Obstructive sleep apnoea (OSA) syndrome is common, and obesity is a major risk factor. Increased peripharyngeal and central adiposity result in increased pharyngeal collapsibility, through increased mechanical loading around the upper airway, reduced tracheal traction on the pharynx, and reduced neuromuscular activity, particularly during sleep. Significant and sustained weight loss, if achieved, is likely to be a useful therapeutic option in the management of OSA and may be attempted by behavioural, pharmacological, and surgical approaches. Behavioural therapy programs that focus on aspects such as dietary intervention, exercise prescription patients and general lifestyle counselling have been tested. Bariatric surgery is an option in the severely obese when nonsurgical measures have failed, and laparoscopic adjustable gastric banding and Roux-en-Y gastric bypass are the most commonly employed techniques in the United Kingdom. Most evidence for efficacy of surgery comes from cohort studies. The role of sibutramine in OSA in the obese patients has been investigated, however, there are concerns regarding associated cardiovascular risk. In this paper the links between obesity and OSA are discussed, and the recent studies evaluating the behavioural, pharmacological and surgical approaches to weight loss in OSA are reviewed.
Obstructive sleep apnoea (OSA) syndrome is common with a prevalence of approximately 4% in middle-aged men and 2% in middle-aged women [
The evidence to support the role of excess weight as a causal factor in the aetiology of OSA is convincing. In a population study involving 2148, prevalence of obesity was significantly higher in those with OSA than those without, whether male (22% versus 8%) or female (32% versus 18%) [
There are several mechanisms by which obesity could result in OSA, and these may act synergistically. It is proposed that increased peripharyngeal fat deposition results in mechanical loading that offsets the maintenance of airway patency by the dilator muscles and that this increase in collapsibility is particularly prominent during sleep when there is a reduction in neuromuscular activity [
A summary of studies of behaviourally and pharmacologically induced weight loss in obstructive sleep apnoea.
Paper | Design | % Male | Group | Intervention | Followup months | ΔBMI | ΔAHI | % With OSA cure | |
---|---|---|---|---|---|---|---|---|---|
Johansson et al., 2011 [ | Prospective cohort | 63 | 100 | BMI 30–40, AHI ≥ 15 | VLCD and weight maintenance programme | 12 | 35→31 | 36→19 | 10 |
Nerfeldt et al., 2010 [ | Prospective cohort | 33 | 73% | BMI ≥ 30, AHI ≥ 10 and/or ODI ≥ 6, OSAS symptoms | LCD and behavioural change support | 24 | 40→35 | 43→28 | ? |
Johansson et al., 2009 [ | Randomised, controlled, and parallel group | 63 | 100 | BMI 30–40, AHI ≥ 15 | VLCD | 2 | 34→29 | 37→12 | 17 v 0 |
Tuomilehto et al., 2009 [ | Randomised, controlled, and parallel group | 72 | 74 | BMI 28–40 kg/m2, AHI 5–15 events/hr | VLCD, supervised lifestyle counselling | 12 | 33→? | 10→6 | 63% |
Barnes et al., 2009 [ | Cohort | 12 | 25 | BMI > 30 kg/m2, AHI 10–50 events/hr | VLCD and exercise programme | 12 | 36→30 | 25→18 | 0 |
Foster et al., 2009 [ | Randomised, controlled, and parallel group | 264 | 41 | BMI ≥ 25 kg/m2, AHI ≥ 5 events/hr, type 2 diabetes | Intensive lifestyle intervention (diet/exercise) | 12 | 37→? | 23→18 | 14 v 4 |
Kemppainen et al., 2008 [ | Randomised, controlled, and parallel group | 52 | 79 | BMI 28–40 kg/m2, AHI 5–15 events/hr | VLCD, supervised lifestyle counselling | 3 | 33→? | 11→8 | ? |
Kajaste et al., 2004 [ | Cohort | 31 | 100 | BMI > 35 kg/m2, ODI > 10 events/hr | VLCD, CBT weight reduction program | 24 | 44→40 | 51→32 (ODI) | ? |
Kansanen et al., 1998 [ | Prospective, cohort | 15 | 93 | Overweight with OSA | VLCD | 3 | 38→35 | 31→19 (ODI) | 20 |
Suratt et al., 1992 [ | Cohort | 8 | 62 | Obese with OSA | VLCD | ? | 54→46 | 90→62 | ? |
Ferland et al., 2009 [ | Nonrandomised, parallel group | 40 | 88 | BMI ≥ 30 kg/m2, OSAS | Sibutramine, diet and exercise v CPAP | 12 | 37→35 | 40→37 | ? |
Phillips et al., 2009 [ | Cohort | 93 | 100 | BMI 30–38 kg/m2, RDI ≥ 15 events/hr | Sibutramine, diet and exercise | 6 | 34→32 | 46→30 (RDI) | ? |
Yee et al., 2007 [ | Cohort | 87 | 100 | BMI 30–38 kg/m2, RDI ≥ 15 events/hr | Sibutramine, diet and exercise | 6 | 34→32 | 46→30 (RDI) | 5 |
Martinez and Basile, 2005 [ | Randomised, double-blind, and controlled group | 19 | 100 | BMI 25–35 kg/m2, AHI ≥ 10 events/hr | Sibutramine v placebo | 1 | ? | 28→27 | ? |
Abbreviations: ΔAHI: apnoea hypopnoea index before and after intervention, ΔBMI: body mass index before and after intervention, CBT: cognitive behavioural therapy, CPAP: continuous positive airway pressure,
Several studies have been performed to evaluate the effects of approaches including dietary modification and exercise, as well as counselling. Two small cohort studies [
Thus, the outcomes of behavioural therapy for weight loss in OSA are mixed. Furthermore, for these programmes to be available in the clinical setting may not be economically feasible due to the costs involved such as employment of dietician, physiotherapist, and nurse. Moreover, drop-out rate is not insubstantial; in the study by Nerfeldt et al. 30% of patients did not complete the two-year program [
Sibutramine is a serotonin and noradrenaline reuptake inhibitor that promotes weight loss by enhancing satiety and increasing energy expenditure through thermogenesis [
A summary of studies of surgically induced weight loss in obstructive sleep apnoea.
Paper | Design | % Male | Group | Procedure | Followup months | ΔBMI | ΔAHI | % with OSA cure | |
---|---|---|---|---|---|---|---|---|---|
Behrens et al., 2011 [ | Retrospective, cohort | 34 (21 OSA) | 3 | BMI > 30 | LSG | 10 | 50.3→39.9 | ? | 76 |
Martí-Valeri et al., 2007 [ | Cohort | 30 (14 OSA) | 90 | Obese with associated respiratory comorbidity requiring NIPPV | RYGB | 12 | 56→32 | 64→17 (RDI) | ? |
Fritscher et al., 2007 [ | Cohort | 12 | 75 | BMI ≥ 35 with obesity-related comorbidity or ≥40 without and AHI ≥ 15 | RYGB | 18 | 56→34 | 46→16 | 25 |
Haines et al., 2007 [ | Prospective, cohort | 101 | ? | Obese, RDI > 5, ESS ≥ 6 | RYGB | 11 | 56→38 | 51→15 (RDI) | ? |
Kalra et al., 2005 [ | Retrospective, cohort | 34 (19 AHI ≥ 5) | ? | Adolescent, BMI ≥ 40, obesity-related comorbidity | RYGB | 5 | 60.8→41.6 | 9.1→0.6 | ? |
Lankford et al., 2005 [ | Retrospective, cohort | 15 | 40 | Obese, OSA | RYGB | 12 | 48→32 | 40→? | 0 |
Guardiano et al., 2003 [ | Retrospective, cohort | 8 | 12 | Obese, OSA | RYGB | 28 | 49→34 | 55→14 (RDI) | 50 |
Peiser et al., 1984 [ | Cohort | 15 | 93 | Morbidly obese with OSA | RYGB | 3 | ? | 82→15 | 80 |
Lettieri et al., 2008 [ | Retrospective, cohort | 24 | 25 | Obese with EDS, AHI ≥ 5 | GB | 12 | 51→32 | 48→24 | 4 |
Rasheid et al., 2003 [ | Prospective, cohort | 11 | ? | Obese, ESS ≥ 6, RDI > 5 | GB | 12 | 62→40 | 56→23 (RDI) | ? |
Rao et al., 2009 [ | Retrospective, cohort | 46 | ? | BMI ≥ 32.5 with obesity-related comorbidity or ≥37.5 without and AHI ≥ 15 | LAGB | 13 | 45→30 | 38→13 | 78 |
Dixon et al., 2005 [ | Prospective, cohort | 25 | 68 | BMI > 35, AHI > 25 | LAGB | 18 | 53→37 | 62→13 | ? |
Busetto et al., 2005 [ | Cohort | 17 | 100 | BMI > 50, AHI > 20 | IGB | 6 | 56→49 | 59→14 | 59 |
Grunstein et al., 2007 [ | Prospective, controlled, nonrandomised | 3023 | 30 | Female: BMI ≥ 38, male: BMI ≥ 34 | Various | 24 | 42→32 | ? | ? |
Valencia-Flores et al., 2004 [ | Prospective, cohort | 29 | 45 | Morbidly obese | Various | 14 | 56→39 | 52→? | 46 |
Poitou et al., 2006 [ | Prospective cohort | 35 | 17 | BMI > 40, AHI > 10 | RYGB, LAGB | 12 | 51→40 | 24→10 | 63 |
Pillar et al., 1994 [ | Cohort | 14 | 79 | Morbidly obese with OSA | RYGB, VBG | 4 | 45→33 | 40→11 | 43 |
Charuzi et al., 1992 [ | Cohort | 47 | ? | Morbidly obese with OSA | RYGB, VBG | 10 | ? | 61→8 | 40 |
Omana et al., 2010 [ | Retrospective, cohorts | 123 | 24 | Obese | LSG (49) | 15 | 52→? | ? | 55 |
Abbreviations: ΔAHI: apnoea hypopnoea index before and after intervention, ΔBMI: body mass index before and after intervention, EDS: excessive daytime somnolence, ESS: epworth sleepiness scale, GB: gastric bypass, IGB: intragastric balloon, LAGB: laparoscopic adjustable gastric banding, LSG: laparoscopic sleeve gastrectomy,
The National Institute for Health and Clinical Excellence Guidelines state that bariatric surgery should be recommended as a treatment option for adults with BMI of 40 kg/m2 or more, or between 35 and 39.9 kg/m2 in the presence of significant comorbidities such as type 2 diabetes or hypertension. In addition, all nonsurgical measures should have been employed with failure to achieve or maintain adequate weight loss over at least 6 months. Further, individuals should be managed within a specialist obesity service, be both fit for anaesthesia and surgery, and willing to commit to long-term followup [
Since the 1980’s, the results of several studies of the effects of surgical weight loss in patients with OSA have been published (see Table
From these reports, it is tempting to conclude that surgical weight loss is the panacea for OSA at least in the obese population. Caution is required in the interpretation of these studies, however, the main criticisms being that they did not include a control group, and in many cases data collection was retrospectively carried out. A large multicentre prospective controlled study with 2-year followup was carried out in Sweden by Grunstein et al. [
In summary, the aforementioned studies report promising results regarding the effects of bariatric surgery on sleep apnoea symptoms and polysomnography at least in the short term up to one to two years. What of the long term are benefits maintained? One study of LAGB surgery for obesity showed progressive loss of weight over the first 2 to 3 years with a plateau in BMI to 6 years [
The morbidity and mortality related to surgical intervention should also be considered in the overall evaluation of the benefits of bariatric surgery. Grunstein et al. reported a perioperative mortality rate of 0.21% and an incidence of other complications (including bleeding, thromboembolism, wound complications, deep infections, pulmonary, and other complications) of 13% [
The mechanisms by which weight loss results in a reduction in severity of OSA or even resolution have been explored in a number of studies. Following weight loss, there is a reduction in nasopharyngeal collapsibility and resistance implying that the calibre of the upper airway increases [
Obesity is a major (and perhaps the leading) risk factor for obstructive sleep apnoea. The prevalence of OSA is increased in the obese patients and vice versa; OSA is related to various anthropometric measures; the severity of OSA increases in association with weight increase. Pathophysiological mechanisms by which obesity can lead to OSA have been identified. It follows that weight loss may lead to an improvement in the severity of OSA and perhaps even it’s resolution. This has been borne out in studies evaluating behavioural and surgical approaches to weight loss. Behavioural methods have focussed on dietary intervention, encouragement of exercise, and support in lifestyle change, and randomised studies have been encouraging. The evidence for benefits in association with bariatric surgery stems from cohort studies, many of them retrospective, but nonetheless persuasive. However, surgical intervention is recommended only after nonsurgical measures have failed and is associated with an albeit low mortality rate and also significant morbidity.
Although the studies of both behavioural and surgical interventions aimed at weight loss in the management of OSA have shown promising results, there are some limitations to these studies that should be noted. The majority of studies reviewed were uncontrolled and involved low subject numbers often with a male predominance. There were varying criteria for inclusion in studies with respect to the severity of obesity and both the presence and severity of OSA; the inclusion criteria for some studies included the presence of excessive daytime somnolence or other obesity-related comorbidities while other studies did not. Likewise, different methodologies were utilised for the confirmation of OSA such as polysomnography, limited sleep studies, and respiratory polygraphy, and there was variation in the outcome measures reported such as apnoea hypopnoea index, respiratory disturbance index, oxygen desaturation index, and other measures of nocturnal oxygenation. For these reasons, it is difficult to compare the different studies and to extrapolate to the wider population. The proportion of subjects lost to followup was often significant and this may lead to bias with overestimation of the improvement associated with the intervention as those not benefitting are more likely not to return for reassessment. The time to reevaluation varied from as little as 2 months to over 24 months. The benefits reported in studies with shorter durations may have been influenced by short-term simultaneous behavioural changes such as reduction of alcohol intake or increase in exercise which may not be maintained in the longer time. Alternatively shorter studies may have underestimated the potential for weight loss and attendant improvements in OSA that may be possible with longer behavioural interventions.
For either approach, behavioural or surgical, to be successful, weight loss has to be maintained in the long term to prevent relapse, and “maintenance of weight loss programmes” necessarily incorporating similar features to the behavioural weight loss programmes may not be economically feasible. Furthermore, success of such programmes or of maintenance of weight loss independently is heavily dependent on the motivation of the individual. Further randomised, controlled trials are required to confirm the beneficial effects of bariatric surgery, and those of the behavioural interventions. Studies are necessary to identify those in which behavioural therapy is likely to be effective so that limited resources are efficiently utilised.