sleep-related breathing disorders. Sleep. 2006; 29( 3): 375– 380. [DOI] [PubMed] [Google Scholar]
- 8. Patil SP, Ayappa IA, Caples SM, Kimoff RJ, Patel SR, Harrod CG. Treatment of adult obstructive sleep apnea with positive airway pressure: An American Academy of Sleep Medicine Clinical Practice Guideline. J Clin Sleep Med. 2019; 15( 2): 335– 343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. West SD, Jones DR, Stradling JR. Comparison of three ways to determine and deliver pressure during nasal CPAP therapy for obstructive sleep apnoea. Thorax. 2006; 61( 3): 226– 231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kohler M, Smith D, Tippett V, Stradling JR. Predictors of long-term compliance with continuous positive airway pressure. Thorax. 2010; 65( 9): 829– 832. [DOI] [PubMed] [Google Scholar]
- 11. Van Ryswyk E, Anderson CS, Antic NA, et al. Predictors of long-term adherence to continuous positive airway pressure in patients with obstructive sleep apnea and cardiovascular disease. Sleep. 2019; 42( 10): zsz152. [DOI] [PubMed] [Google Scholar]
- 12. Lebret M, Martinot JB, Arnol N, et al. Factors contributing to unintentional leak during CPAP treatment: A systematic review. Chest. 2017; 151( 3): 707– 719. [DOI] [PubMed] [Google Scholar]
- 13. Ryan S, Garvey JF, Swan V, Behan R, McNicholas WT. Nasal pillows as an alternative interface in patients with obstructive sleep apnoea syndrome initiating continuous positive airway pressure therapy. J Sleep Res. 2011; 20( 2): 367– 373. [DOI] [PubMed] [Google Scholar]
- 14. Lanza A, Mariani S, Sommariva M, et al. Continuous positive airway pressure treatment with nasal pillows in obstructive sleep apnea: long-term effectiveness and adherence. Sleep Med. 2018; 41: 94– 99. [DOI] [PubMed] [Google Scholar]
- 15. Borel JC, Tamisier R, Dias-Domingos S, et al. ; Scientific Council of The Sleep Registry of the French Federation of Pneumology (OSFP). Type of mask may impact on continuous positive airway pressure adherence in apneic patients. PLoS One. 2013; 8( 5): e64382. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Lebret M, Rotty MC, Argento C, et al. Comparison of auto- and fixed-continuous positive airway pressure on air leak in patients with obstructive sleep apnea: Data from a randomized controlled trial. Can Respir J. 2019; 2019: 6310956. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Bastos HN, Cardoso AV, Castro AS, et al. Randomised short-term trial of high-span versus low-span APAP for treating sleep apnoea. Sleep Breath. 2016; 20( 1): 183– 190, discussion 190. [DOI] [PubMed] [Google Scholar]
- 18. Berry RB, Brooks R, Gamaldo CE, et al. ; for the American Academy of Sleep Medicine. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. Version 2.4. Darien, IL: American Academy of Sleep Medicine; 2017. [Google Scholar]
- 19.Philips Respironics. Interpretation guide for Encore software compliance reports. 2014.
- 20.Philips Respironics. Clinical applications guide, REMstar Auto. 2009. https://www.usa.philips.com/c-dam/b2bhc/us/whitepapers/sleep-therapy/REMstarAuto_ClinAppGuide%20FINAL.pdf
- 21. Weaver TE, Grunstein RR. Adherence to continuous positive airway pressure therapy: the challenge to effective treatment. Proc Am Thorac Soc. 2008; 5( 2): 173– 178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Rowland S, Aiyappan V, Hennessy C, et al. Comparing the efficacy, mask leak, patient adherence, and patient preference of three different CPAP interfaces to treat moderate-severe obstructive sleep apnea. J Clin Sleep Med. 2018; 14( 1): 101– 108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Goh KJ, Soh RY, Leow LC, et al. Choosing the right mask for your Asian patient with sleep apnoea: A randomized, crossover trial of CPAP interfaces. Respirology. 2019; 24( 3): 278– 285. [DOI] [PubMed] [Google Scholar]
- 24. Teo M, Amis T, Lee S, Falland K, Lambert S, Wheatley J. Equivalence of nasal and oronasal masks during initial CPAP titration for obstructive sleep apnea syndrome. Sleep. 2011; 34( 7): 951– 955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Chai CL, Pathinathan A, Smith B. Continuous positive airway pressure delivery interfaces for obstructive sleep apnoea. Cochrane Database Syst Rev. 2006: CD005308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Blanco M, Ernst G, Salvado A, Borsini E. Impact of mask type on the effectiveness of and adherence to unattended home-based CPAP titration. Sleep Disord. 2019; 2019: 4592462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Rotty MC, Suehs CM, Mallet JP, et al. Mask side-effects in long-term CPAP-patients impact adherence and sleepiness: the InterfaceVent real-life study. Respir Res. 2021; 22( 1): 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Valentin A, Subramanian S, Quan SF, Berry RB, Parthasarathy S. Air leak is associated with poor adherence to autoPAP therapy. Sleep. 2011; 34( 6): 801– 806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Eikermann M, Jordan AS, Chamberlin NL, et al. The influence of aging on pharyngeal collapsibility during sleep. Chest. 2007; 131( 6): 1702– 1709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Lebret M, Arnol N, Martinot JB, et al. Determinants of unintentional leaks during CPAP treatment in OSA. Chest. 2018; 153( 4): 834– 842. [DOI] [PubMed] [Google Scholar]
- 31. Turnbull CD, Manuel AR, Stradling JR. Does either obesity or OSA severity influence the response of autotitrating CPAP machines in very obese subjects? Sleep Breath. 2016; 20( 2): 647– 652. [DOI] [PubMed] [Google Scholar]
- 32. Edmonds JC, Yang H, King TS, Sawyer DA, Rizzo A, Sawyer AM. Claustrophobic tendencies and continuous positive airway pressure therapy non-adherence in adults with obstructive sleep apnea. Heart Lung. 2015; 44( 2): 100– 106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Camacho M, Riaz M, Capasso R, et al. The effect of nasal surgery on continuous positive airway pressure device use and therapeutic treatment pressures: a systematic review and meta-analysis. Sleep. 2015; 38( 2): 279– 286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Li HY, Lin Y, Chen NH, Lee LA, Fang TJ, Wang PC. Improvement in quality of life after nasal surgery alone for patients with obstructive sleep apnea and nasal obstruction. Arch Otolaryngol Head Neck Surg. 2008; 134( 4): 429– 433. [DOI] [PubMed] [Google Scholar]
- 35. Nakata S, Noda A, Yagi H, et al. Nasal resistance for determinant factor of nasal surgery in CPAP failure patients with obstructive sleep apnea syndrome. Rhinology. 2005; 43( 4): 296– 299. [PubMed] [Google Scholar]
- 36. Tarabichi M, Fanous N. Finite element analysis of airflow in the nasal valve. Arch Otolaryngol Head Neck Surg. 1993; 119( 6): 638– 642. [DOI] [PubMed] [Google Scholar]
- 37. Hudgel DW, Robertson DW. Nasal resistance during wakefulness and sleep in normal man. Acta Otolaryngol. 1984; 98( 1-2): 130– 135. [DOI] [PubMed] [Google Scholar]
- 38. Nigro CA, González S, Arce A, Aragone MR, Nigro L. Accuracy of a novel auto-CPAP device to evaluate the residual apnea-hypopnea index in patients with obstructive sleep apnea. Sleep Breath. 2015; 19( 2): 569– 578. [DOI] [PubMed] [Google Scholar]
- 39. Coller D, Stanley D, Parthasarathy S. Effect of air leak on the performance of auto-PAP devices: a bench study. Sleep Breath. 2005; 9( 4): 167– 175. [DOI] [PubMed] [Google Scholar]
- 40. Baek JH, Jeon J-Y, Lee S-A. Accuracy of the auto scoring by the S9 CPAP in patients with obstructive sleep apnea. Sleep Med Res. 2016; 7( 1): 26– 32. [Google Scholar]
- 41. Isetta V, Navajas D, Montserrat JM, Farré R. Comparative assessment of several automatic CPAP devices’ responses: a bench test study. ERJ Open Res. 2015; 1( 1): 00031–2015. [DOI] [PMC free article] [PubMed] [Google Scholar]