Weight Loss in the Management of Obstructive Sleep Apnea
Understanding the relationship between weight loss and obstructive sleep apnea (OSA) begins with a central understanding of the pathogenicity of weight gain on OSA. This article first focuses on obesity as a risk factor for OSA and the potential mechanisms that underscore obesity to the pathophysiology of OSA and, finally, summarizes the types of weight loss strategies, including surgical, pharmaceutical, and lifestyle interventions and their effects on OSA.
OSA is estimated to affect 2% to 4% of the general population. Risk factors for OSA include family history, male gender, age, and obesity. In 2009 to 2010, more than 78 million adults and more than 12 million children in the United States were obese. Men, older women, and adolescents represent populations with the highest rates. As the rates of obesity climb, so will rates of its associated diseases, specifically, diabetes, hypertension, and OSA. Most patients with OSA are obese. Data from the Wisconsin Sleep Cohort study, a large population-based study aimed to determine the natural history of cardiopulmonary effects of sleep-disordered breathing and its prevalence, indicate that a one standard deviation increase in the body mass index (BMI) is associated with a 4-fold increase in risk for sleepdisordered breathing and that each percentage change in weight is associated with an approximate mean change in the apnea hypopnea index (AHI) of 3%. Hence, a person who gains 5% of body weight would be expected to increase his/ her AHI by 15% compared with an individual whose weight remains stable. The greatest numbers of OSA are in patients presenting for bariatric surgery and a BMI greater than 35 kg/m2 . It is estimated that 2 of 3 women and greater than 95% of men who present for bariatric surgery have an AHI more than 10 events per hour. Although obesity is a known risk factor for OSA, its exact mechanism is less clear. Obesity likely modulates upper airway control through mechanical effects on structure and function, ventilation, and hypoxemia. OSA is characterized by repetitive occlusions of the upper airway, resulting in oxygen desaturation and/or arousals. Although the causal relationship between obesity and OSA is not well understood, obesity may promote occlusion of the upper airway through decreases in the size of the upper airway, which enhances its collapsibility. Obesity increases fat deposition of the trunk, abdomen, and neck, particularly in the posterior and lateral regions of the oropharynx and the soft palate,leading to increased airway collapsibility in obese patients compared with normal patients. Obesity also increases muscle and fat in the uvula and is significantly related to the severity of sleepdisordered breathing. Population-based studies further support that obesity and larger neck size are associated with greater risk for sleepdisordered breathing. A one standard deviation increase in neck circumference is associated with a 5-fold increase in risk for sleep-disordered breathing. Functional studies of airway critical pressure (the nasal pressure at which airflow ceases) indicates that patients with OSA have increased upper airway collapsibility with a critical pressure of 3.3 cm water, in contrast to normal patients with upper airway critical pressures of 13.3 cm water. Patients with OSA have higher critical pressure, lower airflow, and higher degrees of upper airway collapsibility compared with snorers and normals. Use of positive airway pressure can increase the diameter of the upper airway, improve airflow, and modify the critical pressure. Even small amounts of weight loss may reduce the critical pressure. In one study a 13% reduction in weight resulted in a decrease in critical pressure from 3.1 to 2.4 cm water. An additional study supports that complete resolution of OSA can occur once critical pressure falls to less than 4 cm water. These studies support that upper airway collapsibility improves with weight reduction in patients with OSA. Obesity may also affect mechanical loading of the upper airway through metabolic dysregulation and their effects on the upper airway. Leptin and adiponectin, both manufactured by adipocytes, play a role in this complex interaction on energy consumption and expenditure. Leptin functions to inhibit appetite, promote satiety, and stimulate ventilation. Antagonists of leptin such as C-reactive protein mitigate the physiologic properties of leptin. On the contrary, adiponectin levels increase with weight loss and are inversely associated with overall and central adiposity. Low levels of adiponectin have been associated with the development of type 2 diabetes and insulin resistance, diseases commonly seen in patients with obesity and OSA. Obesity may also affect OSA through pathways targeting ventilation and hypoxemia. Obesity hypoventilation syndrome is defined as obesity (BMI >30 kg/m2 ), daytime hypoventilation with awake PCO2 greater than 45 mm Hg, and sleepdisordered breathing in the absence of other causes of hypoventilation. In obesity hypoventilation, the increase in the work of breathing is related to decreased lung compliance and the efforts to displace the ribs and diaphragm, generating ventilation perfusion mismatch. Muscle impairment may also play a role. Finally, patients with obesity hypoventilation have reduced response to hypoxia and hypercapnia. Data from the Wisconsin Sleep Cohort Study indicate that weight loss is associated with a reduction in the severity and likelihood sleepdisordered breathing. A 10% reduction in weight is estimated to produce a 26% decrease in the AHI. Despite obvious benefits to losing weight in this population, small studies indicate that initial weight loss success is not sustainable in the long term. In 24 subjects who had cured their OSA following weight loss, regain of weight led to recurrence of OSA. Currently, weight loss is not recommended as primary therapy for OSA. It is instead recommended as an adjunct to more proven strategies such as continuous positive airway pressure (CPAP). Despite the untoward and negative outcomes associated with obesity, achieving weight loss can be quite challenging. Comorbid medical illnesses associated with obesity and symptoms of OSA including hypersomnia and fatigue may affect weight loss goals and efforts. Also, patients of advanced age and with comorbid illness may be deemed less suitable for certain weight loss therapies such as surgery. Hence patients with OSA face several challenges when trying to lose weight. A variety of ways to lose weight are available and a review of the literature on weight loss strategies in patients with OSA and its effects on OSA are discussed.
Bariatric surgery has emerged as a treatment option for morbidly obese patients who fail traditional weight loss approaches. In the Swedish Obese Subjects Study, a prospective, nonrandomized intervention trial comparing longitudinal outcomes of more than 4000 patients undergoing bariatric surgery to matched controls, patients who had surgery lost more weight (23% of BMI in the surgical group vs 0.1% in the control group). Weight loss was maintained by the tenth year (16% of BMI in the surgical group) with a lower incidence of obesity-related morbidity at 10-year follow-up. Bariatric surgery induces weight loss by malabsorption, restriction, or hormonal alterations. Considerations for bariatric surgery include a BMI greater than 40 kg/m2 or BMI greater than 35 kg/m2 with one or more comorbid conditions, such as diabetes, dyslipidemia, coronary artery disease, hypertension, OSA, or nonalcoholic steatohepatitis. One of the most common bariatric procedures is a Roux-en-y gastric bypass (RYGB), which consists of a small proximal gastric pouch anastomosed to the roux or alimentary limb of the small intestine. The small intestine is divided at 30 to 50 cm distal to the ligament of Treitz. The proximal portion of the small intestine and the remaining gastrum become the biliopancreatic limb, which transports secretions such as gastric acid once anastomosed 75 to 150 cm from the gastrojejunostomy. The gastrojejunostomy is formed by connecting the proximal gastric pouch to the more distal portion of the divided small intestine. Ninety percent of RYGB are performed laparoscopically. RGYB induces weight loss by restriction (creation of a small gastric pouch), by malabsorption (by diversion of the pancreaticobiliary secretions) to a short common gut, and by affecting hormone levels via a decrease in the orexigenic hormone ghrelin secreted by the foregut. Lengthening the roux limb can promote further malabsorption. Weight loss is rapid the first year after surgery and ranges from 62% to 70% of excess weight lost, but usually stabilizes after 2 years. Another commonly performed surgery for weight loss is the laparoscopic adjustable gastric banding (LAGB). Weight loss is achieved by restriction of the upper part of the stomach by placement of a tight adjustable prosthetic band. The band is made of a locking silicone ring connected to an implanted infusion port located in the subcutaneous tissue. Infusion of saline in the port will decrease the size of the band and promote stomach restriction. Although weight loss after LAGB is more gradual compared with RYGB surgery, both have similar long-term results. Less operative mortality and reversibility are the most significant benefits of the LAGB. Another form of restrictive weight loss surgery is the laparoscopic sleeve gastrectomy (LSG), which consists of creating a small tubular stomach by removing most of the greater curvature. Although initially used as a first-step procedure before RYGB or duodenal switch surgery, it is now widely used as an isolated bariatric procedure. Data are limited about its safety and long-term outcomes; however, a recent report examining 1-year outcomes between LAGB, LSG, and open RYGB and laparoscopic RYGB showed that reductions in BMI and morbidity associated with LSG lie between those of LAGB and RYGB, with no difference in mortality. Leakage of the staple line is one of the major complications of LSG and occurs in 7% of patients. Two endoscopic procedures, the intragastric balloon and the endoluminal vertical gastroplasty, have been used in clinical trials; however, additional research is needed to investigate long-term outcomes. Other types of weight loss surgery include biliopancreatic diversion and biliopancreatic diversion with duodenal switch, which rely primarily on malabsorption to facilitate weight loss, but both are rarely used because of the high incidence of complications, including protein malnutrition, diarrhea, and stomal ulceration. Currently, there are no clear guidelines regarding procedure selection and the procedure largely depends on the expertise of the surgeon, the institution, and patient preference. Several studies have been performed to evaluate outcomes of OSA in patients who have undergone bariatric surgery. Despite significant weight loss after bariatric surgery and improvement in the severity of OSA, it may still persist in some cases. Bariatric surgery by itself is not a cure for OSA, which is defined by a normal AHI and complete discontinuation of CPAP. A metaanalysis reviewed the effect of bariatric surgery in 342 patients with OSA. Although the severity of the OSA had improved, and the AHI had decreased by 38 events from 54 to 15.8, moderate OSA was still observed. This moderate degree of OSA was still present despite a significant amount of weight loss and a reduction of BMI by 17.9 kg/m2 (55.3–37.7 kg/m2 ). These findings highlight that despite large amounts of weight loss and reductions in BMI, a significant portion of patients after bariatric surgery are still moderately obese at 10-years follow-up. A recent randomized controlled trial comparing bariatric surgery (LAGB) to conventional weight loss in obese patients did not find any statistically significant differences in AHI reduction between groups despite greater degrees of weight loss in the bariatric surgery group.
In summary, bariatric surgery as a weight loss therapy is effective in reducing obesity-related comorbidities and mortality; however, obese patients should not expect a complete cure of their OSA but could achieve improvements in severity of disease. OSA persists in many patients despite major weight loss, and use of CPAP should be anticipated.
Anti-obesity drugs are being used frequently to assist in achievement of weight loss by modifying fat digestion and suppressing appetite. However, the effect of these drugs with the primary focus on OSA has been studied in very few trials. Many of these trials involved sibutramine, which has been withdrawn from the market; however, existing trials using sibutramine in patients with OSA have yielded mixed results when OSA outcomes were measured. In an open uncontrolled cohort study of 87 obese subjects with OSA, at 6-month follow-up, the respiratory disturbance index and weight had decreased by 16.4 events per hour and 8.3 kg, respectively. In contrast, in a randomized controlled trial examining the efficacy of sibutramine versus CPAP on sleep-disordered breathing, subjects who received sibutramine did not achieve a statistically significant reduction in OSA severity or AHI at 1 year from 41.5 to 33 events per hour. In part, this may be due to the modest reduction in weight loss of 5% compared with the Yee study in which subjects lost more weight, suggesting that greater degrees of improvement in OSA may be linked to the amount of weight lost.
Orlistat inhibits pancreatic lipase, thereby decreasing dietary fat absorption. Orlistat is approved for use in the United States either by prescription (120 mg) or over the counter (60 mg). In multiple randomized controlled trials and a meta-analysis, the use of orlistat compared with lifestyle changes alone was associated with greater weight loss at 4 years. Use of orlistat has also been shown to decrease the incidence of type 2 diabetes in obese patients. In studies performed specifically in obese patients with sleepdisordered breathing, use of orlistat in conjunction with dietary modification was associated with an average of 3.5 kg of additional weight loss after 1 year but many patients were not able to comply with the dietary restrictions. The major side effects of orlistat are flatus, fecal incontinence, and oily spotting. Patients taking orlistat should be advised to take a multivitamin tablet given the risk of fat-soluble vitamin malabsorption.
Other drugs that suppress appetite include the serotonin agonists, sympathomimetics, antidepressants, and antiepileptic drugs. Their use however is limited by frequent and serious side effects, short-term use, and regain of weight after their discontinuance. Lorcaserin, a selective 5-HT2C serotonin receptor agonist, was recently approved by the Food and Drug Administration (FDA) for the treatment of obesity. Data about safety and efficacy are limited to a few shortterm trials that showed a significant dosedependent weight loss compared with placebo, with headache, nausea, and dizziness being the most common side effects. In a randomized controlled trial of lorcaserin versus placebo, subjects receiving lorcaserin lost 5.8% of body weight compared with 2.8% in the placebo group. Sympathomimetic drugs such as phentermine stimulate the release of norepinephrine or inhibit its reuptake into nerve terminals. It activates the sympathetic nervous system, which suppresses appetite and increases resting energy expenditure. Phentermine is approved by the FDA for shortterm use (12 weeks) and leads to weight loss of 5 to 6 pounds compared with placebo. The weight loss may plateau after a few months with some regain in weight after medication discontinuance. Weight loss is offset by frequent side effects and the potential risk of drug abuse as is the case with all sympathomimetic medications. Phentermine-topiramate, a new medication recently approved by the FDA, leads to significant weight loss at 1 year compared with placebo (10% vs 1.6% of baseline body weight). The most common adverse events with phenterminetopiramate are paresthesia, dry mouth, constipation, dysgeusia, and insomnia. Although studies emphasize its effectiveness as a weight loss medication, long-term safety data are lacking. Its discontinuation should be gradual because of the risk of seizure with topiramate withdrawal. Although topiramate alone has been used in the management of epilepsy, its notable side effect of weight loss has led to its off-label use as a weight-loss medication. In a study be Li and colleagues, patients who received topirimate lost 6.5% of total body weight. Despite positive results, for now topiramate as single therapy is not recommended to treat obesity. Bupropion, a drug used to treat depression, has been shown to produce significant weight loss. A combination of bupropion and naltrexone has been studied in a large randomized, double-blind, placebo-controlled trial with more than 1700 subjects. Nearly half of the subjects randomized to receive combination sustained-release naltrexone (16 or 32 mg) with sustained-release buproprion lost weight as compared with only 16% of subjects receiving placebo. The average weight lost with the combination drug was 5% to 6% of body weight, with the most frequent side effects of nausea and headache reported. Currently, there is no commercial form of the drug combination. In summary, weight loss medications can be added to diet and exercise in obese patient with BMI greater than 30, or BMI greater than 27 with comorbidities. Downsides of the use of weight-loss medications include frequent side effects, short-term benefit, and lack of long-term safety. Expected weight loss with these medications is modest and shortlived and should be combined with lifestyle changes.
Lifestyle interventions include diet modification, behavioral therapy, exercise, and their combinations. Surprisingly, little data exist on the efficacy of weight loss via dietary management on OSA. A Cochrane review called for more randomized controlled trial data with regard to the effectiveness of weight loss, exercise, and sleep hygiene techniques on the treatment of OSA. The available studies in this area have several limitations, including small sample size, lack of a control group, and short-term follow-up. To date, only 2 randomized controlled trials have been performed using a control group. Tuomilehto and colleagues reported the results of a prospective randomized controlled parallel group 1-year follow-up study in 72 patients on a very low calorie diet (VLCD) with supervised lifestyle counseling in patients with mild OSA. Subjects receiving 600 to 800 kcal/d using a commercially available weight loss product lost a mean of 10.7 kg (10.6% of the initial weight) versus 2.4 kg (2.6% of the initial weight) in controls who received lifestyle counseling only. At the end of the study period, nearly twice as many subjects in the intervention group were cured of OSA compared with the control group (61% vs 32%). Furthermore, there was a significant improvement in mean oxygen saturation and less time with oxygen saturation less than 90% during sleep in the intervention group. In addition, subjects in the intervention arm lost a mean of 11.6 cm in waist circumference. A 5-cm reduction in waist circumference was associated with a decrease in AHI of 2.5 events per hour. A larger randomized controlled trial comparing VLCD to controls in obese patients with moderate to severe OSA receiving CPAP showed that subjects in the intervention arm lost 18.7 kg and decreased the AHI by 25 events per hour compared with a 1.1-kg weight gain and no effect on AHI in controls at the end of a 9-week period. Seventeen percent of subjects on the VLCD were cured of OSA (defined as AHI <5) and 50% of subjects were reclassified as mild OSA (AHI <15) by the end of the study period. Although impressive, the major limitation of this study was its short duration; hence, the authors were unable to comment on the sustainability of weight loss beyond 9 weeks. Other studies examining the effect of VLCD on OSA have shown that even modest degrees of weight loss in general reduce the AHI, but do not cure OSA. In a single study of VLCD in patients with OSA, a mean reduction of 9.2 kg led to significantly improved blood pressure and a reduction in the oxygen desaturation index from 31 to 19 events per hour. Interestingly, the authors reported that in 20% of subjects OSA did not improve despite weight loss. Similar results were noted by Surratt and colleagues, who examined 8 obese subjects with OSA who consumed between 400 and 800 kcal/d. Although weight loss from a VLCD led to a significant drop in the BMI (54–46 kg/m2 ), it did not lead to a statistically significant improvement in the AHI probably because subjects were still morbidly obese at the termination of the trial. A study examining a combination of VLCD and exercise in a small group of obese patients with AHI greater than 10 showed significant and partially sustainable weight loss of 12.3 kg over a period of 12 months without significant improvements in the AHI. In general, VLCD may produce significant degrees of weight loss in the short term with variable improvements in the degree of OSA that are not necessarily correlated with weight loss. Although such weight loss programs can be run effectively in the outpatient setting, it should be stressed that several patients will still have OSA despite weight loss and that treatment of OSA may still be needed.
Behavioral therapy refers to strategies that help set and achieve weight loss goals, manage stress, and improve healthy behaviors. Behavioral therapy is effective particularly when combined with additional strategies to lose weight, such as calorie restriction or medications. Unfortunately, there are very few studies that report the impact of behavioral therapy alone on weight loss in patients with OSA. The largest study to date, the Sleep AHEAD Study, included 264 subjects with OSA and type 2 diabetes mellitus, randomizing them to receive either an intensive lifestyle intervention using calorie restriction and physical activity or diabetes support and education (control). The intensive lifestyle intervention led to a mean weight loss of 10.8 kg and a reduction in AHI of 4.6 events per hour compared with a 0.6-kg weight gain and 4.8 events per hour increase in the AHI among controls. Although the average change in AHI seems modest, nearly 40% of subjects in the intervention arm improved the severity of OSA. Remission rates of OSA in both arms were low at 3% to 13%. Behavioral therapy was also reported to generate weight loss by Kajaste and colleagues following a cognitive behavioral weight reduction program. Forty-four percent of patients decreased the oxygen desaturation index to less than 10 and decreased the oxygen desaturation index by more than 50% at 1-year follow-up. Dietary selfmonitoring via food logs may be helpful for some patients in one small study to lose small amounts of weight. These studies indicate that behavioral therapy seems to be most effective when used in conjunction with programs that incorporate VLCD. Exercise is considered a key component in weight loss programs. Exercise efficiently burns calories and, when combined with caloric restriction, weight loss can be achieved. Some studies have been performed examining the impact of exercise on OSA. In a small open trial of exercise in patients with moderate to severe OSA patients using CPAP, after 6 months of physical exercise, the respiratory disturbance index decreased by 9 events per hour on average and was observed even in the absence of any significant change in weight. Data from 2 longitudinal studies lend additional support to these findings. Cross-sectional data from the Wisconsin Sleep Cohort Study showed an association between increased hours of exercise per week and reduced sleep-disordered breathing after adjustment for covariates. Similarly, data from the Sleep Heart Health Study indicate that patients with sleep-disordered breathing perform in less vigorous physical activity compared with patients without sleep-disordered breathing, and that vigorous exercise 3 hours per week is associated with a respiratory disturbance index onethird less than found in sedentary patients in an unadjusted model. It is not entirely understood if exercise exerts an independent positive effect on OSA, and the existing studies are inconclusive. In a randomized controlled trial comparing effects of an aerobic exercise versus stretching on AHI, subjects randomized to 12 weeks of moderateintensity aerobic exercise decreased their AHI by 24% on average without significant change in weight. Meanwhile, longitudinal data also from the Wisconsin Sleep Cohort Study indicate that exercise reduces the incidence of new onset mild or moderate sleep-disordered breathing over a follow-up period of 8 years. Although the observed effect in this study was partially mediated by the change in weight, there also seemed to be some protective effect independent of body habitus. Exercise may affect OSA by increasing in upper airway muscle tone, changing body fat distribution of body fat, and/or altering control of breathing and arousal.
CONTINUOUS POSITIVE AIRWAY PRESSURE
CPAP itself may mediate an effect on weight loss through sleep. Daytime sleepiness and fatigue are commonly reported in patients with OSA and can be both motivational and empowerment barriers to weight loss. CPAP can improve fatigue and reduce sleepiness in OSA and theoretically could reduce some challenges of weight loss. Unfortunately few studies have been published in this area. In patients using a VLCD and cognitive behavioral weight loss program, the use of CPAP did not result in any additional weight loss. Similarly, a retrospective study found that CPAP use of 4 hours per night 70% of nights was not associated with weight loss. There were no significant differences in weight loss between adherent and nonadherant users. Interestingly, women CPAP users gained weight. It is important to recognize that in the Redenius study, sleep duration was not monitored, a significant covariate linked to weight gain. On the contrary, a small study of 32 obese or overweight patients with OSA on CPAP reported that CPAP adherent users were more likely to lose weight than nonadherent users. More studies examining OSA as a risk factor for obesity, accounting for significant covariates, are needed and also whether CPAP adherence has an independent effect. At this point, CPAP alone cannot be recommended as the sole weight loss strategy for obese patients with OSA.
Although obesity is one of the important risk factors for OSA, the mechanism behind its causality is not well understood. In obese patients with OSA, weight loss should not be recommended as the sole treatment of OSA, but should be used in conjunction with other treatments such as CPAP. Although weight loss strategies vary, these can mainly be classified into surgical, pharmacologic, and lifestyle (including caloric restriction, behavioral therapy, and exercise) treatments. In general, weight loss may reduce the number of apneas and hypopneas, but complete remission in most patients is not to be expected. Furthermore, additional challenges with the difficulty of sustainable weight reduction should impart close attention and follow-up to ensure continued evaluation and management of OSA.