A 54-Year-Old Man With Obstructive Sleep Apnea

DR BURNS: Mr J is a 54-year-old man with a history of sleep apnea. He lives in a suburb of Boston with his wife and 2 children, owns a small company, and has managed care insurance.

In 1995, Mr J's wife noted that he stopped breathing at night. Mr J also recalled that he awoke, on occasion, in the middle of the night with palpitations. He mentioned this to his primary care physician, Dr M, who referred him for a sleep study. The study revealed 206 obstructive events, giving an apnea hypopnea index (AHI) of 36 per hour. He had oxygen desaturation to a nadir of 74%. Based on these results, his pulmonologist recommended that Mr J proceed with nasal continuous positive airway pressure (CPAP) titration. The results of his CPAP study showed that 5 to 6 cm of water pressure eliminated the obstructive events and oxygen desaturation. Mr J seemed to have some difficulty falling asleep with the mask, but overall it appeared to be well tolerated during both rapid eye movement (REM) and non-REM sleep.

With the nasal CPAP, Mr J's palpitations ceased, but after about a year he discontinued the nasal CPAP because it interfered with his ability to fall asleep and with his sex life. A trial of clonazepam, 0.25 mg, about a half hour before sleep did not improve his ability to sleep with the CPAP. When he pursued the use of an oral device, Mr J was told that the most effective agent for treatment of sleep apnea was nasal CPAP, but he did not resume its use.

Currently, Mr J states that he feels well. He falls asleep without difficulty and is able to sleep about 8 hours per night. He does not experience shortness of breath and he does not perceive any difficulties with his sleep patterns or daytime sleepiness.

Mr J has a medical history of coronary artery disease (with stent placement), hypertension, hyperlipidemia, diverticulitis, recurrent sinusitis, and L5-S1 radiculopathy. Because of his history of coronary artery disease and hypertension, Dr M strongly encouraged Mr J to use his nasal CPAP. Mr J stated that he could not tolerate CPAP, but he did lose 20 lb and now weighs 205 lb (his body mass index [BMI] is now 28.6). Mr J's current medications include aspirin, atenolol, atorvastatin, gabapentin, and ibuprofen. He does not smoke and rarely drinks.

Dr M and Mr J wonder whether Mr J still has sleep apnea and what evaluation and treatment they should consider at this time.

The first time that it came up was when my wife was ill. She is a sound sleeper but she was awake during the night and noticed I was not breathing as she thought I should.

Sometimes, I would wake up feeling as if I was having a panic attack or something was wrong—my heart would be racing and I didn't know why. But it wasn't frequent enough that I ever put the 2 together prior to the sleep study. I went for one of the overnight sleeping tests. It turned out that in some cases I was borderline and in some I was way over.

The first recommendation was, of course, lose weight and exercise. That wasn't very immediate, so they suggested I try CPAP. The full mask was very uncomfortable—it covers too much of your face, and I never could get one that fit exactly right. They tried fitting me with different sizes and different styles, but it turned out the nasal pillow mask worked better for me than the full mask did.

When I was using the nasal CPAP, I don't remember any time awakening or having that heart-racing condition. I did have more dreams than I did without it. I stopped using it about 2 years ago, because with it on I couldn't sleep comfortably. In the meantime, I had, as I should have done earlier, lost 20 lb. I don't think that the sleep apnea is as bad or occurs as frequently, or at least my wife hasn't mentioned it or poked me and said, "You're not breathing." Having no symptoms that really bother me—I mean, they may be destroying my body but I don't know it—it's not really an issue. If I had symptoms, I think I'd be a lot more inclined to use [the CPAP]. It seems like somebody could design a better mask. The whole headgear and everything is so intrusive, I think, that it probably discourages some people from using it.

Mr J was treated with CPAP after his study showed that he had sleep apnea, and initially he seemed to feel that he was a little more energized, feeling better, and at least for a while when I asked him on his visits he told me that he was wearing it more often than not. Recently, he told me that he had stopped using the CPAP because he found it so uncomfortable and so disruptive to his ability to initiate sleep and to his sex life.

Mr J has the classic cardiac risk factors. He should benefit from adequate treatment of his sleep apnea, and I had hoped that we could find a way to treat him that would be both palatable to him and effective.

Are there ways that nasal CPAP could be better tolerated by this patient? If not, are there alternative treatments such as surgery or medication that we might offer him? And is it necessary or recommended to repeat his sleep study now that he has lost weight?

What are the definition, epidemiology, and underlying pathophysiology of sleep apnea? How do patients present and what are common examination findings? Who should be screened and how? What treatment options are available? What is the evidence that intervening improves the quality or length of life? What does the future hold? What do you suggest for Mr J?

DR KUNA: Mr J has obstructive sleep apnea (OSA) and was unable to tolerate nasal CPAP. He subsequently lost weight and now wonders whether he still has sleep apnea and, if so, whether he should pursue treatment.

Mr J's history is complicated by 2 factors: his lack of subjective daytime sleepiness and his history of hypertension and cardiovascular disease. Mr J is at a crossroads in his medical care because medicine is at the threshold of answering 2 fundamental questions regarding OSA: who really has the disorder and who should be treated?

Obstructive sleep apnea is characterized by frequent cessation of inspiratory airflow during sleep due to closure of the pharyngeal airway despite the persistence of respiratory efforts. These repetitive abnormal respiratory events can be associated with transient oxygen desaturations and arousals that disturb the continuity of sleep and can lead to excessive daytime hypersomnolence.

To diagnose OSA, current standard-of-care guidelines recommend in-laboratory polysomnography.^{1 - 3} A polysomnogram quantifies sleep stages, arousals, abnormal respiratory events during sleep, oxygen saturation, heart rate, body position, and leg movements. The polysomnographic metric used to define OSA is the AHI, the average number of apnea and hypopnea events per hour of sleep.³ An obstructive apnea is defined as a cessation of airflow on repetitive inspiratory efforts for at least 10 seconds due to complete closure of the pharyngeal airway. A hypopnea is defined as an abnormal decrease in inspiratory airflow for at least 10 seconds due to partial closure of the airway associated with transient oxygen desaturation and/or arousal. Current severity criteria identify mild OSA as an AHI between 5 and 15, moderate OSA as an AHI between 15 to 30, and severe OSA as an AHI greater than 30 events per hour.³ Using these criteria, Mr J's AHI of 36 events per hour establishes the diagnosis of severe OSA. Mr J consistently denies daytime sleepiness and therefore does not have OSA syndrome, ie, an AHI of at least 5 with symptoms of excessive daytime sleepiness.

Some caveats need to be considered regarding these definitions. First, controversy exists as to whether the AHI is the best metric to define OSA. In general, AHI is poorly correlated with both objective and subjective measures of daytime hypersomnolence.^{4 - 10} A second caveat is that AHI is dependent on the definition and measurement of hypopneas. Varying definitions of hypopnea in the scientific literature can lead to a 10-fold difference in AHI results.¹¹ To address this problem, recent efforts have been made to improve the techniques used to measure airflow during the polysomnogram and standardize the definition of hypopnea.^{3 ,12 - 13} The reproducibility of results of a first-night polysomnogram is another consideration but is less likely to be a problem in patients such as Mr J with a relatively high AHI.^{14 - 16}

Mr J is not alone—OSA is a relatively common problem. In the United States, OSA is more prevalent than asthma and as prevalent as diabetes mellitus. The cross-sectional Wisconsin Sleep Cohort Study of 602 employed adults aged between 30 and 60 years found that 24% of men and 9% of women have OSA as defined by an AHI of at least 5 events per hour.^{17 - 18} The estimated prevalence of mild to moderate OSA is 12% to 32%.^{17 ,19 - 20} Even if OSA is defined as an AHI of 15 or more, it is estimated to affect 9% of middle-aged men and 4% of middle-aged women.¹⁷ Obstructive sleep apnea syndrome affects about 4% of men and 2% of women.^{17 ,21 - 23} The prevalence of OSA in women increases following menopause.²⁴ Obstructive sleep apnea may become even more common as obesity becomes more prevalent. However, a similar prevalence of OSA has been reported in an Asian population, in which obesity is relatively uncommon.²⁵

Mr J also has hypertension and coronary artery disease, conditions that are associated with OSA. Results of the Sleep Heart Health Study, a multicenter cross-sectional study of more than 6000 men and women (aged >40 years; mean age, 64 years), indicate that OSA is a risk factor for systemic hypertension independent of obesity (odds ratio [OR], 1.37; 95% confidence interval [CI], 1.03-1.83) for those with an AHI of 30 events per hour or higher vs fewer than 1.5 events per hour.²⁶ Participants with an AHI greater than 11 events per hour had a 42% (95% CI, 13%-78%) greater odds of prevalent, self-reported cardiovascular disease (ie, coronary heart disease, stroke, and congestive heart failure) than participants with an AHI of 0 to 1.3 events per hour.²⁷ Risk of cardiovascular disease was independent of BMI and hypertension and occurred even within the lowest AHI quartile (1-10 respiratory events per hour). Patients with OSA are also at increased risk for motor vehicle collisions and insulin resistance.^{28 - 31} After adjusting for BMI and percentage of body fat, an AHI of 5 or more events per hour in middle-aged men is associated with an increased risk of having impaired or diabetic glucose tolerance (OR, 2.15; 95% CI, 1.05-4.38).²⁹ In the Wisconsin Sleep Cohort Study, adults with an AHI greater than 15 (vs no sleep-disordered breathing) were significantly more likely to have multiple motor vehicle collisions in 5 years (OR, 7.3; 95% CI, 1.8->25).³¹ Accident occurrence was not predicted by either objective or subjective measures of daytime sleepiness.

Patency of the pharyngeal airway during inspiration is determined by the balance between 2 counteracting factors: a dilating force exerted by activation of pharyngeal muscles surrounding the airway and a collapsing force due to subatmospheric intraluminal pressure transmitted into the pharyngeal airway from the thoracic cavity.^{32 - 33} The sleep state is associated with a reduction in motor output to the pharyngeal dilating muscles.³⁴ While the airway normally remains open during both wakefulness and sleep, in patients with OSA, imbalances in these forces during sleep favor airway closure. This imbalance appears to arise primarily from anatomic factors that increase airway collapsibility.³⁵ Although there is a poor understanding of how variations in pharyngeal anatomy increase airway collapsibility, pharyngeal airway closure during sleep is associated with conditions that decrease airway size, eg, obesity, retrognathia, and adenotonsillar hypertrophy.^{36 - 38} The latter problem is the most common cause of OSA in children but is far less common in adults.³⁹ The pharyngeal airway also becomes narrower when patients sleep in the supine position due to gravitational forces pulling the soft palate and base of the tongue into the pharyngeal airway. The "balance of forces" theory models pharyngeal airway collapse during static conditions. Another model characterizes the pharyngeal airway under dynamic conditions as a Starling resistor.^{40 - 41} Under dynamic conditions, nasal airway restriction due to nasal congestion, deviated nasal septum, nasal polyps, or other factors predisposes to snoring and OSA by increasing transmission of intrathoracic subatmospheric pressure during inspiration into the collapsible pharyngeal airway.

By the time a patient like Mr J seeks medical attention for OSA, symptoms have typically been present for years. Patients usually present with snoring, waking with gasping and choking, and awakening several times a night.⁴² The bed partner may provide valuable information regarding snoring, snorting, and gasping, but less commonly cessation of breathing. Other presenting symptoms include nocturia and excessive daytime sleepiness. The latter is particularly likely to prompt a referral when the patient has difficulty remaining awake at work or while driving. Mr J noted nocturnal palpitations, which may have been precipitated by oxygen desaturation and/or increased sympathetic activity associated with abnormal respiratory events during sleep. Arrhythmias are not common nor are they frequently observed in patients with OSA during polysomnography.⁴³

The Epworth Sleepiness Scale, an 8-item self-administered questionnaire in which patients rate how likely they are to doze during sedentary activities, is a convenient way to assess subjective daytime sleepiness.⁴⁴ We are not provided with any Epworth scores on Mr J, but even if they are within accepted limits of normal (≤10), one still wonders if he is truly asymptomatic. Patients with longstanding sleepiness may be less likely to recognize this symptom, as supported by preliminary results that report a discordance between subjective and objective measures of daytime sleepiness in 29% of 107 patients with OSA (mean AHI, 66 events per hour).⁴⁵ Alternatively, only 11% of those patients with an Epworth score of 10 or less were found to be objectively sleepy.

The absence of daytime symptoms in patients like Mr J may arise from a lower vulnerability to sleep disruption. Studies indicate that sleep deprivation can cause severe daytime sleepiness and functional impairment in some individuals while only minimally affecting others.^{46 - 47} Whether this differential vulnerability extends to cardiovascular and other consequences of OSA is unknown.

Patients are usually at their maximum weight at initial presentation, but otherwise the physical examination findings are often unremarkable. Examining the oral cavity and oropharyngeal airway may reveal enlarged tonsils, enlarged tongue, elongated soft palate, recessed mandible, micrognathia, and/or a small oropharyngeal airway.^{36 - 37} When possible, the nasal passage should be inspected for evidence of nasal septal deviation, nasal polyps, turbinate hypertrophy, and congestion. However, normal findings of upper airway examination do not rule out OSA. Obstructive sleep apnea is frequently associated with increased neck circumference.⁴⁸ A neck circumference greater than 17 inches in men and 16 inches in women increases the likelihood of OSA. A small minority of patients with severe OSA present with daytime sleepiness and signs of right-sided congestive heart failure, the so-called Pickwickian syndrome.⁴⁹

Obstructive sleep apnea is strongly related to obesity, an important public health problem in its own right. Studies have identified BMI as the strongest predictor of AHI.^{50 - 51} Data from the Wisconsin Sleep Cohort Study reveal that a 1-SD increase in BMI is associated with a 4-fold increase in the risk of having an AHI of 5 or higher.¹⁷ A subsequent 4-year follow-up of this community sample revealed that a 10% weight gain predicted a 32% increase in AHI relative to stable weight and a 6-fold increase in the risk of developing an AHI of 15 or higher.⁵² The link between obesity and sleep-disordered breathing is further supported by data showing a significant decrease in AHI in patients who lose weight.^{52 - 55} Wisconsin Sleep Cohort Study participants who reduced body weight by 10% experienced a 26% reduction in AHI.⁵² In one of the only studies that randomly assigned obese patients with OSA to weight loss (n = 15) or no treatment (n = 8) conditions, a 10% weight loss was associated with a 50% reduction in AHI.⁵⁵ Additional research is needed to determine the relationship between progressive weight loss and AHI, whether sleep apnea resolves with weight loss, and the long-term effect of weight loss treatment on AHI.

While clinicians tend to associate OSA with obesity, patients who are not overweight can also have OSA. When obtaining a medical history, physicians should inquire about factors known to increase the likelihood of OSA: symptoms of nasal airway restriction, alcohol consumption, sedative/hypnotic use, and family history of snoring or OSA.^{56 - 59} The familial predisposition to snoring and sleep apnea may be due to genetically determined differences in pharyngeal airway size. Alcohol and sedatives/hypnotics exacerbate OSA and can convert a person with primary snoring into a patient with OSA.⁵⁷ These agents increase airway collapsibility by differentially suppressing motor output to the pharyngeal dilating muscles relative to respiratory pump muscles.^{58 ,60}

Given the prevalence of OSA, the review of systems in a routine medical history should include sleep-related questions concerning the presence of snoring, daytime hypersomnolence, waking up with snorting/gasping, cessations of breathing during sleep, awakenings during the night, and nocturia. In general, a polysomnogram may be useful if a patient reports snoring and 1 or more of the other symptoms. Simple, widely applicable clinical surveys have been designed to screen patients who are asymptomatic but at risk for OSA, such as those who are obese or hypertensive. However, these are primarily research tools and have not yet been applied generally to clinical decision making.^{50 ,61 - 62}

There are no standardized clinical pathways to guide decision making in selecting among the various treatment options for OSA. Treatment should be tailored to the individual patient. Decisions concerning management should begin with a careful review of the polysomnogram results. In Mr J's case, if his abnormal respiratory events were confined to the supine position, measures could be taken to keep him in the lateral recumbent and prone positions during sleep, eg, by securing tennis balls in the back of a night shirt. Additionally, given his symptoms of palpitations one would also want to know whether Mr J had any cardiac arrhythmia during his polysomnogram and, if so, the type of arrhythmia and the oxygen saturation level at the time of the cardiac event.

All overweight patients with OSA should receive counseling about weight loss. The symptomatic improvement that Mr J experienced is very likely due to his ability to lose weight. Mr J also has a history of recurrent sinusitis. In patients with restricted nasal breathing, appropriate measures should be instituted to treat the underlying cause(s). In addition, patients should be advised not to consume alcohol or take sedatives/hypnotics within 4 hours of going to sleep. Beyond these conservative measures, 3 treatments are available for OSA: nasal CPAP, oral mandibular advancement appliance, and pharyngeal/mandibulofacial surgery. Each therapy has its advantages and disadvantages.

Mr J was prescribed nasal CPAP, which prevents airway closure by acting as a pneumatic splint.^{63 - 69} As with most patients, Mr J's trial of nasal CPAP in the sleep laboratory was successful in eliminating the apneas and hypopneas and he appeared to tolerate the treatment. However, at home he was unable to adhere to this treatment, which he described as intrusive and disruptive. When CPAP adherence is measured objectively, studies have failed to consistently identify predictors of CPAP adherence, but some studies offer suggestive evidence.⁷⁰ One study found that patients with an AHI of 15 or higher are more likely to adhere to treatment compared with patients with an AHI of less than 15 (hazard ratio, 2.48; 95% CI, 1.79-3.46).⁷¹ Patients like Mr J who are prompted by a bed partner to seek evaluation are less likely to adhere to CPAP than self-referred patients. In a group of 80 patients, mean hours of CPAP use per night at 3 and 6 months of therapy was about 1 hour less in partner-referred patients than that in self-referrals.⁷² In addition, the ability to adhere to CPAP treatment has been related to the presence of daytime hypersomnolence, which Mr J lacked. Patients with an Epworth Sleepiness Scale score higher than 10 (maximum score, 24) are more likely to adhere to treatment than patients with a score of 10 or less (hazard ratio, 1.92; 95% CI, 1.41-2.61).⁷¹ Daytime sleepiness is one of the first symptoms to improve with treatment and returns when the treatment is withdrawn.⁷³ This positive feedback to an otherwise sleepy patient reinforces treatment adherence.

Adverse effects of CPAP are primarily related to discomfort from the mask, mucosal dryness, and skin bruising by the mask interface.⁷⁴ While some patients are able to use CPAP whenever they sleep, average CPAP use in patients with OSA is often disappointing.^{75 - 76} I applaud Mr J's honesty in informing his physician that he had stopped using his CPAP. In general, self-reported CPAP use is unreliable. Most current CPAP machines have the ability to store in memory the pressure in the circuit over time. A download of these data allows objective assessment of patient adherence. Of note, long-term adherence to CPAP usually can be predicted based on the first several days of use.⁷⁷ Trained health care professionals in a sleep center or from a home health care company can improve long-term adherence by educating the patient about OSA and CPAP use and providing support prior to and during initiation of treatment.⁷² Patients such as Mr J who have difficulty using CPAP should be questioned carefully regarding why they cannot tolerate the treatment. A comfortably fitting mask is essential for adequate adherence. In-line humidification can also improve adherence by helping prevent mucosal dryness and nasal congestion.⁷⁸ Studies comparing patient adherence to CPAP vs either bilevel positive airway pressure (noninvasive ventilation with positive end expiratory pressure) or autoCPAP (automatically adjusting positive airway pressure) have not yielded consistent results.^{79 - 82}

The oral mandibular advancement appliance for treatment of OSA is becoming more widely used.^{83 - 89} The oral appliances are worn during sleep and pull the mandible forward, thereby dilating, and perhaps stiffening, the pharyngeal airway. Oral appliance treatment needs to be managed by a dentist working in concert with a sleep physician. The mandible is advanced gradually to avoid temporomandibular joint pain, the primary adverse effect of oral device treatment.⁸⁷ As a result, it can take weeks to months to achieve the desired amount of advancement, an unacceptable delay in patients with excessive daytime sleepiness.⁸⁹

The most common surgical procedure for OSA is the uvulopalatopharyngoplasty, a resection of the uvula, rim of the soft palate, tonsils, adenoids, and lateral pharyngeal wall.^{90 - 91} Tracheostomy and surgical advancement of the mandible are other surgical options.⁹² While oral devices and uvulopalatopharyngoplasty can be effective treatments, the overall response to these treatments is disappointing. Studies consistently report that these treatments reduce the AHI less than 5 to 10 events per hour in less than 50% of patients.^{83 - 90} Of the 2 treatments, recent randomized studies report that the oral appliance has a higher success rate, defined as either an AHI of less than 10 or a 50% reduction in pretreatment respiratory indices.^{84 ,88} In contrast, CPAP when set at the proper pressure and used consistently eliminates the upper airway obstruction in almost all patients. Unfortunately, it currently is not possible to identify which patients will have a favorable outcome to these treatment alternatives prior to intervention.^{90 ,92} While surgical and pharmacologic treatments of a restricted nasal airway rarely eliminate OSA, the resulting decrease in nasal airway resistance makes it more likely that treatments such as weight loss, CPAP, oral device use, or upper airway surgery will be successful.^{93 - 98}

Despite Mr J's negative experience with CPAP treatment, most placebo-controlled randomized trials performed in patients with moderate to severe OSA indicate that CPAP treatment improves quality of life, daytime sleepiness, nocturnal symptoms, and daytime function.^{63 ,65 - 69 ,99} Several studies also show improvements in objective sleepiness and tests of cognitive function.^{66 - 67} However, unlike Mr J, the patients recruited into all but one of these studies were subjectively sleepy, based on mean Epworth scores higher than 10. The one small study (n = 54) that recruited patients like Mr J, ie, nonsleepy individuals with severe OSA (AHI >30, Epworth <10), failed to find improvements in quality of life, objective sleepiness, arterial blood pressure, or performance on tests of neuropsychological function.¹⁰⁰ While this study had limited power to detect a difference, these findings highlight the difficulty of attempting to treat nonsleepy patients with severe OSA such as Mr J.

The decision to treat Mr J at the time of presentation with severe OSA and nocturnal palpitations was clear-cut. Now, following his weight loss, it is likely that his AHI has decreased and his oxygen saturation during sleep has improved.^{52 - 55} Data from the Wisconsin Cohort Study predict that Mr J's 8.5% weight loss from baseline would result in about a 23% reduction in AHI, bringing it down from 36 to 27 events per hour, ie, moderate OSA.⁵² Even if his AHI were now less than 15, he would still be diagnosed as having OSA, despite his lack of daytime symptoms. For CPAP treatment to be covered by insurance, current Centers for Medicare and Medicaid guidelines require that patients with an AHI between 5 and 15 have daytime symptoms or cardiovascular disease.¹⁰¹

The decision to treat an asymptomatic patient with mild to moderate OSA is particularly controversial. Five randomized controlled trials in patients with mild to moderate OSA showed some improvement with CPAP treatment in subjective nocturnal symptoms and quality of life measures, but no objective improvements in sleepiness or arterial blood pressure.^{4 ,102 - 105} Although the ability of CPAP treatment to lower daytime blood pressure in most patients with OSA is uncertain, a placebo-controlled study in patients with more severe OSA showed that CPAP use reduced 24-hour mean blood pressure by 3 to 4 mm Hg, with the greatest reductions (mean, 7.9 mm Hg) occurring in patients treated for coexisting hypertension.^{18 ,106} In addition, Peker et al¹⁰⁷ recently reported that patients with relatively poor adherence to CPAP treatment have an 11-fold increase in risk of cardiovascular disease, independent of age, BMI, and blood pressure, whereas effective treatment of OSA significantly reduced this excess risk. At present, it is still unknown whether treatment decreases the risk of cardiovascular disease in all patients with OSA or only a subset, eg, those with more severe disease. Mr J's cardiovascular risk can probably best be reduced through weight reduction, exercise, and pharmacologic control of his hyperlipidemia and hypertension.

If future clinical trials confirm that treatment of OSA reduces systemic blood pressure and the risk of cardiovascular disease even in asymptomatic patients with OSA, we will then have much stronger justification for initiating treatment in asymptomatic patients like Mr J.¹⁰⁸ However, the need to diagnose and treat the large number of asymptomatic patients with mild to moderate OSA with our current "gold standard" diagnostic tests and available treatments would impose a significant burden on existing health care system resources.

Recent technological advances in the development of relatively inexpensive portable monitors to record breathing during sleep should be used to develop innovative, cost-efficient approaches to diagnose patients with OSA.¹⁰⁹ Another approach will be to establish risk-stratification strategies that allow patients to be prioritized according to their risk of having OSA.^{48 ,61 ,110} Primary care physicians could incorporate these algorithms into routine clinical evaluations.

At his current weight, Mr J no longer wakes up with palpitations and continues to deny daytime sleepiness. I recommend repeating the polysomnogram, but preferably after Mr J has lost additional weight, perhaps by enrolling in a structured weight reduction program. If the repeat polysomnogram reveals that Mr J still has OSA, treatment with an oral mandibular advancement device is an option. Many patients like Mr J who are unable to tolerate nasal CPAP are able to wear an oral device to treat their OSA.⁸⁶ However, current evidence is insufficient to justify treating asymptomatic patients with mild to moderate OSA.

A PHYSICIAN: Why do we need sleep labs? It seems that every patient I refer for a sleep study turns out to have sleep apnea. Why shouldn't we just offer these patients nasal treatments and appliances?

DR KUNA: We certainly need novel, innovative, clinical pathways that will allow primary care physicians to treat many of these patients in the office rather than referring them to a sleep laboratory. The field is not there yet, but we are moving in that direction. Clinical pathways such as portable monitoring devices and portable CPAP titration devices have not yet been adequately validated to distribute them widely. However, we are starting to test these methods so that we can eventually offer them to patients.

A PHYSICIAN: These patients have periodic oscillations in their heart rate, at a fairly low frequency, that correspond to the periodic cycles of OSA. Could you comment on the attempt to develop new ways to detect sleep apnea using the electrocardiogram and the Holter monitor?

DR KUNA: To get testing out of the laboratory and into the home, we need very simple methods to give monitors to patients to take home. These types of measurements, such as measuring heart rate variability, hold great promise in allowing us to have a screening test that we can use to identify those people.

A PHYSICIAN: Without using the patients as their own controls, how can you have any idea that the intervention really works at all?

DR KUNA: The randomized controlled trials have 2 designs for the placebo. One is an oral placebo pill, but the most recent placebo has been sham CPAP.^{102 ,111} It has exactly the same circuit and machine, but delivers a suboptimal pressure. The best study design is to bring patients into the sleep laboratory, test them on the sham CPAP, and demonstrate that the AHI is the same as baseline. These sham CPAP devices can be fitted with monitors that track adherence. Randomized controlled trials show similar adherence to sham CPAP and effective CPAP.^{4 ,66 - 67 ,69 ,112 - 113} These studies show that placebo control is necessary because subjective measures improve significantly on the sham intervention. However, a placebo-controlled trial to examine the effects of treatment on cardiovascular consequences would be unethical due to the length of time required. An intriguing recent report by Peker et al¹⁰⁷ suggests an alternative study design based on the fact that patients have very different adherence to nasal CPAP. In this design, all patients are treated and objectively documented CPAP adherence becomes one of the variables in the multiple logistic regression model.

A PHYSICIAN: Given how central obesity is a risk factor for OSA and how crucial and proven effective weight loss can be in terms of improving a patient's AHI, shouldn't we really focus on the issue of patients' massive obesity?

DR KUNA: We need national initiatives to address the epidemic of overweight and obesity in the US population. Body mass index is the strongest predictor of OSA and epidemiological studies clearly demonstrate significant reductions in AHI with weight loss.^{50 ,52} We do need treatments such as CPAP and oral mandibular advancement appliances to acutely manage symptomatic patients with OSA. We need effective upper airway surgeries to correct upper airway anatomic abnormalities and provide symptomatic patients with treatment alternatives. However, if future studies demonstrate that treatment of OSA reduces cardiovascular risk, we will then have justification to treat the very large number of asymptomatic patients with OSA. This huge influx of patients will overwhelm already overburdened resources. To address this looming problem and reduce the prevalence of OSA, we must develop effective strategies to help people lose and control their body weight. Such a logical approach would also lead to improvements in diabetes, hypertension, lipid levels, and osteoarthritis. It is the obvious solution, but it is just not happening.