754 4Thorax 1993;48:754-764 Occasional review Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from Obstructive apnoea

Steven G McNamara, Ronald R Grunstein, Colin E Sullivan

The last 20 years of respiratory medical prac- tion. While an apnoea is agreed upon as a tice have seen many changes. As other sub- cessation of for 10 or more sec- specialties such as oncology, infectious onds,2 there is a wide range in the frequency diseases, allergy/immunology, intensive care, of such events during sleep; how this fre- and occupational medicine increasingly make quency (or "respiratory disturbance index") inroads into the traditional areas of respira- correlates with disease severity and incidence tory practice, new fields requiring the exper- continues to be debated.3 Most researchers tise of the respiratory physician have use a working definition of five apnoeas per developed. Arguably the dominant examples hour to define sleep disordered breathing2 but of these are disorders of breathing in sleep. this is by no means consistent in the litera- Until recently terms such as "REM sleep," ture. This definition was developed at a time "apnoea index," and "nasal CPAP" were for- when OSA was thought to be a rare disorder eign to most people working in respiratory and these arbitrary cut offs allowed medicine. The parallel development of meth- researchers to communicate in a common ods of measuring ventilation and blood gas "language"-an important consideration in levels non-invasively with the standardised reviewing any literature on this subject. As monitoring of sleep has led to a recognition OSA is increasingly recognised as a common that sleep disorders-particularly abnormal disorder such definitions need to be reconsid- breathing during sleep-are both causes and ered. contributors to a broad spectrum of clinical Most clinicians recognise OSA as a dis-

morbidity and mortality. order characterised by repetitive apnoeas, http://thorax.bmj.com/ While researchers and clinicans working in loud , and excessive daytime sleepi- the field agree on the importance of accurate ness. However, in OSA the patient is often diagnosis and appropriate treatment for the last to realise the extent of the mental and patients with sleep breathing disorders, there physical effects of the disorder. Recent stud- is considerable disparity in the availability of ies have shown that some forms of OSA may investigation and treatment for such patients occur without the presence of snoring4 or in different countries. In the USA it has been apnoea5 but with obvious clinical effects. estimated that there is more than one sleep Similarly, excessive daytime sleepiness may laboratory per 250 000 people; in Australia not occur but instead the clinical picture may on September 26, 2021 by guest. Protected copyright. there is at least one per 1 000 000, but in the mimic an anxiety state, especially in women.6 UK the ratio is much lower. While these dif- Adult criteria for OSA may also not be appro- ferences in the availability of facilities may priate in children.7 It is important that the reflect a variation in health care expenditure clinical definition of OSA is kept flexible. The or structure, it is likely that they also reflect a individual with one or two apnoeas per hour, divergence in the relative importance with oxygen desaturation to 60-70%, and which health planners and policy makers view impaired arousal reflexes due to autonomic disorders of breathing in sleep. To provide a neuropathy is far more vulnerable to the con- detailed review of recent advances in the sequences of their OSA8 than a healthy, David Read understanding of sleep and breathing dis- asymptomatic 75 year old with 15 apnoeas Laboratory, hour. The or health Department of orders requires a monograph.' The purpose per epidemiologist Medicine, University of this review is to assess some recent devel- administrator may wish for a more rigid defi- ofSydney and Sleep opments in the most common sleep breathing nition of OSA, but until we have a better Disorders Centre, understanding of the exact "dose" of OSA Department of disorder-obstructive sleep apnoea-with Respiratory Medicine, particular emphasis on recent data on diagno- that produces a specific "clinical effect" it is Royal Prince Alfred sis, prevalence, predisposing factors, clinical better to avoid such rigidity. Hospital, and treatment in the adult patient. Camperdown, Sydney sequelae, 2006, Australia EPIDEMIOLOGY S G McNamara Epidemiological studies in OSA fall into three RR Grunstein categories: firstly, studies based solely on C E Sullivan Diagnosis and prevalence questionnaire data about habitual snoring, or Reprint requests to: Dr R Grunstein, WHAT IS OBSTRUCTIVE SLEEP APNOEA? a history of witnessed apnoeas, or both; sec- Department of Clinical One of the key problems in recognising the ondly, studies in which questionnaires are Pharmacology, Sahlgrens University Hospital, importance of obstructive sleep apnoea validated by full polysomnographic sleep Gothenburg 41345, Sweden (OSA) is the lack of a clearly agreed defini- studies or nocturnal monitoring in Obstructive sleep apnoea 755

Recent epidemiological studies in obstuctive sleep apnoea Study n Age (y) type Prevalence Comment Reference Gislason, Sweden 3201 30-69 B 0-7-1-9% Men only 9 Cirignotta, Italy 1510 30-69 B 2-7% (RDI >10) RDI >10, men only 10 Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from Stradling, UK 893 35-65 C 5% (men) Oximeter dips per hour >5 11 Bearpark, Australia 400 40-65 C 10% men, 7% women MESAM 4 recorder 12 calculated RDI >10 Young, USA 263 30-60 C 7-8% men, 2-3% women RDI >10 on full sleep 13 studies Jennum, Denmark 1504 30-60 B 10-9% men and Inductive plethysmography 14 6-3% women to screen 50% of study had RDI >5 population RDI-respiratory disturbance index (events/hour); study type B-questionnaire with full sleep studies in a subgroup of sleep apnoea positive replies; study type C-overnight screening or sleep studies in entire group.

a random or selected subpopulation; finally, ter understanding of the relationship between studies where all or most patients undergo sleep apnoea and clinical effects. Even a con- full sleep studies or nocturnal respiratory servative view of the more recent OSA epi- monitoring. In the past five years studies of demiological studies, however, would suggest the latter two types (some ongoing) have a potentially huge investigative and thera- shown that OSA is a common finding (table) peutic load for health care systems. but there is a wide range (1-9%) in the reported prevalence of OSA.9'4 These differ- SYMPTOMS AND HISTORY TAKING: CAN ences may reflect disparity in methodologies, CLINICAL ASSESSMENT PREDICT SLEEP population differences in obesity and alcohol APNOEA? consumption, or even genetic variability. For If OSA is as common as epidemiological evi- example, the percentage of the population dence suggests, it is important that simple with heavy snoring doubles when the bed methods of diagnosis are available. Obviously partner contributes to the questionnaire (fig the simplest potential method of diagnosis 1)."1 Some questionnaire data also do not would be by history and physical examina- seem to correlate highly with actual respira- tion. The symptoms associated with OSA are tory monitoring. The high prevalence in some many and varied and may include nocturnal

studies does not correspond with a high fre- choking attacks, morning headaches, gastro- http://thorax.bmj.com/ quency of such symptoms as hypersomno- oesophageal reflux, nocturia, impotence, poor lence. For example, Jennum and Soul'4 memory and concentration, and alteration in reported that 10-9% of men aged 30-60 years mood.'6 However, those considered to be had OSA (more than five apnoeas/hour) but "key" or "major" symptoms are snoring, only 1-9% complained of hypersomnolence. apnoeas witnessed by bed partners, and In our own epidemiological studies in excessive daytime somnolence.'6 Busselton, a Western Australian rural com- The predictive power of these key symp-

munity,'2 10% of men aged 40-65 years had toms has been examined in an Australian on September 26, 2021 by guest. Protected copyright. more than 10 apnoeas per hour. This study study.'7 Apnoeas observed by a bed partner, used a well validated ambulatory monitoring with a lesser contribution from coexisting system (the MESAM 4 recorder)'5 which hypertension, body mass index, and age, pro- measures oxygen saturation, snoring, heart duce a predictive model with high sensitivity rate, and body position. It is important, how- but only moderate specificity. Other workers'8 ever, to confirm whether this high prevalence have found that these "key symptoms" is accompanied by a high prevalence of symp- explained only 36% of the variability in tomatology and clinical consequences. It is apnoea index. The authors suggested that the clear that epidemiological information will presence of key symptoms did not obviate the need to be more sophisticated to allow a bet- need for a properly performed diagnostic sleep study. Additional data provided by such measurements as neck circumference may Figure I Increase in Do you snore or have you been told you help predict OSA, but not to the extent of frequency of "often do? replacing sleep studies."9 snoring" reported ifwife is The nature of these symptoms also empha- actually present at interview wnth subject. sises the importance of obtaining a history Adapted from reference 11 from the spouse, bed partner, and other fam- with permission. ily members in the proper assessment of the patient with OSA. Unless they are told few, if any, patients are aware that they snore or stop breathing during sleep, yet this concerns many bed partners to the point where they initiate the medical review. Excessive sleepi- ness may be recognised by the patient but often, for social or other reasons, this too may be denied by the sufferer of OSA-again 756 McNamara, Gnunstein, Sullivan

underlining the critical importance of confir- development of OSA,2526 particularly if matory history from a family member, friend, accompanied by respiratory failure.26 Other or workmate. studies have failed to find a link between life- time alcohol consumption and OSA.2728

INVESTIGATION OF OSA: DO WE NEED TO Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from MEASURE SLEEP? OBESITY Under ideal circumstances a full sleep study Most patients with OSA are overweight, but is the most appropriate investigation for the relationship between sleep apnoea and assessing OSA. It allows accurate quantifica- weight is not a simple one. Studies employing tion of breathing events and provides infor- different upper airway imaging modalities mation on sleep fragmentation and arousals have shown that patients with sleep apnoea which may be just as important in producing have a decreased pharyngeal cross sectional clinical effects as the respiratory events. area, but the contributing role of upper air- However, such studies are expensive and, if way fat to this observation is controversial.2>" current epidemiological surveys are accurate, External neck circumference is increased in full sleep studies for all suspected cases of OSA and it has been stated that this measure- OSA will be beyond the reach of health care ment explains the link between obesity and systems. A recent EEC consensus report sug- sleep apnoea.'2 The hypothesis suggested by gested that full sleep studies may not be this study is that neck circumference is an appropriate for highly probable cases of index of neck fat deposition, and increased OSA.20 A recent prospective study from fatty tissue in the neck region in turn pro- Scotland2' compared the relative value of motes mass loading and obstruction of the sleep and respiratory monitoring in the upper airway in sleep, leading to sleep assessment of OSA. The authors concluded apnoea. However, in a study of 1464 patients that sleep monitoring was not necessary in presenting for sleep assessment we have the assessment of patients suspected of hav- found waist measurement to be a better pre- ing OSA and that respiratory measurements, dictor of OSA than neck measurement.3 particularly inductive plethysmography, were Nevertheless it is clear that OSA is associated adequate. Oximetry alone was of limited with a central fat distribution. Abdominal value. Whether the findings of such a study obesity may reduce volumes, particularly could be extrapolated to home studies (where in the supine posture, and both may reflexly real cost savings would occur), interpreted by influence upper airway dimensions and lead less experienced physicians, is not known. In to impaired respiratory muscle force.3436 The contrast to the previous study, Gyulay et a122 link between central obesity and sleep apnoea found that home oximetry was a sensitive may also be related to abnormal upper airway

screening test for OSA but specificity was muscle function. A reduction in type I and http://thorax.bmj.com/ low. There is now a plethora of devices avail- Ilb muscle fibres in the middle pharyngeal able for sleep apnoea "screening" but their constrictor muscle has been found in non- reliability and cost-benefit need to be proven, obese habitual snorers." Similar changes in particularly for home use. If the desired end muscle fibre have been noted in other skeletal point is correct diagnosis of OSA and exclu- muscles in obesity.'8 Moreover, studies of sion of other sleep disorders plus successful patients with OSA before and after weight treatment of OSA in appropriate cases, how loss have shown changes in upper airway

much is saved by screening studies? A key function rather than structure,'9 supporting a on September 26, 2021 by guest. Protected copyright. issue is the sophistication of the physician hypothesis of abnormal upper airway muscle interpreting data; this has not been assessed function in obese patients with OSA. The in studies of sleep apnoea screening.20 association of central obesity with increased cardiovascular risk indicates that this measure of adiposity should be controlled for in future Predisposing factors studies of health risk in sleep apnoea and A number of conditions or associations will arguably the reverse is also true. predispose an individual to OSA. Only recent data in this area will be reviewed and the ENDOCRINE AND METABOLIC DISORDERS reader will be referred to more detailed Several endocrine and metabolic disorders reviews where appropriate. other than obesity are associated with an increased prevalence of OSA. Hypo- FAMILIAL thyroidism may lead to sleep apnoea by Sleep apnoea aggregates in families. The risk reducing chemosensitivity,40 myxoedematous of having OSA increases progressively with infiltration of the upper airway, and upper increasing numbers of affected relatives and airway myopathy.4' Interestingly, hypothyroid this risk is independent of age, obesity, and rats have a similar upper airway muscle fibre alcohol consumption.2' Such risk may be the structure as that reported in habitual snor- result of similarities in facial structure affect- ers.42 It is controversial whether treatment ing upper airway dynamics in sleep. with thyroxine will cure sleep apnoea in hypothyroidism.4' 43 Sleep apnoea may also ALCOHOL provoke cardiovascular complications when Acute alcohol ingestion promotes the devel- initiating thyroid hormone replacement.4' opment of apnoea during later sleep.24 Some Over 50% of patients with acromegaly have studies have suggested that lifetime alcohol sleep apnoea and there is a higher than consumption may be a risk factor for the expected prevalence of central apnoea Obstructive sleep apnoea 757

Figure 2 High rate of important to recognise that it may occur in a prevalence ofsleep 50r disordered breathing in a range of sleep disorders which may coexist series of53 patients with with OSA.56 Sleepiness in OSA is predomi- acromegaly. A higher than 40 nantly related to repetitive arousal and sleep expectedprevalence of central apnoea was fragmentation, but a direct effect of hypox- Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from observed. 30 aemia is possible." OSA is also characterised by a range of excessive daytime sleepiness 20 from simply increased sleep time in a previ- ously short sleeper to obtundation. Sleepiness 10 may lead to both impairment of work perfor- mance58 and driving.59 One problem is that patients themselves may not be aware of their 0 Central Obstructive No apnoea degree of sleepiness and information from apnoea apnoea family is often helpful. There is also a rela- tively poor correlation between severity of OSA and daytime sleepiness4 and no simple (fig 2).44 Increased biochemical activity test will accurately quantify daytime sleepi- (growth hormone and insulin-like growth fac- ness. Such a test would be useful to identify tor 1 levels) are associated with the presence the high risk patients for treatment, assessing of central sleep apnoea.44 Treatment of ability to work and drive and evaluating acromegaly with the somatostatin analogue response to therapy. The standard test for octreotide reduces the severity of sleep quantfying sleepiness-the Multiple Sleep apnoea45; it is also associated with Cushing's Latency Test (MSLT)-may not be useful in disease.46 OSA and a modification-the Multiple One important observation in OSA is the Wakefulness Test (MWT)-has been shown male predominance of the disorder. to be more sensitive for detecting sleepiness Interestingly, androgen treatment has been before and after treatment for OSA.60 reported to provoke sleep apnoea47 48 and Patients with sleep apnoea perform worse oestrogen treatment may reduce mild sleep on driving simulator tasks.61 62 Haraldsson apnoea in women.49 However, anti-androgen et a161 reported that 15 patients with OSA ran treatment does not appear to reduce the offthe road 101 times in a 60-90 minute sim- severity of OSA in men with severe sleep ulated highway drive compared with only apnoea,50 and oestrogen therapy does not twice in 10 controls. Although poor perfor- alter more severe forms of sleep apnoea in mance in simulation tasks may be overcome postmenopausal women.5' by greater vigilance in the real life situation,

A recent report has identified a strong link data from a number of centres show a higher http://thorax.bmj.com/ between OSA and the connective tissue dis- actual accident rate between patients with order, Marfan's syndrome.52 Nearly two OSA and controls.5963 Interestingly, self- thirds of patients with Marfan's syndrome reported accident rates may not differ have OSA and such patients, being tall and between patients with OSA and controls, thin, do not have the typical body habitus of questioning the reliability of patient history in the sleep apnoea patient. The reason for the this area.63 Loss of driving privileges can be link between Marfan's syndrome and OSA is economically and socially disastrous to to likely be the abnormal compliance of the patients and they may be reluctant to admit on September 26, 2021 by guest. Protected copyright. upper airway resulting from abnormal con- problems.63 nective tissue. One important consideration is Accidents where the driver falls asleep are that the changes in intrathoracic pressure and likely to cause fatalities.64 65 There are legal blood pressure in OSA may provoke aortic precedents both in US and British law where, dilatation and rupture, a common mode of if a patient drives a car with the knowledge death in Marfan's syndrome. that he or she frequently falls asleep while driving, such an action may be negligent and Clinical sequelae lead to either civil or criminal liability. One ENDOCRINE AND METABOLIC EFFECTS recent case occurred in the UK when a trans- Patients with sleep apnoea are also charac- port driver, with a history of sleepiness whilst terised by a reversible neuroendocrine defect driving, drove a semitrailer into a stationary in growth hormone and testosterone secre- queue of vehicles and killed six people. The tion, probably due to central effects of sleep driver was sentenced to three years imprison- fragmentation and hypoxaemia.53 Growth ment and was subsequently confirmed as hav- hormone deficiency in OSA may explain ing severe sleep apnoea (R Wilkinson, British impaired growth seen in children with upper Sleep Society Meeting, Leicester, 1992, airway obstruction which often improves fol- unpublished). The responsibility of the physi- lowing adenotonsillectomy.54 In adults cian dealing with sleepy patients who con- impaired growth hormone secretion leads to tinue to drive will vary between countries. It central adiposity, reduced muscle and bone is often frustrating to leave such patients mass.55 untreated because of resource limitations in provision of appropriate treatment. This is NEUROPSYCHOLOGICAL AND SOCIAL especially the case as evidence exists that CONSEQUENCES OF OSA treatment with nasal continuous positive air- Excessive daytime sleepiness is characteristic way pressure (CPAP) dramatically improves but not pathognomonic of sleep apnoea. It is daytime sleepiness56 60 66 and even driving 758 McNamara, Grunstein, SuUlivan

simulator performance.62 oxaemia is controversial, particularly regard- Several studies have found that patients ing the blood pressure rise in phase 3. It has with OSA perform poorly on psychometric been suggested that arousal is the primary tests compared with controls and a variable stimulus for this rise in blood pressure.76

degree of improvement occurs after nasal Others have provided evidence supporting a Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from CPAP therapy.67-70 Whether this is an effect contributory role for hypoxaemia.7778 What- of impaired concentration or actual deteriora- ever the mechanism, the considerable tion in cognition and memory is uncertain. changes in cardiorespiratory behaviour, Follow up tests may produce practice effects together with reported changes in cerebral that need to be controlled, as does education blood flow, provide an environment for level and alcohol use. Few standard psycho- increasing the risk of various vascular disease metric tests are designed to test for the subtle end points. differences in cognition that are often reported by patients with OSA. Most studies Chronic effects have looked at small groups of patients with It is unknown to what degree these acute severe disease and it is unknown what cogni- cardiovascular changes produce chronic tive impairment exists in milder forms of effects. A number of groups have reported OSA. that patients with OSA are characterised by markedly elevated sympathetic nerve traffic CARDIOVASCULAR SEQUELAE OF OSA while awake which may promote chronic ele- The cardiovascular consequences of OSA can vation of blood pressure.78-0 In addition we be considered from two aspects. Firstly, the have recently observed that patients with acute cardiovascular changes that occur dur- OSA have a potent pressor response to ing an apnoea with the associated hypoxia, eucapnic hypoxia compared with controls (fig hypercapnia, acidosis, and arousal from sleep, 3).81 This pressor response was related to dis- and secondly, the chronic cardiovascular ease severity and likely to be the result of morbidity and mortality associated with OSA, exposure of the cardiovascular system to namely hypertension, myocardial infarction, intermittent hypoxia in sleep. In normal sub- stroke, and death. jects the central effects of hypoxaemia tend to increase blood pressure but these effects are Acute effects counterbalanced by peripheral vasodilation.81 Obstructive apnoeas are accompanied by pro- The pressor response in OSA may be due to found haemodynamic changes.71 Cyclical impaired vasodilatory mechanisms, as there is increases in systemic and pulmonary arterial preliminary evidence of an attenuated fore- blood pressure occur coincidently with arm blood flow response to acetylcholine in

obstructive events.72 Each apnoea can be con- patients with OSA (a Carlson, J Hedner, per- http://thorax.bmj.com/ sidered in three phases with respect to the sonal communication). This in turn indicates observed effects on blood pressure, heart rate, impaired endothelial dependent (nitric oxide sympathetic and parasympathetic nerve activ- mediated) vasodilation. Interestingly, the ity, and cardiac output.71 Phase 1 (recovery vasodilatory response to hypoxia in chemod- from the previous apnoea typically during the enervated rats is mediated by nitric oxide.82 early part of the next apnoea) is characterised Apart from nitric oxide, other vasoactive by minor pleural pressure swings, minimal mediators appear to be altered in OSA. These changes in heart rate and muscle sympathetic include eicosanoids,8' endothelin,84 and nerve activity (MSNA), and modest changes adenosine.8' Intrathoracic pressure swings in on September 26, 2021 by guest. Protected copyright. in oxygen saturation. As the apnoea pro- OSA may alter volume regulating hor- gresses (phase 2) there is progressive hypox- mones,8687 although some of these changes aemia, increasing pleural pressure swings, may potentially attenuate rises in blood pres- bradycardia (and possibly bradyarrhythmias), heightened MSNA, and an overall rise in blood pressure. With arousal and resumption saturation of ventilation (phase 3) oxygen * Dl 75, Sao2min 72% returns to normal. There are significant (? 170 o Di 6, Sao2min 91% . increases in heart rate and blood pressure ' 160 may rise to levels ranging from 200 to 300 (D 150 QI 140 0 0 mm Hg. In phase 3 MSNA increases, but this E 130 appears to be rapidly interrupted before the *- t E 120 0.~~~~~~ peak in blood pressure following apnoea. X 110 There is a fall in stroke volume during a)co 100 E 90 apnoea, particularly at its termination.7'74 95 90 85 80 75 70 The combination of fall in stroke volume and 100 rise in blood pressure suggests a substantial Arterial oxygen saturation (%) increase in total peripheral resistance. Figure 3 Mean aterialpressure at different oxygen The potentiation of MSNA activity during saturation levels during a hypoxic ramp test in two patients with different severity ofsleep apnoea. Patient 1 (dosed apnoea is likely to be the result of a combina- circles) had severe sleep apnoea (respiratory disturbance tion of apnoea and hypoxaemia. Somers and index (DI) 75 events per hour, minimum oxygen coworkers were able to demonstrate a 12 fold saturation in sleep (Sao2min) 72%) and patient 2 (open circles) had minimal obstructive sleep apnoea (DI 6 events increase in the MSNA response to hypoxia by per hour and Sao2min 91%). Valuesfrom two tests in voluntary apnoea in normal subjects.75 each patient are shown and thefitted polynomial regression However, the relative contribution of hyp- lines. Adaptedfrom reference 81 with permission. Obstructive sleep apnoea 759

sure. In addition, changes in intracranial an association between snoring (reported on pressure may influence blood pressure.88 questionnaire) and hypertension, but these Other evidence for a potential effect of OSA studies are typically also confounded by coex- on development of hypertension has been isting obesity or interpretation is difficult

provided by an elegant series of studies in rats because there are only a few cases of OSA or Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from exposed to seven weeks of intermittent hypertension.7' In a recent study of an hypoxia.89-9' These animals developed persis- Oxfordshire general practice, for example, the tent elevation of daytime mean arterial pres- number of overnight oxyhaemoglobin desatu- sure. Sectioning of the carotid sinus nerve rations was correlated with arm cuff pressure, and chemical peripheral sympathectomy with but this was not independent of age or body 6-hydroxydopamine eliminated the persistent mass index on multivariable analysis.28 elevation of blood pressure.90 9' However, only 5% of the subjects were obese Increased left ventricular wall thickening, and the actual numbers of individuals with independent of resting daytime blood pres- hypertension was not reported. In another sure, has been observed in OSA.92 This may study, examining the relationship between result from acute increases in left ventricular blood pressure and OSA in a sleep clinic pop- afterload leading to left ventricular wall stress ulation, patients with known hypertension or and increased sympathetic tone, causing a on antihypertensive agents were excluded.99 direct trophic effect on the myocardium.9' Despite these methodological limitations, Interestingly, both left and right ventricular such studies indicate that the independent hypertrophy have been found after repetitive association between classic measures of OSA hypoxia in the rat.8993 Left ventricular hyper- severity, such as apnoea index or minimum trophy has an adverse prognosis in itself and oxygen saturation in sleep, and hypertension may also lead to impaired myocardial con- is unlikely to be strong. We observed recently tractility, perhaps exacerbated by an adverse that morning but not evening blood pressure effect of tissue hypoxia on myocardial muscle was related to severity of OSA, independent metabolic demand.7' Recently, Malone and of central obesity and age. Nevertheless, these coworkers reported that nasal CPAP three variables only explained 18% of the improved left ventricular function in men variance in blood pressure, with the largest with OSA and idiopathic dilated cardio- contribution from central obesity.3" Future myopathy.94 Withdrawal of CPAP led to studies correlating blood pressure with other deterioration in myocardial function. Left measures, such as arousal index or even the ventricular hypertrophy also occurs more magnitude of the blood pressure response to frequently in patients who do not have the individual apnoeas, may prove an important normal nocturnal "dip" in measurements of line of investigation.

ambulatory 24 hour blood pressure.95 Patients Prospective studies examining the longitu- http://thorax.bmj.com/ with OSA certainly have episodic rises in dinal development of hypertension in blood pressure at night, but the overall pat- untreated patients with OSA compared with tern of 24 hour ambulatory blood pressure matched controls are unlikely to be per- shows a preserved diurnal pattern.96 formed due to the ethical issues of not treat- Despite a number of potential mechanisms ing sleep apnoea. One potential method of for the development of sustained hyperten- avoiding the problems of confounding vari- sion, there is no irrefutable evidence that ables and interpretation of cross sectional sta- OSA directly causes daytime hypertension.7197 tistical data is effectively to treat OSA and Sleep apnoea is a common finding in patients examine the subsequent response of blood on September 26, 2021 by guest. Protected copyright. attending hypertension clinics,7'9798 but this pressure. Factors such as weight loss and may be due to shared confounding factors alcohol consumption need to be controlled. such as central obesity33 and increasing age. Patients should be withdrawn from anti- Large scale epidemiological studies often find hypertensive medication ideally, but this may not be possible, particularly in patients with OSA.100 Since Coccagna and colleagues reported that blood pressure fell in five 170 patients with OSA after tracheostomy,1°' sev- 160 eral studies have looked at the blood pressure 150 response to treatment of sleep apnoea.98102103 140 1°°° 0 i Most of these studies are limited by small 130 numbers, confounded by parallel weight * z; I I change and antihypertensive medications or 120 ° 1° ° ° 10 1 E O°1I use of single cuff measurements as an index E 110 0~ I I of blood pressure. More recently two groups a- 100 0 m 4o 6 8 8 have reported decreased nocturnal and morn- 90 I L L0 ing awake intra-arterial blood pressure follow- 8 1 80 5 w ing CPAP therapy.104 105 Naughton and Pierce 70 I9 reported a fall in automated daytime blood 60 pressure readings in hypertensive patients 50 with OSA after one week of CPAP. 106 0 2 4 6 8 10 12 14 16 18 20 22 24 However, some of these patients were also receiving antihypertensive drugs. In a Figure 4 Fall in systolic and diastolic ambulatory blood pressure before (closed circles) study of and after (open circles) eight weeks of nasal CPAP in 14 men not using antihypertensive 24 hour ambulatory blood pressure in 19 agents. Adaptedfrom reference 107 with permission. men with OSA before and after eight weeks 760 McNamara, Grunstein, Sullivan

of nasal CPAP'07 there was a moderate fall in respiratory failure with hypercapnia."17 They 24 hour mean systolic and diastolic blood clearly represent the severe end of the pressure during the day but only in the sys- spectrum of OSA, and it is evident that upper tolic blood pressure at night (fig 4). This fall airway obstruction during sleep in the will produce the occurred in the 14 patients with good CPAP presence of other factors Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from compliance but not in those unable to use "Pickwickian" patient. What are the factors CPAP. This is the first study to report a that will transform a patient with OSA into a decrease in ambulatory daytime blood pres- "Pickwickian" patient? Firstly, there is likely sure following treatment of OSA. Such a day- to be a complex interrelationship between time fall may not be unexpected in view of these factors. Some patients develop hyper- the high sympathetic nerve traffic during capnia without evidence of right heart failure wakefulness observed in untreated patients and the reverse is also true. Most studies with OSA7880 and the decrease in sympathetic examining this interrelationship are limited in activity seen following nasal CPAP size and vary in the method of assessing right treatment.71 Recent preliminary data have heart failure. The key factors suggested are also suggested that patients with OSA may presence of chronic airflow limitation, obe- have abnormally high responses to daytime sity, awake hypoxaemia, hypercapnia, pro- stressors such as mental arithmetic which are found nocturnal hypoxaemia, and alcohol reversible with treatment (JH Peter, personal consumption."'1122 A recent preliminary communication). One criticism of this 24 report has shown that daytime Pco2 and hour ambulatory blood pressure study is the FEV1 explain 29% of the variability in pul- lack of a placebo treated control group to monary artery pressure in sleep apnoea. avoid "regression to the mean." This phe- Daytime hypoxaemia had no independent nomenon has not, however, been observed in contribution. Of 78 patients with pulmonary 24 hour ambulatory blood pressure studies'08 hypertension and FEV, >1-51, 30% had a and "placebo" CPAP has not been success- Po2 >80 mmHg. These data suggest that fully used in any short or long term study. patients with OSA can often develop pul- Such a technique would be obvious to the monary hypertension in the absence of day- patient and bed partner as snoring and upper time hypoxaemia.12' Treatment of airway obstruction would continue and the "Pickwickian" patients and other patients expected improvement in daytime sleepiness with OSA and leads with CPAP treatment would not occur. to a normalisation of both hypercapnia and pulmonary artery pressure. 120 122 OSA, STROKE, MYOCARDIAL INFARCTION AND DEATH

The advent of nasal CPAP has prevented Treatment http://thorax.bmj.com/ large studies investigating the natural history Despite many drug trials and attempts at of untreated OSA. However, in certain sleep electrical stimulation of the upper airway,'23 disorder centres established in the 1970s long the most accepted treatments for OSA are term data are available which strongly suggest nasal CPAP and surgery. that the mortality risk is increased in untreated sleep apnoea.109-"12 He et al 109 CPAP observed an increased cumulative mortality in The advent of nasal CPAP revolutionised untreated patients with an apnoea index positive airway pressure ventilation and above 20 compared with those having an allowed a wider range of patients to be on September 26, 2021 by guest. Protected copyright. index below 20. Tracheostomy or CPAP treated. Nasal CPAP is an effective treatment treatment, but not uvulopalatopharyngoplasty but compliance is variable.'24 When accurate (UPPP), reduced the mortality risk. These hours of use are measured, patient compli- authors, however, did not provide informa- ance is 70%.125 The improvements in mask tion on the cause of death, but studies from and machine technology are likely to further Stanford"0 and Haifa"2 suggest an excess of increase compliance as will humidification cardiovascular deaths in OSA. A number of devices aimed at reducing nasal side effects. groups have reported an increased risk of It is not known whether the variation of inspi- myocardial infarction"'3 114 and stroke"5 in ratory and expiratory pressure'26 with the use sleep apnoea. Snoring is a strong risk factor of the more expensive bi-level positive airway for sleep related strokes while sleep apnoea pressure devices will influence patient com- symptoms (snoring plus reported apnoeas or pliance. Indeed, some authors have suggested excessive daytime sleepiness) increase the risk that reduction of expiratory positive airway of cerebral infarction with an odds ratio of pressure may lead to incomplete improve- 80.1"' Cerebral blood flow changes associated ment in upper airway calibre.'27 Although the with sleep apnoea may certainly be linked to severity of OSA may influence compliance, these observations."16 However, there is no patient symptoms and their response to information currently available on an CPAP provide a powerful reinforcement to increased thrombotic tendency or accelerated machine use. atherosclerosis in sleep apnoea. Patients with OSA and daytime respiratory failure may benefit from nasal intermittent OSA, RESPIRATORY FAILURE AND PULMONARY positive pressure ventilation (nIPPV). Such HYPERTENSION patients may require high CPAP pressures The "Pickwickian" patients described in and, in the short term, may prefer nasal venti- older literature were often characterised by lation.'28 These treatment periods of nasal marked obesity, right heart failure, and awake ventilation may be brief and, after improve- Obstructive sleep apnoea 761

ment in blood gas levels, nasal CPAP can be "cures" will still leave a substantial degree of commenced. sleep disordered breathing. Results are fre- quently reported without detailed examina- SURGERY tion of sleep quality, hypopnoea, or arousals. Tracheostomy Polo et al,141 while observing a decrease in Before the introduction of nasal CPAP as a obstructive apnoea following UPPP, found an Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from treatment for OSA tracheostomy was the increase in partial upper airway obstruction main method of treatment. The current indi- thus stressing the importance of close postop- cation for tracheostomy is in patients with erative follow up. This may also explain why severe OSA who have been unable to comply investigators reporting apnoea reduction fol- with CPAP. Our clinical practice is to fully lowing UPPP find no correlation between evaluate such patients as hospital admissions subjective and objective improvement."5 with repeat polysomnography, intensive sup- One of the problems in examining the long port for nasal CPAP treatment including term efficacy of UPPP is that most papers ENT review, humidification of inspired air report short term results or have incomplete through the CPAP machine and, if necessary, follow up. Recently Larsson et al36 reported a customised CPAP masks. Tracheostomy can two year follow up of 50 consecutive patients produce significant morbidity and may be a who underwent UPPP. After six weeks 60% problem in the morbidly obese, fat n-ecked of patients were classified as "responders" individual. However, skilful minimalist (50% decrease in apnoea index or number of surgery and regular nursing care allow tra- apnoeas per hour). At two years the cheostomy to be a therapeutic option in some "response rate" had fallen to 36%. In addi- patients. tion there are data"35 137-138 suggesting an appreciable morbidity and mortality for Facial reconstructive surgery patients treated by UPPP, while the increased Many patients with OSA have abnormalities mortality of OSA is not modified by UPPP.109 in facial structure on cephalometry"29 and it There is also some evidence that the vibration has been suggested that correction of such of snoring may protect against further upper factors by maxillofacial surgery will lead to airway obstruction,'42 suggesting that removal cure in sleep apnoea.130 However, the correla- of the vibrating tissues (plate, uvula) may tion of mechanical, characteristics of the critically alter important upper airway affer- pharyngeal airway with cephalometry is only ent receptors, thereby causing narrowing and indirect and, moreover, such surgery is airway closure during sleep at other sites expensive, requiring several operative proce- postoperatively. dures. Virtually all cases have been reported Recently a novel technique using naso- by one group130 and -therefore -efficacy will pharyngoscopy during sleep, simultaneous need to be shown in wider clinical trials using use of nasal CPAP to manipulate intra- http://thorax.bmj.com/ the expertise of several surgeons. pharyngeal pressure, and digitised computer assessment of airway lumen size has been Uvulopalatopharyngoplasty (UPPP) employed to select patients who may respond This operation was developed by Ikematsu to UPPP.140 The authors could identify for the treatment of heavy snoring in the early patients with exclusively nasopharyngeal nar- 1950s.131 A similar operation was used by vet- rowing of the passive airway. Such patients erinary surgeons to treat bothersome breath- had 50% improvement in severity of sleep ing in bulldogs, a breed of dog potentially apnoea four months after UPPP compared on September 26, 2021 by guest. Protected copyright. useful as a large animal model of OSA.132 The with no change in the non-exclusively introduction of UPPP for the treatment of nasopharyngeal occluders. There was, how- OSA into North America occurred in 1981,133 ever, a deterioration in the severity of OSA at the same year as the efficacy of nasal CPAP 14 month follow up. Moreover, "responders" in OSA was first reported.'34 Initially there still had a substantial degree of disordered was great enthusiasm for performing UPPP, breathing in sleep at 14 months after surgery often with minimal patient assessment. The (mean apnoea index 58, with 35% of total accumulated data from many studies over the sleep time spent in disordered breathing). past decade, however, suggest extreme cau- Such pre-UPPP assessment techniques are tion in performing this form of surgery for complex and expensive and current results do OSA.131-138 There is still no consensus on the not seem to warrant more widespread use. It most appropriate methods to assess the likely is possible that better results may be pro- success of UPPP in an individual duced in patients with milder OSA. It is patient.139-'40 Awake tests such as computed therefore important to stress that UPPP tomographic scanning, nasopharyngoscopy, remains an experimental treatment method in and even multilevel upper airway pressure OSA. Careful objective preoperative and measurement in sleep have been used."39140 postoperative assessment of breathing during Often areas that narrow on preoperative sleep with long term follow up should be per- assessment are still narrowed postoperatively, formed. or new areas of upper airway narrowing are found.-35-139 Moreover, the response to surgery is highly variable. Most papers define Conclusion a "cure" as a 50% reduction in the number of Obstructive sleep apnoea is a common dis- apnoeas per hour of sleep; in many patients order which will become an increasingly with repetitive apnoea, however, these important part of respiratory practice. The 762 McNamara, Grunstein, Sullivan

best methods of diagnosis are costly and Allen K, Saunders NA. A comparison of clinical assess- ment and home oximetry in the diagnosis of obstructive increased sophistication of screening devices sleep . Am Rev Respir Dis 1993;147:50-3. and interpreting physicians are required to 23 Redline S, Tosteson T, Tishler PV, Carskadon MA, Millman RP. Studies in the genetics of obstructive sleep reduce the need for full polysomnography. apnea: familial aggregation of symptoms associated with There is a wide range of clinical disorders sleep related breathing disturbances. Am Rev Respir Dis associated with OSA which affect the decision 1992;145:440-4. Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from 24 Issa FG, Sullivan CE. Alcohol, snoring and . to treat patients. However, in some cases the J Neurol Neurosurg Psychiatry 1982;45:353-9. link between OSA and these clinical disorders 25 Tan ETH, Lambie DG, Johnson RH, Robinson BJ, Whiteside EA. Sleep apnoea in alcoholic patients after has not been proved convincingly. Studies withdrawal. Clin Sci 1985,69:655-61. clarifying these issues will allow better selec- 26 Chan CS, Grunstein RR, Bye PTB, Woolcock AJ, tion of patients requiring treatment. Nasal Sullivan CE. with severe chronic airflow limitation-comparison of hypercapnic CPAP is the standard treatment for OSA and and eucapnic patients. Am Rev Respir Dis 1989;140: tracheostomy should only be considered in 1274-8. 27 Jalleh R, Fitzpatrick MF, Mathur R, Douglas NJ. Do patients truly unable to tolerate CPAP. patients with the sleep apnea/hypopnea syndrome drink Uvulatopalatopharyngoplasty and facial more alcohol? Sleep 1992;15:319-21. reconstructive surgery should only be consid- 28 Stradling JR, Crosby JH. Relation between systemic hypertension and sleep or snoring: analysis ered as part of objective controlled trials. in 748 men drawn from a general practice. BMJ 1990; 300:75-8. Dr McNamara is a recipient of a National Health and 29 Homer RL, Mohiaddin RH, Lowell DG, Shea SA, Medical Research Council of Australia postgraduate scholar- Burman ED, Longmore DB, et al. Sites and sizes of fat ship and Dr Grunstein is a recipient of a National Heart deposits around the in obese patients with Foundation of Australia overseas travelling fellowship. obstructive sleep apnea and weight matched controls. EurRespirJ 1989;2:613-22. 30 Eller K, Woodson H, Gay SB, Suratt PM. Pharyngeal fat in obstructive sleep apnea. Am Rev Respir Dis 1992; 145:214. 1 Saunders NA, Sullivan CE (eds). Sleep and breathing. New 31 Schwab RJ, Prasad A, Gupta KB, Pack Al, Listerud J, York: Marcel Dekker, 1993. Hoffman EA, et al. Fat and water measurements of 2 Guilleminault C, Tillian A, Dement W. The sleep apnea upper airway soft tissues in normal subjects and patients syndromes. Annu Rev Med 1976;27:465-84. with sleep disordered breathing using magnetic reso- 3 Guilleminault C, Partinen M. Obstructive sleep apnea syn- nance proton spectroscopy. Am Rev Respir Dis 1992; drome. Clinical research and treatment. New York: Raven 145:214. Press, 1990. 32 Davies RJO, Stradling JR. The relationship between neck 4 Guilleminault C, Stoohs R, Clerk A, Simmons J, circumference, radiographic pharyngeal anatomy, and Labanowski M. From obstructive sleep apnea syndrome obstructive sleep apnea. Eur RespirJ 1990;3:509-14. to upper syndrome: consistency of 33 Grunstein RR, Wilcox I, Yang TS, Gould Y, Hedner JA. daytime sleepiness. Sleep 1992;15(Suppl):Sl3-6. Snoring and sleep apnea in men-association with cen- 5 Gould GA, Whyte KF, Rhind GB, Airlie MA, Catterail tral obesity and hypertension. IntJ Obesity (in press). JR, Shapiro CM, et al. The sleep hypopnea syndrome. 34 Begle RL, Badr S, Skatrud JB, Dempsey JA. Effect of lung Am Rev Respir Dis 1988;137:895-8. inflation on puimonary resistance during NREM sleep. 6 Ambrogetti A, Olsen LG, Saunders NA. Differences in Am Rev RespirDis 1990;141:854-60. the symptoms of men and women with obstructive sleep 35 Hoffstein V, Zamel N, Phillipson EA. Lung volume apnea. Aust NZJ'Med 1991;21:863-6. dependence of pharyngeal cross-sectional area in 7 Rosen CL, D'Andrea L, Haddad GG. Adult criteria for patients with obstructive sleep apnea. Am Rev Respir Dis http://thorax.bmj.com/ obstructive sleep apnea do not identify children with seri- 1984;130:175-8. ous obstruction. Am Rev RespirDis 1992;146:1231-4. 36 Lopata M, Onal E. Mass loading, sleep apnea, and the 8 Guilleminault C, Tilkian A, Lehrman K, Forno L, pathogenesis of obesity hypoventilation. Am Rev Respir Dement WC. Sleep apnea syndrome: states of sleep and Dis 1982;126:640-5. autonomic dysfunction. J Neurol Neurosurg Psychiatry 37 Smirne S, lannaccone S, Ferini-Strambi L, Comola M, 1977;40:718-24. Colombo E, Nemni R. Muscle fibre type and habitual 9 Gislason T, Ahlmqvist M, Ariksson G, Taube A, Boman snoring. Lancet 1991;337:597-9. G. Prevalence of sleep apnea syndrome amongst 38 Wade AJ, Marbut MM, Round JM. Muscle fibre type and Swedish men. J Clin Epidemiol 1988;41:571-6. the aetiology of obesity. Lancet 1990;335:805-8. 10 Cirignotta F, d'Alessandro R, Partinen M, Zucconi M, 39 Rubinstein I, Colaptino N, Rotstein LE, Brown IG, Cristina E, Gerardi R, et al. Prevalence of every night Hoffstein V. Improvement in upper airway function snoring and obstructive sleep among 30-69 year after weight loss in patients with obstructive sleep on September 26, 2021 by guest. Protected copyright. old men in Bologna, Italy. Acta Neurol Scand 1989; apnea. Am Rev Respir Dis 1988;138:1192-5. 79:366-72. 40 Zwillich CW, Pierson DJ, Hofeldt FD, Lufkin EG, 11 Stradling JR, Crosby JH. Predictors and prevalence of Weil JV. Ventilatory control in myxedema and hypo- obstructive sleep apnoea and snoring in 1001 middle thyroidism. NEnglJMed 1975;292:662-5. aged men. Thorax 1991;46:85-90. 41 Grunstein RR, Sullivan CE. Hypothyroidism and sleep 12 Bearpark H, Elliott L, Cullen S, Grunstein R, Schneider apnea. mechanisms and management. Am J Med 1988; H, Althaus W, et al. Home monitoring demonstrates 85:775-9. high prevalence of sleep disordered breathing in men in 42 Petroff BJ, Kelly AM, Rubinstein NA, Pack AI. Effect of the Busselton population. Sleep Res 1991;20A:41 1. hypothyroidism on myosin heavy chain expression in rat 13 Young T, Zaccaro D, Leder R, et al. Prevalence and cor- pharyngeal dilator muscles. J Appl Physiol 1992;73: relates of sleep disordered breathing in the Wisconsin 179-87. sleep cohort study. Am Rev Respir Dis 1991 :143:A380. 43 Rajagopal KR, Abbrecht PH, Derderian SS, Pickett C, 14 Jennum P, Soul A. Epidemiology of snoring and obstruc- Hofeldt F, Tellis CJ, et al. Obstructive sleep apnea in tive sleep apnoea in a Danish population age 30-60. hypothyroidism. Ann Intern Med 1984;101:471-4. J Sleep Res 1992;1:240-4. 44 Grunstein RR, Ho KY, Sullivan CE. Sleep apnea and 15 Stoohs R, Guilleminault C. MESAM 4: an ambulatory acromegaly. Ann Intern Med 1991;115:527-32. device for the detection of patients at risk for obstructive 45 Grunstein RR, Stewart DA, Handelsman DJ, Ho KY, sleep apnea syndrome (OSAS). Chest 1992;101:1221-7. Sullivan CE. Growth hormone and sex steroids in sleep 16 Sullivan CE, Issa FG. Obstructive sleep apnea. Clin Chest apnoea. In: Home J, ed. Sleep 90. Bochum: Pontenagel Med 1985;6:632-4. Press, 1990:422-5. 17 Crocker BD, Olson LG, Saunders NA, Hensley MJ, 46 Shipley JE, Starkman MN. EEG sleep in Cushing's dis- Mckeon JL, Muree Allen K, et al Estimation of the ease. Sleep Res 1990;19:102. probability of disturbed breathing during sleep before a 47 Schneider BK, Pickett CK, Zwillich CW, Weil JV, sleep study. Am Rev RespirDis 1990;142:14-8. McDermott MT, Santen RJ, et al. Influence of testo- 18 Hoffstein V, Szalai J. Predictive value of clinical features sterone on breathing during sleep. 7 Appl Physiol 1986; in diagnosing obstructive sleep apnea. Sleep 1993;16: 61:618-23. 118-22. 48 Johnson MW, Arch AM, Remmers JE. Induction of the 19 Davies RJO, Ali NJ, Stradling JR. Neck circumference obstructive sleep apnea syndrome in a woman by exoge- and other clinical features in the diagnosis of obstructive nous androgen administration. Am Rev Respir Dis 1984; sleep apnoea syndrome. Thorax 1992;47:101-5. 129:1023. 20 Stradling J. Consensus report: sleep studies for sleep- 49 Pickett CK, Regensteiner JG, Woodard WD, Hagerman related breathing disorders. J Sleep Res 1992;1:265-73. DG, Weil JV, Moore LG. Progestigin and estrogen 21 Douglas NJ, Thomas S, Jan MA. Clinical value of reduce sleep-disordered breathing in post-menopausal polysomnography. Lancet 1992;339:347-50. women. JAppl Physiol 1989;66:1656-61. 22 Gyulay SG, Olson LG, Hensley MJ, King MT, Muree 50 Stewart D, Grunstein RR, Berthon-Jones M, Handelsman Obstructive sleep apnoea 763

DJ, Sullivan CE. Androgen blockade does not affect on 02 saturation and systemic and pulmonary arterial sleep disordered breathing or chemosensitivity in men pressure in anesthetised dogs. Am Rev Respir Dis 1992; with obstructive sleep apnea. Am Rev Respir Dis 1992; 146:1402-10. 146:1389-93. 78 Leuenberger UA, Jacob EM, Sweer LW, Zwillich CW, 51 Cistulli P, Barnes D, Grunstein R, Sullivan CE. Effect of Sinoway LI. Hypoxemia and surges of sympatheric short-term hormone replacement in post menopausal activity mediate blood pressure oscillations in obstruc-

women with obstructive sleep apnoea. Am Rev Respir tive sleep apnea. Circlation 1992;86(Suppl):I-637. Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from Dis 1993;147:A686. 79 Hedner J, Ejnell H, Sellgren J, Hedner T, Wallin G. Is 52 Cistulli P, Sullivan CE. Sleep apnea in Marfan's syn- high and fluctuating muscle nerve sympathetic activity drome. Am Rev Respir Dis 1993;147:645-8. in the sleep apnea syndrome of pathogenic importance 53 Grunstein RR, Handelsman DJ, Lawrence SJ, Blackwell for the development of hypertension. J Hypertnsion C, Caterson ID, Sullivan CE. Neuroendocrine dysfunc- 1988;6:S529-31. tion in sleep apnea. Reversal by continuous nasal posi- 80 Becker MJ, Strohl KP, Thomas MD. Augmented tive airway pressure. J Clin Endocrinol Metab 1989; peripheral chemoreflex gain in patients with obstructive 68:352-8. sleep apnea syndrome. Circulation 1992;86(Suppl):I- 54 Stradling JR, Thomas G, Williams P, Warley AH, 636. Freeland A. Effect of adenotonsillectomy on noctumal 81 Hedner J, Wilcox I, Laks L, Grunstein RR, Sullivan CE. hypoxemia, sleep disturbance and symptoms in snoring A specific and potent pressor effect of hypoxia in children. Lancet 1990;335:249-53. patients with sleep apnea. Am Rev Respir Dis 1992; 55 Rudman D, Feller AG, Nagraj HS, Gergans GA, Lalitha 146:1240-5. PY, Goldberg AF, et al. Effects of human growth hor- 82 Sun M-K, Reis DJ. Evidence nitric oxide mediates the mone in men over 60 years old. N Engl J Med vasodepressor response to hypoxia in sino-denervated 1990;323:1-6. rats. Life Sci 1992;50:555-65. 56 Roth T, Roehrs T, Carskadon M, Dement W. Daytime 83 Krieger J, Benzoni D, Sforza E, Sassard J. Urinary excre- sleepiness and alertness. In: Kryger MA, Roth T, tion of prostanoids during sleep in obstructive sleep Dement WC, eds. )inciples and practice ofsleep medicine. apnea patients. Clin Exp Pharmacol Physiol 1991;18: Philadelphia: Saunders, 1989:14-23. 511-5. 57 Bedard MA, Montplaisir J, Richer F, Malo J. Nocturnal 84 Ehlenz K, Herzog P, von Wichert P, Kaffarnik H, Peter hypoxemia as determinant of vigilance impairment in JH. Renal excretion of endothelin in obstructive sleep sleep apnea syndrome. Chest 1991;100:367-70. apnea syndrome. In: Gautier C, Escourrou P, Curzi- 58 Kales A, Caldwell AB, Cadieux RJ, Vela-Bueno A, Ruch Dascalova, eds. Colloque INSERM. Vol. 217. John LG, Mayes SD, et al. Severe obstructive sleep apnea-II: Libby Eurotext, 1991:403. associated psychological and psychosocial conse- 85 Lloyd HGL, Laln L, Grunstein R, et al. Plasma adeno- quences. J Chron Dis 1985;38:426-37. sine concentrations during episodes of obstructive sleep 59 Findley L, Unverzagt M, Suratt P. Automobile accidents apnea. In: Jacobson KA, Daly JW, Manganiello V, eds. in patients with obstructive sleep apnoea. Am Rev Respir Purines in celular signalling: targets for new drugs. New Dis 1988;138:337-40. York: Springer Verlag, 1990:385-6. 60 Poceta JS, Timms RM, Jeong D-U, Ho S-L, Erman MK, 86 Krieger J, Lals L, Wilcox I, Grunstein RR, Costas LJV, Mitler MM. Muliple wakefulness test in obstructive McDougall JG, et al. Atrial natriuretic peptide release sleep apnea syndrome. Chest 1992;101:893-7. during sleep in patients with obstructive sleep apnea 61 Haraldsson P, Carenfelt C, Laurell H, Tornross J. Driving before and during treatment with nasal continuous posi- vigilance simulator test. Acta Otolaryngol (Stockh) tive airway pressure. Clin Sci 1989;77:407-1 1. 1990;110: 13640. 87 Ehlenz K, Peter J, Kaffarnik H, Von Wichert P. 62 Findley L, Fabrizio M, Knight H, Norcross BB, La Forte Disturbances in the volume regulating hormone sys- AJ, Suratt PM. Driving simulator performance in tems: a key to the pathogenesis of hypertension in patients with sleep apnea. Am Rev Respir Dis 1989; obstructive sleep apnea? Pneumonologie 1991;45(Suppl 1): 140:529-30. 239-45. 63 Bearpark H, Fell D, Grunstein RR, Leeder S, Berthon- 88 Jennum P, Borgesen SE. Intracranial pressure in Jones M, Sullivan CE. Snoring and sleep apnea have obstructive sleep apnea. Chest 1989;95:279-83. adverse effects on driving: implications for road safety. 89 Fletcher EC, Lesske J, Qian W, Millar CC, Unger T. MedJAust (in press). Repetitive, episodic hypoxia causes diurnal elevation of http://thorax.bmj.com/ 64 Findley L, Weiss J, Jabour E. Serious automobile crashes blood pressure in rats. Hypertension 1992;19:555-61. caused by undetected sleep apnea. Arch Intern Med 90 Fletcher EC, Lesske J, Behm R, Millar CC, Stauss H, 1991;151:1451-2. Unger T. Carotid , systemic blood pres- 65 Parsons M. Fits and other causes of loss of consciousness sure and chronic episodic hypoxia mimicking sleep whilst driving. QJtMed 1986;227:295-303. apnea. JApplPhysiol 1992;72:1978-84. 66 Rajogobal KR, Bennett LL, Dillard TA, Tellis CJ, 91 Fletcher EC, Lesske J, Culman J, Miller CC, Unger T. Tenholder MF. Overnight nasal CPAP improves hyper- Sympathetic denervation blocks blood pressure eleva- somnolence in sleep apnea. Chest 1986;90:172-6. tion in episodic hypoxia. Hypertnsion 1992;20:612-9. 67 Greenberg GD, Watson RK, Deptula D. Neuro- 92 Hedner JA, Ejnell H, Caidahl K. Left ventricular hyper- psychological dysfunction in sleep apnea. Sleep 1987; trophy independent of hypertension in patients with

10:254-62. obstructive sleep apnea. JHypertnsion 1990;8:941-6. on September 26, 2021 by guest. Protected copyright. 68 Bearpark H, Grunstein R, Touyz S, Channon L, Sullivan 93 Nattie EE, Bartlett D, Johnson K. Pulmonary hyperten- CE. Cognitive and psychological dysfunction in sleep sion and right ventricular hypertrophy caused by inter- apnea before and after treatment with CPAP. Sleep Res mittant hypoxia in the rat. Am Rev Respir Dis 1978; 1987;17:303. 118:653-7. 69 Cheshire K, Engeiman H, Deary I, Shapiro C, Douglas 94 Malone S, Liu PP, Holloway R, Rutherford R, Zie A, NJ. Factors impairing daytime performance in patients Bradley TD. Obstructive sleep apnoea in patients with with sleep apnea/hypopnea syndrome. Arch Intern Med dilated cardiomyopathy: effects of continuous positive 1992;152:538-41. airway pressure. Lancet 1991;338:1480-4. 70 Montplaisir J, Bedard MA, Richer F, Rouleau I. 95 Verdecchia P, Schillaci G, Guerrieri M, Gatteschi C, Neurobehavioural manifestations of obstructive sleep Benemio G, Boldrini F, et al. Circadian blood pressure apnea syndrome before and after treatment with contin- changes and left ventricular hypertrophy in essential uous positive airway pressure. Sleep 1992;15:S17-9. hypertension. Circulation 1990;81:528-36. 71 Working Group on OSA and Hypertension. Obstructive 96 Wilcox I, Grunstein RR, Collins FL, Doyle JM, Kelly sleep apnea and blood pressure. What is the relation- DT, Sullivan CE. Circadian rhythm of blood pressure ship? Blood Pressure 1993 (in press). in patients with obstructive sleep apnea. Blood Pressure 72 Coccagna G, Mantovani M, Brignani F, Parchi C, 1992;1:219-22. Lugaresi E. Continuous recording of pulmonary and 97 Stradling JR. Sleep apnoea and systemic hypertension. systemic arterial pressure during sleep in syndromes of Thorax 1989;44:984-9. hypersomnia with periodic breathing. Bull Physiopathol 98 Fletcher EC, DeBehnke RD, Lavoi MS, Gorin AB. Respir 1972;8:1159-72. Undiagnosed sleep apnea in patients with essential 73 Tolle FA, Judy WV, Yu PL, Markand ON. Reduced hypertension. Ann Intern Med 1985;103:190-5. stroke volume related to pleural pressure in obstructive 99 Hoffstein V, Mateika J. Evening to morning blood pres- sleep apnea. JAppl Physiol 1983;55: 1718-24. sure variations in snoring patients with and without 74 Garpestad E, Katayama J, Parker JA, Ringler J, Lilly J, obstructive sleep apnea. Chest 1992;101:379-84. Yasuda T, et al. Stroke volume and cardiac output 100 Isaksson H, Svanborg E. Obstructive sleep apnea syn- decrease at the termination of obstructive apneas. JAppl drome in male hypertensives refractory to drug therapy. Physwil 1992;73- 1743-8 _ Nocturnal automatic blood pressure measurements an 75 Somers VK, Mark AL, Abboud FM. Sympathetic activa- aid to diagnosis? Clin Exp Hypertens 1991;13:1195-212. tion by hypoxia and hypercapnia-implications for sleep 101 Coccagna G, Mantoveni M, Brignoni F, Pardu C, apnea. Clin Exp Hyperns 1988;10:413-22. Lugaresi E. Tracheostomy in hypersomnia with periodic 76 Ringler J, Basner RC, Shannon R, Schwartzstein R, breathing. Bull Eur Physiopathol Respir 198 1;8:117-28. Manning H, Weinberger SE, et al. Hypoxia alone does 102 Guilleminault C, Simmons FB, Motta J, Cummiskey J, not explain blood pressure elevations after obstructive Rosekind M, Shroeder JS, et al. Obstructive sleep apnea apneas. JAppl Physiol 1990;69:2143-8. syndrome and tracheostomy: long term follow up expe- 77 Iwase N, Kikuchi Y, Hida W, Miki H, Taguchi 0, rience. Arch Intern Med 1981;141:985-8. Satoh M, et al. Effects of repetitive airway obstruction 103 Burack B, Pollack C, Borowiecki B, Weitzman E. The 764 McNamara, Grunstein, Sullivan

hypersomnia-sleep apnea syndrome: a reversible major 123 Edwards LC, Daniels BK, Stanson AW, Sheedy PF, cardiovascular hazard (Abstract). Circulation 1977; Shepard JW. The effect of transcutaneous electrical 56:17. stimulation during wakefulness and sleep in patients 104 Jennum P, Wildschiotz G, Christensen NJ, Schwartz T. with obstructive sleep apnea. Am Rev Respir Dis 1992; Blood pressure, catecholamines and pancreatic polypep- 146:1030-6. tide in obstructive sleep apnea with and without nasal 124 Sullivan CE, Grunstein RR. Continuous positive airway

continuous positive airway pressure (nCPAP) treat- pressure in sleep disordered breathing. In: Kryger MH, Thorax: first published as 10.1136/thx.48.7.754 on 1 July 1993. Downloaded from ment. AmJ Hypertens 1989;2:847-52. Dement WC, Roth TP, eds. Principles and practice of 105 Mayer J, Becker H, Brandenburg U, Penzel T, Peter JH, sleep medicine. Philadelphia: Saunders, 1989:559-70. v. Wichert P. Blood pressure and sleep apnea: results of 125 Reeves-Hoche MK, Meck R, Zwillich CW. Long term long-term nasal continuous positive airways pressure compliance with nasal continuous positive airway pres- therapy. Cardiology 1991;79:84-92. sure (CPAP). Am Rev Respir Dis 1992;145:A169. 106 Naughton M, Pierce R. Effects of nasal continuous posi- 126 Sanders MH, Kern N. Obstructive sleep apnea treated tive airway pressure on blood pressure and body mass by independently adjusted inspiratory and expiratory index in obstructive sleep apnoea. Aust NZ J Med positive airway pressures via nasal mask: physiological 1991;21:917-9. and clinical implications. Chest 1990;98:317-24. 107 Wilcox I, Grunstein RR, Hedner JA, Doyle JM, Kelly 127 Gugger M, Vock P. Effect of reduced expiratory pressure DT, Fletcher PJ, et al. Effect of short term treatment on pharyngeal size during nasal positive airway pressure with nasal continuous positive airway pressure on sys- in patients with sleep apnoea: evaluation by continuous temic blood pressure in obstructive sleep apnea. Sleep computed tomography. Thorax 1992;47:809-13. (in press). 128 Piper AJ, Sullivan CE. Effects of short-term NIPPV in 108 Staessen JA, Fagard RH, Lijnen PJ, Thijs L, Van Hoof the management of patients with severe obstructive R, Amery AK. Ambulatory blood pressure monitoring sleep apnea and hypercapnia. Chest 1993 (in press). in clinical trials. Jf Hypertens 1991;9(Suppl 1):S 13-9. 129 Shepard JW, Gefter WB, Guilleminault C, Hoffman EA, 109 He J, Kryger M, Zorick F, Conway W, Roth T. Hoffstein V, Hudgel DW, et al. Evaluation of the upper Mortality and apnea index in obstructive sleep apnea. airway in patients with obstructive sleep apnea. Sleep Experience in 385 patients. Chest 1988;94:9-14. 1991;14:361-71. 110 Partinen M. Guilleminault C. Daytime sleepiness and 130 Riley RW, Powell NB, Guilleminault C. Maxillofacial vascular morbidity at seven-year follow-up in obstruc- surgery and obstructive sleep apnea. A review of 80 tive sleep apnea patients. Chest 1990;97:27-32. patients. Otolaryngol Head Neck Surg 1989;101:353-61. 111 Thorpy MJ, Lederich PS, Glovinsky PB, Burack B, Sher 131 Ikematsu T. Study of snoring. 4th report: therapy. JJ_ap AE, McGregor P, et al. Nocturnal death in patients with Otorhinolaryngol 1964;64:434-5. obstructive sleep apnoea: the Montefiore long term 132 Hendriks JC, Kline LR, Kovalski RJ, O'Brien JA, follow up study. Sleep Res 1989;18:316. Morrison AR, Pack AI. The English bulldog: a natural 112 Lavie P, Berger I, Yoffe N, Robin A-H, Chillag A, model of sleep-disordered breathing. J Appl Physiol Zomer J, et al. Long-term morbidity and mortality of 1987;63: 1344-50. SAS patients. I Sleep Res 1992;1:(Suppl 1): 131. 133 Fujita S, Conway W, Zorich F, Roth T. Surgical cor- 113 Hung J, Whitford EG, Parsons RW, Hillman DR. rection of anatomic abnormalities in obstructive sleep Association of sleep apnea and myocardial infarction in apnea syndrome: uvulopalatopharyngoplasty. Otolaryngol men. Lancet 1990;336:261-4. Head Neck Surg 1981;89:923-34. 114 Koskenvuo M, Kaprio J, Telakivi T, Partinen M, 134 Sullivan CE, Berthon-Jones M, Issa FG, Eves L. Heikkila K, Sarna S. Snoring as a risk factor for Reversal of obstructive sleep apnoea by continouous ischemic heart disease and stroke. BMJ7 1987;294: positive airway pressure applied through the nose. 16-19. Lancet 1981;i:862-5. 115 Palomaki H, Partinen M, Erkinjuntti T, Kaste M. 135 Rodenstein DO. Assessment of uvulopalatopharyngo- Snoring, sleep apnea syndrome, and stroke. Neurology plasty for the treatment of sleep apnea syndrome. Sleep 1992;42(Suppl 6):75-81. 1992;15(suppl):S56-62. 116 Klingelhofer J, Hajak G, Sander D, Schulz-Varszegi M, 136 Larsson H, Carlsson-Nordlander B, Svanborg E. Long- Ruther E, Conrad B. Assessment of intracranial hemo- time follow-up after UPPP for obstructive sleep apnea dynamics in obstructive sleep apnea syndrome. Stroke syndrome. Acta Otolaryngol (Stockh) 1991;111:582-90. 1992;23: 1427-33. 137 Harmon JD, Morgan W, Chaudhary B. Sleep apnea: http://thorax.bmj.com/ 117 Burwell C, Robin E, Whaley R, Bikelman A. Extreme morbidity and mortality of surgical treatment. South obesity associated with alveolar hypoventilation: a MedJ 1989;82:161-4. Pickwickian syndrome. Am J Med 1956;21:811-8. 138 Walker EB, Frith RW, Harding DA, Cant BR. 118 Bradley TD, Rutherford R, Grossman RF, Lue F, Uvulatopharyngoplasty in severe idiopathic obstructive Zamel N, Moldofsky H, et al. Role of daytime hypox- sleep apnoea syndrome. Thorax 1989;44:205-8. emia in the pathogenesis of right heart failure in the 139 Shepard JW, Thawley SE. Evaluation of the upper air- obstructive apnea syndrome. Am Rev Respir Dis 1985; way by computed tomography in patients undergoing 131:835-9. uvulopalatopharyngoplasy for obstructive sleep apnea. 119 Whyte KF, Douglas NJ. Peripheral edema in the sleep Am Rev RespirDis 1989;140:711-6. apnea/hypopnea syndrome. Sleep 1991;14:354-6. 140 Launois SH, Feroah TR, Campbell WN, Issa FG,

120 Fletcher EC, Schaaf JW, Miller J, Fletcher JG. Long Morrison D, Whitelaw WA, et al. Site of pharyngeal on September 26, 2021 by guest. Protected copyright. term cardio-pulmonary sequelae in patients with sleep narrowing predicts outcome of surgery for obstructive apnea and chronic lung disease. Am Rev Respir Dis sleep apnea. Am Rev RespirDis 1993;147:182-9. 1987;135:525-33. 141 Polo 0, Brissaud L, Sales B, Besset A, Billiard M. Partial 121 Laks L, Krieger J, Podszus T. Pulmonary hypertension upper airway obstruction in sleep after uvulopalato- in obstructive sleep apnea: retrospective multicentre pharyngoplasty. Arch Otolaryngol Head Neck Surg 1989; study. Am Rev Respir Dis 1992;145:A865. 115:1350-4. 122 Berthon-Jones M, Sullivan CE. Time course of change 142 Plowman L, Lauff DC, Berthon-Jones M, Sullivan CE. in ventilatory response to CO2 with long-term CPAP Waking and genioglossal muscle responses to upper air- therapy for obstructive sleep apnea. Am Rev Respir Dis way pressure oscillation in sleeping dogs. J Appl Physiol 1987;135: 144-7. 1990;68:2564-73.