Current Treatment Options in Neurology

, 16:305

Advances in the Treatment of Obstructive Sleep Apnea


  • David Young
    • Emory University
    • Emory University
Sleep Disorders (S Chokroverty, Section Editor)

DOI: 10.1007/s11940-014-0305-6

Cite this article as:
Young, D. & Collop, N. Curr Treat Options Neurol (2014) 16: 305. doi:10.1007/s11940-014-0305-6
Part of the following topical collections:
  1. Topical Collection on Sleep Disorders

Opinion statement

This article focuses on the treatment of obstructive sleep apnea (OSA), using the most recent available data. The first choice of treatment for patients with moderate or severe obstructive sleep apnea is continuous positive airway pressure (CPAP), which was first described in 1981 (Sullivan et al. Lancet 1(8225):862-5, 1981) and works by splinting the airway open to facilitate proper airflow. For patients with mild OSA, other treatments may be considered including positional therapy, weight loss, or oral appliances. Oral appliances are continuing to become more mainstream, and may be a reasonable first-line treatment even for some patients with moderate OSA, such as those who cannot tolerate or do not want to use CPAP. Some evidence suggests that adherence to mandibular advancement devices (MADs), a type of oral appliance, may be superior to that of CPAP. Recent evidence has suggested that the MAD may be similar to CPAP in preventing cardiovascular mortality in OSA, though objective measurement of MAD adherence was not available in the study. Objective adherence monitors are now available for oral appliances and should prove valuable for clinicians. Pharmacotherapy has not been shown to be significantly effective in the treatment of OSA and should be considered as an adjunctive treatment class, though some emerging evidence may support pharmacotherapy for specific purposes, such as acetazolamide for high-altitude travelers and zonisamide for weight loss. Upper airway surgical intervention remains a second- or third-line treatment class for moderate to severe OSA, though multiple case series of maxillomandibular advancement (MMA) have shown considerable, statistically significant improvements in AHI. Weight loss should always be recommended for patients with OSA who are overweight or obese, as weight loss may result in improvement in OSA. Bariatric operations are effective for obesity and are reasonable considerations for obese patients, although a recent randomized controlled trial found that bariatric intervention failed to achieve superiority over conventional weight loss therapy in terms of apnea-hypopnea index (AHI) reduction.


Obstructive sleep apnea (OSA)Continuous positive airway pressure (CPAP)Positional therapyMandibular advancement devices (MADs)Maxillomandibular advancement (MMA)Apnea–hypopnea index (AHI)


Obstructive sleep apnea (OSA) is a sleep disorder in which the upper airway partially or completely collapses during sleep, blocking airflow and resulting in arousals and awakenings. The sleep disruption often leads to unrefreshing sleep and excessive daytime sleepiness. Obstructive sleep apnea is more common in obese and older individuals [1, 2], and weight gain is associated with worsening of OSA [3]. For the period of 2007–2010, the overall United States prevalence of mild to severe sleep-disordered breathing (defined as an apnea-hypopnea index, or AHI, ≥5) has been estimated by one study to be 26 % for the population aged 30–70 [2]. The same group estimated the current prevalence of moderate to severe sleep-disordered breathing (AHI ≥15) to be 13 % for men and 6 % for women aged 30–70 [2].

Obstructive sleep apnea is a risk factor for stroke and death from any cause, independent of certain other risk factors including hypertension, hyperlipidemia, smoking, body mass index, diabetes mellitus, atrial fibrillation, gender, race and age [4]. It has also been associated with systemic and pulmonary hypertension, atrial fibrillation, gastroesophageal reflux disorder and coronary artery disease. In children, it is associated with attention-deficit hyperactivity disorder, failure to thrive, and poor scholastic performance.

The diagnosis of OSA is typically made by polysomnography (PSG) in a sleep laboratory, although oligosomnography (OSG), also sometimes called “home sleep testing,” is increasingly becoming the first sleep test performed for patients deemed to have a high pre-test probability of OSA.

This article focuses on the treatment of OSA (Table 1). No current treatment for OSA can claim complete efficacy and tolerability. Continuous positive airway pressure (CPAP) is the standard treatment for most patients with OSA and is effective across the severity spectrum, but its effectiveness is often blunted by suboptimal adherence.
Table 1

Current modalities for the treatment of obstructive sleep apnea

Positive airway pressure (PAP) devices

 Continuous PAP

 Bi-level PAP

 Adaptive servo-ventilation (ASV)

Oral appliances

Surgical interventions

 Upper airway procedures, including tracheostomy

 Bariatric interventions

Weight loss

Other modalities

 Nasal expiratory resistance devices

 Oral pressure therapy

 Upper airway muscle training

 Upper airway muscle stimulation

 Positional therapy

 Pharmacotherapy (mostly adjunctive)

 Transnasal insufflation

Diet and lifestyle

Obesity is a strong risk factor for OSA, and weight loss is associated with improvements in indices of sleep-disordered breathing. A recent study of overweight and obese patients with OSA [5] found that at five-year follow-up, patients with at least a 5 % weight loss experienced a significant change in AHI compared with those who did not achieve at least a 5 % weight loss (−3.5 vs +5.0, p = 0.002). Weight loss should therefore be recommended for all patients with OSA who are overweight or obese.

Bariatric surgical procedures are effective treatments for obesity, regardless of the specific type of procedure performed. A recent systematic literature review on obese patients with OSA [6] examined the efficacy of different bariatric interventions in the treatment of OSA. The study found that over 75 % of patients undergoing bariatric intervention achieved some degree of improvement in OSA, with biliopancreatic diversion being the most successful and laparoscopic adjustable gastric banding (LAGB) the least. There was considerable overlap in the percentages of excess weight loss achieved by each type of procedure. A randomized controlled trial [7] of obese patients with OSA compared weight loss and AHI reduction in one group undergoing bariatric surgical therapy for weight loss against a second group undergoing conventional weight-loss therapy. LAGB was the only bariatric procedure included in this study. At two-year follow-up the bariatric patients had lost considerably more weight than those in the conventional weight-loss group (27.8 kg vs 5.1 kg, p < 0.001). Both groups experienced a reduction in AHI as well (bariatric group 25.5, conventional group 14). However, the difference between AHI reduction for each group was not statistically significant (p = 0.18).

Pharmacologic treatment

Various pharmacologic trials have targeted different possible mechanisms of OSA, but no pharmacologic treatment has yet been shown to consistently treat OSA effectively. A review by the American College of Physicians on the treatment of OSA [8] reviewed mirtazapine, xylometazoline, fluticasone, paroxetine, pantoprazole, steroid plus CPAP, acetazolamide, and protriptyline, and concluded that the overall evidence was not sufficient for recommendation of any of these agents.

Zonisamide is associated with weight loss in obese patients. A recent randomized placebo-controlled trial [9] compared zonisamide to CPAP regarding AHI reduction and body weight. At 24-week follow-up, the zonisamide group experienced weight loss while the CPAP group experienced a similar amount of weight gain (−2.7 kg vs +2.3 kg, p < 0.001). Zonisamide reduced AHI, but less than CPAP even after adjusting for compliance.

Patients on CPAP who travel to high altitudes may be unable or unwilling to use CPAP during travel, due to convenience issues including inconsistent availability of electricity. Acetazolamide has been recently studied in patients with OSA normally on long-term CPAP, at high altitude settings, in a randomized, placebo-controlled double-blind trial [10]. In the trial, patients discontinued CPAP at high altitude and received acetazolamide or placebo. The acetazolamide group had lower AHI and higher oxygen saturations compared with placebo, suggesting that patients with OSA who cannot or do not wish to use CPAP at high altitudes may benefit from acetazolamide. Of note, the study period was very brief; each patient received acetazolamide for only two nights.

Interest in serotonergic medications stems from their effects on upper airway musculature. Mirtazapine is an antidepressant that antagonizes alpha-2, 5-HT2 and 5-HT3 receptors, and enhances 5-HT1A-mediated serotonergic transmission [11]. A short, placebo-controlled crossover study [12] showed that mirtazapine significantly reduced AHI at 4.5-mg and 15-mg doses. However, two subsequent randomized, placebo-controlled trials [13, 14] did not replicate the findings of the first study, as the AHI did not improve with the use of mirtazapine in either trial. Mirtazapine was also associated with weight gain, a common and notorious side effect. Given the nondefinitive evidence for benefit in OSA, and the association of weight gain with worsening of OSA, mirtazapine cannot be recommended as a treatment for OSA.

Donepezil is an acetylcholinesterase inhibitor used in the treatment of Alzheimer’s dementia. In a small trial [15], donepezil improved AHI in patients with both Alzheimer’s and OSA, independent of BMI. In a subsequent study with similar sample size [16], this time in patients with OSA but not Alzheimer’s dementia, a statistically significant reduction in AHI and ESS score was seen with donepezil at one-month follow-up, compared with placebo, suggesting that donepezil may confer some benefit to patients with OSA without Alzheimer’s dementia. AHI on baseline polysomnography was 26.4 ± 14.3 for the placebo group and 42.2 ± 19.4 for the donepezil group, a difference that did not reach statistical significance (p = 0.06). The placebo and donepezil groups did not significantly differ in age, BMI, neck circumference, or Epworth scores.

OSA and hypertension are both cardiovascular risk factors, and OSA is a well-established cause of secondary hypertension. A medication or combination of medications that treats both conditions simultaneously would be beneficial in reducing cardiovascular morbidity. A randomized controlled trial is underway [17] comparing chlorthalidone plus amiloride versus amlodipine monotherapy, in patients older than 40 with moderate OSA and stage I hypertension (140 to 159/90 to 99 mmHg). The primary outcomes are AHI and blood pressure improvement.

Surgical treatment

Upper airway surgical interventions for obstructive sleep apnea are listed in Table 2. A recent clinical practice guideline [18] evaluated surgical interventions including UPPP, laser-assisted uvulopalatoplasty, radiofrequency ablation, and various combinations of procedures, and reported that the evidence was insufficient to conclude that the surgical interventions studied are superior to control treatment, as outcomes were not consistent. No surgical intervention was shown to be more effective than CPAP. This guideline did not examine tracheostomy, the most consistently effective surgical intervention for OSA. Tracheostomy is generally considered a last-resort treatment for various reasons including risks of pneumonia, vocal cord paralysis, and aspiration. It also precludes swimming and has an unsightly appearance. A recent meta-analysis of tracheostomy in adult patients with OSA [19] found a statistically significant decrease in the mean apnea index (AI) from 73.0 ± 27.1 to 0.2 ± 1.2, regardless of BMI; some hypopneas persisted although the decrease was still dramatic, with the mean AHI decreasing from 92.0 ± 34.8 to 17.3 ± 20.5. Overall mortality was reduced with tracheostomy compared with no treatment. Some patients developed central apneas postoperatively, a well-described phenomenon, but this essentially normalized after 14 weeks to a central apnea index of 2.1 ± 3.5.
Table 2

Upper airway surgical interventions for obstructive sleep apnea


Uvulopalatopharyngoplasty (UPPP)

Laser-assisted uvulopalatoplasty (LAUP)

Genioglossus advancement and hyoid myotomy

Inferior sagittal mandibular osteomyotomy

Mandibular/maxillomandibular advancement

Nasal operations



Glossectomy/partial glossectomy

Radiofrequency ablation of palate, tongue, or turbinates

Palatal implants

Uvulopalatopharyngoplasty (UPPP) was first described as a surgical option for OSA in 1981 [20]. A study of 54 patients who underwent UPPP [21] found a significant reduction in AHI from 20.39 ± 11.68 to 12.70 ± 13.67 postoperatively, as measured by PSG with a mean time of 18.7 months elapsing between the operation and PSG. Cure was defined as both an ≥50 % AHI reduction and a postoperative AHI <5; 42.6 % of patients were considered cured. Isometric strength of the anterior portion of the tongue, as measured by a dynamometer, was found to be significantly higher in the UPPP responders (p < 0.0001). Various cephalometric variables were measured, and orofacial myofunctional evaluations performed, and there were no statistically significant differences between responders and nonresponders apart from anterior tongue strength. BMI and preoperative AHI also were not predictive of surgical success in this study. Of note, only patients with a Friedman class of I or II, BMI <40, and tonsils graded 2–4 were considered candidates for UPPP and included in this study, and thus the study population excludes a significant portion of patients typically seen in sleep medicine clinics.

In a literature review by a different group [22], UPPP resulted in an overall AHI reduction of 33 %, but with residual AHI remaining elevated (mean of 29.8). This paper also examined the literature on laser-assisted uvulopalatoplasty (LAUP), upper airway radiofrequency ablation, and maxillomandibular advancement (MMA). Nine case series on MMA showed an overall reduction in AHI of 87 % (95 % CI 80–92 %) with a mean postoperative AHI of 7.7. Most of the series reported a residual postoperative AHI of <10. It is important to note that most patients who undergo MMA for OSA comprise a highly select population. Such patients are typically selected based on specific anatomic criteria including hypopharyngeal narrowing and skeletal hypoplasia, which in combination may be expressed as retrognathia. The literature on LAUP and radiofrequency ablation reported inconsistent results; the observational studies slightly favored the surgical procedures, but the randomized controlled trials did not.

In one study which examined all VA inpatient operations over a ten-year period [23], the 30-day mortality rate for UPPP was 0.2 % and the nonfatal serious complication rate was 1.5 %, with no effect of patient age or year of operation on the risk of death or serious complication. A more recent paper on a group of Swedish patients treated with UPPP between 1997 and 2005 [24] reported a serious complication rate of 3.7 % and no deaths in 3,572 patients.

Assistive devices

Positive airway pressure devices

The standard treatment for most patients with OSA is continuous positive airway pressure (CPAP). CPAP functions by pressurizing air to splint a patient’s airway open, preventing collapse and thus facilitating airflow [25]. Although CPAP is effective for most patients with OSA, adherence is often challenging for patients. Potential adverse effects include skin irritation, dry mouth or nasal passages, claustrophobic reactions, and minor epistaxis.

A recent American College of Physicians clinical practice guideline based on a systematic review of decades of literature [18] found CPAP to be superior to no treatment, and superior to sham CPAP, in reductions of AHI. CPAP was superior to no treatment in terms of minimum oxygen saturation (no comparison to sham CPAP was available). Auto-CPAP was similar to fixed CPAP in terms of AHI reduction, and although adherence was statistically significantly higher with auto-CPAP, the difference of 0.19 hours is unlikely to be clinically meaningful. Evidence comparing Bilevel PAP to fixed CPAP was not sufficient to conclude superior efficacy of one over the other, and no differences in adherence were found.

CPAP and weight management

The evidence has been conflicting regarding whether patients with OSA who use CPAP tend to gain or lose weight with PAP usage, with studies generally being retrospective or small. A recent, large six-month randomized, controlled trial [26] of patients with OSA (patients with AHI ≥10 only) compared CPAP and sham CPAP in terms of weight changes, finding that patients who used CPAP actually gained a modest amount of weight (0.35 ± 5.01 kg) while patients on sham CPAP lost weight (0.70 ± 4.03 kg, p = 0.001). These results suggest that overweight and obese patients with OSA who use CPAP should not assume that they will lose weight with CPAP use, and need to adopt or maintain healthy lifestyle choices for weight loss, in addition to using CPAP.

CPAP and hypertension

A recent meta-analysis [27] found that CPAP reduced blood pressure (BP) by statistically significant, but modest amounts compared with various control groups (sham CPAP, pill placebo, or standard care). The difference in diurnal systolic BP (SBP) was −2.58 mm Hg (95 % CI −3.57 to −1.59 mm Hg) and that of diastolic BP (DBP) was −2.01 mm Hg (95 % CI −2.84 to −1.18 mm Hg). A follow-up, patient-level meta-analysis by the same group [28] aimed to identify factors most strongly associated with BP reduction with CPAP. Uncontrolled hypertension in this study was defined as SBP ≥140 and/or DBP ≥90 regardless of medication usage. The study found that patients with uncontrolled hypertension were likely to experience the greatest benefit in BP reduction from CPAP use, after controlling for OSA severity and daytime sleepiness. In a randomized controlled crossover trial [29] comparing CPAP with valsartan in patients with hypertension and OSA (both conditions treatment-naïve), both CPAP and valsartan resulted in statistically significant 24-hour mean blood pressure (MBP) reductions over eight weeks of treatment. However, the reduction with valsartan was significantly greater: when patients with poor CPAP adherence (<3 hours/night) were excluded, the difference in MBP change between the two treatments was −5.94 mmHg in favor of valsartan (p = 0.019).

Adaptive servo-ventilation (ASV)

This is a relatively recent mode of PAP treatment that has been used to treat patients with congestive heart failure (CHF), as these patients often have both central and obstructive events. Such devices work by continuously monitoring airflow, maintaining a constant expiratory pressure to prevent airway obstruction, and varying the inspiratory pressure if necessary to maintain a constant minute ventilation. A small study [30] (n = 11) of male patients with stable CHF and sleep apnea (five with OSA, six with CSA) found that at six-month follow-up, ASV significantly reduced AHI from 49 to 7.6 (p = 0.001), and improved LVEF by +5.7 (versus −4.0 for a comparison group not receiving ASV, p = 0.04). Another small study [31] included 22 patients with CHF and central sleep apnea with Cheyne-Stokes respiration, who were randomized to ASV or no ASV after cardiac resynchronization therapy. This study found a lower rate of cardiac death and rehospitalization after discharge in the ASV group (p > 0.05); the follow-up times varied but were for a mean of 349 days.

Positional therapy

Positional sleep apnea has been described as an AHI ≥5 with supine AHI at least twice that of non-supine AHI [32, 33]. One study of 574 patients with OSA found that 55.9 % had positional sleep apnea, and these patients tended to be younger and thinner than patients with non-positional OSA [34]. Positional therapy involves the use of assistive devices to help a patient avoid the supine position during sleep, as the frequency and severity of obstructive events are exacerbated by supine sleep. The concept was described as early as 1948 [35]. However, there is not much literature on the effectiveness of positional therapy, definitions for positional sleep apnea have not been consistent in the available literature, and many patients have difficulty with long-term adherence to certain positional therapies such as the tennis ball technique [36]. Positional therapy typically causes arousals or awakenings, disrupting sleep architecture and quality. Two recent, small, short-term (3–4 week) studies [37, 38] have investigated the use of novel positional devices that vibrate when the patient assumes a supine position, prompting the patient to move. Both studies found reductions in AHI and high adherence (85 % in one, 92.7 % in the other). Long-term studies on the effects of positional therapy on cardiovascular sequelae have not been attempted [39]. More research is needed to assess the efficacy of positional therapy before it can be strongly recommended.

Oral appliances

The most common type of oral appliance (OA) used for treatment of obstructive sleep apnea is the mandibular advancement device (MAD), which has various commercially available models. The AASM has published practice parameters [40] on the use of OAs for OSA. Potential side effects of MADs include teeth pain, teeth loosening, TMJ discomfort, headaches, dry mouth, and excessive salivation.

Recently, the American College of Physicians [8] has recommended that MADs be considered as an alternative to CPAP in patients who prefer a MAD or do not tolerate CPAP. In the ACP’s review of five studies comparing MADs versus sham, MADs reduced AHI by 14.04 and Epworth Sleepiness Scale (ESS) scores by almost 2 compared with sham therapy. One such study [41] included an MAD group, a sham group, and a no-intervention group and found that sham therapy had no placebo effect.

One study [42] compared MAD to CPAP after a one-night PSG titration of each treatment, in patients with AHIs 10–60 and at least two symptoms of OSAHS. The study found that, as measured on home sleep testing, CPAP was superior to MAD in terms of AHI, snoring, desaturation index, and mean SaO2. There were no significant differences between the two treatments on measures of subjective sleepiness (as measured by the Epworth Sleepiness Scale) or objective sleepiness (as measured by the Oxford Sleep Resistance Test). On the administered health-related quality of life questionnaire, MAD produced significant improvement versus baseline on four of six domains, compared with two for CPAP. A total of 71.2 % of patients reported preference of MAD over CPAP. Patients also reported higher nightly hours of usage with MAD versus CPAP (7.0 vs 6.0, p < 0.001), and usage for more nights (98 % vs 90 %, p < 0.01).

A study of patients with mild to moderate OSA [43] found that at one-year follow-up, nasal CPAP (nCPAP) improved AHI more than MAD, but more patients stopped nCPAP due to side effects. ESS scores gradually improved in both groups and there were no differences in ESS scores between the two groups. At one-year follow-up, the initial improvements in AHI were sustained for both groups. There are currently no randomized controlled trials comparing CPAP to MAD in terms of cardiovascular morbidity [44]. A non-concurrent cohort study [45] of patients with severe OSA found no difference in cardiovascular mortality between patients treated with CPAP and those treated with MAD, even though the residual AHI was higher in the patients treated with MAD (16.3 ± 5.1 vs. 4.5 ± 2.3; P < 0.001). Untreated patients had higher cardiovascular mortality than patients of either treated group. Of note, in that study the patients on MAD had all been previously intolerant of CPAP, and all patients non-adherent to either therapy were excluded from final analysis. Patients were somewhat more adherent to MAD than CPAP, though the overlap was considerable (5.8 ± 1.6 h for CPAP and 6.5 ± 1.2 h for MAD).

There is need for further research on how often patients require follow-up for MAD readjustment or replacement.

Advancements in MAD technology have occurred recently [44]. The ability to electronically monitor MAD adherence, rather than relying on patient self-report, will allow for more accurate comparison of MAD with other modalities. A recent prospective, three-month study [46] of MAD adherence in OSA patients found a mean use of 6.6 ± 1.3 hours per night, and 84 % of patients were “regular users”, defined as using the MAD at least four hours per night on at least 70 % of nights (a common standard used for CPAP adherence as well).

Novel therapies

Transnasal insufflation

Transnasal insufflation (TNI) is the delivery of warm, humidified air through an open nasal cannula at a continuous, high flow rate [47]. One group studied polysomnographic predictors of response to TNI in OSA [48]. The main therapeutic mechanism of TNI is a small increase in end-expiratory pharyngeal pressure, and thus the study hypothesis was that TNI would be more effective in patients with OSA consisting mostly of hypopneas rather than apneas. Prior to study commencement, all subjects had undergone baseline PSG demonstrating an RDI of >5 (22.6 ± 15.6), and a CPAP titration demonstrating a nasal pressure between 8–11 cm H2O. TNI reduced baseline RDI from 22.6 ± 15.6 to 17.2 ± 13.2 (p < 0.01). A therapeutic response, defined as a reduction in RDI to both <10 events/hour and at least 50 % less than baseline, was observed in 27 % of subjects. Higher response rates were seen in patients with higher proportions of hypopneas and respiratory effort-related arousals (RERAs).

Oral pressure therapy

A novel treatment modality, oral pressure therapy (OPT) involves application of a vacuum to the mouth, pulling the soft palate forward and stabilizing the tongue for improved airway patency. Because it has been very recently developed, literature evaluating OPT is limited. Colrain et al. [49] studied OPT in 63 subjects with mild to severe OSA. The subjects were generally male (69.8 %), middle-aged (53.6 ± 8.9 years) and overweight or obese (BMI 32.3 ± 4.5 kg/m2). Subjects were assessed by PSG at the beginning and end of a four-week treatment period, and each subject also had a control PSG (i.e., without treatment). Twenty of 63 subjects experienced a reduction in AHI to ≤10/hr and ≤50 % of control values. Further research is needed on the efficacy of OPT and how it compares with other OSA treatment options.

Nasal expiratory resistance device

The nasal expiratory resistance device, also known as nasal expiratory positive airway pressure (nasal EPAP), was described by Colrain et al. in a 2008 pilot study [50] that found significant reduction in AHI overall, and normalization of AHI in some subjects with mild OSA. A small study [51] on a specific form of this device, Provent®, found statistically significant reductions in AHI, ESS, and snoring, and a small but statistically significant increase in mean oxygen saturation. At one-month follow-up, 41 % of subjects had at least a 50 % AHI reduction compared to baseline.

A larger, sham-controlled study [52] of nasal EPAP in patients with pre-study AHI of ≥10 per hour found a significant decrease in AHI with nasal EPAP versus sham therapy (42.7 % vs 10.1 %, p < 0.0001) at three-month follow-up. ESS scores in the nasal EPAP group were reduced from 9.9 at baseline to 7.2 at three months (p < 0.0001). This study defined treatment success as a ≥50 % AHI reduction, or an AHI reduced to <10 (for subjects with a baseline AHI of ≥10). At three-month follow-up, treatment success was seen in 50.7 % of patients on nasal EPAP and 22.4 % of patients on sham therapy (p = 0.001). Of note, both this and the aforementioned study used patient report to determine adherence; there were no objective measures of adherence in either study.

Upper airway muscle training/stimulation

Several studies have explored the effects of increasing upper airway muscle tone as a treatment for obstructive sleep apnea, whether by exercise training or by direct electrical stimulation. All studies found significant improvements in AHI, though normalization of AHI did not occur with any of the therapies.

A recent prospective 12-month study in the New England Journal of Medicine [53] investigated upper airway stimulation in patients with moderate to severe OSA, using surgically implanted upper airway stimulator devices. The co-primary outcomes were responses in AHI and oxygen desaturation index (ODI) scores. An AHI response was defined as an AHI score ≥50 % lower than baseline as measured by polysomnography at two, six, and 12 months, and an AHI of <20 at 12-month follow-up. ODI was measured using the criterion of ≥4 % desaturation from baseline, and a response was defined as ≥25 % reduction in ODI from baseline. At 12 months, the median AHI score decreased from 29.3 to 7.4 (p < 0.001) and the ODI score decreased from 25.4 to 7.4 (p < 0.001). Subjects showing a response at 12 months were then randomized to either therapy maintenance or withdrawal. The AHI in the maintenance group remained similar (7.2 to 8.9), while the AHI in the withdrawal group rose significantly (7.6 to 25.8). The difference in the changes in mean scores was 16.4 (p < 0.001). A similar effect occurred with the ODI.

One study [54] examined the use of exercises developed by speech pathologists in an attempt to strengthen the upper airway muscles. These oropharyngeal exercises targeted the soft palate, tongue, and facial muscles. In addition, “stomatognathic” function exercises were performed daily, with once-weekly visits with a speech pathologist. This group was compared with a control group that performed sham exercises. Sleep studies, anthropomorphics, and questionnaires were administered at baseline and after three months. The exercise group had a reduction in both neck circumference and OSA severity (AHI, 22.4 ± 4.8 vs 13.7 ± 8.5) when compared with the control subjects, who showed no significant change.

An older study explored the effects of didgeridoo playing on moderate OSA [55]. A didgeridoo is a long, thin wind instrument that requires circular breathing techniques, and complex interactions between the lips and vocal tract, so that vibrations in the upper airway are more readily transmitted to the lower airways. Twenty-five patients with moderate OSA learned to play the didgeridoo over four months, and the AHIs of this group decreased significantly versus a control group, though normalization of the AHI did not occur.

Pediatric considerations

The major factor in the development of OSA in children is adenotonsillar hypertrophy, and thus adenotonsillectomy (AT) has been the generally accepted standard of treatment for pediatric obstructive sleep apnea. The recent Childhood Adenotonsillectomy Trial (CHAT) was a large randomized, controlled trial of early AT (defined as operation within four weeks after randomization) versus watchful waiting for OSA [56]. The trial included patients aged 5–9. Children with a BMI z-score of >3 (i.e., very obese children) were excluded, as were children with recurrent tonsillitis and children on medication for attention-deficit hyperactivity disorder. Baseline AHI was similar in the two groups. At seven-month follow-up, the AHI decreased by 3.5 in the AT group versus 1.6 in the watchful waiting group (p < 0.001). Normalization of the OSA syndrome was defined as both a reduction of the AHI to <2 and the obstructive apnea index (OAI) to <1, and this occurred in 79 % of the AT group versus 46 % in the watchful waiting group (p < 0.001). The authors considered that the normalization in the watchful waiting group may have been related to lymphoid tissue regression, airway growth, routine medical care or regression to the mean. In addition to the improvements in polysomnographic parameters, scores on inventories assessing sleep-related breathing disorder symptoms, behavior, and quality of life were significantly improved in the AT group compared to the watchful waiting group. The primary outcome was attention and executive function as measured by the Developmental Neuropsychological Assessment, and unlike the above findings, the primary outcome did not differ between the two groups at seven months.

In a 12-week double-blind, placebo-controlled study [57] of children with polysomnographically diagnosed nonsevere OSA (defined as AHI < 10), oral montelukast significantly reduced the obstructive apnea index (AI) compared with placebo (3.9 ± 1.6 to 1.7 ± 1.0, p < 0.01). The montelukast group also experienced a reduction in the obstructive AHI, but the results were just short of statistical significance (6.0 ± 3.2 to 3.6 ± 2.3, p = 0.07). The placebo group had no significant changes in obstructive AI or AHI. Of note, children with a history of tonsillectomy and adenoidectomy were excluded, as were obese children (obesity defined as BMI >95th percentile) and children with craniofacial, neuromuscular, syndromic, or defined genetic abnormalities.

A randomized, double-blind, crossover trial of intranasal budesonide for six weeks [58] found significant improvements in AHI and adenoid size compared to placebo in children with mild OSAS, and the benefits persisted for at least eight weeks after therapy cessation. Children older than six and younger than 12 who had not had adenotonsillectomy in the preceding year were included. Mild OSAS was defined as an AHI >2 but ≤7, or an AHI ≤2 but with a respiratory arousal index of ≥2 and oxygen saturation nadir >85 %. This appears to be the most recent study of an intranasal steroid for treatment of OSA in children, though a study shortly afterward found that corticosteroids suppress in-vitro proliferation of adenoid and tonsillar tissue [59].

Compliance with Ethics Guidelines

Conflict of Interest David Young and Nancy Collop declare that they have no conflicts of interest.

Human and Animal Rights and Informed Consent This article does not contain any studies with human or animal subjects performed by any of the authors.

Copyright information

© Springer Science+Business Media New York 2014