UEF//eRepository
DSpace https://erepo.uef.fi
Rinnakkaistallenteet Luonnontieteiden ja metsätieteiden tiedekunta
2020
Prevalence and characteristics of positional obstructive sleep apnea (POSA) in patients with severe OSA
Oksenberg, A
Springer Science and Business Media LLC
Tieteelliset aikakauslehtiartikkelit
© Springer Nature Switzerland AG 2019 All rights reserved
http://dx.doi.org/10.1007/s11325-019-01897-1
https://erepo.uef.fi/handle/123456789/24846
Downloaded from University of Eastern Finland's eRepository
1
Reviewer Information Page:
1 2 3
4
Number of tables: 4
5
6
Number of figures: 3
7
8
Abstract word count: 248
9
10
Word count: 3116
11
12
References: 31
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
39
2
Prevalence and characteristics of positional obstructive sleep apnea
1
(POSA) in patients with severe OSA
2 3 4
Arie Oksenberg 1, Natan Gadoth 1, Juha Töyräs 2,3,4, Timo Leppänen 2,3 5 6
7
8
1Sleep Disorders Unit, Loewenstein Hospital – Rehabilitation Center, Raanana, Israel.
9
2Department of Applied Physics, University of Eastern Finland, Kuopio, Finland.
10
3Department of Clinical Neurophysiology, Diagnostic Imaging Center, Kuopio University Hospital, 11
Kuopio, Finland.
12
4School of Information Technology and Electrical Engineering, The University of Queensland, Brisbane, 13
Australia.
14 15 16
17 18 19 20 21 22 23
24 25 26 27 28
29 Address for correspondence:
30 Arie Oksenberg, PhD 31 Sleep Disorders Unit,
32 Loewenstein Hospital - Rehabilitation Center, 33 POB 3 Raanana - ISRAEL
34 Phone: 972- 9 -7709122 35 Fax: 972- 9 - 7709123 36 E-Mail: arieo@clalit.org.il 37
38
3 Abstract
1
Purpose We assessed the prevalence of Positional Patients (PPs) and the main predictors of 2
positional dependency in severe Obstructive Sleep Apnea (OSA). A simulated effect of 3
Positional Therapy (PT) vs. Continuous Positive Airway Pressure (CPAP) was also assessed.
4
Methods Polysomnographic recordings of 292 consecutive patients with severe OSA (Apnea- 5
Hypopnea Index (AHI) ≥ 30) who slept > 4h and had ≥ 30 min sleep in both supine and lateral 6
positions were assessed. PPs had a supine AHI / lateral AHI ratio ≥ 2 and Non-Positional 7
Patients (NPP) had a supine AHI / lateral AHI ratio < 2.
8
Results 35.3% of the severe OSA patients were PPs. They were less obese and had less 9
severe OSA (p<0.001) compared to NPPs. The percentage of total apnea-hypopnea time from 10
total sleep time (AHT%) was the most significant predictor for positional dependency. By 11
sleeping in the lateral posture, 78 (75.7%) patients obtained significant improvement of their 12
OSA severity and 9 (8.7%) of them became “Non-OSA”. Moreover, if CPAP was used only for 13
50% of total sleep time, 53 patients (18.2%) gained more benefit from avoiding the supine 14
posture than from CPAP therapy.
15
Conclusions More than a third of the studied severe OSA patients were PPs. These patients 16
could achieve a significant decrease in the number and severity of apneas and hypopneas by 17
adopting the lateral posture, suggesting that PT may be a valuable therapy for a significant 18
portion of these severe OSA patients who for whatever reason are not been treated by CPAP.
19 20
Key words: Obstructive Sleep Apnea (OSA), severe OSA patients, positional patients, 21
positional therapy, supine posture, lateral position, OSA treatment.
22 23
Trial Registry: ClinicalTrials.gov Identifier: NCT03232658 24
25
26 27 28 29
4 Introduction
1
Changes in sleeping position affect the occurrence and severity of obstructive sleep apnea 2
(OSA). OSA patients who have at least double frequency of breathing abnormalities while sleeping in 3
the supine versus the lateral position are considered positional patients (PP) and suffer from positional 4
OSA (POSA). Conversely, Non-Positional Patients (NPP) refers to those with high frequency of 5
breathing abnormalities in all positions1. Since the severity of OSA among PPs is totally dependent on 6
the sleep time spent in the supine posture, PPs obtain a substantial benefit from Positional Therapy 7
(PT), i.e. the avoidance of sleeping in the supine posture. Therefore, the distinction between PP and 8
NPP has important treatment implications.
9
The prevalence of PPs determined previously in our sleep disorders unit among all OSA patients 10
was about 56%, and even higher for mild-moderate OSA patients (65% to 69%)1. In the general 11
population, showed by a recent Swiss study, the prevalence it is even greater2. Among 1719 middle 12
age and older men and women, 53% were PPs, and among the 71.2% with OSA, 75% were PPs.
13
These results suggest that a large proportion of PPs could be successfully treated by PT and 14
therefore, significant effort should be invested in developing effective, simple, cheap, comfortable 15
devices for this treatment.
16
NPPs are the OSA patients with the most severe phenotype and for them Continuous Positive 17
Airway Pressure (CPAP) is the treatment of choice. Unfortunately, CPAP is not comfortable and many 18
OSA patients either refuse to try CPAP or reject it after a short trial period3, 4. Weaver and Grunstein, 19
20085 reported a non-adherence rate of 46%-83%, while in a recent study from Singapore, 21.5%
20
patients with severe OSA were non-adherent6. The overall CPAP non-adherence rate based on a 7-h / 21
night sleep time reported in studies conducted over the twenty-year period was 34.1%, with no 22
significant improvement over time 7. Moreover, the CPAP non-adherence of mild or asymptomatic 23
patients could be even worse8, 9. Consequently, many OSA patients that should be treated are not, a 24
fact which can have severe health consequences10. For CPAP-intolerant OSA patients, oral appliance 25
therapy, ENT and/or maxilla-mandibular surgery and upper airway stimulation are other potentially 26
beneficial therapies11. 27
PT diminishes disease severity and in certain patients might significantly improve their general 28
well-being, since in the lateral posture both the frequency and the severity of apnea and hypopnea 29
events is significantly reduced in severe OSA patients12. In the supine position, apneas and arousals 30
5
are longer, desaturations are deeper, snoring is louder, and tachy-bradycardia changes immediately 1
after apneas and hypopneas are more severe compared to the lateral position12. 2
Severe OSA patients are the main group of OSA patients seeking treatment. They are also those 3
at highest risk of the detrimental effects of this disease. If a significant portion of severe OSA patients 4
are PPs, PT may become as a valuable alternative therapy for patients that are not using CPAP 5
treatment.
6
A consensus regarding the prevalence of PPs in severe OSA has not yet been reached, and 7
there is a need for better understanding of the characteristics of severe OSA, the differences between 8
PPs and NPPs, and the main predictors of PPs. We hypothesize that the prevalence of PP in severe 9
OSA patients is relatively low and the factors defining their positionality are similar to those found 10
among all OSA patients. In this study we aimed to investigate the prevalence of PPs among patients 11
with severe OSA, to assess the differences between PPs and NPPs and to assess the main predictors 12
of positional dependency. We also performed a simulation to assess the effect of using PT in severe 13
OSA patients versus the simulation effect of CPAP usage.
14 15
Methods 16
Patients 17
456 adult patients (≥18 years old) underwent a complete polysomnographic (PSG) recording at 18
the Sleep Disorders Unit, Loewenstein Hospital – Rehabilitation Center (Raanana, Israel) between 19
March 2015 and October 2016, and were diagnosed to have severe OSA (Apnea Hypopnea Index 20
(AHI), AHI ≥ 30). Patients were excluded from the further analysis if they had slept less than 30 21
minutes in the supine position (n=110), less than 30 minutes in lateral positions (n=37) or had slept 22
less than 4 hours during the PSG (n=17). Thus, 292 patients were finally included for the further 23
analysis of the present retrospective study. Changes in body position were determined using a video 24
system and scored by a sleep technician. Patients were classified to be either PPs or NPPs according 25
to the Cartwright definition13: PPs have AHI≥5 and a ratio of supine AHI / lateral AHI≥2. NPPs have 26
AHI≥5 and supine AHI / lateral AHI ratio< 2.
27
Overnight Polysomnographic 28
6
All PSG recordings were performed with the Embla Sleep Diagnostic System, (Flaga, Reykjavik, 1
Iceland) and the analysis included conventional diagnostic parameters1. Sleep stages were scored 2
manually by the same polysomnographic technicians during the entire study, following standard 3
criteria14 using the RemLogic software, version 3.1.1. Respiratory events were scored based on 4
scoring criteria provided by the American Academy of Sleep Medicine in 201215. An apnea was scored 5
if the amplitude of oronasal thermistor signal decreased ≥90% from the baseline for at least 10 6
seconds. A hypopnea was scored if the amplitude of oronasal airflow decreased ≥30% from the 7
baseline for 10 seconds or longer and was followed by an arousal or a drop of at least 3% in SaO215. 8
For each patient we calculated conventional PSG parameters as well as Total Apnea Time (TAT), 9
Total Hypopnea Time (THT) and TAT+THT as percentage of Total Sleep Time (AHT%). The Epworth 10
Sleepiness Scale (ESS)16 was used for the subjective daytime sleepiness estimation.
11
Treatment simulation 12
The effects of treatment modalities on the severity of OSA were simulated in the present study. The 13
effect of PT was simulated assuming that the patients did not sleep in the supine position, therefore, the 14
lateral AHI was considered the total AHI. The simulation of the CPAP effect was based on the assumption 15
that the frequency of obstructive apnea/hypopneas events is constant throughout the night. Furthermore, 16
it was assumed that CPAP with optimal pressure prevents all obstructive events, and no apneas or 17
hypopneas occur during simulated CPAP usage. CPAP adherence was considered optimal only for 4 18
hours or for 50% of the total sleep time17. 19
Statistical analysis 20
The statistical analysis was performed with SPSS 22.0 (IBM Corporation, Chicago, USA) with 21
p<0.05 set as the limit of statistical significance. Statistical significance of differences in demographic and 22
sleep data between NPPs and PPs was evaluated with Mann-Whitney U test (continuous variables) and 23
with Chi-square test (categorical variables). Prevalence trends of PPs along AHI, AHT%, and along BMI 24
and age categories were assessed with Chi-square test. Stepwise binomial logistic regression analysis 25
(forward likelihood ratio) was performed to find the most significant predicting factors for positional 26
dependency. The built model was based on the most significant predictive factors, and the sensitivity, 27
specificity, as well as the negative and positive predictive values were calculated.
28
7 1
Results 2
The study group included substantially more males than females (80.8% vs. 19.2%), with an overall 3
median age of 60.4 years. Patients were mostly obese (median BMI 32.1 kg/m2) and 46.2% of them had 4
hypertension (Table 1). Although the patients had many breathing abnormalities in supine and lateral 5
postures (supine AHI = 69.5 (23.6-153.7); lateral AHI = 43.1 (2.3-179.4), (median (range)), respectively, 6
severity of OSA was not reflected by their ESS score.
7
Among the study group, 35.3% were PPs (Table 2). There were no significant differences in gender 8
distribution (p=0.939) or age (p=0.962) between PPs and NPPs but NPPs were more obese (p<0.001, 9
Table 2).
10
NPPs had higher AHI during REM sleep, NREM sleep, and while sleeping in lateral position leading 11
to higher total AHI. TAT was also higher in NPPs compared to PPs (p<0.001) while no statistically 12
significant difference was observed in THT (p=0.819, Table 2). Furthermore, obstructive apnea events 13
were slightly longer (p=0.044) and the AHT% was higher in NPPs (p<0.001, Table 2). In the NPP group, a 14
higher percentage of patients were hypertensive, but the difference did not reach statistical significance 15
(p=0.119). No significant difference (p=0.386) in ESS score was observed between PPs and NPPs.
16
The prevalence of PPs in relation to the 5 AHI categories is presented in Figure 1. A significant 17
(p<0.001) negative trend was observed, i.e. the prevalence of PPs decreased with increasing AHI. A 18
similar negative trend (p<0.001) was observed for AHT% and BMI (Table 3). Moreover, the same trend 19
(p<0.001) was also observed when the patient population was divided into two categories based on 20
parameter in question: increasing AHI, AHT% or BMI led to decrease in the prevalence of PPs. In 21
contrast, there were no statistically significant differences in the prevalence of PPs between the age 22
categories or between genders (Figure 2).
23
The stepwise binomial logistic regression model showed the most significant predicting factors for 24
positional dependency were AHT% (β=-0.147, p<0.001), BMI (β=-0.128, p<0.001), and total apnea time 25
(β=0.011, p=0.040) (Table 4).This indicates that when AHT% (odd ratio (OR)=0.863) and BMI 26
(OR=0.880) increases the probability to be classified as PP decreases. In contrast, when total apnea time 27
increases the probability to be classified as PP increases (OR=1.011).
28
The stepwise binomial logistic regression model (including AHT%, BMI, and total apnea time as 29
predicting variables) classified 77.2% of the patients correctly to be either NPP or PP. The model’s 30
8
sensitivity, specificity, positive and negative predictive values were 81.7%, 68.2%, 83.6%, and 65.2%, 1
respectively. The area under a ROC curve (AUC) was 0.857 (95% CI: 0.815-0.899, p<0.001) indicating 2
good predictive accuracy of the model.
3
When the effects of simulated PT and CPAP therapies were compared, certain patients gained 4
greater benefits from PT than CPAP (Figure 3). When adherence to CPAP therapy was considered 4 5
hours, PT was more effective for 6.5% (n=19) of the patients. When adherence of CPAP was considered 6
50% of total sleep time, 18.2% (n=53) of the patients gained greater benefit from PT than from CPAP 7
(Figure 3).
8
In addition, after completion of the simulated PT for the whole group of PP, 8.7% (n=9) of the 9
patients were judged to have no OSA with median AHI of 3.8 (range: 2.3-4.9). Similarly, 24.3% (n=25, 10
median AHI (range): 12.7 (5.2-14.8)) and 42.7% (n=44, median AHI (range): 23.5 (15.3-29.7)) of the 11
patients were re-classified to have mild and moderate OSA, respectively. Only 24.3% (n=25) of the 12
patients remained in the severe OSA category after simulated although their median AHI was reduced 13
from 40.4 (range: 30.2-91.6) to 33.1 (30.0-47.5).
14 15 16
Discussion 17
Several findings of this study are of interest. First, 35.3% of these severe OSA patients were PPs.
18
The highest prevalence of PP in OSA is found in mild-moderate OSA ranging between 49.5-87% of OSA 19
patients1. Few studies have reported the prevalence of PP in severe OSA which is lower and ranging 20
between 6.5-44.9%18. This variability is consequence of several factors including the number and 21
ethnicity of the patients, but also due the several definitions for positional OSA patients. The prevalence 22
of PP in severe OSA found in this study is consistent with the prevalence in our previous study1 and 23
others19 using the same criterion for positional dependency.
24
Second, if we assume that PPs use Positional Therapy (PT) successfully, 44 (42.7%) of these 25
severe OSA patients will become moderate, 25 (24.3%) will become mild, and 9 (8.7%) will be free of 26
OSA. By sleeping in the lateral posture, 75.7% of these PPs will improve their condition significantly as a 27
consequence of an important decrease in OSA severity while the remaining 24.3% will still suffer from 28
severe OSA but with decreased severity.
29
9
Third, the simulated comparison of CPAP vs. PT treatments revealed that if CPAP is used only for 1
50% of the total sleep time, 53 patients (18.2%) would benefit more from PT than from CPAP therapy.
2
Also, if we consider adherence to CPAP usage as 4 hours per night, 19 patients (6.5%) would benefit 3
more from PT than from CPAP. Thus, even for some PPs with severe OSA, avoiding the supine posture 4
during sleep could be a better treatment alternative than sub-optimal CPAP adherence.
5
A recent study17 using a similar simulation procedure showed that severe OSA patients need 6.5 6
hours of continuous CPAP use in order to normalize the AHI, and by using this therapy for only 4 hours 7
(the still accepted minimal threshold for CPAP use) leaves many OSA patients with a significant residual 8
AHI. This could explain the failure of some clinical trials to obtain significant cardiovascular benefits with 9
CPAP usage20. In addition, CPAP adherence of only 4 hours leaves significant REM–related OSA 10
untreated, which may have important cardiovascular implications21. 11
12
What are the main differences between PPs and NPPs in this group of patients with severe 13
OSA?
14
While there were no statistically significant age or gender differences between PPs and NPPs, 15
the NPPs were significantly more obese. PPs had not only a significantly lower total AHI, but also 16
lower AHI during NREM and REM sleep, and obviously while sleeping in the lateral position.
17
Additionally, PPs had shorter obstructive apnea events, shorter total apnea-hypopnea time and lower 18
percentage of apnea and hypopnea time out of total sleep time (AHT%) than NPPs. In other words, in 19
this group of severe OSA patients, similarly to the total OSA patients, PPs had significantly less severe 20
OSA than NPPs.
21
Similar to what has been found in the OSA population diagnosed both at the Sleep Disorders 22
Units and at home in the general OSA population1, 2, the prevalence of PPs decreases significantly 23
when AHI increases (Figure 1) and when BMI increases (Table 3). However, by using a stepwise 24
binomial logistic regression model we found that the most significant predicting factors for positional 25
dependency in this severe OSA group were AHT%, BMI, and total apnea time (Table 4). When AHT%
26
and BMI increased, the probability to be classified as PP decreased but when total apnea time 27
increased, the probability increased. It is logical that when AHT% increases the probability to be 28
classified as PP decreases as in NPPs more respiratory events occur in lateral position compared to 29
10
PPs. Furthermore, the finding that increase in total apnea time increases the probability to be 1
classified as PP is also in line with our previous findings31. It has been shown that in severe OSA, 2
proportion of apnea events from all respiratory events is higher in supine compared to lateral positions 3
(43.5% vs. 24.1%, p<0.001) and that apneas are longer in supine position compared to lateral 4
positions31. This is, more a patient with severe OSA sleeps in supine position the higher is the total 5
apnea time due to increased number and duration of apnea events.
6
The simulated effects of PT and CPAP on OSA severity in PPs with severe OSA 7
The gold standard treatment for severe OSA is the use of CPAP devices. The use of CPAP 8
dramatically reduces the morbidity and improves quality of life of patients with severe OSA but 9
unfortunately, CPAP adherence is far from optimal5. For many severe OSA patients’ adaptation to 10
CPAP is difficult and consequently the adherence to CPAP is unsatisfactory.
11
Based on the present results, PPs with severe OSA, who refuse or have difficulties using CPAP, 12
could be treated with PT and they may obtain a significant amelioration in their sleep disorder. Thus, 13
these patients should be informed that by sleeping in the lateral posture either with or without a PT 14
device, they could obtain an improvement of their condition and achieve considerable benefits.
15
Many patients with severe OSA may have voluntarily adopted the lateral posture during sleep 16
after they or their bed partner noted an improvement when shifting from supine to lateral sleep 17
posture. It is also possible that they just are unable to sleep on their back because they simply cannot 18
breathe, or they wake up frequently with frightening choking sensation22. Although this has never been 19
quantified, this description has been often reported by suspected OSA patients in clinical settings.
20
Although adherence to PT using the old tennis ball technique was low23, the long-term adherence 21
has improved significantly with the use of new generation devices that appear to be much more 22
comfortable24-27. In current clinical practice, PT is mainly offered to mild-moderate OSA patients who 23
form the biggest group of PPs among OSA patients1, 18. The results of the present study suggest that 24
the avoidance of the supine posture during sleep could be a possible treatment alternative even for 25
some PPs with severe OSA.
26
Limitations of the present study 27
The data assessed were obtained from 292 severe OSA patients who underwent only a single 28
night PSG recording; assessing several nights of PSG recordings could strengthen the findings28. 29
11
Furthermore, the finding that 35.3% of the severe OSA patients had positional OSA was based on the 1
relatively low number of patients (n=292) and thus, future studies are needed with a larger patient 2
population from multiple centers to verify this result. In addition, the effect of body position was based 3
on supine versus lateral posture and the potential effect of head/neck position on the occurrence and 4
severity of breathing abnormalities during sleep was not assessed29. Additionally, most of our 5
participants were men (80.8%) and thus further studies including more women are needed. We used 6
the most common criterion for defining positional dependency for OSA patients, but the use of other 7
definitions could have provided different results30. The possible difference in positional dependency of 8
apneas/hypopneas between REM and NREM sleep were not assessed in this study. This topic is of 9
clinical interest since theoretically PT could result in an increase of REM–related OSA in some 10
patients, which could have important cardiovascular implications. We are aware that only a few studies 11
have assessed the long-term PT usage, and although the adherence to new devices based on using 12
vibration at the neck or chest is encouraging24-27 additional research is certainly needed. Our 13
simulation protocol also needs to be re-evaluated; we assumed that by using PT supine sleep time is 14
zero, which in practice could be erroneous. A recent systematic review and meta-analysis of three 15
prospective cohort studies and four randomized controlled trials on the efficacy of new generation of 16
devices for PT found that their use produced 84% reduction in supine sleep time26. Nevertheless, 17
despite these limitations our results indicate that PPs are a significant subgroup of severe OSA 18
patients for whom PT could represent a valuable therapeutic alternative if CPAP is not used.
19
Summary 20
In severe OSA, many breathing abnormalities occur in both supine and lateral postures and for 21
patients with severe OSA; CPAP is the treatment of choice. In the present study, we found that 35.3%
22
of severe OSA patients were PPs and by sleeping in the lateral posture, 75.7% of them can improve 23
their condition substantially. These PPs are less obese and have less severe disease than NPPs. The 24
simulated Positional Therapy in severe OSA showed great benefits but randomized controlled 25
prospective studies are needed.
26 27 28
12
Authors’ Contributions: AO conceived the research, TL carried out the analyses and prepared the 1 Figures, JT proposed the simulation analysis. All authors assisted with writing and interpretation of the 2 findings. All authors approved this manuscript in its final form.
3
Funding Information: This research was supported by the Academy of Finland (decision number 313697), 4 Finnish-Norwegian Medical Foundation, the Research Committee of the Kuopio University Hospital 5 Catchment Area for the State Research Funding (projects 5041767 and 5041768), Tampere Tuberculosis 6 Foundation, and the Respiratory Foundation of Kuopio Region.
7
Conflict of Interest: The authors certify that they have no affiliations with or involvement in any 8 organization or entity with any financial interest (such as honoraria; educational grants; participation in 9 speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and 10 expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or 11 professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in 12 this manuscript.
13
Informed consent: This study was approved by the Helsinki ethical committee of the Loewenstein Hospital 14 Rehabilitation Center. (protocol 0006-17-LOE). For retrospective chart reviews, formal consent is not 15 required.
16
Ethical approval: All procedures performed in studies involving human participants were in accordance 17 with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki 18 declaration and its later amendments or comparable ethical standards.
19 20
ABBREVIATIONS:
21
AHI = Apnea Hypopnea Index 22
AHT% = Apnea + Hypopnea Time as Percentage of TST 23
AUC = Area under the Curve 24
BMI = Body Mass Index 25
CPAP = Continuous Positive Airway Pressure 26
ESS = Epworth Sleepiness Scale 27
NPP = Non-Positional Patients 28
NREM = Non - Rapid Eye Movement Sleep 29
OSA = Obstructive Sleep Apnea 30
POSA = Positional Obstructive Sleep Apnea 31
PSG = Polysomnography 32
PP = Positional Patients 33
PT = Positional Therapy 34
REM =Rapid Eye Movement Sleep 35
ROC – Receiver Operating Characteristic 36
TST = Total Sleep Time 37
38
13 References
1
1. Oksenberg A, Silverberg DS, Arons E, Radwan H. (1997). Positional vs. non-positional 2
obstructive sleep apnea patients. Anthropomorphic, nocturnal polysomnographic and multiple sleep 3
latency test data. Chest, 112: 629–639 4
2. Heinzer R, Petitpierre NJ, Marti-Soler H, Haba-Rubio J. (2018). Prevalence and characteristics of 5
positional sleep apnea in the HypnoLaus population based - cohort. Sleep Medicine, 48:157-162.
6
3. Weaver TE, Sawyer AM. (2010). Adherence to continuous positive airway pressure treatment for 7
obstructive sleep apnoea: implications for future interventions. Indian J Med Res. 131:245-58.
8
4. Yang MC, Lin CY, Lan CC, Huang CY, Huang YC, Lim CS, Liu YC, Wu YK. (2013). Factors 9
affecting CPAP acceptance in elderly patients with obstructive sleep apnea in Taiwan. Respir Care;
10
58:1504-13.
11
5. Weaver TE, Grunstein RR. (2008). Adherence to continuous positive airway pressure therapy:
12
the challenge to effective treatment. Proc Am Thorac Soc, 5:173-8.
13
6. Tan B, Tan A, Chan YH, Mok Y, Wong HS, Hsu PP. (2018). Adherence to Continuous Positive 14
Airway Pressure therapy in Singaporean. Am J Otolaryngol, 39:501-506.
15
7. Rotenberg BW, Murariu D, Pang KP. (2016). Trends in CPAP adherence over twenty years of 16
data collection: a flattened curve. J Otolaryngol Head Neck Surg, 45:43: 1-9.
17
8. Barbé F, Mayoralas LR, Duran J, et al. (2001). Treatment with continuous positive airway 18
pressure is not effective in patients with sleep apnea but no daytime sleepiness. A randomized, controlled 19
trial. Ann Intern Med, 134:1015–23.
20
9. Rosenthal L, Gerhardstein R, Lumley A, Guido P, Day R, Syron ML, Roth T. (2000). CPAP 21
therapy in patients with mild OSA: implementation and treatment outcome. Sleep Med, 1:215-20.
22
10. Kendzerska T, Mollayeva T, Gershon AS, Leung RS, Hawker G, Tomlinson G. (2014). Untreated 23
obstructive sleep apnea and the risk for serious long-term adverse outcomes: a systematic review. Sleep 24
Med Rev, 18:49-59.
25
11. Lorenzi-Filho G, Almeida FR, Strollo PJ. (2017). Treating OSA: Current and emerging therapies 26
beyond CPAP. Respirology, 22:1500-1507 27
12. Oksenberg A, Khamaysi Y, Silverberg DS, Tarasiuk A. (2000). Association of body position with 28
severity of apneic events in patients with severe non-positional obstructive sleep apnea. Chest, 118:
29
1018-24 30
13. Cartwright RD. (1984). Effect of sleep position on sleep apnea severity. Sleep, 7:110-4.
31
14
14. Iber C, Ancoli-Israel S, Chesson A, Quan S. (2007). The AASM Manual for the Scoring of Sleep 1
and Associated Events: Rules, Terminology, and Technical Specifications. Westchester Il: American 2
Academy of Sleep Medicine pp. 17-59 3
15. Berry RB, Budhiraja R, Gottlieb DJ, et al. (2012). Rules for scoring respiratory events in sleep:
4
update of the 2007 AASM manual for the scoring of sleep and associated events. J Clin Sleep Med, 8:
5
597-619.
6
16. Johns MW (1991). A new method for measuring daytime sleepiness: the Epworth Sleepiness 7
Scale. Sleep, 14: 540–545.
8
17. Kulkas A, Leppänen T, Nikkonen S, Oksenberg A, Duce B, Mervaala E, Töyräs J. (2018).
9
Required CPAP usage time to normalize AHI in obstructive sleep apnea patients: a simulation study.
10
Physiol Meas, 39(11):115009. doi: 10.1088/1361-6579/aae7fe.
11
18. Yingjuan M, Siang WH, Leong Alvin TK, Poh HP. (2019). Positional therapy for Positional 12
Obstructive Sleep Apnea. Sleep Med Clin 14:119-133.
13
19. Ravesloot MJ, Frank MH, van Maanen JP et al, (2016). Positional OSA part 2: retrospective 14
cohort analysis with a new classification system (APOC). Sleep Breath 20:881-8 15
20. McEvoy RD, Antic NA, Heeley E et al. (2016). CPAP for prevention of cardiovascular events in 16
obstructive sleep apnea New Engl. J. Med, 375 919–31 17
21. Alzoubaidi M, Mokhlesi B. (2016). Obstructive sleep apnea during rapid eye movement sleep:
18
clinical relevance and therapeutic implications Curr Opin Pulm Med, 22(6):545-54.
19
22. Gadoth N, Oksenberg A. (2015). Positional Therapy in Obstructive Sleep Apnea: For Whom and 20
for Whom Not. N. de Vries et al. (eds.), Positional Therapy in Obstructive Sleep Apnea. Springer 21
International Publishing Switzerland; pp 383-394 22
23. Bignold JJ, Deans-Costi G, Goldsworthy MR, Robertson CA, McEvoy D, Catcheside PG, Mercer 23
JD. (2009). Poor long-term patient compliance with the tennis ball technique for treating positional 24
obstructive sleep apnea. J Clin Sleep Med, 5:428-30 25
24. de Ruiter MHT, Benoist LBL, de Vries N, de Lange J (2018) Durability of treatment effects of the 26
Sleep Position Trainer versus oral appliance therapy in positional OSA: 12-month follow-up of a 27
randomized controlled trial. Sleep Breath. 22:441-450.
28
25. Levendowski D, Cunnington D, Swieca J, Westbrook P. (2018). User Compliance and Behavioral 29
Adaptation Associated With Supine Avoidance Therapy. Behav Sleep Med 16:27-37 30
26. Ravesloot MJ, White D, Heinzer R, Oksenberg A, Pépin JL. (2017). Efficacy of the New 31
Generation of Devices for Positional Therapy for Patients with Positional Obstructive Sleep Apnea: A 32
Systematic Review of the Literature and Meta-Analysis. J Clin Sleep Med, 13:813-824 33
15
27. Beyers J, Vanderveken OM, Kastoer C, Boudewyns A, De Volder I, Van Gastel A, Verbraecken 1
JA, De Backer WA, Braem MJ, Van de Heyning PH, Dieltjens M. (2019).Treatment of sleep-disordered 2
breathing with positional therapy: long-term results. Sleep Breath. 2019 doi: 10.1007/s11325-019-01792- 3
4 9.
28. Joosten SA, O'Donoghue FJ, Rochford PD, Barnes M, Hamza K, Churchward TJ, et al. (2014).
5
Night-to-night repeatability of supine-related obstructive sleep apnea. Annals of the American Thoracic 6
Society, 11:761 7
29. Van Kesteren ER, van Maanen JP, Hilgevoord AA, Laman DM, de Vries N. (2011). Quantitative 8
effects of trunk and head position on the apnea hypopnea index in obstructive sleep apnea. Sleep, 9
34:1075-81.
10
30. Levendowski DJ, Oksenberg A, Vicini C, Penzel T, Levi M, Westbrook PR. (2018). A systematic 11
comparison of factors that could impact treatment recommendations for patients with Positional 12
Obstructive Sleep Apnea (POSA). Sleep Med. 50:145-151. doi: 10.1016/j.sleep.2018.05.012 13
31. Leppänen T, Töyräs J, Muraja-Murro A, Kupari S, Tiihonen P, Mervaala E, Kulkas A. (2016).
14
Length of Individual Apnea Events Is Increased by Supine Position and Modulated by Severity of 15
Obstructive Sleep Apnea. Sleep Disorders. 2016:9645347.
16 17 18 19 20 21 22 23 24 25 26 27 28 29
16
Table 1. Demographic and sleep data (median (range)) in the entire cohort of severe OSA 1
patients.
2
Patients (n) 292
Male (n (%)) 236 (80.8)
Female (n (%)) 56 (19.2)
Age (years) 60.4 (20.6-89.5)
BMI (kg/m2) 32.1 (21.2-57.6)
TST total (min) 361 (243-532)
TST supine (min) 141 (31-421)
TST lateral (min) 202 (33-409)
TST REM (min) 60 (0-180)
TST Non-REM (min) 298 (174-503)
AHI total (1/h) 50.9 (30.1-170.2)
AHI supine (1/h) 69.5 (23.6-153.7)
AHI lateral (1/h) 43.1 (2.3-179.4)
AHI REM (1/h) 54.9 (0.0-129.7)
AHI nonREM (1/h) 51.1 (13.5-175.5)
Supine AHI / Lateral AHI ratio 1.4 (0.3-25.4) Total apnea time (min) 84.4 (2.5-354.3) Total hypopnea time (min) 40.2 (0.0-163.3) Total apnea + hypopnea time (min) 136.5 (42.8-369.5)
AHT % 38.2 (13.6-76.3)
Obstructive apnea duration (sec) 25.0 (13.6-60.8) Central apnea duration (sec) 16.4 (10.0-40.8) Mixed apnea duration (sec) 32.4 (12.8-69.3) Hypopnea duration (sec) 26.1 (11.8-71.6)
ESS (points) 9 (0-24)
Hypertension (n (%)) 135 (46.2)
BMI = body mass index, TST = total sleep time, REM = rapid eye movement sleep, AHI = apnea- 3
hypopnea index, AHT% = percentage of total apnea + hypopnea time from TST, ESS = Epworth 4
sleepiness scale 5
6 7 8 9 10 11 12 13 14 15
17
Table 2. Demographic and sleep data (median (range)) in non-positional (NPP) and positional (PP) 1
severe obstructive sleep apnea patients. Positional obstructive sleep apnea was defined based on 2
the Cartwright classification (i.e. supine AHI at least twice the non-supine AHI).
3
Non-Positional
(NPP) Positional
(PP) p-value
Patients (n (%)) 189 (64.7) 103 (35.3) -
Male (n (%)) 158 (81.0) 83 (80.6) 0.939 *
Female (n (%)) 36 (19.0) 20 (19.4)
Age (years) 60.4 (20.6-89.5) 60.8 (22.4-86.5) 0.962
BMI (kg/m2) 33.6 (21.2-57.6) 29.8 (21.9-43.3) <0.001
TST total (min) 361 (243-532) 361 (246-468) 0.174
TST supine (min) 133 (31-421) 152 (32-340) 0.238
TST lateral (min) 203 (33-409) 200 (33-393) 0.200
TST REM (min) 58 (0-152) 66 (0-180) 0.072
TST nonREM (min) 303 (174-503) 289 (182-405) 0.005
AHI total (1/h) 61.3 (30.1-170.2) 40.4 (30.2-91.6) <0.001 AHI supine (1/h) 69.1 (23.6-153.7) 69.5 (35.7-136.6) 0.927 AHI lateral (1/h) 61.3 (20.0-179.4) 21.4 (2.3-47.5) <0.001 AHI REM (1/h) 59.4 (0.0-129.7) 44.3 (0.0-107.9) <0.001 AHI nonREM (1/h) 61.7 (13.5-175.5) 39.8 (16.4-78.0) <0.001 Supine AHI / non-supine AHI ratio 1.09 (0.31-2.00) 3.13 (2.05-25.44) <0.001 Total apnea time (min) 111.9 (4.9-354.3) 58.6 (2.5-212.4) <0.001 Total hypopnea time (min) 41.2 (0.0-163.3) 39.6 (4.4-125.6) 0.819 Total apnea + hypopnea time (min) 160.0 (54.6-369.5) 100.0 (42.8-223.8) <0.001
AHT % 45.9 (17.2-76.3) 28.7 (13.6-57.2) <0.001
Obstructive apnea duration (sec) 25.7 (14.6-60.8) 24.7 (13.6-38.4) 0.044 Central apnea duration (sec) 16.6 (10.0-40.8) 16.0 (11.6-25.1) 0.748 Mixed apnea duration (sec) 33.0 (12.8-69.3) 30.3 (19.0-55.0) 0.171 Hypopnea duration (sec) 25.6 (11.8-71.6) 27.4 (16.1-53.0) 0.078
ESS (points) 9 (0-24) 9 (0-21) 0.386
ESS>10 (n (%)) 74 (39.2) 38 (36.9) 0.440
Hypertension (n (%)) 94 (49.7) 41 (39.8) 0.119 *
4
Statistical significance of differences in variables between NPP and PP was evaluated with Mann- 5
Whitney U test. An asterisk denotes that statistical significance of difference was evaluated with 6 Chi-square test. Bolded p-values denote statistically significant differences (p<0.05) between the 7 variables.
8
BMI = body mass index, TST = total sleep time, REM = rapid eye movement sleep, AHI = apnea- 9
hypopnea index, AHT% = percentage of total apnea + hypopnea time from TST, ESS = Epworth 10 sleepiness scale
11 12 13 14 15
18
Table 3. Prevalence of Positional Patients (PPs) according to different categories of the 1
percentage of total apnea and hypopnea time from total sleep time (AHT%) and body mass index 2
(BMI) in severe OSA patients.
3
Variable No. of severe OSA
patients No. of PP (%) AHT %
<30 84 58 (69.0)
30 to <40 71 32 (45.1)
40 to <50 56 8 (14.3)
50 to <60 37 5 (13.5)
≥60 44 0 (0)
BMI <25 12 6 (50.0)
25 to <30 83 46 (55.4)
30 to <35 98 31 (31.6)
35 to <40 58 18 (31.0)
≥40 40 2 (5.0)
4
The prevalence of PP shows statistically significant reduction as AHT% (X2 = 87.4, df = 4, 5
p<0.001) and BMI (X2 = 32.9, df = 4, p<0.001) increase.
6 7
19
Table 4. Stepwise binomial logistic regression suggests that the most significant predicting factors for positional dependency were the percentage of total apnea and hypopnea time from total sleep time (AHT%), body mass index (BMI), and total apnea time.
Parameter β SE Wald df OR (95 % CI) p-value
BMI (kg/m2) -0.128 0.033 14.919 1 0.880 (0.825-0.939) <0.001 Total apnea time (min) 0.011 0.005 4.233 1 1.011 (1.001-1.022) 0.040
AHT % -0.147 0.027 29.534 1 0.863 (0.819-0.910) <0.001
20
Figure 1 - Effect of OSA severity (i.e. AHI) on the prevalence of Positional Patients (PPs). n denotes the number of PPs. The prevalence of PPs decreases statistically significantly (Chi-square test, X2 = 70.8, df = 4, p<0.001) as AHI increases.
0 10 20 30 40 50 60 70
1
n= 4 n = 4 n = 14
n = 32 n= 49
65.3
47.8
32.6
10.0
6.0
Posi tion al P at ie nt s ( PP ), %
30 to <40 40 to <50 50 to <60 60 to <70 ≥70
A pnea-hypopnea index (AHI)
21
Figure 2 - Effect of AHI, AHT%, BMI, age and gender on the prevalence of positional patients (PPs). n denotes a number of PPs. An asterisk (*) denotes a statistically significant difference (p < 0.001) in the prevalence of positional OSA.
0 20 40 60 80
AHI<50 AHT%<40 BMI<30 Age<60 Male
35.2 35.7 58.1
35.2 34.5
Posi tion al P at ie nt s ( PP ), %
AHI<50 AHI≥50 AHT % <40 AHT % ≥ 40 BMI<30 BMI≥30 Age<60 Age≥60 Male Female
n=81 n=22 n=90 n=13 n=52 n=51 n=45 n=48 n=83 n=20
57.0
14.7*
9.5*
54.7
26.0*
22
Figure 3 - Bland-Altman plot of the differences in effectiveness of simulated Positional Therapy (PT) vs. CPAP. A. Adherence of CPAP therapy was considered to be 4 hours. In this cohort of severe OSA patients, 19 patients (6.5%, i.e. the circles below the red line) would benefit more from PT than CPAP therapy. B. Adherence of CPAP therapy was considered to be 50 % from total sleep time. In this case, 53 patients (18.2%, i.e. the circles below the red line) would benefit more from PT than CPAP therapy.
A B