Sleep and Biological Rhythms 2008; 6: A1–A8

doi:10.1111/j.1479-8425.2008.00359_1.x

Symposia Healthy Function of Sleep – Neurobehavioural Function S01 HEALTHY SLEEP – MEMORY CONSOLIDATION AND BEYOND R STICKGOLD Associate Professor of Psychiatry, Harvard Medical School Department of Psychiatry, Beth Israel Deaconess Medical Centre, Boston, USA The last decade has seen a dramatic increase in our understanding of sleep-dependent memory consolidation, moving it from a generally discredited (or at best ignored) concept to a largely accepted tenet among both memory and sleep researchers, even found in undergraduate psychology textbooks. This work has established a firm connection between sleep and memory function. But given the complexity of both sleep and memory, each of which comprise multiple stages and subtypes, even the most basic characterizations of this phenomenon remain unclear. To date, most studies have sought to assign consolidation of particular classes of memory to discrete sleep stages, for example consolidation of procedural memories to REM sleep or declarative memories to slow wave sleep (SWS). But exceptions to this simple dichotomy come close to outnumbering supporting studies. Several new studies, presented here, lead us to now propose a unified sleep-dependent memory consolidation hypothesis, to describe how sleep stages contribute to memory consolidation. We propose that sleep plays a more nuanced role in memory processing than previously considered, with sleep stages being selectively involved not with specific memory types, but with specific components of the memory consolidation process. Specifically, we propose that SWS stabilizes recently acquired declarative and procedural memories, while REM and Stage 2 nonREM sleep subsequently enhance them, selectively reinforcing their most valuable components and integrating them into pre-existing networks of stored information. This new hypothesis is consistent with the growing literature of sleep-dependent consolidation across a range of memory functions, and is strongly supported by new results presented in this talk. These new results, in concert with the unified memory consolidation hypothesis, move the study of sleep and memory beyond a discussion of classical memory consolidation (i.e., stabilization), into the realm of more powerful and valuable forms of sleep-dependent memory processing that (i) enlarge the neural networks in which memories are stored, (ii) extract patterns and rules from large bodies of encoded information (iii) integrate them with other, older memories into rich semantic networks, and, perhaps as a consequence of these other processes, (iv) selectively enhance those aspects of memories of greatest value to the organism.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

S-02 RECENT ADVANCES IN THE ASSESSMENT OF NEUROCOGNITIVE FUNCTION CR CLARK Flinders University, Adelaide Neurocognitive assessment continues to serve as a primary method to study the development and integrity of brain function and improvements in the testing process have helped to secure its role in the field of neuropsychology. The introduction of standardized methods of test administration, scoring and interpretation represents one of the most important advances in the past fifty years. At present, scientists and clinicians have a remarkable range of standardized tests from which to choose, but individual tests are often constrained by a lack of adequate normative data or limited coverage of domains of function, leaving significant opportunities for further improvement. The presentation will report the development and use of a standardized test battery designed to permit sensitive measurement of function across the full lifespan. The battery is extensive and permits, as required, an integration of demographic, behavioural, cognitive, affective and psychophysiological measures to provide an informative characterization of brain and mental function. Subsets of the battery permit targeted assessments that include both local touchscreen and remote web-based protocols. Discussion will also address a number of professionally sensitive issues including computer-based assessment, test standardization (equipment, test personnel, testing environment and automated scoring) and the use and validity of electrophysiological measurements of brain state and mental functioning.

S-03 NORMAL SLEEP IN CHILDREN T OLDS University of South Australia, Adelaide Background: There are few data on the sleep patterns of Australian children, or the associations between sleep and sociodemographic factors [age, sex, socio-economic status (SES)], temporal context (type of day, season), type of day (school day, weekend, holiday), and individual characteristics (weight status). Methods: Between 2001 and 2007, 4,033 9–18 year old Australians reported the time of falling asleep and waking up on 9,053 individual nights. Using a computerised use-of-time recall, the adolescents also reported all activities performed on that day, choosing from 250 different activities and reporting in time slices as fine as 5 minutes. Results Day type: The major source of inter-subject variance in sleep time was day type. Sleeps were categorised into four types: bed on a school day, wake-up on a school day (S-S); bed on a school day, wake-up on a non-school day (S-NS); bed on a non-school day, wake-up on a nonschool day (NS-NS); and bed on a non-school day, wake-up on a school day (NS-S). Adolescents slept least on NS-S nights (553 min) and most on NS-NS nights (603 min).

A1

Healthy Function of Sleep

Age: Sleep time decreased by about 9 min/night with each year of age (p ⱕ 0.0001), but declines were much greater (15–16 min/night) on S-S and NS-S nights, suggesting that older adolescents stayed up later and later, but were obliged to get up at the same time to get to school. Sex: Girls slept on average 5 min more each night than boys (p = 0.003), the largest differences occurring on S-NS and NS-NS days. Weight status: There was a strong association between weight status and sleep duration (p = 0.0004), with obese children sleeping 20 min less each night than underweight children. The differences across weight status categories were most marked on NS-S nights (65 min). SES: Sleep time increased by 4 min/night with every 1 SD increase in SES. Discussion: Day type is an both an important predictor of sleep duration, and a major moderator variable in the relationships between sleep and sociodemographic characteristics. Although they are all significant predictors of sleep duration, taken together, age, sex, weight status, SES and day type explained only about 10% of the inter-individual variance in sleep time.

S-04 SLEEPINESS – MEASUREMENT AND DRIVING CONSEQUENCES M HOWARD Institute for Breathing and Sleep, Melbourne, Victoria, Australia New indicators of the state of sleepiness and long term sleep propensity have the potential to be useful clinical and research measures. Clinical assessment of sleepiness has focused on subjective assessment of an individuals’ sleep propensity over time (trait) or objective assessment via measurement of EEG defined sleep latency in a standardised setting (MSLT/MWT). However, the relationship between the current measurements and real world outcomes, such as road crashes, is weak at best. An individuals’ instantaneous level of sleepiness (state) is more closely linked to crash risk, with specific sleepiness symptoms potentially being good indicators of sleepiness state. EEG defined indicators of drowsiness (increased alpha and theta activity), behavioural changes, slow eyelid movements and prolonged eye closure have been demonstrated in sleepy subjects prior to sleep onset in both the laboratory and real world settings, including driving. These variables can be readily measured and are related to cognitive outcomes in the laboratory, however their relationship to long-term outcomes remains unknown and warrants further exploration.

Orofacial Special Interest Group Symposia S-05 ORAL APPLIANCES IN THE MANAGEMENT OF SLEEP-DISORDERED BREATHING: THE PHYSICIANS’ PERSPECTIVE A CHAN1,2, R LEE1,2, K SUTHERLAND1,2, P CISTULLI1,2 1 Centre for Sleep Health and Research, Department of Respiratory Medicine, Royal North Shore Hospital, St Leonards, NSW, Australia, 2 Woolcock Institute of Medical Research, University of Sydney, NSW, Australia Oral appliances are an alternative to continuous positive airway pressure (CPAP) for the treatment of obstructive sleep apnoea (OSA). They are

A2

worn during sleep, and aim to maintain the patency of the upper airway by increasing its dimensions and reducing its collapsibility. Although CPAP is a highly efficacious treatment, its obtrusive nature limits clinical effectiveness. Oral appliances are a simpler form of treatment and are often considered by patients to be more acceptable compared to CPAP. Treatment of OSA with an oral appliance requires a multidisciplinary approach. The clinical practice parameters of the American Academy of Sleep Medicine recommend their use for mild to moderate OSA; or for patients with severe OSA who are unable to tolerate or refuse treatment with CPAP. Oral appliances have been shown to have a beneficial impact on a number of important clinical end-points, including the polysomnographic indices of OSA, subjective and objective measures of sleepiness, blood pressure, aspects of neuropsychological functioning and quality of life. Elucidation of the mechanism of action of oral appliances has provided insight into the factors which predict treatment response and may improve the selection of patients for this treatment modality. Although not as efficacious as CPAP for improving the polysomnographic indices of OSA, oral appliances are often considered by patients to be a more acceptable treatment. This has the potential to translate to better treatment compliance and may provide a similar level of effectiveness in clinical practice.

Paediatric Special Interest Group Symposium S-07 VENTILATION OF CHILDREN WITH NEUROMUSCULAR DISORDERS M RYAN1,2 1 Royal Children’s Hospital, Melbourne, Australia, 2Murdoch Children’s Research Institute, Melbourne, Australia Respiratory failure, and complications of respiratory impairment, are a common cause of morbidity and mortality in children with neuromuscular disorders. Respiratory impairment in these patients results from weakness of the respiratory muscles and a reduction in chest wall and lung compliance, which cause a decreased vital capacity (VC), while expiratory muscle weakness can increase the residual volume. The ability to cough depends on VC, expiratory muscle strength and bulbar muscle function. Respiratory function in children with NMD should be regularly assessed using techniques to assess lung function (static lung volumes, flows and indices of gas exchange), nocturnal respiration, cough and symptoms and signs related to hypoventilation. The course of respiratory impairment in Duchenne muscular dystrophy, spinal muscular atrophy and other common paediatric NMDs will be reviewed. Sleep disorders in children with neuromuscular disorders include sleep-disordered breathing (SDB), obstructive and central sleep apnoea, which place these children at risk of pulmonary hypertension, cor pulmonale and neurocognitive dysfunction. SDB commonly precedes overt respiratory failure. The VC and maximum inspiratory pressure predict SDB and nocturnal hypoventilation. The peak cough flow is useful in assessing the effectiveness of airway clearance. Polysomnography is a vital part of the assessment and should be undertaken in all children with neuromuscular disorders at baseline and at intervals determined by the respiratory function. Non-invasive ventilatory support (NIV) is increasingly used in acute and chronic respiratory failure caused by neuromuscular disorders. In several small, non-controlled paediatric series NIV has been shown to improve SDB, gas exchange, symptoms, quality of life and survival in patients with acute deterioration of chronic respiratory failure and

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

symptomatic daytime hypercapnia. Relative indications for introduction of NIV include symptomatic nocturnal hypoventilation without daytime hypercapnia, symptoms of chronic respiratory impairment, and asymptomatic nocturnal hypercapnia. The place of invasive ventilation is less clear, but tracheostomy and ventilation should be considered with severe bulbar involvement, recurrent aspiration or retention of secretions, and inability to tolerate or failure to respond to NIV.

Respiratory Special Interest Group Symposium

and weight of numbers and individuals. Many patients have both conditions and the clinical effects, particularly in relation to cardio-vascular disease, overlap and exacerbate each other. Diabetes care has developed into a multidisciplinary approach involving general practice, medical specialists and a variety of allied health professionals with a view to deliver optimal care. Education of all of these groups is paramount to allow endocrinologists to see only the more advanced and problematic cases. The methods used in the Southern Adelaide Health Service to further these aims and the potential application to sleep medicine will be outlined.

S-10 EXPOSURE TO SLEEP MEDICINE IN THE AUSTRALIAN POPULATION. WHAT HAVE WE GOT RIGHT NOW?

Insomnia and Sleep Health Special Interest Group Symposium

N MARSHALL1,2 1 NHMRC Centre for Clinical Research Excellence in Respiratory and Sleep Medicine, Sydney, NSW, Australia, 2Woolcock Institute of Medical Research, Sydney, NSW, Australia

S-13

The prevalence of OSA in the community has been reasonably well described as has the treatment of patients with OSA seen in tertiary sleep disorders clinics. Most people with OSA in the community have mild severity with unobtrusive symptoms. Population-based prevalence studies in Australia indicated that sleep apnea (AHI > 5/hour) was present in about 25% of men and around 7–8% of middle-aged women in the early 1990s. The rise in obesity since then is expected to have increased the prevalence of OSA. What is less clear is how much of the overall population with treatable sleep disordered breathing have had access to any of the various treatment modalities as the few available data sources we are aware of are not sufficiently precise, comprehensive, or detailed enough. Australian Medicare billing data indicate that no more than 2% of the population could possibly have been investigated with polysomnography between 1994 and 2004 and many of these billing records will be repeat studies on the same patient. A populationbased mail survey of registered voters in NSW conducted in 2000 indicated that around 6.2% of the sample complained to their doctor of snoring or sleep apnea during their lifetime (SSA) and 3.4% had had a ‘sleep study’ (which may include testing modalities other than standard PSG). The most commonly reported treatment for SSA was surgery and most people treated with surgery did not report a sleep study. Patients who did not report a sleep study were more likely to be treated. Much of the treatment for SSA is probably not being conducted by or in conjunction with sleep physicians but rather by sleep or ENT surgeons. Whether these combinations of treatment approaches are sufficient, insufficient or over-provided in order to cope with the burden of sleep apnea is unclear. From a population health perspective the appropriate approach to treating sleep apnea is unknown because none of the available screening and treatment combinations have been shown to be effective in the majority of potential patients.

S OLSEN University of Queensland, Brisbane, Qld, Australia

S-11 DIABETES CARE – A RELEVANT MODEL FOR A MULTIDISCIPLINARY APPROACH TO SLEEP APNOEA S STRANKS Repatriation General Hospital, Adelaide Management of diabetes mellitus and sleep apnoea share many of the same difficulties, particularly complexity, multi-system complications

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

PSYCHOLOGICAL APPROACH TO IMPROVING CONTINUOUS POSITIVE AIRWAY PRESSURE THERAPY (CPAP) ADHERENCE

The treatment of choice for Obstructive Sleep Apnoea (OSA) is Continuous Positive Airway Pressure therapy (CPAP). However, the effectiveness of this treatment is limited by suboptimal adherence rates. This presentation will outline an ongoing research program aimed at identifying and validating modifiable psychological predictors of CPAP adherence and acceptance. We present a new, inclusive Health Beliefs Model (HBM) of CPAP acceptance and adherence, as well as data supporting the utility of constructs from this model in predicting CPAP adherence before experience with CPAP. We describe the development of measures of Cues to Action and Barriers to Action designed to supplement existing measures of the Health Beliefs Model. In addition, we describe our assessment battery designed to identify individuals who are likely to accept and adherence to CPAP in the longer term. Finally, we outline the development of a theory driven Motivational Interviewing (MI) intervention which utilises measures from the HBM and targets patient’s specific motivations to accept CPAP as a treatment and then continue to adhere to it. We describe nurse training, session content and assessment procedures. Preliminary statistics regarding the outcomes of this trial will also be presented.

S-14 THE “NUTS AND BOLTS” OF INCREASING ADHERENCE TO CONTINUOUS POSITIVE AIRWAY PRESSURE (CPAP) WITH A GROUP COGNITIVE BEHAVIOURAL THERAPY (CBT) INTERVENTION D BARTLETT Woolcock Institute of Medical Research, Sydney, NSW 2050, Australia Obstructive Sleep Apnea (OSA) is a common sleep disorder. The accepted treatment for OSA is continuous positive airway pressure (CPAP) which improves neurocognitive function and reduces the associated risks of daytime sleepiness and cardiovascular disease. CPAP has been shown to be an efficacious treatment when used on a nightly basis.

A3

Healthy Function of Sleep

However CPAP is often ineffective due to initial refusal uptake and or poor adherence rates which range from 5–50% over 1-week to a 6-month period. A few studies have assessed behavioural interventions to improve adherence with mixed findings, often involving labour intensive protocols. The use of social cognitive therapy, which explores how individuals make choices and their perceptions of being able to successfully accomplish a future behaviour (self efficacy), resulted in improved adherence in a small randomised study1. More recently, 100 individuals diagnosed with CPAP, were randomised to either two 1-hour CBT interventions (including a video of real CPAP users) plus treatment as usual (TAU (mask fitting and information)) or treatment as usual only2. Hours of CPAP usage were measured at 7 nights and 28 nights. At 28 nights the CBT group used their CPAP for 2.9 hours longer; were 6.9 times more likely to adhere to CPAP therapy and had higher scores in relation to self efficacy compared with the TAU group. These significant and positive outcomes may also have resulted from the additional time that the researchers spent with these CPAP patients. An NHMRC grant is currently underway exploring in a randomised trial the role of CBT with social cognitive therapy compared with social reciprocity where an equal amount of time is spent with patients in a social setting. Both groups initially undergo a TAU which now contains substantial educational material relating to OSA and CPAP usage. Data will not be presented as this randomised trial has not been completed. However previous research findings on CBT interventions to increase CPAP adherence will be discussed in conjunction with the methodology used in the above mentioned grant. We aim to explore the implementation of this ‘real life’ intervention for the health professionals attending this session. References (1) Aloia M, Lina Di Dio M, Ilniczky M et al., Improving compliance with nasal CPAP and vigilance in older adults with OSAHS. Sleep Breath 2001;5:13–21. (2) Richards D, Bartlett DJ, Wong K et al., Increased adherence to CPAP with a group Cognitive Behavioural Treatment Intervention: A Randomised Trial. Sleep 2007;30:635–640.

participate in group 2 (in-laboratory PSG and home recordings for two 3 night sequences; three nights on nasal flow and three on oximetry in random order). For group 1, 199 had evaluable data for both devices and for group 2, 93 subjects had evaluable data for three nights and 70 had evaluable data for the first night for both devices. Results: For PSG AHI ⱖ 10, the accuracy assessed by AUC (95% CI) for in-laboratory nasal flow RDI (0.88 (0.83–0.92)) was lower than oximetry 3%DR (0.93 (0.89–0.96); 95% CI of diff 0.01 to 0.1, P = 0.03). For PSG AHI ⱖ 30, there was no significant difference in the AUC of in-laboratory recording between devices (95% CI of diff -0.02 to 0.09, P = 0.17). For PSG AHI ⱖ 10, the AUC of home nasal flow RDI for the first night and for three nights of recording were 0.83 (0.74–0.92) and 0.89 (0.83–0.96) and for home oximetry 3%DR were 0.81 (0.71–0.91) and 0.85 (0.77–0.93) respectively. For PSG AHI ⱖ 30, the AUC of home nasal flow RDI for a single night and over 3 nights were 0.95 (0.90–1.00) and 0.97 (0.93–1.00) and for home oximetry 3% DR the AUC for a single night and over three nights were 0.90 (0.81–0.99) and 0.91 (0.83–0.98) respectively. There was no significant difference in the AUC of both devices measured over a single night or over three nights with respect to PSG AHI ⱖ 10 and ⱖ30. For PSG AHI ⱖ 10, the AUC for nasal flow RDI over 3 nights tended to be higher than the first night of recording (P = 0.09). However, the AUC for oximetry 3% DR over 3 nights was higher than the first night (P = 0.04). Conclusion: Both nasal flow and oximetry have a high accuracy in diagnosing OSA at home with respect to PSG AHI ⱖ 10 and ⱖ30 thresholds. With 3 nights of recording the number of subjects with evaluable data increased and the accuracy of oximetry with respect to PSG AHI ⱖ 10 also increased. The assessment of device accuracy in the laboratory may not reflect how the device performs in the home setting.

Sleep and Cardiovascular Health S-18

Unattended Sleep Monitoring S-15 COMPARING THE DIAGNOSTIC UTILITY OF NASAL FLOW AND OXIMETRY IN THE LABORATORY AND FOR SINGLE AND THREE NIGHTS AT HOME IN OBSTRUCTIVE SLEEP APNOEA LM ROFAIL, K WONG, G MARKS, R GRUNSTEIN Woolcock Institute of Medical Research, Sydney, Australia Introduction: Studies investigating utility of a single channel monitors for OSA detection have predominantly involved oximeters with limited data available for single channel nasal flow. Our aim was to assess and compare the validity of nasal flow and oximetry recordings using separate single channel monitors in the laboratory on the same night as PSG and at home for a single and three nights each with respect to in-laboratory PSG. Methods: Between July 2005 and October 2007, 382 of 746 participants who were referred for the evaluation of OSA to the Sleep Disorders Clinic were approached for the study limited by device availability. 254 subjects were asked to participate in group 1 (in-laboratory PSG and parallel nasal flow recording and oximetry) and 128 were asked to

A4

CARDIOVASCULAR COMPLICATIONS OF SLEEP APNEA MSM IP Division of Respiratory Medicine, Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Hong Kong, SAR, China The strong associations of cardiovascular diseases (CVD) with obstructive as well as central sleep apnea have been well documented. Retrospective case controlled studies suggest that subjects with CVD and OSA may be more at risk of sudden cardiac death during sleeping hours compared to those without OSA or in the general population. Apart from comorbidity of CVD and obstructive sleep apnea (OSA) due to the common link of obesity, various pathophysiologic mechanisms in OSA may contribute to cardiovascular pathogenesis. Repetitive episodes of hypoxemia-reperfusion, hypercarbia, sympathetic activation, and intrathoracic pressure swings may trigger cellular and biochemical processes which predispose to atherosclerosis, the underlying pathology to many CVD such as hypertension, ischemic heart and related diseases and stroke. Several biochemical or functional markers of endothelial dysfunction, which precedes or indicates early atherosclerosis, and carotid artery atherosclerosis, have been reported to be independently associated with OSA. OSA may also be causally related to dysfunction in insulin/glucose or lipid metabolism which are risk factors for atherosclerotic CVD. A strong association of OSA and hypertension has long been observed. Epidemiologic and randomized controlled studies have shown that increase in apnea-hypopnea index or oxygen desaturation

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

are independently related to increase in blood pressure, and CPAP treatment of OSA can decrease blood pressure in both normotensive and hypertensive subjects. Recent reports suggest that daytime sleepiness may be a phenotypic marker for blood pressure lowering response to CPAP and requires further exploration. Other than contributing to atherosclerosis and thus ischemic heart disease and related disorders such as heart failure and arrhythmias, OSA exerts acute effects on oxygenation and hemodynamics that may predispose to myocardial ischemia or heart failure. Nocturnal ECG changes of ischemia during post-apneic periods have been demonstrated in patients with coronary artery disease, and myocardial infarct patients with OSA appeared to have worse prognosis. Both systolic and diastolic dysfunction, independent of hypertension, have been described in OSA, and treatment of OSA has been shown to improve ventricular function in left heart failure. Central sleep apnea (CSA) occurs commonly in ventricular dysfunction and heart failure, Presence of CSA is reported to be a poor prognostic factor of the outcome of heart failure. It is not yet clear if CPAP treatment alters heart function, although some studies have suggested that use of nocturnal CPAP could improve cardiac function and reduce CSA (8). References Ip M et al. Circulating nitric oxide is suppressed in obstructive sleep apnoea and is reversed by nCPAP. Am J Respir Crit Care Med 2000. Ip M et al. Endothelial function in obstructive sleep apnea and response to treatment. Am J Respir Crit Care Med 2004. Somers VK et al. Day-night pattern of sudden death in OSA. NEJM 2005. Robinson GV et al. Obstructive sleep apnoea/hypopnoea syndrome and hypertension. Thorax 2004. Bradley TD et al. Sleep apnea and heart failure Part I: Obstructive sleep apnea. Circulation 2003. Ip M et al. OSA is independently associated with insulin resistance. Am J Respir Crit Care Med 2000. Tan K et al. HDL-dysfunction in obstructive sleep apnea. Atherosclerosis 2005. Bradley TD et al. Sleep apnea and heart failure Part II: Central sleep apnea. Circulation 2003.

S-21 HEART FAILURE AND THE LUNG A BERSTEN1,2 1 Flinders Medical Centre, Adelaide, 2Flinders University, Adelaide A sudden rise in pulmonary microvascular pressure (Pmv) can result in hydrostatic pulmonary edema manifest as sudden onset dyspnea, diaphoresis, hypertension and tachycardia, increased work of breathing, and acute (hypoxemic and hypercapnic) respiratory failure. In addition to measures aimed at reducing Pmv, positive airway pressure techniques quickly provide symptomatic relief. More subtle changes include a spike in the inflammatory cytokine TNF-a and bidirectional protein flux across the alveolocapillary membrane. It is likely that the sudden increase in Pmv leads to elevated serum proteins in the epithelial lining fluid resulting in elevated surface tension, alveolar collapse, impaired gas exchange and elevated work of breathing through inhibition of normal surfactant function. In chronic heart failure prolonged elevation of Pmv leads to potentially maladaptive changes in the pulmonary circulation and lung parenchyma. Pulmonary hypertension is a well known effect, and animal models suggest a reduction in filtration coefficient due to thickening of the alveolocapillary membrane. In two models of chronic heart failure

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

we have found both absent pulmonary edema despite elevated Pmv, and increased dry lung weight due, at least partly, to increased collagen in the parenchyma. Concurrently, despite impaired tissue mechanics, lung mechanics are normal due to lower than normal surface tension associated with type II cell hyperplasia and increased surfactant content. These newly described compensatory changes provide a new paradigm for novel interventions in chronic heart failure, and for understanding acute decompensation. The implications for sleep disturbance has not been examined.

Surgical Management of OSA S-22 IS SURGERY INDICATED IN THE TREATMENT OF SLEEP APNOEA? E WEAVER University of Washington, Washington, United States Obstructive sleep apnoea is prevalent in adults, and it is associated with significant morbidity and mortality. Unfortunately, its treatment remains complicated. Continuous positive airway pressure (CPAP) therapy is the first line treatment because it is highly efficacious and safe. However, CPAP is limited by inadequate use in a significant proportion of patients. Surgery is available as another treatment option, but it rarely cures sleep apnea. For this reason, some have challenged whether there is a role for surgery in sleep apnea management. The overall objective of this presentation is to answer the question, “Is surgery indicated in the treatment of sleep apnoea?” The presentation will define and frame the question specifically to avoid vagueness, confusion, and misunderstanding, all of which can confound the issue of whether surgery is indicated for the treatment of sleep apnoea. The appropriate roles and goals of surgery will be highlighted, which is necessary to assess the evidence for or against surgery. A sample of studies examining a wide variety of relevant outcomes will be summarized in an attempt to answer the overarching question. Finally, the presenter will raise and answer several common challenges about the role and evidence for surgery in the treatment of sleep apnoea. Upon completion, audience members will have a more thorough appreciation of the appropriate roles, goals, and expectations of surgery. They will have a better handle on the evidence and whether surgical treatment evaluation is appropriate in some of their sleep apnea patients.

S-23 SLEEP AND PSYCHIATRIC DISORDERS – CAUSE OF EFFECT? ROBERT STICKGOLD Associate Professor of Psychiatry, Harvard Medical School Department of Psychiatry, Beth Israel Deaconess Medical Centre, Boston, USA Most of the major psychiatric disorders, including all of the DSM-IV axis one disorders, list dysregulations of sleep as one of their defining characteristics. Invariably, the assumption has been that the sleep disorder is a consequence of the psychiatric disorder. But a range of studies now raises the question of the extent to which independent sleep disorders can contribute to, or even cause, psychiatric illness.

A5

Healthy Function of Sleep

Studies of children with both sleep apnea and ADHD indicate that correcting the apnea with tonsillectomy and/or adenoidectomy can reduce or eliminate the ADHD. Similar studies in adults with sleep apnea and major depression report that treatment of the apnea with continuous positive airway pressure (CPAP) can reduce or eliminate the depression. Anecdotal reports provide striking evidence of a role for sleep in bipolar disorder as well. In these instances, sleep’s influence is not on the development or resolution of the disorder, but on the switching between states, with sleep deprivation triggering the onset of episodes of mania. But perhaps the most provocative hypothesis is the claim that posttraumatic stress disorder (PTSD) is a memory disorder, and, more specifically, a consequence of defective sleep-dependent memory processing. The hypothesis makes two major claims. First, it argues that PTSD develops when normal mechanisms of post-encoding memory processing fail to convert an isolated, emotionally charged, episodic memory of a traumatic event into an integrated, semanticized, and emotionally modulated narrative story of the trauma. Second, it proposes that these mechanisms that normally process traumatic memories are largely sleep dependent, and that it is the failure of these sleepdependent memory processes that leads to PTSD. Recent studies demonstrating sleep’s role in the normal processing of emotional memories, as well as in the extraction of gist representations and “rules” from larger sets of memories, provide evidence for the proposed role of sleep in normal trauma processing, and comparisons of dream studies in normal subjects and patients with PTSD support the argument that the classic repetitive dreams of PTSD, which replay the traumatic event in near-veridical form, are a reflection of this failed sleep-dependent memory processing. Together, these wide ranging investigations suggest that sleep should be considered more seriously both in the etiology and treatment of psychiatric disorders.

Neuro-cognitive function in sleep disorders S-26 NEUROCOGNITIVE FUNCTION IN INSOMNIA AND CIRCADIAN DISORDERS G KENNEDY Victoria University, Melbourne, Australia The synchronisation or entrainment of the brain to the local environment is managed by a system of internal ‘clock like’ cellular mechanisms (circadian system) that control the temporal organisation of the individual organism. Dysfunction of the circadian system is reflected in terms of temporal disorganisation observed in many aging individuals, and in people with sleep disorders, jet-lag, and shift-lag. Research on the effects of chronic insomnia and circadian rhythms disorders, including; delayed sleep phase insomnia, jet lag and shift lag has shown that a range of serious pathophysiological changes may occur over time. People suffering from insomnia and circadian sleep disorders have been shown to be at increased risk for psychiatric disorders, particularly anxiety and depression, decreased quality of life, increased healthcare utilization and costs, drug and alcohol abuse, decreased occupational performance and increased accident rates. It has been well documented that a range of neurocognitive changes occur in people suffering from these sleep disorders. Neurocognitive changes appear to be mediated via increased levels of sleepiness and fatigue present in those suffering chronic insomnia and/or circadian sleep disorders. Fatigue and sleepi-

A6

ness are dependent on factors such as time of day (circadian), duration of prior wakefulness, work loads and previous periods of sleep. Increasing levels of fatigue and sleepiness result in reduced vigilance, increased reaction time and impairment of performance. These neurocognitive changes translate into reduced occupational capacity and increased rates of serious accidents. The full extent of the neurocognitive changes that occur has not yet been fully documented and little is known about the underlying neurobiological mechanisms mediating these negative changes in cognition and performance. This paper will explore some of the issues related to studying neurocognitive dysfunction in people with sleep disorders.

S-27 NEURO-COGNITIVE FUNCTION IN SLEEP DISORDERS: NEUROBEHAVIOURAL DYSFUNCTION IN CHILDREN K LUSHINGTON University of South Australia, Adelaide South Australia, Australia In children, sleep is thought to be important for healthy brain function. It is not surprising therefore that any pathology leading to disordered sleep will have neuropsychological, behavioural and emotional sequelae. Much of our present understanding of the impact of disordered sleep on brain function in children has come from studies undertaken in children with a sleep breathing disorder. For example, current evidence from studies in children with upper airway obstruction suggest that, alone and in combination, hypoxia and sleep fragmentation are associated with reduced cognitive capacity, learning, memory, attention and concentration and poorer academic success. A similar pattern of findings in children with other types intrinsic sleep disorders have also been reported. Despite these gains, the role of sleep in child development nonetheless remains poorly understood. The impact of specific sleep factors, such as nocturnal hypoxia, sleep architectural abnormalities, sleep fragmentation and sleep restriction, on child functioning remains unclear. In addition, little is known about the potential factors which may potentiate or conversely protect a child with disordered sleep from daytime deficits. Finally, little is known in children about the interaction between sleep, the developing brain and daytime functioning – e.g. the impact of disordered sleep on brain function and perhaps development is likely to differ in a 6 month old versus an adolescent. In summary, in children with disordered sleep there is a clear need for more research at the interface between basic and clinical studies to illuminate possible mechanisms that underlie dysregulation of neurocognition, behaviour and emotion. In this review we examine our current knowledge on the impact of disordered sleep on neurocognitive and behavioural functioning in children with disordered sleep and propose new directions for investigation.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

Australasian Sleep Technologists Association Symposium S-28 INFLUENCE OF REVISED AASM CRITERIA ON DIAGNOSTIC OUTCOMES W RUEHLAND1, P ROCHFORD1, F O’DONOGHUE1, R PIERCE1, P SINGH2, A THORNTON2 1 Institute for Breathing and Sleep, Austin Health, Heidelberg, Victoria, Australia, 2Royal Adelaide Hospital, Adelaide, SA, Australia The new AASM Manual for the Scoring of Sleep and Associated Events contains differences in PSG scoring criteria compared to previous published standards – most notable are differences in the hypopnoea definition. In the manual the “recommended” hypopnoea definition requires a ⱖ4% desaturation; the “alternative” definition requires ⱖ3% desaturation or arousal. These requirements are stricter than “Chicago” criteria, which are currently recommended for clinical use in Australia. This is likely to impact on the magnitude of AHI and have implications for OSA diagnosis and research. We examined the impact of hypopnoea definition differences on AHI in 328 consecutive patients investigated for OSA in two laboratories. The median AHI using the “recommended” hypopnoea definition (AHIRec) was approximately 30% of the median AHI using the “Chicago” definition (AHIChicago) whereas the median AHI using the “Alternative” definition (AHIAlt) was approximately 60% of the AHIChicago; large, AHI-dependent, patient-specific differences were observed. Failure to adjust diagnostic criteria for the new rules would result in approximately 40% of patients previously classified as positive for OSA using AHIChicago being negative using AHIRec and 25% being negative using AHIAlt. These results demonstrate that using different published standard hypopnoea definitions leads to marked differences in AHI. This has implications for disease identification, severity grading, comparability of research and clinical results, treatment decisions, treatment funding by third parties, OSA prevalence estimates, estimates of the public health impact of OSA and establishment of links between OSA and co-morbidities. The results provide insight to clinicians and researchers in interpreting results obtained using different published standard hypopnoea definitions. Professional organisations and accrediting bodies should consider the adoption of a single standardised definition across all laboratories.

S-29 PSG SCORING CONCORDANCE AND THE EFFECT OF TRAINING A THORNTON1, W RUEHLAND2, R PIERCE2, P SINGH1, P ROCHFORD2 1 Royal Adelaide Hospital, Adelaide, 2Institute of Breathing & Sleep, Austin Health, Heidelberg PSG measurements such as AHI inform treatment decisions and are critical in research studies. This study was undertaken to understand the inter-scorer and inter-laboratory variation in PSG measurement and to investigate options to reduce this variation. A randomised controlled trial was used to examine the impact of a novel external proficiency testing (EPT) program on inter-scorer variation in sleep, respiratory and arousal scoring. Fifteen laboratories were randomised into three groups, one group receiving no feedback about performance (control), one group receiving EPT reports (EPT only) and one group receiving EPT

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

reports and training sessions (EPT plus training). The groups scored the same 15 studies at 0, 3 and 6 months to measure changes in inter-scorer variation over time. At baseline, epoch by epoch agreement on sleep scoring was 82% with a statistically significant improvement of 5% and 3% in the EPT groups. Respiratory scoring agreement, measured by the coefficient of variation in AHI between scorers was poor at baseline with values of 37% in the control group, 27% in the EPT only group and 25% in the training group. EPT achieved small improvements to 25% in the ET group and 21% in the training group. Arousal index was highly variable at baseline averaging 36%. The training group achieved significant improvement in arousal index from 38% to 27%. Inspection of individual data suggested that poorer scorers achieved greater improvements but even by 6 months had failed to reach the levels of agreement of the better scorers. The control group showed no change In any measure over the 6 months of the study. The level of discordance demonstrated by the 30 scorers at baseline should be of significant concern to clinicians who base treatment decisions on PSG measures such as AHI. In some of the 15 studies the coefficient of variation of AHI exceeded 50%. EPT achieves modest but useful improvements in agreement between scorers and it is possible that enhancements of the methodology could achieve further improvement. Despite the modest improvements, EPT does provide individual scorers, laboratories and professional bodies with an objective measurement that can be used to measure performance. Acknowledgments: The authors gratefully acknowledge the significant commitment and contribution of the participating laboratories and scorers. The study was funded by Australasian Sleep Trials Network.

Sleep & Metabolic Health S-30 METABOLIC ASSOCIATIONS OF OBSTRUCTIVE SLEEP APNOEA MSM IP Division of Respiratory Medicine, Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Hong Kong, SAR, China It has been observed in clinical practice and confirmed through epidemiologic and clinic-based studies that obstructive sleep apnoea (OSA) is highly associated with various features of the metabolic syndrome, comprising of abdominal obesity, insulin resistance or glucose intolerance, dyslipidaemia, and hypertension. Most research has focused on the association of OSA and various individual metabolic parameters, independent of obesity, the most common metabolic confounder in these subjects. Studies have reported that OSA and periodic breathing are commonly seen in type 2 diabetic subjects. There is also increasing data to support an adverse causal effect of OSA on insulin sensitivity/ resistance in both adults and children, although rigorous evidence from interventional studies are still needed to show that treating OSA can lead to improved insulin sensitivity/resistance in non-diabetics or better glycemic control in diabetics. Another metabolic disorder, dyslipidaemia, is also commonly seen in subjects with OSA although the exact lipid parameter that has reported to be deranged vary from total cholesterol to HDL-cholesterol to triglycerides. Data on an independent effect of OSA on dyslipidaemia has not been consistent, and effect of CPAP treatment have been mostly derived from observational studies. There are multiple pathophysiologic mechanisms in OSA which may potentially give rise to metabolic dysregulation, including intermittent hypoxia, sleep fragmentation and sympathetic activation, which may

A7

Healthy Function of Sleep

in turn affect the expression of intermediary mediators. Translational studies have also shown dysregulation of various inflammatory cytokines or adipokines which may potentially lead to adverse metabolic effects. Animal experiments exposed to intermittent hypoxia, as a model of OSA, have shown deleterious effects on glucose and lipid regulation. However, results of these studies are not always consistent and causality of OSA on metabolic dysfunction remains a controversial issue. Meanwhile, healthcare professionals should be aware of the high association and encouraged to look for metabolic derangements in subjects with OSA and vice versa, and treatment given to each disorder as appropriate.

S-34 OBESITY IN AUSTRALIA P CLIFTON 1 CSIRO, Adelaide, 2University of Adelaide, Adelaide MONICA international data showed that in the mid 80’s 9% of men and 8% of women aged 35–64 years were obese and that Australia ranked in the middle of the MONICA countries. The AusDiab study showed that in 1999–2000 about 60% of Australians were overweight or obese about 2.5 times more than in 1980. The number of people with diabetes had trebled over this period of time. In a follow up study 5 years later the

A8

average increase in weight over this period of time in people less than 65 was 1.8 kg. 4% of those in the overweight category had moved in to the obese category. The Australian Longitudinal Study of Women showed that in 1996 79% of young women (aged 28–33) had a normal weight while 15% were overweight and 6% were obese. Eight years later at survey 4, 24% were overweight and 16% were obese. 40% of middle aged women were overweight or obese and 8 years later this had increased to 60%. Obesity was relatively stable at 13% in older women with 52–53% overweight or obese. In children there was a rapid increase in obesity from the mid 80s to the mid 90s with a tripling in the rate of obesity but this may have flattened off in the last few years. 15.3% of boys and 16.0% of girls (aged 7–15 years), were overweight in 1995, compared to 9.3% of boys and 10.6% of girls in 1985 while 4.7% of boys and 5.5% of girls (aged 7–15 years) were obese in 1995, compared to 1.4% of boys and 1.2% of girls in 1985. Data from the most recent national survey is not yet available. A 2000 survey in NSW schools reported an obesity rate of 9.9% in boys and 7.1% in girls aged 7–11 years and similar rate was found in a 2003 survey in Victoria. A further survey in 2004 in NSW children showed no changes from the 2000 survey. In 4 your olds in SA obesity rose from 3.5% for girls and 3.2% for boys in 1995 to 5.8% for girls and 4.1% for boys in 2002. No increases have occurred since 2002. In conclusion weight increase still appears to be a major problem for young men and women but it appears to be less dramatic in other age groups.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Sleep and Biological Rhythms 2008; 6: A9–A20

doi:10.1111/j.1479-8425.2008.00359_2.x

Oral Presentations New Investigator Session O-01 DEVELOPMENT OF A SCREENING INSTRUMENT TO ASSIST IN THE DIAGNOSIS OF OBSTRUCTIVE SLEEP APNOEA

O-02 EVIDENCE FOR OSA-DEPENDENT VULNERABILITY TO SLEEP RESTRICTION AND ALCOHOL: SIMULATED DRIVING PERFORMANCE

A FEDSON1,3, S MUKHERJEE3,4, M COOPER1, G LOVE3,4, J LEE1, L SIMPSON1,3, K WARD1,3, J HUNG2, L PALMER1,2,3, D HILLMAN3,4 1 Centre for Genetic Epidemiology and Biostatistics, University of Western Australia, Perth, WA, Australia, 2School of Medicine and Pharmacology, University of Western Australia, Perth, WA, Australia, 3Western Australian Sleep Disorders Research Institute, Perth, WA, Australia, 4Department of Pulmonary Physiology Sir Charles Gairdner Hospital, Perth, WA, Australia

A VAKULIN1, S BAULK1, P CATCHESIDE1, J DORRIAN3, C VAN DEN HEUVEL4, N ANTIC1, D MCEVOY1 1 Adelaide Institute for Sleep Health, Repatriation General Hospital, Adelaide, SA, Australia, 2School of Molecular and Biomedical Science, Discipline of Physiology, University of Adelaide, Adelaide, SA, Australia, 3 Centre for Sleep Research, University of South Australia, Adelaide, SA, Australia, 4Child & Adolescent Sleep Health Unit, Discipline of Paediatrics, University of Adelaide, Women’s and Children’s Hospital, Adelaide, SA, Australia

Introduction: Accurate prediction of the risk of Obstructive Sleep Apnoea (OSA) would enable development of a triage system for highrisk OSA patients who could undergo expedited diagnostic procedures and treatment. This would be helpful in rural areas of Australia where there is limited access to laboratory-based Polysomnography (PSG). The aims of this study were to determine which questionnaire and anthropometric factors collected as part of the WA Sleep Health Study correlated most strongly with OSA severity, and to derive models for prediction of OSA patients presenting to a sleep clinic in Western Australia. Methods: OSA was quantified by the Apnoea-Hypopnoea Index (AHI), measured during laboratory-based PSG. We included 915 patients (536 males aged 48.9 ⫾ 13.9 years, 379 females aged 50.3 ⫾ 13.5 years) who underwent detailed evaluation including a comprehensive questionnaire, clinical examination, and PSG. Various parameters were compared between the cases (AHI ⱖ 15/h) and controls (AHI < 15/h). Using multivariate stepwise regression analyses, models for prediction of OSA cases were derived. Subsequently, the newly derived diagnostic models were tested on 460 subjects (261 males aged 50.9 ⫾ 12.7 years, 199 females aged 44.3 ⫾ 13.8 years) for the purpose of validation. Results: Neck circumference, pharyngeal grade 4, habitual snoring, habitual gasping and elevated diastolic blood pressure were significant independent predictors of OSA in males (r2 = 28%), after adjustment for age and body mass index (BMI). In females, neck circumference and habitual snoring were found to be significant independent predictors of OSA (r2 = 34%), after adjustment for age and BMI. The models developed were associated with sensitivity, specificity, positive predictive value, and negative predictive values of 95%, 30%, 83%, and 63% in the male validation sample, and 83%, 68%, 77%, and 76% in the female validation sample. The percentage of male and female patients correctly classified by the derived models were 81% and 76% respectively. Discussion: We have developed a useful risk model for determining OSA (AHI ⱖ 15/h) prior to PSG. The diagnostic models derived from these easily measured parameters may prove useful for predicting the presence of OSA in patients referred from primary care. Inclusion of this risk model in a primary care setting may lead to triaging of high risk OSA patients that would assist in allocation of patients for laboratory based PSG, leading to better management of patients with sleep disorders in rural areas.

Introduction: Driver drowsiness is an important under-recognised factor contributing up to 33% of motor vehicle accidents (MVAs) with an annual cost of up to $38 billion in the US. OSA is an important factor, with patients showing a 2–7 fold increased risk of MVA. Sleepiness is increased by sleep restriction and exacerbated by alcohol resulting in poor driving performance in healthy young drivers. However, vulnerability to these insults in middle-aged moderate-severe OSA patients has not been investigated. We compared simulated driving performance in the mid-afternoon in moderate to severe, middle aged OSA patients and healthy age matched controls following sleep restriction and low-dose alcohol. We hypothesised that OSA patients would show greater decrements in performance under these conditions. Methods: All participants completed a 90-minute mid-afternoon monotonous simulated driving task (AusEd) following baseline (8h), restricted (4h) sleep, and alcohol (BAC 0.05 g/dL) conditions. Driving simulator data from 22 OSA patients (mean[sd]: age = 52.8[12.5]y; BMI = 36.3[9.2]kg/m2; RDI = 56.5[20.7]/hour), and 17 healthy controls (age = 49.7[13.3]; BMI = 24.3[2.5]; RDI = 8.2[4.1]/hour) were examined using a mixed model analysis. Results: There were significant group (P < 0.001), condition (P < 0.001), time (P < 0.001) and group ¥ condition interaction effects (p = 0.012) in steering deviation. Decrements in performance became more apparent after 30–40 minutes on task. More patients than controls crashed following sleep restriction (p = 0.009) and alcohol (p = 0.02). Compared to baseline condition, breaking reaction time was slower in both groups following sleep restriction and alcohol (p = 0.016) but was not different between groups. Conclusions: Compared to controls, OSA patients showed increased vulnerability to sleep restriction and alcohol on their ability to maintain lane position. Under these conditions a greater number of patients had a complete performance failure (crash) compared to controls. Braking reaction time (although slower following sleep loss and alcohol) was not different between the two groups, suggesting that OSA patients are more vulnerable to sleep restriction and alcohol effects during sustained, vigilance-demanding, but not reaction tasks. Our findings support the need to advise untreated moderate-severe OSA patients to avoid further sleep loss and alcohol when driving for extended periods. Support: NH&MRC project grant #390400, AusEd driving simulator developers.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

A9

Healthy Function of Sleep

O-03 A MOTIVATIONAL INTERVENTION TO IMPROVE PATIENT ADHERENCE TO CONTINUOUS POSITIVE AIRWAY PRESSURE THERAPY S OLSEN1, S SMITH2, TPS OEI1, J DOUGLAS3 1 University of Queensland, St Lucia, Queensland, Australia, 2University of Queensland, Kelvin Grove, Queensland, Australia, 3The Prince Charles Hospital, Chermside, Queensland, Australia Introduction: The treatment of choice for Obstructive Sleep Apnoea (OSA) is Continuous Positive Airway Pressure therapy (CPAP). However, the effectiveness of this treatment is limited by suboptimal adherence to the treatment. This study incorporated Health Beliefs constructs found to predict CPAP adherence before experience with CPAP into a theory driven Randomised Control Trial using a Motivational Interviewing (MI) intervention. Methods: 79 consecutive patients (70% male) newly diagnosed with OSA who were naive to CPAP, were randomly assigned to a nurse-led MI + Standard Care intervention (39 participants, 54% male) or to Standard Care alone (40 participants, 84% male). MI patients received 2, 45 minute sessions of MI whilst commencing CPAP and a 30 minute booster session after 1 month of CPAP experience. Objective CPAP adherence was assessed at 1, 2 and 3 months post-treatment initiation. Projected total recruitment at the conclusion of this randomised control trial is 120 patients. Results: Preliminary findings are based on 21 MI patients and 29 standard care patients who had reached the 1 month post-treatment initiation time-point. Patients who received 3 sessions of Motivational Interviewing used CPAP 2.02 hours more per night than patients who received standard care at 1 month post-treatment initiation (5.76 versus 3.74 hours per night, p = 0.02). Patients who received MI also demonstrated more willingness to accept the treatment initially, with only 5% of patients declining to use the treatment at home as compared to 31% of patients in standard care who did not accept the treatment. Moreover, 67% of MI recipients used CPAP more than 6 hours per night, while only 24% of standard care patients adhered to treatment at this level. Conclusion: Preliminary results indicate a positive trend towards more CPAP use after 3 brief sessions of nurse-led Motivational Interviewing as compared to standard care alone. The intervention can be delivered by nurses according to a manualised protocol, and the intervention is well accepted by the patients. Preliminary data supports an improvement in mean CPAP hours, consistent the Health Beliefs Model theory.

O-04 CHANGES IN LUNG VOLUME AND UPPER AIRWAY DILATOR MUSCLE ACTIVITY AT SLEEP ONSET IN OBESE MALE OBSTRUCTIVE SLEEP APNOEA D STADLER1,2, P CATCHESIDE1,2, D PAUL1, J BRADLEY1, D MCEVOY1,2,3 1 Adelaide Institute for Sleep Health, Repatriation General Hospital, Daws Road, Daw Park, SA, Australia, 2School of Molecular and Biomedical Science, Discipline of Physiology, University of Adelaide, Adelaide, SA, Australia, 3Department of Medicine, Flinders University, Bedford Park, SA, Australia Introduction: Obstructive sleep apnoea (OSA) is a common respiratory disease characterised by repetitive periods of upper airway (UA) collapse during sleep. While decrements in UA dilator muscle activity is

A10

thought to be a major contributor of UA collapse, decreasing lung volume (LV) is also likely to result in UA instability. Since UA collapse is common immediately after sleep onset in OSA patients, and LV compressive effects of obesity may be partially compensated during wake, we sought to determine the magnitude of LV changes at sleep onset in obese male OSA patients. We also explored changes in genioglossus muscle (EMGGG) activity following the wake-sleep transition. Methods: Eight obese (BMI; 34.3 ⫾ 1.0 kg m-2) male OSA patients aged 47.9 ⫾ 3.5 years with moderate-to-severe OSA (ApnoeaHypopnoea Index; 65.7 ⫾ 10.2 events hour-1) participated. Patients slept supine throughout the night. Changes in LV were assessed by two pairs of magnetometer coils placed anterior-posteriorly on the chest and abdomen while EMGGG muscle activity was assessed by two intramuscular wires. Sleep onsets were separated into three categories; stable breathing (SB), hypopnoea (HYP) or apnoea (APN) on the basis of no respiratory events, a hypopnoea or an apnoea within 30 secs following sleep onset respectively. For each category, changes in LV and EMGGG (inspiratory and expiratory activity) for the first three post-sleep onset breaths were investigated. Results: 44.7 ⫾ 10.3, 28.1 ⫾ 3.9 and 27.2 ⫾ 8.5% of sleep onsets were accompanied by SB, HYP and APN respectively. The average onset time of a hypopnoea and an apnoea for HYP and APN sleep onsets was 5.0 ⫾ 0.9 and 6.5 ⫾ 1.3 sec respectively. Mixed model analysis revealed a significant sleep onset category and breath effect with respect to changes in LV following sleep onset (category; p < 0.001 and breath number; p < 0.001). LV declined from sleep onset by 28.2 ⫾ 22.4, 64.1 ⫾ 42.0 and 155.2 ⫾ 30.6 mL by the third post sleep onset breath for SB, HYP and APN categories respectively. There were also significant category effects in EMGGG muscle activity, with inspiratory and expiratory EMGGG activity decreasing by ~41% and ~25% by the third post sleep onset breath, for APN transitions respectively (p < 0.001). Conclusion: In obese OSA patients, LV and EMGGG muscle activity frequently decline immediately following sleep onset with the magnitude of fall appearing to be related to the severity of the developing respiratory events.

O-05 BLOOD PRESSURE AND HEART RATE DURING WAKEFULNESS AND SLEEP IN PRIMARY SCHOOL CHILDREN WITH AND WITHOUT SLEEP DISORDERED BREATHING R PATTERSON1, D O’DRISCOLL1, A FOSTER1, J YANG1, G NIXON1, M DAVEY2, A WALKER1, V ANDERSON3, J TRINDER4, R HORNE1 1 Ritchie Centre for Baby Health Research, MIMR, Melbourne, Australia, 2 Department of Sleep and Respiratory Medicine, Southern Health, Melbourne, Australia, 3Department of Psychology University of Melbourne and Royal Children’s Hospital, Melbourne, Australia, 4Department of Psychology, University of Melbourne, Melbourne, Australia Introduction: Sleep disordered breathing (SDB) in adults has been linked to elevated blood pressure (BP) during wakefulness and sleep. There is emerging evidence that SDB in children is also associated with hypertension. However there are limited studies on BP during sleep in children with SDB. We aimed to compare BP and heart rate (HR) during quiet wakefulness (QW) and sleep in primary school aged children with SDB and healthy controls. Methods: 20 children (10M/10F) aged 7–13 y with SDB (Obstructive Apnoea Hypopnoea Index (OAHI) >1 events/h, median 4.3/h, range: 1.1 to 26.1) and 20 healthy control (CTRL) children (10M/10F) with no history of snoring were studied. The two groups were matched for age and BMI. All children were free of significant medical conditions and

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

were not on any medication that affects sleep or BP. All children underwent routine overnight polysomnography with the addition of continuous non-invasive BP recording (FinometerTM, FMS, BV Arnhem, Netherlands). Beat-by-beat BP and HR were averaged across 30 s epochs (minimum of 5 epochs) of quiet wakefulness (QW) prior to lights out and the first cycle of slow wave sleep (SWS) prior to the first REM cycle. Mean arterial BP (MAP) and HR in QW and SWS were compared between groups using two-way repeated measures ANOVA. Results:

GROUPS CONTROL SDB

QW MAP (mmHg)

SWS MAP (mmHg)

QW HR (bpm)

SWS HR (bpm)

63 ⫾ 2 69 ⫾ 2

58 ⫾ 2 65 ⫾ 2

81 ⫾ 2 88 ⫾ 3

75 ⫾ 3 81 ⫾ 3

MAP decreased from QW to SWS in both groups (p = 0.008). MAP was higher in the SDB group in QW and SWS (p = 0.03). The magnitude of the MAP reduction from QW to SWS was not different between the groups (p = 0.85). HR also decreased from QW to SWS in both groups (p < 0.001). However, HR was not significantly different between groups in QW or SWS (p = 0.08). Similar to MAP, reduction in HR from QW to SWS was not different between groups (p = 0.72). Conclusions: This study demonstrates that both control children and those with SDB exhibit a fall in BP and HR from wakefulness to the first cycle of SWS. Children with SDB have higher MAP in wakefulness, but a similar sleep-related fall as controls. Therefore MAP is also elevated in sleep in the SDB group in comparison to controls.

O-06 SELF-REPORTED MOTOR VEHICLE CRASHES (MVC) AND OBSTRUCTIVE SLEEP APNOEA (OSA) K WARD1, E HUDSON1, M COOPER1, J LEE1, G LOVE3, A FEDSON1, L SIMPSON1, D HILLMAN2, L PALMER1, S MUKHERJEE3 1 Centre for Genetic Epidemiology & Biostatistics, University of WA, Perth, WA, Australia, 2Department of Pulmonary Physiology, Sir Charles Gairdner Hospital, Perth, WA, Australia, 3Western Australian Sleep Disorders Research Institute, Perth, WA, Australia Background: Previous studies have shown that patients with OSA are up to 7 times more likely to have a MVC than those without sleep apnoea1. Driver sleepiness has been identified as one of the major causes of highway accidents and fatal crashes 2. This study aims to describe the prevalence of self-reported MVC in a case series of sleep clinic patients and to identify risk factors associated with MVC. Methods: Self-reported questionnaire data including Epworth Sleepiness Score (ESS) and MVC history were collected from a case series of new sleep clinic patients between Jan 2006 to May 2008. Apnoea Hypopnoea Index (AHI) was calculated from laboratory polysomnography (PSG). Crash rates were compared for OSA severity groups according to AHI. Non-OSA was defined as an AHI of 0 to 5 events per hour. Results: Questionnaire and sleep study data were available for 1759 patients (62% male). Mean AHI was 34.7 events/hour (SD = 29.60). Mean ESS was 9.8 (SD = 5.44). 94.3% were drivers and the mean number of years driving was 30.2 (SD = 13.26). The majority (n = 1120; 67%) of patients reported having had at least one MVC. The mean crash rate for our patients (0.075 MVC/year/person) is significantly higher than reported crash rates for the general community in

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

WA3 (0.019 MVC/year/person) (P = 2.2 ¥ 10-16). In our patients, 124 reported having a MVC because they felt sleepy or fell asleep behind the wheel (11% of crashes). 423 (26%) reported having at least one ‘near miss’ car crash due to sleepiness. A high proportion (28%) of patients reported having fallen asleep whilst behind the wheel. Mean MVC rates were higher in the severe group compared with non-OSA for ever having had a car crash and for car crashes caused through falling asleep at the wheel, but this difference was not statistically significant. However, stratification by gender showed females had significantly more ‘near miss’ (P = 0.01) and ‘fall asleep’ (P = 0.04) crashes across AHI categories. There was a trend towards a higher incidence of falling asleep behind the wheel in females as AHI increases, but not for males. Discussion: This study demonstrates that sleep clinic patients report MVCs significantly more often than the general population, as well as near miss car crashes due to sleepiness and falling asleep at the wheel. There was a trend of patients with severe OSA having increased rates of MVCs. Female patients were more likely to report driver fatigue and had more fatigue related crashes.

Sleep and Breathing Oral Presentations O-07 A SIMPLIFIED METHOD FOR IDENTIFYING OBSTRUCTIVE SLEEP APNEA IN GENERAL PRACTICE CL CHAI1, N ANTIC1, S ROWLAND1, P CATCHESIDE1, A ESTERMAN3, R REED2, H WILLIAMS4, S DUNN5, D MCEVOY1 1 Adelaide Institute for Sleep Health, Adelaide, South Australia, Australia, 2 Flinders University, Adelaide, South Australia, Australia, 3University of South Australia, Adelaide, South Australia, Australia, 4Southern Division of General Practice, Adelaide, South Australia, Australia, 5Charles Darwin University, Darwin, Northern Territory, Australia Introduction: To address the escalating burden of disease and growing waiting lists for sleep services in Australia, we aimed to develop a simple two-step method for identifying patients in general practice with moderate to severe obstructive sleep apnea (OSA), via an initial screening questionnaire followed by home monitoring with a portable twochannel device. Methods: Patients attending six general practice clinics were asked to complete sleep surveys. Based on the Berlin Questionnaire, patients were stratified into high and low risk for OSA. A sample of patients were selected at a predetermined ratio of 4:1 (high risk : low risk) to undergo simultaneous home polysomnography (PSG) and monitoring with a two-channel device (ApneaLink™, ResMed) which records oximetry and airflow/AHI. Survey data were analysed using Chi-square automatic interaction detection (CHAID) to identify variables potentially predictive of moderate to severe OSA (apnea-hypopnea index, AHI > 30), for inclusion in a simplified questionnaire and scoring algorithm. Receiver operating characteristic (ROC) analyses were used to assess the accuracy of the two steps in identifying patients with OSA. Results: 80 patients successfully completed home sleep studies. 31 patients (39%) were found to have moderate to severe OSA on PSG. Four variables (snoring, waist circumference, witnessed apneas and age) were found to be significantly predictive of OSA and were incorporated into a screening questionnaire and scoring algorithm (ROC area under curve (AUC) = 0.85 ⫾ 0.05 [SEM], p < 0.001). Oximetry was highly predictive of OSA (3% dip rate ROC AUC = 0.96 ⫾ 0.03, p < 0.001) and proved to be more accurate and technically reliable than the airflow/ AHI. Using appropriate cut-offs (3% oxygen desaturation index of ⱖ16/

A11

Healthy Function of Sleep

hour in patients identified at risk on questionnaire), the two-step diagnostic process showed 90.3% sensitivity and 98.0% specificity. Discussion: A two-step method consisting of a 4-item screening questionnaire followed by portable home oximetry appears to allow for simple and accurate identification of patients in general practice with moderate to severe OSA. Whilst still requiring validation, simplified diagnosis has the potential to significantly improve access to and the cost-effectiveness of care for patients with this common sleep disorder.

O-08 SLEEP APNEA IS AN INDEPENDENT RISK FACTOR FOR ALL-CAUSE MORTALITY IN A COMMUNITY-BASED COHORT: THE BUSSELTON HEALTH STUDY N MARSHALL1, K WONG1, P LIU2, S CULLEN3, M KNUIMAN4, R GRUNSTEIN2 1 National Health and Medical Research Council Centre for Clinical Research Excellence in Respiratory and Sleep Medicine, Australia, Australia, 2Woolcock Institute of Medical Research, University of Sydney, Sydney, Australia, 3Western Australian Sleep Disorders Research Institute, Perth, Australia, 4School of Population Health, University of Western Australia, Perth, Australia Background: Previously published clinical cohort studies have suggested that obstructive sleep apnea (OSA) is a risk factor for cardiovascular disease associated mortality. However, it is unknown whether sleep apnea is an independent risk factor for all-cause mortality in a community-based sample free from clinical referral bias. Methods: Residents of the Western Australian town of Busselton were investigated with a home sleep apnea monitoring device (MESAM IV) in 1990. OSA was quantified via the respiratory disturbance index (RDI). Data matching from the Australian National Death Index and the Western Australian Death Register were used to ascertain mortality in 397/400 participants (99.3%) after up to 14 years (mean follow-up 13.4 years). Univariate analyses and multivariate Cox proportional hazards modelling was used to ascertain the association between sleep apnea and mortality after adjustment for age, gender, body mass, mean arterial pressure, total cholesterol, high-density lipoprotein cholesterol, diabetes and doctor diagnosed angina in those free from heart attack or stroke at baseline (n = 380). Results: Among the 380 participants, 18 had moderate-severe OSA (RDI ⱖ 15/hour, 6 deaths) and 77 had mild OSA (RDI < 5/hour, 5 deaths). After adjustment for other mortality risk factors moderate-tosevere OSA was associated with a greater risk of all-cause mortality (Fully adjusted Hazard Ratio = 6.24, 95% CL 2.01, 19.39) than no-OSA (n = 285, 22 deaths). Mild OSA (RDI 5-<15/hour) was not an independent risk factor for higher mortality (HR = 0.47, 95% CL 0.17, 1.29). Conclusion: Moderate-severe sleep apnea is associated with a large increased risk of all-cause mortality after controlling for other mortality risk factors in this community-based sample.

O-09 THE INDEPENDENT ASSOCIATION OF CARDIOVASCULAR DISEASE OUTCOMES WITH OBSTRUCTIVE SLEEP APNOEA A FEDSON1,3, J HUNG2, M COOPER1, G LOVE3,4, J LEE1, K WARD1,3, L SIMPSON1,3, L PALMER1,2,3, D HILLMAN3,4, S MUKHERJEE3,4 1 Centre for Genetic Epidemiology and Biostatistics, University of Western Australia, Perth, WA, Australia, 2School of Medicine and Pharmacology, University of Western Australia, Perth, WA, Australia, 3Western Australian Sleep Disorders Research Institute, Perth, WA, Australia, 4Department of Pulmonary Physiology Sir Charles Gairdner Hospital, Perth, WA, Australia Background: Several studies have demonstrated a link between Obstructive Sleep Apnoea (OSA) and Cardiovascular Disease (CVD). It has been suggested that OSA may be a risk factor for CVD, but both may simply coexist due to common risk factors such as increasing age, male gender, obesity, smoking status, hypertension and diabetes. The aim of this study was to investigate the association between OSA and CVDassociated phenotypes including prevalent CVD, hypertension and diabetes using cross-sectional data from a sleep clinic cohort: the WA Sleep Health Study. Methods: We examined the association between OSA and self-reported CVD in referred patients who underwent overnight polysomnography (PSG) at a sleep clinic in WA. CVD was defined by prevalent myocardial infarction, angioplasty, bypass, TIA, stroke or carotid surgery. OSA was quantified by the Apnoea-Hypopnoea Index (AHI) and patients were grouped into quartiles [AHI < 13.8 (Q1), 18.8 < AHI < 26.5 (Q2), 26.5 < AHI < 49.1 (Q3), AHI > 49.1 (Q4)]. Logistic regression analysis was conducted to characterize the association between OSA and prevalent CVD adjusted for age, sex, BMI, history of hypertension, history of diabetes and smoking status. Results: We included 1196 patients (743 males aged 50.9.6 ⫾ 13.6 years, 433 females aged 51.1 ⫾ 13.0 years) in our analyses: 119 patients (9%) reported a history of CVD; 211 (16%) reported a history of diabetes; and 482 (37%) reported a history of hypertension. Multivariate analyses indicated that only hypertension was independently associated with OSA after adjustment for covariates and other CVD outcomes. Patients in AHI quartile 4, AHI quartile 3 and AHI quartile 2 compared to AHI quartile 1 had an odds ratio of hypertension of 1.4 (p-value > 0.05), 1.6 (p-value < 0.05) and 2.3 (p-value < 0.001) respectively. Conclusion: OSA was strongly associated with hypertension independently of age, gender and obesity, and there was a linear association between OSA severity and hypertension risk. Further investigation will use known CVD genes and longitudinal data in the WASHS population in order to investigate the aetiological pathways underlying these associations.

O-10 NURSE LED CLINICS – A NEW WAY FORWARD FOR MANAGING PATIENTS WITH CPAP N DUNN, J DOUGLAS The Prince Charles Hospital, Brisbane, Australia Introduction: Given the epidemic of obesity and increased recognition of Obstructive Sleep Apnoea Hypopnoea Syndrome, an exponential rise in the numbers of patients requiring review at Sleep Clinics is likely. A new approach to discharging stable patients on CPAP is needed. A Nurse Led Discharge Clinic is one way we are addressing this issue.

A12

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

Methods: Patients who had been on CPAP for 12 months, had an Epworth Sleepiness Score <11 at last review and no driving issues were invited to attend the Nurse Led Discharge Clinic. Patients with a listed mobile phone number were reminded of the clinic appointment by SMS. Using a specifically designed assessment and intervention tool, the nurse assesses the patient’s sleep pattern, treatment effectiveness and symptom management. When clinical conditions warrant physician evaluation, patients can immediately access the sleep physician. Patient education and information on health promotion are provided to encourage long term CPAP compliance. Feedback from patients, participating nurses and physicians was assessed. Results: Outcomes of the 51 patients who were booked for the clinic between January and May 2008 are as follows:

Booked 51

Attended

Discharged

Rebooked nurse clinic

40

28 (70%)

7

Physician review

Failed to attend

FTA & rebooked

5

11 (22%)

2

Patients who fail to attend are sent a letter with a copy to GP, advising of failure to attend & offering to rebook an appointment. Other clinic FTA rates are about 13%. Of these 11, five did not have a listed mobile number. The practice of SMS message as a reminder of appointment has improved FTA rates in our other clinics. Feedback from patients has been positive. Initial reactions from nurses were mixed. Problems with processes such as chart availability and database updates have been addressed. Physicians’ responses to the process have been very favourable. Discussion: We have demonstrated Nurse Led Discharge Clinics are effective in discharging patients who are stable on CPAP. Waiting times for physician appointments should decrease. Expanding the role of the nurse has increased skills and work satisfaction. Factors behind the higher fail to igher than expected failure to attend rate are being examined. Data collection on the above parameters is ongoing.

O-11 DRIVING PERFORMANCE IN OBSTRUCTIVE SLEEP APNOEA: DOES VULNERABILITY RESOLVE WITH CPAP TREATMENT S BAULK1, A VAKULIN2, P CATCHESIDE3, J DORRIAN4, N ANTIC3, C VAN DEN HEUVEL5, D MCEVOY3 1 Centre for Sleep Research, University of South Australia, Adelaide, SA, Australia, 2Department of Physiology, University of Adelaide, Adelaide, SA, Australia, 3Adelaide Institute for Sleep Health, Repatriation General Hospital, Adelaide, SA, Australia, 4School of Psychology, University of South Australia, Adelaide, SA, Australia, 5Sleep Disorders Unit, Women’s & Children’s Hospital, Adelaide, SA, Australia Introduction: Estimates suggest that OSA may be a contributing factor in up to 50,000 motor vehicle accidents per year in Australia alone. We have demonstrated that OSA patients are more vulnerable than healthy controls to sleep restriction and alcohol, resulting in impairment to simulated driving performance in the mid-afternoon. While CPAP is considered the gold standard treatment for OSA, many patients demonstrate incomplete recovery in some neurocognitive and subjective domains. The aim of this study is to determine if vulnerability to sleep restriction and alcohol is resolved following 3 months of CPAP treatment.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Methods: To date, 8 OSA patients (7m, 1f; mean[sd]: age = 53.9[15.2]y; BMI = 35.5[11.1]kg/m2; RDI = 76.0[19.2]/hour), and 8 healthy controls (6m, 2f; age = 51.4[10.3]; BMI = 23.7[2.3]; RDI = 6.8[2.8]/hour) have completed a 90-minute mid-afternoon simulated driving task (AusEd) following normal (8h) and restricted (4h) sleep, and alcohol (BAC 0.05 g/dL) conditions both before and after a 3 month-period including CPAP treatment in the patient group only (CPAP used >4hrs on 86.0 [⫾16]% of nights). Results: These preliminary data show significant pre-treatment effects of group, condition (p < 0.05), time on task and group ¥ condition interactions, indicative of greater vulnerability to sleep restriction and alcohol in patients vs. controls (p < 0.05). These effects appear to be reversed following treatment. Conclusions: Preliminary findings suggest that 3 months of CPAP treatment improves driving performance and removes the vulnerability effects to sleep restriction and alcohol. However, further data are required to more firmly establish these effects, and to investigate the potential influence of treatment compliance. Support: We acknowledge the support of NH&MRC project grant #390400. We also thank the AusEd driving simulator developers.

O-12 REGIONAL FAT DEPOSITION AND SEVERITY OF OBSTRUCTIVE SLEEP APNOEA L SIMPSON1, L PALMER1, S MUKHERJEE2, D HILLMAN2, P EASTWOOD3, J POTTER2, G LOVE2, M COOPER1, K WARD1, A FEDSON1, J LEE1, K MADDISON3, J WALSH3, J KIRKNESS3 1 Centre for Genetic Epidemiology and Biostatistics, University of Western Australia, Perth, Western Australia, Australia, 2Department of Pulmonary Physiology, Sir Charles Gairdner Hospital, Perth, Western Australia, Australia, 3Western Australian Sleep Disorders Research Institute, Queen Elizabeth Medical Centre, Perth, Western Australia, Australia Background: Regional fat deposition is implicated in the pathogenesis of Obstructive Sleep Apnoea (OSA) but is difficult to assess using BMI or other body surface measures. Dual-energy x-ray absorptiometry (DXA) is a validated method of assessing regional distribution of body fat. Aim: To investigate which regions of fat deposition, assessed by anthropometric measurements and DXA, are associated with disease severity. Methods: This study utilized a case series from the WA Sleep Health Study (n = 97). Patients were assessed using questionnaire, overnight PSG study and standard anthropometry (including waist, hip and neck circumference). The amount (g) and percentage of fat in the trunk, android and gynoid regions were measured using a Luna Prodigy DXA whole body scan. Univariate analyses investigated differences between the study subset and the general clinic population. Multivariate analyses investigated associations between regional obesity measures from conventional anthropometry (waist, hip and neck circumference) and DXA (trunk, android and gynoid fat) and OSA-related measures (loge(AHI), Epworth Sleepiness Score, nadir oxygen saturation, smoking status, alcohol intake, BMI). Results: Participants comprised 61 males and 36 females, 67% had BMI > 30 and 71% had moderate or severe OSA. Mean age was 51 (SD = 14) years. Participants were not significantly different to the general clinic population (n = 1662) with respect to loge(AHI), gender, age, neck circumference, height, weight, BMI and Epworth score. Loge(AHI) correlated with all regional obesity measures (p < 0.05), independently of gender, except hip circumference in men and gynoid fat. Total android fat and neck circumference had the highest correlation coefficients in both men (r = 0.36 & r = 0.43) and women (r = 0.44 &

A13

Healthy Function of Sleep

r = 0.39). Multivariate regression determined that neck circumference was the best independent predictor of loge(AHI) (r2 = 0.26). Standard DXA regions did not contribute to the prediction of loge(AHI) once neck circumference was in the model. Conclusion: Neck circumference was the most effective regional predictor of AHI in sleep apnoea patients. This suggests that the direct compressive effect of neck fat, rather than the lung volume reduction effect of abdominal fat, mediates OSA severity. Further analyses based on custom DXA analyses of the neck region will be used to ascertain whether fat mass rather than overall mass is involved in OSA severity.

Paediatric Oral Presentations O-13 THE NATURAL HISTORY AND OUTCOMES OF CHILD SLEEP PROBLEMS ACROSS SCHOOL TRANSITION: LONGITUDINAL POPULATION STUDY J QUACH1, H HISCOCK3, L CANTERFORD2, M WAKE3 1 University of Melbourne, Department of Paediatrics, Melbourne, Victoria, Australia, 2Murdoch Children’s Research Institute, Melbourne, Victoria, Australia, 3Centre for Community Child Health, Royal Children’s Hospital, Melbourne, Victoria, Australia Introduction: Adequate sleep is important for developing children’s learning and behaviour. This study aimed to determine: (1) the natural history of sleep problems over a two-year period spanning school entry and (2) associations of current sleep problems with children’s cognition, behaviour, HRQoL and language at age 6.5–7.5 years. Methods: Data were drawn from the first two waves of the nationally representative Longitudinal Study of Australian Children, collected at 4–5 and 6–7 years of age. Measures: Proportions with no, persistent, resolved and incident (i.e. new) sleep problems over the 2 waves. Parent-reported HRQoL (PedsQL); parent- and teacher-reported behaviour (SDQ); non-verbal (WISC-4 reasoning) and verbal (PPVT vocabulary) cognition; and teacher-reported learning (literacy and mathematics, ECLS). Analysis: The effects of current child sleep problems were quantified using linear regression, adjusted for age, gender and social demographic variables. Results: Sleep data were available at both waves for 4,464 children. At wave 2, sleep problems were present in 22.7% (17.0% mild, 3.8% moderate and 1.9% severe). From wave 1, 2.9% of children had a persistent and 2.8% developed a moderate/severe problem, while a further 10.1% resolved. Sleep problems were associated with worse child HRQoL, behaviour and learning (all p < 0.001), for each of which scores worsened in a stepwise fashion with increasing sleep problem severity. Sleep was not associated with either cognitive measure. Discussion: We found that sleep problems were still common at school entry (22.7%) and that persistence or development of new sleep problems was associated with significantly poorer child behaviour, learning, and quality of life. Intervention trials are needed to determine whether intervention for sleep problems around school entry reduces their prevalence and impact.

A14

O-14 PREVALENCE OF OBSTRUCTIVE SLEEP APNOEA IN OVERWEIGHT AND OBESE CHILDREN AND ADOLESCENTS AT AN AUSTRALIAN TERTIARY PAEDIATRIC SLEEP UNIT D TAYLOR2, S SURESH1, G WILLIAMS1, C PARSLEY1, M-A HARRIS1 1 Mater Children’s Hospital, Brisbane, Queensland, Australia, 2Mackay Base Hospital, Mackay, Quensland, Australia Objective: To determine the prevalence of obstructive sleep apnoea (OSA) in symptomatic overweight and obese children at an Australian tertiary sleep unit, and compare this with the prevalence of obstructive sleep apnoea in symptomatic children with a normal body mass index. Methods: A retrospective review of patients who had a polysomnogram performed for risk of sleep disordered breathing at a tertiary paediatric sleep unit over a 2 year period between January 2005 and December 2006 was performed. Polysomnograms from children and adolescents with craniofacial abnormalities, neuromuscular conditions, laryngomalacia and other structural airway abnormalities were excluded from the analysis. BMI data and polysomnogram results were analyzed. Results: After exclusions, a total of 291 subjects’ data were analyzed. The mean age was 7.89 (range 2 to 17.15 years). Two thirds of the subjects were male. 54.3% were less than 8 years of age. The prevalence of OSA in the obese and overweight symptomatic children was 29.8%, while that in the symptomatic normal weight group was 21.3%. There was no statistically significant difference in the incidence of OSA between the normal weight group and the overweight and obese groups, analyzed either separately or together. However, there was a statistically significant difference in the incidence of OSA in the obese children more than 8 years of age compared with the normal weight group (OR 3.278, 95% CI 1.235–8.661), but not in those less than 8 years. Obese males were also found to be significantly more likely to have obstructive sleep apnoea than obese females in the entire group (p = 0.0344) and in the children over 8 years of age (p = 0.0461). Conclusions: The prevalence of OSA in obese and overweight children and adolescents referred to an Australian tertiary paediatric sleep unit with sleep related symptoms over a 2 year period was 29.8%. OSA was more likely to be associated with obesity in children greater than 8 years of age. Obese males were more likely to have OSA than obese females. Prospective population based data is needed to fully understand the ramifications of increasing obesity on sleep disordered breathing in children.

O-15 PRIMARY SNORING IN CHILDREN IMPACTS CARDIOVASCULAR FUNCTIONING A JACKMAN1, A PETERS1, C NICHOLAS1, A FOSTER2, D O’DRISCOLL2, R HORNE2, J TRINDER1 1 University of Melbourne, Melbourne, VIC, Australia, 2Monash University, Melbourne, VIC, Australia Introduction: Primary snoring (PS) affects up to 30% of children. PS is the mildest form of sleep-disordered breathing (SDB), OSA being more severe SDB. OSA affects children’s cardiovascular health. PS, however, has traditionally been considered benign, and its impact on cardiovascular functioning has not been reported. Methods: Participants were 40 children (22F/18M; mean age 8.3 ⫾ 2.2 y) referred for clinical assessment of SDB and matched controls. Overnight polysomnography and wrist actigraphy data were collected. Heart rate (HR) and HR variability (HRV) data from the total

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

sleep period and from periods of uninterrupted sleep were analysed using both time and frequency domain methods. Participants were grouped by clinical diagnosis: PS (OAHI < 1; n = 11), mild OSA (OAHI 1–5; n = 7), moderate OSA (OAHI 5–10; n = 7), severe OSA (OAHI > 10; n = 8), and controls (n = 7). Results: Significant overall differences were found in HR (p < 0.05) over the total sleep period and during stable sleep, whereby HR was highest in severe OSA (92 ⫾ 13 bpm), followed by PS (82 ⫾ 15 bpm), moderate OSA (76 ⫾ 5 bpm), mild OSA (74 ⫾ 8 bpm), and controls (71 ⫾ 8 bpm). Significant overall differences were also found in all measures of HRV (p < 0.05) in both spectral and time domain analyses, whereby the standard deviation of R-R intervals in the time domain was lowest in severe OSA (53 ⫾ 23 ms), followed by PS (59 ⫾ 28 ms), moderate OSA (94 ⫾ 37 ms), mild OSA (107 ⫾ 45 ms), and controls (115 ⫾ 46 ms). In time domain analyses, HRV was significantly lower in PS than in controls (p < 0.05). With values corrected for total HRV, there were no significant overall differences reflective of differences in autonomic balance. Conclusion: PS was shown to have an impact on cardiovascular functioning equivalent to that of moderate OSA, challenging the notion that it is a benign condition. These data have implications regarding the management of PS, particularly when considered together with mounting evidence of neurocognitive, behavioural, and academic dysfunction.

O-16 SLEEP DISORDERED BREATHING ELEVATES BOTH WAKING AND SLEEPING BLOOD PRESSURE IN PRIMARY SCHOOL CHILDREN R HORNE1, A FOSTER1, R PATTERSON1, D O’DRISCOLL1, J YANG1, S WONG1, M NG1, F BASHIR1, G NIXON2, M DAVEY2, A WALKER1, J TRINDER3 1 Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Melbourne, Vic, Australia, 2Melbourne Children’s Sleep Unit, Southern Health, Melbourne, Vic, Australia, 3Department of Psychology, University of Melbourne, Melbourne, Vic, Australia Introduction: Sleep disordered breathing is common in children and ranges in severity from primary snoring (PS) to obstructive sleep apnoea (OSA). OSA has been associated with elevated blood pressure, however the effects of severity of SDB have not been investigated. This study aimed to measure blood pressure non-invasively and continuously during sleep in children with a range of severity of SDB and non-snoring controls. Methods: 101 children aged 7–13 y were studied. 81 were referred for assessment of SDB and 20 non-snoring controls were recruited from the community. Routine polysomnography (PSG) was performed and mean arterial pressure (MAP) recorded continuously using a FinometerTM (FMS, BV Arnhem, The Netherlands). None of the children had any significant medical conditions or were on no medication. Children were divided into groups according to obstructive apnoea/hypopnoea index (OAHI). Control children OAHI < 0 and no history of snoring (N = 20), PS OAHI < 1 event/h (N = 51), mild OSA OAHI 1–5 events/h (N = 16) and moderate/severe OAHI > 5 events/h (N = 14). MAP data were grouped into quiet awake (recorded before sleep onset), NREM 1/2, SWS and REM. Data were compared with 2-way ANOVA with Student Newman Kuels post hoc analyses. Results: There was no difference in BMI, total sleep time, sleep efficiency or sleep latency between groups. Overall MAP was lower in the control group compared with all SDB groups (p < 0.001). Awake MAP was lower in the control group (63 ⫾ 3 mmHg) than the PS group (74 ⫾ 2 mmHg, p < 0.01) and the moderate/severe group

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

(73 ⫾ 3 mmHg, p < 0.05). NREM 1/2 MAP was lower in the control group (61 ⫾ 2 mmHg) than in the PS (67 ⫾ 2 mmHg, p < 0.05), mild OSA (72 ⫾ 3 mmHg, p < 0.05) and moderate/severe (73 ⫾ 3 mmHg, p < 0.1). SWS MAP was also lower in the control group (61 ⫾ 2 mmHg) than the mild OSA group (73 ⫾ 3 mmHg, p < 0.05) and moderate/ severe (71 ⫾ 3, p < 0.05). REM MAP was lower in the control group (65 ⫾ 2 mmHg) compared to the PS group (72 ⫾ 2 mmHg, p < 0.05), the mild OSA group (82 ⫾ 4 mmHg, p < 0.01), and the moderate/ severe OSA group (80 ⫾ 4 mmHg, p < 0.001). Conclusion: This study recorded MAP continuously overnight and found that SDB was associated with increased MAP during sleep compared to non snoring control children, regardless of the severity of SDB. These findings highlight the importance of considering the long term cardiovascular effects of any severity of SDB in children.

O-17 THE HEART RATE RESPONSE TO AROUSAL FROM SLEEP IS REDUCED IN CHILDREN WITH DOWN SYNDROME D O’DRISCOLL1, R HORNE1, M DAVEY2, S HOPE3, G NIXON1 1 Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Monash University, Victoria, Australia, 2Melbourne Children’s Sleep Unit, Department of Sleep and Respiratory Medicine, Southern Health, Victoria, Australia, 3Monash Cardiovascular Research Centre, Monash Heart, Southern Health, Victoria, Australia Introduction: Arousal from sleep in healthy young individuals is associated with a transient increase in heart rate (HR) beyond functional requirements. Individuals with Down Syndrome ((DS) Trisomy 21) have attenuated cardiovascular responses to autonomic tests (challenges) during wakefulness which may reflect autonomic dysfunction. We tested the hypothesis that the HR response to arousal from sleep is reduced in children with DS compared with healthy children. Methods: 7 children with DS (2M/5F) and 7 otherwise healthy (OH) children (4M/3F) referred for investigation of sleep disordered breathing (SDB) were studied. All subjects underwent routine paediatric overnight polysomnography (PSG). All children were classified as having moderate/severe obstructive sleep apnoea (obstructive apnoea/ hypopnoea index (OAHI) >5 events/h). Children were matched for age and body mass index (BMI) z-score (mean ⫾ SEM, Age; DS: 9.0 ⫾ 1.7 y, OH: 7.8 ⫾ 1.4 y, p > 0.05. BMI z-score; DS: 1.14 ⫾ 0.56, OH: 1.08 ⫾ 0.31, p > 0.05). Beat by beat HR was analysed from 15 s pre to 15 s post spontaneous ASDA defined arousals in NREM sleep. The HR change at arousal was determined as the difference between the value prior to arousal and the average of the peak 3 consecutive beats within 15 s. Values were averaged for each child to ensure they contributed equally to the group mean. For each group, the HR pre and post arousal was compared using paired Student’s t-tests. PSG variables and%HR change between groups were compared using Student’s t-tests. Data are presented as mean ⫾ SEM. Results: The mean OAHI and arousal index were not significantly different between groups (OAHI; DS: 20.5 ⫾ 6.4 events/h, OH: 10.7 ⫾ 2.1 events/h, p > 0.05. Arousal index; DS: 28.6 ⫾ 4.4 events/h, OH: 19.8 ⫾ 3.1 events/h, p > 0.05). Arousals were associated with a significant increase in HR in both groups (DS: 100.0 ⫾ 4.5 to 110.6 ⫾ 5.5 bpm, p < 0.001. OH: 89.1 ⫾ 4.0 to 107.7 ⫾ 1.2 bpm, p < 0.001). The magnitude of the HR change at arousal was significantly reduced in the DS group compared with the OH group (DS: 10.6 ⫾ 1.2%, OH: 21.7 ⫾ 1.6%, p < 0.001). Conclusion: The HR response to spontaneous arousal from sleep is reduced in children with DS compared with healthy children.

A15

Healthy Function of Sleep

This attenuated cardiovascular response could be due to reduced sympathetic activation or blunted vagal withdrawal, and may have implications for children with DS given the increased prevalence of SDB in this group.

O-18 COMPARISON OF USING DIFFERENT DESATURATION ASSOCIATION CRITERIA C PARSLEY, G WILLIAMS, C DAKIN, S SURESH Mater Children’s Hospital, South Brisbane, QLD, Australia Introduction: The new AASM scoring rules for adults stipulate two types of respiratory events based on the percentage drop in desaturation (ⱖ3 or ⱖ4%), whereas the paediatric guidelines state an event is associated with an arousal, awakening, or ⱖ3% desaturation. This study was conducted to see the impact of the different event association desaturation values on the quality of data obtained during an overnight PSG. The ambiguity of desaturation criteria may cause differences in event classification, thereby changing the AHI and the apparent severity of sleep-disordered breathing. Methodology: A convenience sample of 20 young children aged 3 mo– 2 yrs with chronic neonatal lung disease weaning from supplemental oxygen were studied. They underwent full overnight PSG (Embla ® N7000) with respiratory events (excluding arousals) scored based on two desaturation criteria: ⱖ3% or ⱖ4% desaturation (Masimo Radical SET™ oximeter, 2 second averaging time). Desaturations associated with moving or arousal events were excluded. Results were analyzed with paired t-tests for detection of a respiratory event with each desaturation criteria. Results: There were more events scored using the ⱖ3% desaturation index. The number of respiratory events with ⱖ3% desaturation was 48.8 mean (range 8–167) and 4% desaturation, 31.2 mean (range 7–97). Compared with a ⱖ3% criteria, the percentage of events missed with a criteria of ⱖ4% desaturation was 28.9% mean (range 0–54.5). There is a significant difference between the percentage of events detected with ⱖ3% vis-à-vis ⱖ4% desaturation (p < 0.005). Conclusions: Though the AASM rules for paediatrics state the use of ⱖ3% desaturation in association with a respiratory event, it is important to be aware of two factors: 1. More respiratory events are included using the lower desaturation value. 2. The oximeter manufacturer accuracy can have an error margin of 3%.

Insomnia and Sleep Health Oral Presentations O-19 SLEEPINESS AND HAZARD PERCEPTION LATENCY S SMITH1, M HORSWILL2, G PARKER2 1 Queensland University of Technology, QLD, Australia, 2The University of Queensland, QLD, Australia Introduction: Sleepiness is the major contributor to at least 20% of all road crashes. Sleepiness-related crashes are likely to be more severe, more often fatal, and more often involve young drivers, than other types of crash. Safe driving is a complex task, in which drivers require higher-

A16

order perceptual and cognitive skills to interact safely with the traffic environment. Hazard perception (HzP) is a critical driving skill. HzP requires scanning of the road environment, fixation on appropriate stimuli, and a ‘holistic’ interpretation of the salience of hazards. HzP latency is the time taken to detect and respond appropriately to a road hazard. Unlike vehicle control skill, hazard perception is reliably related to crash risk (see Horswill & McKenna, 2004 for a review), and a number of states now mandate a hazard perception test as part of licensing for novice drivers. Sleepiness can produce deficits in attention, vigilance and information processing. The impact of sleepiness on hazard perception skill is not known. Method: Two groups of 32 participants (aged 18–30) completed a 90-minute version of the Queensland Hazard Perception test (QHPT; a PC-based, real video, simulator task), a simple reaction time (RT) test and a hazard change-detection test on one of two occasions. One test session took place in the laboratory at 10am (low sleepiness), the other at 3am (high sleepiness). Subjective sleepiness was assessed by the Karolinska Sleepiness Scale at intervals throughout the test session. Objective drowsiness (eyeblink velocity) was assessed continuously with the Optalert system (Sleep Diagnostics). Results: Participants were subjectively sleepier during the night-time test session. Significant correlations were found between (1) mean QHPT score and the Johns Drowsiness Scale (JDS) score during the QHPT test period (r = 0.32, p = 0.01); (2) mean simple reaction time and JDS score during the RT test period (r = 0.38, p < 0.01). This relationship was not found for the hazard change detection task. Discussion: Previous studies have determined a relationship between sleepiness and simulator parameters such as lane excursion. This is the first study to demonstrate a relationship between drowsiness indicated by the JDS score, and performance on a HzP task. Impairment of HzP when sleepy is a potential contributor to road crash, particularly in the early hours of the morning.

O-20 INTENSIVE SLEEP RETRAINING (ISR) AND STIMULUS CONTROL THERAPY (SCT) FOR CHRONIC PRIMARY INSOMNIA; A RANDOMISED CONTROLLED TREATMENT TRIAL J HARRIS2, L LACK1, K KEMP3, H WRIGHT1, R BOOTZIN4 1 Flinders University, Adelaide, South Australia, Australia, 2Adelaide Institute for Sleep Health, Repatriation General Hospital, Adelaide, South Australia, Australia, 3Adelaide Insomnia Clinic, Adelaide, South Australia, Australia, 4University of Arizona, Tucson, Arizona, United States Introduction: Traditional cognitive behavioural interventions, in particular Stimulus Control Therapy (SCT), are effective non-drug treatment options for insomnia. Preliminary research also supports Intensive Sleep Retraining (ISR), a rapid conditioning treatment, for chronic insomnia (Harris et al., 2007). Method: The current RCT compared Intensive Sleep Retraining with SCT, a combination of ISR and SCT, and a Sleep Hygiene control group. Participants were 79 adults (18–65 years) with chronic insomnia, presenting with sleep onset insomnia with a mean sleep diary SOL of 81.7 minutes (SD = 50.3). Results: The active treatment groups (ISR, SCT, ISR + SCT) all reported significant improvements in mean sleep diary sleep onset latency (decreased by 23–36 min) and sleep efficiency (8–16% increase), with moderate to large effect sizes, from pre-treatment to post-treatment. Wake time after sleep onset decreased significantly in the SCT (M = 29.7 min, SD = 37.5) and ISR + SCT (M = 44.8 min, SD = 25.3) groups. Total sleep time increased significantly in the ISR

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

(M = 34.6 min, SD = 55.37) and ISR + SCT (M = 54.6 min, SD = 50.9) treatment groups. Participants receiving ISR (ISR, ISR + SCT) experienced rapidly improved SOL and TST during treatment, suggesting the advantage of rapid improvement in sleep for those receiving ISR treatment. ISR + SCT resulted in consistently larger effect sizes of change than other treatments, including questionnaire measures of depression, fatigue, daytime functioning, vigour, sleep quality, self efficacy for sleep, and dysfunctional sleep cognitions. Mixed model analyses indicated that active treatment groups consistently outperformed the control group. However, there were few statistically significant differences between active treatment groups on individual variables. The ISR + SCT treatment group also showed trends to outperform other active treatment groups with 61% reaching ‘good sleeper’ status. Treatment gains achieved by post treatment in the active treatment groups were largely maintained throughout follow up periods, up to 6 months. Discussion: These results indicate that ISR produces a comparable treatment response to SCT. The benefit of ISR seems to be the rapid improvements in sleep. SCT appears to produce advantages in reductions of wake time overnight that outperform ISR alone. ISR+SCT administered together produced both a rapid response and a trend for superior response over either treatment alone on subjective sleep variables. Given the treatment response, further exploration of ISR treatment is warranted, especially when combined with traditional CBT interventions.

O-21 ASSESSING THE RELATIONSHIP BETWEEN SUBJECTIVE AND OBJECTIVE SLEEPINESS M SHORT, L LACK, H WRIGHT Flinders University, Adelaide, South Australia, Australia Introduction: Sleepiness impairs efficient functioning and contributes to domestic and workplace accidents and deaths. We rely on subjective feelings of sleepiness to warn us of impending sleep. However, prior research has shown variability in the correlation between subjective sleepiness and objective PSG latency. The present study evaluated whether eliciting subjective judgements of sleepiness after a one-minute period with eyes closed would increase correlations between subjective and objective measures. Methods: Subjective judgements of sleepiness were collected from twelve young adult good sleepers following three 1-minute conditions (eyes closed, fixed gaze and reaction time task) These were presented in a balanced order immediately prior to a measure of PSG sleep latency. Subjective and objective measures were obtained a total of 12 times from each participant half-hourly across the period 2000h to 0130h. Results: Within-subjects correlations across an evening period of increasing sleepiness were generally high (means about 0.60) and significant (P < 0.05). Contrary to expectation, there were no differences in correlations between the three conditions. However, when the correlations were calculated across subjects at each testing time, correlations were largely non-significant, with means in most cases close to zero. Discussion: In controlled laboratory conditions, high correlations between subjective and objective measures of sleepiness were found across a large range of sleepiness (2000h to 0130h) in each condition. When clock time was controlled there were no significant correlations across subjects. In other words, individuals are reasonably accurate in detecting changes of objective sleepiness. However, when time of day is controlled, it is not possible to reliably identify those who are objectively sleepy. These results also suggest that the variability of prior research findings may be due, at least partly, to the way in which correlations are derived and indicate that careful methodological procedures are needed to accurately assess this relationship.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

O-22 ASSESSING HORMONE SECRETION DURING SLEEP P LIU1, D KEENAN2, J VELDHUIS3 1 Woolcock Institute of Medical Research, Camperdown NSW, Australia, 2 University of Virginia, Charlottesville VA, United States, 3Mayo Clinic, Rochester MN, United States Background: Assessing hormone secretion during sleep is important to assess circadian variations in melatonin secretion and because androgen and growth hormone deficiency in obstructive sleep apnea may contribute to obesity and altered ventilatory responses. Quantifying pulsatile secretion from peripheral hormone concentrations (deconvolution analysis) requires automated, objective and accurate detection of pulse times and maximum-likelihood estimation of secretion and elimination parameters. In addition to parameter realisability, pulse detection must be experimentally validated. This has never been done. Validation paradigms: To define deconvolution pulse-detection accuracy, 4 paradigms were used: (a) Sprague-Dawley rats had simultaneous recordings of hypothalamic arcuate-nucleus multiunit electrical activity and systemic LH pulses; (b) ovariectomized ewes underwent simultaneous sampling of hypothalamo-portal GnRH and peripheral LH pulses; (c) healthy young men aged 18–30 years were infused with trains of biosynthetic LH pulses under GnRH-receptor blockade; and (d) computer simulations of pulsatile LH profiles. Outcomes: Sensitivity, specificity and receiver-operating characteristic (ROC). Results: Sensitivity and specificity were 0.93 and 0.97, respectively, for the combined rat, sheep and human data (N = 156 pulses) and 0.94 and 0.92 for computer simulations (N = 1632 pulses). For simulated data, the Akaike information coefficient yielded slightly higher sensitivity and the Bayesian information coefficient higher specificity. False-positives errors occurred primarily at low integrated burst mass (<2.0 IU/L/ burst), whereas false-negative errors emerge principally when pulses were justaposed within 45 min. Random variability in the data and high pulse frequencies decreased pulse-detection sensitivity more than specificity (P < 0.02 for each, Spearman rank correlation). Conclusion: This is the first automated, objective and accurate pulsedetection deconvolution procedure validated in independent in vivo and computer-simulation paradigms. This is a valuable tool to assess the interface between endocrine and sleep disorders.

O-23 SLEEP AND OXYGEN SATURATION AT 8000 FEET M MUHM1, L SIGNAL2, P ROCK3, S JONES1, K O’KEEFFE2, M WEAVER3, S ZHU1, P GANDER2, G BELENKY4 1 The Boeing Company, Seattle, WA, United States, 2Massey University, Wellington, New Zealand, 3Oklahoma State University, Tulsa, OK, United States, 4Washington State University, Spokane, WA, United States Introduction: The introduction of aircraft capable of ultra-long range operations (flight times in excess of 16 hours) creates new challenges for flight crew alertness. As the duration of flights extend the amount and quality of in-flight sleep obtained by flight crew is expected to become increasingly important to maintaining their functionality. As with the rest of the aircraft, on-board rest facilities are pressurised to approximately 8000 ft (or lower). Previous research indicates that sleep at higher altitudes is disturbed but it is presently unknown what effects cabin altitudes might have on sleep duration and structure.

A17

Healthy Function of Sleep

Methods: Using a within-subjects design, 20 healthy males (mean age 44 ⫾ 9 years) completed an 18 hour protocol in a hypobaric chamber. The protocol included a 7-hour sleep opportunity (2300–0700) at either 1000 ft or 8000 ft. Sleep was recorded from C4/A1; C3/A2; EOG-R/A1; EOG-L/A2; EMG using an A10 ambulatory recorder (Embla). Oxygen saturation was recorded via a finger probe was placed on the tip of each participant’s finger and data recorded using a Nonin PalmSat 2500 oximeter (sampling rate 0.25 Hz). Sleep and arousals from sleep were scored by two independent, experienced sleep scorers according to standard criteria [1, 2]. Using nVision software a desaturation event was recorded if SpO2 decreased by ⱖ4% from the preceding baseline for a minimum duration of 10 seconds. Results: Raw data suggested a trend for less TST, more wake, more S1 and S2, less REM and a larger number of awakenings and arousals during sleep at 8000 ft compared to sleep at 1000 ft. However, mixed model ANCOVAs indicated that these differences were not statistically significant. In contrast, there were relatively large and significant changes in SpO2 values during sleep at 8000 ft, with an increased desaturation index (median = 14.4 vs 1.9), increased amount of time below 90% saturation (median = 44.4% vs 0%) and decreased mean SpO2 (mean = 85.9 vs 92.1) compared to during sleep at 1000 ft. Discussion: The finding that there were no marked changes observed in polysomnographic sleep duration and quality between acute exposure to 1000 ft and 8000 ft was unexpected given the concurrent changes in oxygen saturation. The types of measures used in this study did not permit exploration of the relationships between and sleep and respiratory variables, but in our view, these findings warrant further careful investigation.

O-24 AUDITORY AROUSAL THRESHOLDS AS A FUNCTION OF SOUNDS OF DIFFERENT PITCHES AND PATTERNS D BRUCK, M BALL, I THOMAS Victoria University, Melbourne, Victoria, Australia Our previous research has tested various sounds (voice, fire cues and beeps) to determine which sound will wake people in slow wave sleep at the lowest decibel levels (auditory arousal threshold, AAT). Participants included children, sober and alcohol impaired young adults, adults with mild/moderate hearing loss and adults over 65 years. Lowest AATs were yielded most consistently by beeps of a 520 Hz square wave signal. Square waves have one fundamental frequency followed by frequency peaks at the 3rd, 5th, 7th etc harmonics. Aim: To measure AATs across signals with a range of pitches and patterns to determine the most effective waking signal. Methods: A sample of 29 young adults aged 18–26 years participated over three nights. Signals sounded for 30 seconds (at levels from 35–95 dBA), increasing every 30 sec by 5 dBA. In Part 1, nine signals with beeps of 0.5 sec duration were tested: (i)–(iv) four square waves (each one with a fundamental frequency of either 400, 520, 800 or 1600 Hz), (v) white noise, (vi) 400–1600 Hz whoop, (vii) 400–800 Hz whoop, (viii) three pure tones of 400, 800 and 1600Hz, and (ix) 3 square waves of 520, 800 and 1200 Hz. Part 2 manipulated the temporal pattern of the Part 1 signal with the lowest AAT, inserting silences of 0, 10 and 21 seconds after each 12 second sequence of beeps (e.g. beeps ON for 12 sec, OFF for 10 sec, ON for 12 sec). Results: Part 1: A one way ANOVA yielded a highly significant difference across the nine signals (F(1,8) = 9.6, p = 0.000). Analyses showed that the 520 Hz square wave yielded the lowest mean AAT, and this AAT was significantly different to any other signal presented in Part 1, with

A18

the exception of the 400 Hz square wave. No significant differences were found across the three temporal manipulations of Part 2. Discussion: The low frequency (400 and 520 Hz) square waves yielded lower AATs than white noise, pure tones and whoops. There was no advantage in presenting ongoing beeps with 12–21 sec of intervening silence. These findings support our earlier research showing that the best sound for awakening from deep sleep is a low frequency (e.g. 520 Hz) square wave. The results have implications for smoke alarm signals and other alarms sounding during the sleep period (e.g. to treat bedwetting).

Chronobiology Oral Presentations O-25 COMPARING THE VIEWS OF DRIVERS AND MANAGERS/DISPATCHERS IN TRUCKING COMPANIES ENTERING A FATIGUE MANAGEMENT SYSTEM TRIAL P GANDER, L SIGNAL, S GARDEN, B SWEENEY Massey University, Wellington, New Zealand Introduction: In 2000, Land Transport New Zealand invited 9 trucking companies to participate in Fatigue Management Systems (FMS) trial, as a regulatory alternative to the prescriptive work hours regime, for managing the safety risk associated with driver fatigue. The present baseline evaluation was undertaken prior to these companies having their FMS accredited by LTNZ. Methods: A 4-page questionnaire for managers/dispatchers and a 7-page questionnaire for drivers were distributed (January–May 2007) that included a series of matched questions taken from the Queensland FMS Pilot evaluation. Follow-up included a reminder postcard, and then a second complete study package sent to non-responders. Participant confidentiality was assured and the study received independent ethical approval. Results: Completed questionnaires were received from 48 managers and dispatchers (71% response rate) and 100 drivers (60% response rate). Driver fatigue was rated as “not a problem” by 24% managers and 27% of dispatchers (in the last 3 months), and by 44% of drivers (in the last 2 weeks). Regarding the causes of driver fatigue, the groups agreed on the 3 most important being: long work days; not enough sleep before trips; and irregular, insufficient, or poor quality sleep during trips. However they disagreed about the relative importance of other factors, with drivers giving higher ranking to work-related factors (irregular meals, poor roads, not enough breaks), and managers giving higher rankings to factors in life outside work (family problems and drinking alcohol the night before driving). Dispatchers gave higher rankings to heavy traffic and family problems, than did drivers. Regarding company policies for managing fatigued drivers, no managers or dispatchers said that they would tell a driver calling in too tired to drive that he must do the trip. However, 12% of drivers believed that this would happen. While no manager said that calling in too tired to drive would count as a sick day, 21% of drivers believed that it would, as did 13% of dispatchers. Discussion: This study represents a unique snapshot of companies selected as responsible operators. It demonstrates that, even in these companies, drivers, managers and dispatchers cannot be assumed to have similar understanding about the size of the driver fatigue problem, the causes of driver fatigue, or company policies to manage it. Education and regular communication are essential for a co-ordinated approach to

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

fatigue management, regardless of the regulatory regime. Limitations include subjective and retrospective information.

O-26 THE EFFECTS OF DIFFERENT ROSTER SCHEDULES ON SLEEP IN MINERS G PAECH, S JAY, N LAMOND, G ROACH, D DAWSON, S FERGUSON The Centre for Sleep Research, University of South Australia, Adelaide, South Australia, Australia Introduction: It has been well documented that shiftwork involving early morning starts and night work can affect both sleep and fatigue. The current study aimed to assess the impact of different rostering schedules at an Australian mine site on sleep and subjective sleep quality. Methods: Participants (n = 51) worked one of four rosters; 4 ¥ 4 (n = 14), 7 ¥ 4 (n = 10), 10 ¥ 5 (n = 17), or Fly-In Fly-Out (FIFO) (n = 12). Sleep (wrist Actigraphy and sleep diaries) was monitored for a full roster cycle including rostered days off (RDO). Results: For all rosters total sleep time (TST) was longer on RDO (7.1 ⫾ 1.6) compared to sleep when on dayshifts (6.0 ⫾ 1.0) and nightshifts (5.9 ⫾ 1.2). Subjective sleep quality ratings on RDO differed depending on whether they followed dayshifts or nightshifts for the 7 ¥ 4 and 4 ¥ 4 rosters. The FIFO roster showed improved subjective sleep quality across dayshifts. Participants working the 10 ¥ 5 roster reported better sleep quality for the shift-change sleep compared to RDO. Discussion: Despite an increase in TST on RDO, this may not be enough to recover from the severe sleep restriction that occurs during work times. The 24h shift-change period provides limited opportunity for recovery. Restricted sleep and quick shift-change periods may lead to long-term sleep loss and associated fatigue.

O-27 DRIVER FATIGUE LEVELS IN TRUCKING COMPANIES ENTERING A FATIGUE MANAGEMENT SYSTEM TRIAL P GANDER, S GARDEN, L SIGNAL, B SWEENEY Massey University, Wellington, New Zealand Introduction: In 2000, Land Transport New Zealand invited 9 trucking companies to participate in a trial of Fatigue Management Systems (FMS), as a regulatory alternative to the prescriptive work hours regime, for managing the safety risk associated with driver fatigue. The present baseline evaluation was undertaken prior to these companies having their FMS accredited by LTNZ. Methods: Managers and dispatchers from each company completed a 4-page questionnaire, while drivers completed a 7-page questionnaire and a 14-day sleep/work diary. Follow up included a reminder postcard, and then a second complete study package sent to non-responders. Participant confidentiality was assured and the study received independent ethical approval. Results: Completed questionnaires were received from 48 managers and dispatchers (71% response rate) and 100 drivers (60% response rate). Ninety-four drivers completed sleep/work diaries (56% response rate).

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Compared to a representative national sample of men, drivers were more likely to report getting enough sleep and waking refreshed often/ always, and more likely to have an Epworth Sleepiness Score ⱕ10 (88% versus 83%). In their diaries, 66% of drivers recorded some work between midnight-5 am (median = 5 night work periods in 14 days), most often due to early starts. For days off, mean sleep = 10.4 hrs, with sleep ⱕ 5 hrs on 1% of days. For days with day work, mean sleep = 7.6 hrs, with sleep ⱕ 5 hrs on 3% of days. For days with night work, mean sleep = 6.2 hrs, with sleep ⱕ 5 hrs on 22% of days. Over the last 2 weeks, 17% of drivers rated their fatigue as a moderate problem and 3% rated it as a big or very big problem. On at least some trips, 24% of drivers reported starting already fatigued, 39% became fatigued while driving, 11% reported feeling so fatigued that their driving was impaired, and 12% reported changing their intended schedule because of feeling fatigued. Over the last 3 months, 28% of managers/dispatchers rated driver fatigue as a moderate problem (none rated it as a big or very big problem). However, 28% rated their company’s performance with regard to the number of fatigue-related accidents as below average or poor. Discussion: This study represents a unique snapshot of companies selected as responsible operators and prepared to undergo close regulatory oversight as participants in the FMS trial. Even in companies at this end of the spectrum, the findings indicate that driver fatigue is not uncommon. Limitations include subjective information and that the questionnaires are retrospective.

O-28 AN INITIAL EVALUATION OF THE INSTITUTE FOR BREATHING AND SLEEP DRIVER AWARENESS QUESTIONNAIRE G KENNEDY1, H GILL1, M MCMAHON2, P SWANN3, R PIERCE2, M HOWARD2 1 Victoria University, Melbourne Victoria, Australia, 2Institute For Breathing & Sleep, Austin Health, Melbourne Victoria, Australia, 3Department Road Safety, Vic Roads, Melbourne Victoria, Australia Sleepiness, fatigue and risky driving behaviours play a significant role in many accidents involving young drivers. Greater knowledge about sleepiness, fatigue and risky behaviours would help reduce serious accident rates for young drivers. The aim of this study was to examine young drivers’ self-reported attitudes, behaviours, knowledge of sleepiness, fatigue and time of day effects in relation to road safety. It was hypothesised that men would have more driving exposure and accidents, and higher scores for risk-taking behaviour and road rage. It was also hypothesised that women would be more knowledgeable about sleep behaviour, circadian rhythms, and sleepiness and fatigue symptoms. The participants were 96 (32.3%) men and women (67.7%) aged 18 to 26 years (M = 21.58, SD = 2.61) with Australian drivers licences. The Institute for Breathing and Sleep Driver Awareness Questionnaire (IBAS-DAQ) was distributed to students at campuses of Victoria University in Melbourne, Australia. Cronbach’s alpha (0.87) indicated good internal consistency for the IBAS-DAQ. Statistical analyses showed differences in knowledge relating only to fatigue symptoms, with women significantly more knowledgeable. Knowledge of circadian rhythms showed the most variance in scores and was identified as an area in which young drivers’ lack adequate knowledge. The findings suggest that there is a need for greater public awareness about the risks of driving while sleepy and/or fatigued. This is particularly important for young drivers who would benefit from educational programs designed

A19

Healthy Function of Sleep

to promote good sleep habits, and highlight the dangers of risky driving behaviours, sleepiness, fatigue and circadian rhythm effects.

O-30 MEASURING DROWSINESS FOLLOWING MODERATE PARTIAL SLEEP RESTRICTION

O-29 ASSESSING FATIGUE-RELATED RISK BEYOND THE ROSTER: A DEFENCES IN DEPTH ASSESSMENT OF RISK IN HEALTHCARE S FERGUSON, S JAY, A WEISSENFELD, D DAWSON, M THOMAS University of South Australia, Adelaide, SA, Australia Introduction: The management of fatigue-related risk in safety-critical workplaces has traditionally focused on hours of work. However, this may not be an effective risk management strategy as research clearly demonstrates increased risk of incident and error associated with reduced sleep, periods of extended wakefulness, and higher subjective ratings. Recently, a more systematic “Defences in Depth” approach has been adopted, which extends beyond the roster to include assessment of the sleep opportunity provided by the work hours, actual sleep and wake, and behavioural signs and symptoms. While long hours are reported to be the norm in Australian hospitals, the effect on prior sleep, prior wake and subjective fatigue ratings has not been examined. Methods: Twelve doctors at a medium sized hospital in Queensland volunteered for the study. Participants wore activity monitors and completed sleep and work diaries for a 14-day period. In addition to sleep and work times, the diaries also recorded subjective fatigue ratings using the Samn-Perelli Fatigue Checklist. Results: The shift pattern worked by the doctors included day shifts (10.0 ⫾ 1.6h), Night shifts (10.3 ⫾ 0.5h) and On-call periods. In general, the roster provided sufficient sleep opportunity between work periods. Sleep in the 24 hours prior to beginning a shift was 6.9 ⫾ 1.3h on day shifts and 6.9 ⫾ 2.1h on Night shifts. A total of 14 out of 111 shifts (12.6%) were associated with actual sleep in the prior 24 hours being less than 6 hours. On average, prior wake at the end of shifts was between 10.7 and 23.3 hours. Subjective fating ratings increased from the beginning to the end of shifts. Eight percent of reports at the beginning of the shift included scores of 6 (extremely tired, very difficult to concentrate) or 7 (completely exhausted, unable to function effectively). In comparison, 23% of reports had a score 6–7 at the end of the shift. Discussion: Data collected this facility indicates that in the majority of instances, fatigue risk is being managed. However, there remain instances of elevated risk associated specifically with inadequate sleep and/or periods of extended wakefulness. To this end, the “Defences in Depth” approach is suggested as the most effective way to manage these specific instances where fatigue risk is actually elevated. This approach to fatigue management is new to many industries, including healthcare and will ensure that risk is assessed and managed at a number of levels, using a scientifically defensible model.

A20

T SLETTEN1, A SEGAL1, J REDMAN1, S LOCKLEY2, S RAJARATNAM1 1 Sleep and Chronobiology Group, School of Psychology, Psychiatry and Psychological Medicine, Monash University, Melbourne, Victoria, Australia, 2 Division of Sleep Medicine, Department of Medicine, Brigham and Women’s Hospital and Division of Sleep Medicine, Harvard Medical School, Boston, MA, United States Sleep restriction is associated with reduced daytime alertness and considerable decrements in performance. The aim of this study was to examine the effects of short-term, partial sleep restriction on subjective and objective measures of daytime sleepiness, and to assess the temporal pattern of sleepiness across the waking day. Healthy young adults participated in a between-subject design. Participants maintained a regular sleep-wake routine (2230–0630 or 0030–0830) for at least 3 weeks prior to a laboratory visit, confirmed with sleep diaries, actigraphy and calling a time-stamped answering machine. Sleep restricted subjects were scheduled to 5 hours sleep on the night before the laboratory visit and to 3 hours on the night in the laboratory with the same wake time as the 8 hour sleep. Control subjects maintained the 8 hour sleep schedule throughout. During the laboratory visit (~26 hours), participants were woken at their habitual wake time and remained in low light conditions (<2 lux) and constant posture during 9 hours of testing. Each hour, sleepiness was assessed objectively as electrooculographic (EOG) slow eye movements and electroencephalographic (EEG) delta and theta activity during the Karolinska Drowsiness Test (KDT, 3 min). In a subset of participants, eye and eyelid movements were also monitored using the Optalert Drowsiness Measuring System (ODMS) with a drowsiness score calculated for each minute of the KDT. Subjective sleepiness was assessed before and after each KDT using the Karolinska Sleepiness Scale (KSS) and an additional 9-point scale anchored with very alert and very sleepy. To date, 20 participants (8M, 12F) aged 22.9 ⫾ 2.8 (mean ⫾ SD) years have completed the protocol; 11 control and 9 sleep restricted. Preliminary analysis of the first hours of testing (2–4 hours after waking) confirms sleepiness was higher following sleep restriction as assessed by the KSS (p < 0.005, n = 20) and the ODMS drowsiness score (p < 0.05, n = 10). Correlations between the different measures of sleepiness demonstrate significant associations between ODMS score and slow eye movements (p < 0.05, R2 = 0.50, n = 9), ODMS score and KSS (p < 0.05, R2 = 0.41, n = 10) and the two subjective scales (p < 0.0001, R2 = 0.83, n = 20). The findings reveal the negative effects of moderate partial sleep restriction on both objective measures and subjective perceptions of sleepiness. This analysis will be extended to examine the temporal changes in these measures of sleepiness across the day. Supported by NHMRC grant # 436758, Compumedics Limited, Sleep Diagnostics Pty Ltd.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Sleep and Biological Rhythms 2008; 6: A21–A58

doi:10.1111/j.1479-8425.2008.00359_3.x

Poster Presentations Paediatric Sleep – Clinical P-01 CLINICAL PRESENTATION DOES NOT PREDICT SYMPTOMS OR SEVERITY OF SLEEP DISORDERED BREATHING IN INFANTS WITH CLEFT LIP AND/OR PALATE J MACLEAN1, D FITZGERALD1, D FITZSIMONS2, P HAYWARD2, K WATERS1 1 Respiratory Medicine, The Children’s Hospital at Westmead; Discipline of Paediatrics & Child Health, University of Sydney, Sydney, NSW, Australia, 2 Cleft Clinic, The Children’s Hospital at Westmead, Westmead, NSW, Australia Rationale: Infants with cleft lip and/or palate demonstrate noisy breathing both awake and during sleep but not all are referred for investigation of these symptoms. Little information is available on the risk of sleep disordered breathing (SDB) in infants with cleft palate lip and/or palate. Objective: To characterize the symptoms and polysomnography (PSG) findings for the clinical presentation of SDB in infants with cleft lip and/or palate. Methods: Infants with cleft lip and/or palate were recruited prospectively to participate in a longitudinal study of sleep and breathing. Families with a new baby with cleft lip and/or palate presenting to a single tertiary Cleft clinic were approached for participation. Infants underwent PSG with clinical information collected using questionnaires. Infants were classified as a clinical presentation if a referral for review of sleep or breathing symptoms was made prior to dissemination of PSG results. Results: PSG has been completed for 24 infants to date. Twelve infants were classified as having a clinical presentation. Males comprised 46% of the sample with 21% of children having an identified syndrome. Witnessed apnoea was reported for 21% of infants with struggling to breath during sleep reported in 47%. Snoring more than 1/2 the time was reported in 21% of infants with heavy or loud breathing during sleep reported in 79%. Age at PSG was 2.9 ⫾ 3.2 months and the sleep duration studied was 4.84 ⫾ 1.98 hours. The apnoea-hypopnoea index was 21.9 ⫾ 19.6 events/hour with a range from 3.9 to 83.2 events/hour. Obstructive apnoea comprised 41% of events with central apnoea accounting for 30% of events. The arousal index was 20.1 events/hour and the majority of arousal events were spontaneous arousals. Excluding two patients with known cardiac lesions, the mean awake O2 saturation was 95.5 ⫾ 2.7% with a minimum O2 saturation during sleep of 68.2 ⫾ 15.5%. Reported symptoms of SDB were similar in those infants with and without a clinical presentation. Apnoea-hypopnoea index, arousal index and minimum O2 saturation during sleep in infants with a clinical presentation did not differ from infants without a clinical presentation. Conclusion: Infants with cleft demonstrate a high frequency of symptoms of SDB with PSG results supporting a high incidence of breathing abnormalities during sleep. There were no clinical symptoms or polysomnographic characteristics that distinguished infants with a clinical presentation from those without clinical presentation. Based on these preliminary results, screening for SDB should be considered in all infants with cleft lip and/or palate.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

P-02 SLEEP QUALITY AND PSYCHOLOGICAL WELLBEING IN MOTHERS OF CHILDREN WITH DEVELOPMENTAL DISABILITY J CHU, A RICHDALE RMIT University, VIC, Australia Introduction: Sleep problems are common in children with a developmental disability (DD) but their impact on mother’s sleep and wellbeing has not been explored. The aim of the study was to investigate the impact of sleep problems and behaviour difficulties in children with a DD on their mothers’ sleep quality and psychological wellbeing. Method: Participants were mothers of children with a DD aged between 2 and 12 years. Participants were recruited from support organisations including Autism Victoria, Autism Behavioural Intervention Association, Asperger’s Syndrome Support Network and Convoy for Kids, and two early intervention centres, five Special Developmental Schools, and four Special Schools from the northern, southern, and eastern regions of Victoria. Results: The sample consisted of 46 mothers of children with a DD. Twenty-six children had an autism spectrum disorder (7.0 ⫾ 4.0 years) and 20 children had another DD (6.5 ⫾ 5.5 years). Pearson correlations were used to assess any associations between the CSS, SDQ total score, PSQI total score, DASS depression, DASS anxiety and PHS frequency and intensity stress scores. Based on the significant correlations, a series of simple regression were conducted, the significant predictors were then entered into a series of stepwise regressions to determine any multiple predictors of maternal sleep quality and maternal psychological wellbeing. Children’s sleep difficulties were significantly associated with maternal sleep quality, r (45) = .45, p < 0.01, depression, r (46) = .37, p < 0.01, stress intensity, r (46) = .46, p < 0.01, and stress frequency, r (46) = .49, p < 0.01. Discussion: The results showed that greater sleep and behaviour problems in children were significantly associated with more disturbed sleep, increased depression, anxiety and stress levels in mothers. Both mothers’ sleep disturbance and children’s behaviour problems were found to significantly predict increased stress intensity and hassle frequency in mothers. Maternal sleep disturbance also significantly predicted increased depression and anxiety levels in mothers. These findings suggest that children’s sleep and behaviour difficulties may be contributory factors to mothers’ sleep and consequently mothers’ wellbeing.

P-03 CARDIOVENTILATORY COUPLING IN PAEDIATRIC PATIENTS REFERRED FOR POLYSOMNOGRAPHY D ELDER, A CAMPBELL, P LARSEN, D GALLETLY University of Otago, Wellington, New Zealand Introduction: Cardioventilatory coupling is the triggering of inspiratory onset by a preceding heartbeat, and is seen most obviously when

A21

Healthy Function of Sleep

extrinsic inputs on ventilatory control are reduced (sleep, sedation and anaesthesia). The aims of this study were: 1. To demonstrate cardioventilatory coupling during sleep in a paediatric clinical population, 2. To look for differences in cardioventilatory coupling according to sleep state and clinical diagnosis. Methods: Paediatric patients referred for polysomnography (PSG) for clinical indications were recruited. Two selections of Nasal pressure and electrocardiogram (ECG) data were downloaded from Stage 2, Stage 4 and REM sleep. Data were entered into a purpose built programme in Labview. Inspiratory onsets (I) were determined for each breath and the timing of the R waves on the corresponding ECG trace were determined and R-R intervals calculated. The time interval between each R wave and the onset of the immediately following inspiration was measured (RI interval). As a measure of constancy of the relationship between R waves and inspiratory onset, the RI–1 interval (time between inspiration and the immediately preceding R wave) dispersion was measured using proportional Shannon Entropy of the RI–1 interval (SHa). Data were log transformed and analysed in a general linear model using repeated measures according to sleep state (SS) with covariates of diagnosis, age and BMI. Correlations between BMI and Shannon entropy in each sleep state were calculated using Spearman’s coefficient. Results: There were 113 subjects recruited: 89 had data suitable for analysis, 41 female and 48 male. Median age was 6 years (range 0.3 to 15 years). For 85, data were available in all three sleep states. Diagnostic groups where Central Nervous Syndrome – 16 subjects, Down syndrome – 13 subjects, Prader Willi Syndrome (PWS) – 7 subjects and Normal – 49 subjects. Shannon entropy values indicating high levels of cardioventilatory coupling were significantly related to SS (p < 0.0001) and to both BMI and diagnosis in association with sleep state (p = 0.007 and p = 0.0003 respectively). Shannon entropy was lower in Stage 2 and Stage 4 sleep in PWS patients and correlated significantly with BMI in Stage 4 sleep (p = 0.001) and Stage 2 sleep (p = 0.003) but not in REM. Discussion: Cardioventilatory coupling is seen more frequently in slow wave sleep where it is associated in particular with the presence of increased BMI and PWS. This implies that there is an augmentation of cardiac influence on ventilatory rhythm in these clinical groups.

was developed and presented at Childcare Centres by the author to provide information for parents and childcare staff, about why we sleep, how sleep works, the changes in sleep needs and structure across age, recognising common sleep problems and sleep disorders in infants, toddlers and children, where to seek help if needed and what are healthy sleep practices. From October 2004 to April 2008, 263 Childcare Centres in 11 different council regions in a major city were contacted about their interest in holding a SHE workshop at their Childcare Centre. A small questionnaire was given out to attendees to provide feedback about the information, presentation, and usefulness of the SHE workshop. Results: By the end of June 2008, 66 SHE workshops had been completed in 64 Childcare Centres with 389 questionnaires being collected from 46 centres. The responses to the questionnaire were evaluated and indicated that the workshops were well received and that 74% of participants reported a change in their attitude toward sleep issues in their family as a result of attending the SHE workshop. Discussion: During these SHE workshops, concerns of Childcare staff about some sleep practices occurring for children attending Childcare Centres were discussed as well as the concerns and questions from parents. These concerns and the positive results of the questionnaire highlight the need and benefits of sleep health education in the community for parents and childcare staff.

P-05 CARDIOVASCULAR VARIABILITY DURING PERIODIC LEG MOVEMENTS IN SLEEP IN CHILDREN L WALTER1, A FOSTER1, R PATTERSON1, G NIXON1, M DAVEY2, V ANDERSON3, J TRINDER4, A WALKER1, R HORNE1 1 Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Monash University, Clayton, Victoria, Australia, 2Melbourne Children’s Sleep Unit, Department of Respiratory and Sleep Medicine, Monash Medical Centre, Clayton, Victoria, Australia, 3Department of Psychology, Royal Children’s Hospital, Melbourne, Victoria, Australia, 4 Department of Psychology, University of Melbourne, Melbourne, Victoria, Australia

P-04 EDUCATING THE COMMUNITY ABOUT SLEEP HEALTH S CRANAGE Sleep Health Education, Chadstone Victoria, Australia Introduction: The community at large have limited awareness about sleep as an important health issue, as indicated by the 2004 report “Wake up Australia: The Value of Healthy Sleep”, organised by Sleep Health Australia. The importance of nutrition and exercise for health and wellbeing are well publicized, but community education about the importance of achieving adequate healthy sleep every day, especially for children, is lacking. The 2004 Sleep in America Poll organised by the National Sleep Foundation found that infants and children in the community were achieving less sleep than recommended. It also found that parents and caregivers were not always aware of the best sleep practices for their children. There is an urgent need to put sleep health education in the health curriculum of every child, their family and their community. Methods: A program called Sleep Health Education (SHE) began in 2004 with the aim of increasing sleep health awareness in the community, especially for parents and caregivers of infants, toddlers and preschoolers. A SHE workshop entitled “The Sleeping Baby and Toddler”

A22

Background: Sleep related Periodic Leg Movements Disorder (PLMD) consists of episodes of repetitive and stereotypic leg movements occurring in sleep. The disorder is known to have a significant impact on sleep quality and duration [1]. In adults, periodic leg movements (PLMS) are associated with a shift to greater sympathetic balance [2]. However, little research has been performed to investigate the effects of PLMS on children. This study aimed to quantify changes in heart rate variability occurring during periods of PLMS, using power spectral analysis. We hypothesised that there would be a change in sympathovagal balance in PLMS+ compared with PLMS-periods. Methods: Ten children (8M/2F; 7–12 y) underwent overnight polysomnography (PSG) and PLMS were scored by trained sleep technologists according to World Association of Sleep Medicine/International Restless Legs Syndrome Study Group standards [3]. Mean periods of 5 min with PLMS (PLMS+) and 5 min without PLMS (PLMS-) for each subject, were analysed by frequency-domain methods during stable stage 2 N-REM sleep. ECG data from the PSG were analysed using Chart software. PLMS+ and PLMS-periods were compared using MannWhitney U-tests. Results: There were no significant changes in very low frequency (VLF), low frequency (LF), high frequency (HF) or Total Power between PLMS+ and PLMS-periods. The LF/HF ratio was significantly higher in the PLMS+ periods compared to PLMS-periods (Table 1).

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

Table 1 VLF power [<0.04Hz] LF power [0.04–0.15Hz] HF power [0.15–0.40Hz] Total power LF/HF ratio

PLMS+ periods Group Mean ⫾ SE

PLMS– periods Group Mean ⫾ SE

P

2850 ⫾ 900 msec2 2371 ⫾ 820 msec2 2492 ⫾ 1202 msec2 8040 ⫾ 2881 msec2 1.35 ⫾ 0.18

792 ⫾ 200 msec2 1183 ⫾ 417 msec2 2970 ⫾ 1259 msec2 5172 ⫾ 1901 msec2 0.53 ⫾ 0.09

n.s. n.s. n.s. n.s. 0.002

Discussion: The results suggest that PLMS in children occurs with significant changes in sympathovagal balance, similar to that observed in adults. However, the trend for total power to increase during leg movement periods suggests that conclusions should be interpreted with caution. References [1]. Picchietti D (2005) Sleep. [2]. Sforza et al. (1999) Neurology. [3]. Zucconi et al. (2006) Sleep Medicine.

only children with a confirmed cleft palate, tonsillar grading correlated with both apnoea-hypopnea index (0.63, p = 0.036) and obstructive index (0.65, p = 0.023). Conclusion: In children referred for sleep evaluation in a multidisciplinary cleft clinic, snoring is the most common sleep related complaint. Tonsil size correlates with the severity of OSA in children with cleft though standard grading may under estimate the significance of tonsil grading.

P-07 THE IMPACT OF SLEEP QUALITY AND QUANTITY ON DAYTIME BEHAVIOUR AND PERFORMANCE IN AUSTRALIAN PRIMARY SCHOOL CHILDREN AGED 6–8 YEARS D SANDFORD, J DORRIAN, S BLUNDEN University of South Australia, Adelaide, SA, Australia

P-06 SLEEP CONCERNS IN A MULTI-DISCIPLINARY CLEFT CLINIC J MACLEAN1, S TAN1, D FITZGERALD1, C SETON2, K WATERS1 1 Respiratory Medicine, The Children’s Hospital at Westmead; Discipline of Paediatrics & Child Health, University of Sydney, Sydney, NSW, Australia, 2 Respiratory Medicine, The Children’s Hospital at Westmead, Sydney, NSW, Australia Rationale: Children with cleft palate are at increased risk of obstructive sleep apnoea (OSA) but symptoms are often not recognized and investigated. Improved awareness of OSA symptoms and signs in children with cleft palate may improve recognition. Objective: To examine the characteristics of a series of children from a multi-disciplinary cleft clinic who presented for review of sleep related symptoms. Methods: From medical records, we obtained referral information, clinical history, polysomnography results, and outcome. Results: Thirty-two children were reviewed over a 10 month period. Sleep review was initiated by plastic surgeons for 47%, by a speech pathologist for 34%, by otolaryngologists for 16% and by a nurse for 3% of children. The primary reason for referral was snoring in 65%, breathing concerns other than snoring for 22%, a prior history of OSA in 9% and headache in 3% of children. The mean age at referral was 6.0 ⫾ 4.4 years. Syndromes had been identified in 25% of children. Speech problems were present in 31% of children and current or past respiratory disease was present in 31%. Snoring was reported by 81% of parents. Awakenings (46%), restless sleep (31%), and struggling to breathe (31%) were more common complaints than witnessed apnoea (6%) or sleep disturbance (6%). Few children had difficulties falling asleep or awakening in the morning (9% each), 16% had current or past symptoms of a parasomnia, and 6% had nocturnal enuresis. daytime sleepiness was reported in 28% of children with 12.5% of parents reporting school, concentration or attention problems. Six percent had undergone surgery for OSA in the past and 25% were currently on medical treatment for snoring or OSA. Significant tonsillar tissue (ⱖgrade 3) was seen in 23% of children. Polysomnography was indicated for 25 children with results available to date from 16 children. All but 1 child had an apnoea-hypopnoea index >1 event/hour with 62% demonstrating an obstructive index >1 event/hour (1–5 events/hour in 3 children; 5–10 events/hour in 4 children; >10 events/hour in 3 children). Including

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Introduction: The aim of this study was to examine in detail the sleep habits of 83 (39 boys and 44 girls), 6–8 year old South Australian school children over a 2 week period. The reported sleep was analysed to identify differences in behaviour reported by parents and teachers. This study is the first to examine Australian children’s sleep over a 2 week period in this age group, as well as attempt to establish optimal sleep thresholds rather than mean sleep durations. Methods: Children were recruited from 10 randomly nominated public (n = 6) and private (n = 4) schools in the Southern Fleurieu region of South Australia. Participating children’s parents were given a sociodemographic questionnaire, Sleep Disturbance Scale (SDSC), Conners’ Parent rating Scale – Revised (short version) (CPRS-R (S)), and a specifically modified 14 day sleep diary. The participants main teacher was also invited to participate in the study by completing a Conners’ Teacher Rating Scale – Revised (short version) (CTRS-R (S)). A series of repeated measures analysis of variance (ANOVA) was used to analyse differences between weekday and weekend sleep. Exploratory analyses (Univariate ANOVA) were conducted to in order to explore differences in demographic variables. Univariate analyses of covariance (ANCOVA) using child age as the covariate were conducted to test differences between behavioural groups and sleep measures. Results: This study found that sleep duration was not significantly different for children who had reported elevated behavioural profiles compared to children reported as having ‘normal’ behaviour profiles. In contrast, sleep quality and feeling refreshed in the morning were found to be significantly different for children displaying elevated profiles than children reported as having ‘normal’ profiles. Also children with elevated behaviour profiles were found to have significantly higher symptoms of initiating and maintaining sleep, sleep-wake transition problems, excessive morning and daytime sleepiness and reported snoring. Discussion: This pilot study has shown that consistent with established sleep averages, Australian children in the 6–8 year old age group are reported by parents to be getting recommended average sleep durations. Although optimal sleep durations for children of this age group; were not established, this study has identified that, optimal sleep for children may not merely rely upon sleep duration but may need to be assessed in conjunction with the quality of the child’s sleep architecture.

A23

Healthy Function of Sleep

P-08 DEVELOPMENTAL CHANGES IN THE ASSOCIATION BETWEEN OBESITY AND SLEEP DISORDERED BREATHING IN CHILDREN AND ADOLESCENTS M KOHLER1, S THORMAEHLEN2, D KENNEDY1, C VAN DEN HEUVEL1, K LUSHINGTON2, J MARTIN3 1 University of Adelaide, Adelaide, Australia, 2University of South Australia, Adelaide, Australia, 3Women’s and Children’s Hospital, North Adelaide, Australia Obesity is a major risk factor for sleep disordered breathing (SDB) in adults; however, whether a similar risk exists for children is unclear with some studies reporting a strong association but not others. A possible explanation for the inconsistent findings is age, with studies in older children more likely to report a relationship than those in younger children. To examine this further we undertook a retrospective study and separately explored the relationship between obesity and OSAS severity in young and older children. This study reviewed overnight polysomnography (PSG) and body mass results from 220 children and adolescents aged 2 to 18 years who were referred for suspected SDB but were otherwise normal patients. Patients were divided into six age groups (2–4 y, 4–6 y, 6–8 y, 8–10 y, 10–12 y and 12–18 y). Groups were matched for gender, ethnicity, body mass, socioeconomic status (SES), snoring, SDB severity, and family history of snoring and smoking. There was a greater proportion of overweight/obese patients amongst adolescents (13–18 y) (71%) compared to all other age groups (29%–45%). Body mass was not significantly associated with either obstructive apnoea/hypopnoea index (OAHI) (r(220) = 0.10) or central apnoea index (CAI) (r(220) = 0.02) across the entire group. When considering each age group separately, a significant association was found amongst adolescents only for both OAHI (r(23) = 0.55, p < 0.01) and CAI (r(23) = 0.63, p < 0.005). Accordingly, both OAHI and CAI were greater amongst overweight/obese adolescents compared to normal weight adolescents. No difference in OAHI or CAI was found between younger overweight/obese compared to normal weight patients. The increased risk for SDB due to obesity seen amongst adults appears to also be evident amongst adolescents. However, this association was not found amongst younger children, suggesting the association is dependent on developmental or age-related changes. While detrimental to general health and associated with similar adverse outcomes, obesity should not be viewed as a major risk factor for SDB in young Caucasian children. However, excess body mass amongst older children and adolescents with SDB should be considered important in determining risk and treatment approach.

have a higher arousal threshold and may thus be more resilient to sleep disturbance in hospital than adults. Methods: Children admitted to hospital for an expected duration of one night were recruited to the study. All children were cared for in a 4-bed cubicle in a paediatric ward. Sleep duration was estimated using an actigraph (Actiwatch AW64, Mini Mitter Co.) worn on the nondominant wrist for the hospital night and 1 week after discharge. Actiwatch data was analysed using Actiwatch-Sleep® v5.38. Mean sleep duration in hospital was compared with normal values and the first night at home using paired t tests (results as mean ⫾ SE). Normal values for mean sleep duration by age are from Iglowstein et al (2003)1. Linear regression and t tests were used to investigate the relationship between sleep duration and age and gender respectively. Results: 7 children have been recruited to date (5 males, mean age 5.3 ⫾ 1.3 y), including 5 children admitted for medical treatment (e.g. an exacerbation of asthma) and 2 for adeno-tonsillectomy. Sleep duration in hospital was significantly less than the normal mean duration for age (8.8 ⫾ 0.57 vs 11.1 ⫾ 0.1 h, p = 0.006). No association was seen between sleep duration in hospital and age (r2 = 0.07, p = 0.56), and sleep duration was not different between male and female subjects (9.1 ⫾ 0.47 vs 8.1 ⫾ 2.0 h, p = 0.49). 5 subjects had successful actigraphy at home. Overall, sleep duration on the first night at home was not different from the hospital night (10.0 ⫾ 0.18 vs 8.6 ⫾ 0.71 h, p = 0.14), but on preliminary analysis the difference was greater in the medical group (mean 10.2 vs 7.7 h in the medical group and 9.7 vs 9.8 h in the surgical group). Discussion: When spending one night in hospital, mean sleep duration in children is significantly shorter than the mean normal sleep duration for age. Preliminary analysis suggests that the reduction in sleep duration in hospital is greatest in a sub-group of children receiving treatment for an acute medical condition, but further recruitment is underway to confirm this. Increased sleep disturbance from administration of medication in the medical group, or sleep rebound due to prior sleep disturbance related to obstructive sleep apnoea in the surgical group, may contribute to this difference. Reference 1. Iglowstein et al. Pediatrics 2003 Feb;111(2):302–7.

Sleep & Breathing – Clinical 1 P-10 PREVALENCE OF POSITIONAL (ROTATIONAL AND VERTICAL) OSA A GILLMAN1, T ROEBUCK1, E VAN BRAAK1, M NAUGHTON1,2 1 The Alfred Hospital Sleep Disorders Unit, Melbourne, VIC, Australia, 2 Monash University, Melbourne, VIC, Australia

P-09 SLEEP DURATION IS REDUCED IN CHILDREN SPENDING ONE NIGHT IN HOSPITAL S YORKSTON, K SOUTHWELL, D O’DRISCOLL, R HORNE, G NIXON Ritchie Centre for Baby Health Research, Melbourne, Victoria, Australia Introduction: Adults have reduced sleep in hospital, leading to rebound of total sleep and REM sleep on return home, which may have medical consequences. Hospitalisation is associated with subsequent behavioural disturbance in children, but the contribution of sleep deprivation in hospital to this disturbance has not previously been studied. Children

A24

Background: The relative effects of (rotational and vertical) position upon OSA severity is not well known. Aims: To assess prevalence of (1) rotational position in OSA using an old traditional1 (supine : non-supine event ratio >2, ie S : NS > 2) and a new stricter functional definition of S : NS > 2 plus a NS-AHI < 15 and <52 eph (2) vertical position3 in patients attending for diagnostic polysmonography (PSG). Methods: Retrospective study of consecutive diagnostic in-laboratory PSGs over 4 months of patients with OSA (AHI > 5 events/hour) and at least 30 minutes in both NS and S rotational positions. Rotational and vertical position was assessed by positional sensor (Compumedics) with digital video correction and pillow count.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

Results (mean ⫾ sd): 50 subjects, age 56 ⫾ 12 years, BMI 32 ⫾ 3 kg/ m2, ESS 9 ⫾ 5 and AHI 46 ⫾ 25 eph were studied. 70% patients used >1 pillow. (** = p < 0.05 compared with non-positional) Definition Prevalence (%) Age (yr): Gender (%씹) BMI (kg/m2) ESS Pillows (no) Mean SpO2 (%) Min SpO2 (%)

NonPP 36 59 ⫾ 12 79 36 ⫾ 10 8⫾5 1.8 ⫾ 0.4 95 ⫾ 2 85 ⫾ 8

Trad PP 64 54 ⫾ 12 77 30 ⫾ 6** 9⫾5 1.6 ⫾ 0.6 95 ⫾ 2 83 ⫾ 7

FunPP ⫾ NS AHI ⱕ 15 44 53 ⫾ 11 73 29 ⫾ 4** 11 ⫾ 5 1.5 ⫾ 0.5** 96 ⫾ 1** 87 ⫾ 6**

FunPP ⫾ NS AHI ⱕ 5 24 51 ⫾ 9** 83 27 ⫾ 4** 11 ⫾ 6 1.6 ⫾ 0.5 97 ⫾ 1** 88 ⫾ 7**

The prevalence of patients with REM SDB (REM AHI : NREM AHI > 2) was 36% (55 ⫾ 13 years, 14/18% males, BMI 32 ⫾ 7 kg/m2 and ESS 9 ⫾ 6). Discussion: In OSA patients, between 24–64% have rotational positional OSA, 70% patients request vertical positional therapy during a sleep study and 36% have REM SDB. Thinner and younger patients have more positional OSA and may be a target group for positional therapy4 or possibly auto-titrating CPAP. Laboratory quality assurance is required to ensure accurate rotational and vertical positionality is documented with as great vigor as has been shown to sleep staging. Reference 1. Carwright et al Sleep 1984;7:110–114. 2. Mador Chest 2005;128:2130–2137. 3. Neill et al. AJRCCM 1997;155:199–204. 4. Cartwright. Sleep 1985;8:87–94.

P-11 DEVELOPMENT OF AN INTELLIGENT DECISION SUPPORT SYSTEM FOR SLEEP DISORDER DIAGNOSIS J BLAKE1, S SMITH2, J DOUGLAS3, R SELLERS3 1 University of Sunshine Coast, Queensland, Australia, 2Queensland University of Technology, Queensland, Australia, 3Queensland Health, Queensland, Australia Diagnosis of sleep disorders requires complex integration of a wide range of data types, and consistent and transparent application of decision criteria. This process may be helped by an intelligent decision support system (IDSS). An IDSS is a decision support system coupled with built in expertise and an inference engine. The aim of this study was to develop and test a new decision support system for sleep disorders. Method: Standard IDSS approaches were used explore the issues associated with the diagnosis of sleep disorders by physicians in 2 clinical sleep disorders services. This included qualitative study of how information systems could be developed to assist both patients and physicians access information in a format which facilitates understanding. The first stage of the IDSS involved the development of an online sleep diary. The diary presents a graphical representation of sleep patterns, to enable quick comprehension of sleep trends by patients and physicians. Standard sleep dairy statistics (SOL, Duration, WASO etc.) are automatically produced, presented in a statistic summary and stored in the Diary database. This allows flexible use of the diary data for use by both the physician and patients. The second stage of the IDSS involves an online admission questionnaire. This enables the patient to gather information at home or anywhere Internet access is available. This questionnaire can be filled in over several sessions, wherever the patient feels comfortable. This data is available in summary and full statistical form

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

for the physician prior to the initial consultation. An online tool associated with both these applications will collate patient data and provide an evidence base for current diagnostic criteria, and for development of the IDSS. An IDSS can assist the physician to provide a consistent diagnosis by establishing a protocol, which may lead different physicians to the same diagnoses. The innovation of this study is that it will provide a tool to increase the efficiency and best practise of a physician, at a time when the burden of these disorders is increasing.

P-12 PREDICTION OF OBSTRUCTIVE SLEEP APNOEA WITH CRANIOFACIAL PHOTOGRAPHIC ANALYSIS R LEE1, P PETOCZ2, T PRVAN2, A CHAN1, R GRUNSTEIN3, P CISTULLI1 1 Centre for Sleep Health and Research, Department of Respiratory Medicine, Royal North Shore Hospital, St Leonards, NSW, Australia, 2 Department of Statistics, Macquarie University, Sydney, NSW, Australia, 3 Woolcock Institute of Medical Research, University of Sydney, Sydney, NSW, Australia Introduction: To develop models based on craniofacial photographic analysis for the prediction of obstructive sleep apnoea (OSA). Methods: Prospective cohort study undertaken in a university teaching hospital sleep investigation unit. One hundred and eighty subjects referred for the initial investigation of OSA were recruited consecutively. Clinical assessment and frontal-profile craniofacial photographic analyses were performed prior to polysomnography. Prediction models for determining the presence of OSA (apnea-hypopnoea index [AHI] ⱖ10) were developed using logistic regression analysis and classification and regression trees (CART). Results: Obstructive sleep apnoea was present in 63.3% of subjects. Using logistic regression, a model with four photographic measurements (face width, eye width, cervicomental angle and mandibular length 1) correctly classified 76.1% of subjects with and without OSA (sensitivity 86.0%, specificity 59.1%, area under the receiver operating characteristics curve [AUC] 0.82). Combination of photographic and other clinical data improved the prediction (AUC 0.87); whereas prediction based on clinical assessment alone was lower (AUC 0.78). The optimal CART model provided a similar overall classification accuracy of 76.7%. Based on this model, 59.4% of the subjects were classified as either high or low risk with positive predictive value of 90.9% and negative predictive value of 94.7%, respectively. The remaining 40.6% of subjects have intermediate risk of OSA. Discussion: Craniofacial photographic analysis provides detailed anatomical data useful in the prediction of OSA. This method allows OSA risk stratification by craniofacial morphological phenotypes.

P-13 THE ASSOCIATION BETWEEN INSULIN RESISTANCE AND OBSTRUCTIVE SLEEP APNEA IN JAPANESE R HIRAMATSU, T KASAI, K-I MAENO, Y TOMO, M KATO, F KAWANA, K NARUI Toranomon Hospital, Tokyo, Japan Background: The insulin resistance (IR) is frequently observed in patients with obstructive sleep apnea (OSA) and it was suggested that

A25

Healthy Function of Sleep

the IR in OSA patients was associated with coexisting obesity. However, the repetitive episodes of apnea and hypopnoea during sleep result in sympathetic overactivation, and there was several reports showed significant association between IR and OSA independent from obesity. However, it remains controversial whether an independent association between IR and OSA exists and in particular, there are no data among Japanese population. Therefore, we investigated the association between OSA and IR expressed as homeostasis model assessment method (HOMA-IR) among Japanese patient population. Methods: We analyzed the data about consecutive patients diagnosed as OSA. Multiple regression analysis was used to determine the relation between several parameters which showed the severity of OSA and HOMA-IR. Log transformed HOMA-IR (because distribution of HOMA-IR was skewed) was used as a dependent variable and as independent variables, age, gender, body mass index (BMI), waist circumference, presence of hypertension, severity of OSA [apnea–hypopnea index (AHI) or 3% oxvgen desaturation index (3% ODI) or lowest oxyhemoglobin saturation (SO2) or% of times SO2 < 90% or arousal index was included in the separate models] were included. Results: The data about 483 consecutive patients (439 males), were analyzed. Patients with severe OSA were predominantly included (AHI 44.0 ⫾ 21.5/h, 3% ODI 32.8 ⫾ 23.5/h, lowest SO2 73.8 ⫾ 12.6%, % of times SO2 < 90% 27.3 ⫾ 29.0%) and their median HOMA-IR was 2.02. The results of multiple regression revealed that the severity of OSA excepting for arousal index was significantly associated with the HOMA-IR [coefficient: AHI 0.02 (P = 0.007), 3%ODI 0.001 (P = 0.03), lowest SO2 -0.003 (P = 0.007),% of times SO2 < 90% 0.002 (P = 0.0014), arousal index 0.001 (P = 0.09)] in addition to the age, waist circumference, hypertension. Conclusion: The severity of OSA was significantly associated with the IR expressed as HOMA-IR and especially only parameters associated with hypoxia showed significant association among Japanese patient population. This may indicate OSA per se, particularly its induced hypoxia, is the determinants among Japanese patients with OSA.

P-14 FRIEDREICH’S ATAXIA AND SLEEP DISORDERED BREATHING J COPLAND, M HO Monash Sleep Centre, Monash Medical Centre, Melbourne, Victoria, Australia Introduction: Friedreich’s Ataxia (FRDA) is a hereditary degenerative neuromuscular disorder with an incidence of about 1 in 30,000. The symptoms often begin at puberty with progressive ataxia and dysarthria which leads to musculoskeletal deformities and immobility. Survival beyond early adulthood is rare, with cardiomyopathy as the main cause of death. We hypothesize that patients with FRDA may have sleep disordered breathing (SDB). We report 8 cases of FRDA who presented with symptoms of snoring, unrefreshing sleep and daytime sleepiness. Methods: All subjects underwent NPSG to assess sleep quality and to document type and severity of sleep disorder. Results: Eight subjects (2M, Age = 36.6 ⫾ 11.3 years, BMI = 26.1 ⫾ 9.4 kg/m2), were studied. As a consequence of the disease process, five subjects had moderately severe functional impairment (SFI) and disability and two had mild functional impairment (MFI). The Sleep Architecture for the group was variable with no clear trends observed: Sleep Latency ranged from 4–59.5 mins, Sleep Efficiency ranged from 55.9– 91.2% and the mean WASO was 77.8 ⫾ 50.8 min (mean ⫾ sd), with a mean NREM percentage of 81.9 ⫾ 5.2% and REM percentage 18.1 ⫾ 5.2% (mean ⫾ sd).

A26

The mean RDI for the group was 18.0 ⫾ 17.0/hour (mean ⫾ sd) with a range of 0–42.6/hour. Four subjects had moderate OSA; two subjects had mild OSA in REM only and two had no OSA. Rare central apnoeas were observed in only three subjects and no hypoventilation was observed. TcCO2 was within normal limits. The mean PLMI was 1.7 ⫾ 3.2/hour (mean ⫾ sd) with mild PLM seen in two subjects. Conclusion: No clear trends were observed in the NPSG variables in the FRDA subjects. But those with symptoms of snoring and unrefreshing sleep should be referred to a Sleep Centre.

P-15 DO COLLATERAL ACTIVITIES INCREASE WITH SLEEP RESTRICTION? A BEHAVIOURAL ANALYSIS DURING DRIVING IN SLEEP APNEA PATIENTS T DUGGAN1, K HARRIS1, A VAKULIN2, C VAN DEN HEUVEL2, N ANTIC2, P CATCHESIDE2, D MCEVOY2, C LITCHFIELD1, S BAULK2 1 School of Psychology, University of South Australia, Adelaide, Australia, 2 Adelaide Institute for Sleep Health, Repatriation General Hospital, Adelaide, Australia Introduction: Obstructive Sleep Apnea (OSA) is a common disorder involving partial and/or complete airflow interruption during sleep, with significant negative impacts on health. As well as increased cardiovascular risk, patients may experience cognitive dysfunction, declines in vigilance, and excessive daytime sleepiness – resulting in increased risk of motor vehicle accident (MVA). Previous studies have used simulated driving to establish a relationship between sleepiness and heightened collateral activities (e.g. postural adjustments) in healthy sleepers. However, behavioural analysis in OSA is yet to be explored. We hypothesise that as patients become sleepy, they will engage in an increased frequency of collateral activities whilst driving. Methods: As part of a repeated measures behavioural analysis, initial data from six patients with severe OSA (5 m, 1 f; mean[sd]: age = 57.8 [9.7]; BMI = 32.2 [7.2]; AHI = 67.8[19.6]/hour was examined. All completed a 90-minute simulated drive following nights of normative (8 hrs) and restricted (4 hrs) sleep. Steering deviation was recorded as a measure of driving performance, while frequency totals of collateral activities (defined by ethogram: postural adjustments; eye movements; self-centred gestures; facial activities; ludic activities) were recorded via video, and later scored in 10-minute epochs as measures of driver behaviour. Results: Initial ANOVA results reveal significant time (p < 0.05) and condition (p < 0.05) effects for steering deviation. Trends of increasing self-centered gestures, postural adjustments and eye activities were found over time in the SR condition. Discussion: Despite the limited number of patients examined thus far, it appears that a relationship may exist between the presence of collateral activities, deteriorated driving performance and sleepiness in OSA patients. This data has important implications, suggesting that collateral activities may represent an alternative observable measure of driver fatigue. Specifically, the data is expected to contribute towards road safety initiatives concerning OSA patients and further high-risk driving populations (i.e. abnormal sleepers, shift-workers, professional drivers).

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

P-16 AWARENESS OF SLEEPINESS AND PERFORMANCE DURING SIMULATED DRIVING IN OBSTRUCTIVE SLEEP APNOEA K HARRIS1, T DUGGAN1, A VAKULIN3, C VAN DEN HEUVEL5, N ANTIC6, D MCEVOY2, J DORRIAN4, S BAULK4 1 School of Psychology, University of South Australia, South Australia, Australia, 2Adelaide Institute for Sleep Health, Repatriation General Hospital, South Australia, Australia, 3School of Molecular and Biomedical Science, Discipline of Physiology, University of Adelaide, South Australia, Australia, 4Centre for Sleep Research, University of South Australia, South Australia, Australia, 5Child and Adolescent Sleep Health Unit, Discipline of Paediatrics, University of Adelaide, South Australia, Australia, 6School of Medicine, Flinders University, South Australia, Australia Introduction: Obstructive Sleep Apnoea (OSA) affects up to 10% of the Australian population and is emerging as a major public health problem – linked to cardiovascular risk, cognitive dysfunction and increased motor-vehicle accident risk. Research has used driving simulation to examine objective sleepiness and performance, but subjective ratings, which are important in the decision to drive, are less well investigated. We hypothesise that OSA patients will perform worse, and report being sleepier than controls, but that controls will give more accurate selfestimates of performance. Methods: As part of a larger study examining driving and neurocognitive impairment in OSA, we examined preliminary data from 6 severely affected OSA patients, (mean [sd]: age = 47.3 [14.3]y; BMI = 41.11 [8.51]; AHI = 72.19 [21.16]/hour), and 6 healthy controls (age = 45.17 [7.33]y; BMI = 24.70 [3.10]; AHI = 11.52 [2.89]/hour). Participants completed a mid-afternoon 90 minute simulated drive after a regular night’s sleep. Performance (steering deviation), subjective performance, and subjective sleepiness were measured. Results: Repeated measures ANOVA demonstrated a significant effect of group (p = 0.012) on steering deviation, but no significant time or interaction effects. Subjective driving performance showed significant effects of group (p = 0.03) and time (p < 0.001), but no interaction effect. Subjective sleepiness showed a significant effect of time (p < 0.001), but no group or interaction effects. Conclusions: OSA patients demonstrated greater lane deviation, and reported greater impaired performance compared with controls. Both groups reported becoming sleepier over time during driving. Further analyses are needed with a larger sample to explore the relationships between all three variables. This research has important implications for determining links between subjective and actual performance measures – thus aiding OSA patients’ awareness of sleepiness while driving.

P-17 ASSOCIATION BETWEEN METABOLIC SYNDROME AND OBESE PATIENTS WITH SEVERE OBSTRUCTIVE SLEEP APNEA WHO REVEALED HYPERCAPNIA DURING AWAKENING K NARUI, T KASAI Toranomon Hospital Sleep Centre, Tokyo, Japan Background: In general, sleep disordered breathing among obesity population is divided into two disorders; sleep apnea-hypopnea syndrome (OSAHS) and obesity hypoventilation syndrome (OHS), according to the presence or absence of hypercapnia during awakening.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

However, the difference in the characteristic and features between these two disorders was not fully elucidated. Methods: One hundred and twenty four consecutive patients who complaint about snoring and obesity (body mass index: BMI ⱖ 30 kg/m2) were prospectively enrolled. Among them, 97 patients with severe OSAHS (AHI ⱖ 30/h) were divided into two groups according to their PaCO2 level during awakening (PaCO2 ⱖ 45Torr: OHS group, <45Torr OSAHS group). These two groups were compared. Results: There were 72 patients with OSAHS and 25 patients with OHS (26%). There were no significant difference in age (45 ⫾ 11 versus 46 ⫾ 12), gender (female 7% versus 4%), BMI (34 ⫾ 4 versus 34 ⫾ 3 kg/m2), AHI (61 ⫾ 21 versus 66 ⫾ 26/h) and Epworth sleepiness scale (11 ⫾ 5 versus 11 ⫾ 5). However, OHS group revealed significantly lower PaO2 level and of course higher PaCO2 level (75 ⫾ 7 versus 80 ⫾ 11, P = 0.04, 47 ⫾ 3 versus41 ⫾ 3 Torr, P < 0.01). Additionally, patients with OHS were more frequently complicated with metabolic syndrome (72 vs 49%, P = 0.04) and reveled higher metabolic score (the number of satisfied criteria of metabolic syndrome) (3.2 vs. 2.7, P = 0.04). In multivariate logistic regression analysis (including age, gender, BMI, AHI, desaturation time, PaO2 level and metabolic syndrome as an independent variables) the presence of metabolic syndrome was associated with OHS (odd ratio: 3.00, 95% confidence interval: 1.04–8.65, P = 0.04). Conclusions: Among severe OSAHS, 26% of patients revealed OHS. The presence of metabolic syndrome was the independent predictor for OHS.

P-18 CLINICAL ASSESSMENT PATHWAY: REPORTING PHYSICIAN AGREEMENT AND EFFECT OF SLEEP STUDY LOCATION ON DIAGNOSIS AND TREATMENT RECOMMENDATIONS A CAMPBELL, A NEILL Otago University Wellington, Wellington, New Zealand Unattended home based sleep studies are being increasingly used to evaluate suspected OSA. Previous research has focused primarily on technical reliability but differences in clinical interpretation are an important source of potential variation. The aim of this study was to determine the level of agreement of two physicians reporting sleep studies undertaken in the laboratory and in the home environment. Methods: Consecutive adult patients with suspected OSA were recruited and underwent 3 polysomnography (PSG) studies, one in the laboratory followed by one home based PSG and another lab based PSG in random order. Two physicians blinded to study location and subject identity were given relevant information (ESS, BMI, sex, age, medications, occupation, summary clinical data) and the results of each sleep study (AHI, AI, average desaturation, nadir SaO2, PLMS index). Physicians were asked: 1. Does this patient have sleep apnoea? 2. Should this patient be treated? 3. If yes with what? CPAP, MAS, conservative, position modification, upper airway surgery, other (one to be chosen). Results were analysed for agreement between physicians and agreement between study location (laboratory and home) for each subject. Results: 30 subjects completed all 3 studies but one home study was unsuccessful and removed from analysis (n = 29). Agreement between physicians was high with both physicians agreeing on diagnosis for 100% of studies, need for treatment 91% and need for CPAP 92.2%. Study location results (home versus laboratory 2): Doctor 1: Diagnosis agreed 25/29 (86%), Treatment agreed 26/29 (90%), CPAP agreed 25/29 (85%).

A27

Healthy Function of Sleep

Doctor 2: Diagnosis agreed 26/29 (86%), Treatment agreed 26/29 (90%), CPAP agreed (90%). Where reporting doctors changed the diagnosis, they both changed for the same 4 patients, where doctors changed their recommendation for CPAP 4 of 5 patients were the same for each. Conclusion: Agreement between two physicians reporting sleep studies for diagnosis and CPAP recommendation is high and indicates adherence to local diagnostic criteria and treatment protocols with both Physicians interpreting the results in a similar way. In a minority the sleep study location affected the AHI at a key diagnostic threshold and influenced treatment recommendation. Home based PSG equipment provided by Compumedics Ltd. Funded by an Otago University Research Grant.

Respiratory Physiology and Respiratory Failure P-19 IS THE VOLUME OF THE PASSIVE PHARYNGEAL LUMEN GREATEST AT SUPINE FUNCTIONAL RESIDUAL CAPACITY IN HEALTHY HUMANS? P WU1, T AMIS1,2, S LEE1, M VERMA1, J WHEATLEY1,2 1 Ludwig Engel Centre for Respiratory Research, Westmead Millennium Institute, NSW, Australia, 2University of Sydney at Westmead Hospital, Westmead, NSW, Australia Introduction: Lung volume is thought to influence pharyngeal airway patency via passive elastic forces transmitted to pharyngeal wall tissues. Reduced lung volume during sleep has been implicated as a potential contributive factor for upper airway narrowing/closure in patients with obstructive sleep apnoea. In the present study we used conditions of depressed upper airway muscle activity to examine effects of lung volume change on passive pharyngeal airway lumen volumes in awake, supine humans. Methods: Five healthy male subjects (age: 25–50 years) inspired from functional residual capacity (FRC) to total lung capacity (TLC) and then expired to residual volume (RV), all while in the supine posture with controlled head and neck position. Acoustic pharyngometry (applied via a mouthpiece) was used to measure pharyngeal lumen volume during a 5 sec breath-hold at each lung volume (lung volume held constant by voluntary glottic closure). All measurements were made during voluntary relaxation of both pharyngeal airway and diaphragm/ abdominal muscles (confirmed by minimal electromyographic activity detected by submental and lower rib cage surface electrodes). Pharyngeal airway volume (PAV; 10 cm segment distal to the oro-pharyngeal junction [as defined at FRC]) was calculated from the cross-sectional area versus distance output of the pharyngometer (Eccovision Acoustic Pharyngometer P/N 3100, Hood). Data were expressed as group mean ⫾ SD and analysed using repeated measures ANOVA with a Bonferroni post hoc test. Results: At supine FRC the group mean PAV was 26.3 ⫾ 6.7 mL, a value significantly greater than at supine RV (20.8 ⫾ 5.0 mL; p < 0.01) but not significantly different to that at supine TLC (23.4 mL ⫾ 6.6 mL, p > 0.05). PAV values at RV and TLC were not significantly different (p > 0.05). Discussion: We conclude that in the supine posture, lung volumes below FRC are associated with reduced passive pharyngeal airway lumen volume, whereas lung volumes above FRC are not necessarily associated with increased lumen volume. These findings suggest that, for supine subjects, FRC may be characterised by linkage of lung

A28

volume to optimal pharyngeal airway volume and that shifts in lung volume away from FRC have the potential to negatively impact on upper airway patency. Supported by NHMRC of Australia.

P-20 VENTILATION-PERFUSION MISMATCH ACCELERATES ARTERIAL OXYGEN DESATURATION DURING APNOEA: A THEORETICAL STUDY OF THE PRETERM INFANT S SANDS, B EDWARDS, P BERGER, M WILKINSON Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Monash University, Melbourne, Victoria, Australia Introduction: Arterial oxygen desaturation during apnoea is common in preterm infants and often surprisingly rapid, at rates up to and beyond 10%sec-1 (Upton et al, 1991; Poets & Southall, 1991). Gas exchange deficiency due to ventilation-perfusion (V/Q) mismatch accounts for a substantial proportion of neonatal hypoxaemia (Koch, 1968; Quine et al, 2008). Currently there is no knowledge of how V/Q mismatch affects the dynamics of deoxygenation or reoxygenation during or following apnoea. We examined the role of V/Q mismatch in  O ) and hypothesised that V/Q the rate of desaturation during apnoea (Sa 2  O. mismatch exacerbates Sa 2 Method: We used a multi-compartment lung-body mathematical model, with parameters for the preterm infant at term age, to compare  O during simuthe degree of V/Q mismatch (inhomogeneity) with Sa 2 lated apnoeas. 50 lung units were distributed log-normally using s = logSD to define the degree of V/Q mismatch (Farhi & Rahn, 1955; West, 1969). A normal s = 0.62 for the preterm infant was calculated based on an end-tidal–arterial PO2 difference of 30 mmHg (in air). Simulations of apnoea were contrasted with those of a singlecompartment (homogenous) model (s = 0), and a pathological model  O was quantified by the fall in saturation in 10 s of apnoea (s = 1.23). Sa 2  O , 10s). (ΔSa 2 Results: The major findings of this study are four-fold. 1) Compared to  O , 10s is reduced in the homogenous model normal conditions, ΔSa 2 (14.6% vs 22.4%), and elevated in the pathological model (30.0%). 2) The difference between arterial saturation and the saturation predicted from alveolar PO2 increases during apnoea, from 2.2% at baseline up to a maximum of 10.7%. 3) Paradoxically, a more rapid instantaneous  O occurs with homogeneity than with normal V/Q mismatch. 4) V/Q Sa 2 mismatch also impairs arterial reoxygenation following apnoea. Discussion: V/Q mismatch accelerates arterial oxygen desaturation during apnoea. This finding implies a potential for more rapid desaturation with apnoea in infants or older patients with acute or chronic respiratory conditions. However, V/Q mismatch alone does not explain the extremely rapid rates observed clinically.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

P-21

P-22

RECRUITED MOTOR UNITS IN THE GENIOGLOSSUS: THE KEY TO UPPER AIRWAY PATENCY AT AROUSAL?

CONTROL OF BREATHING IN SLOW WAVE COMPARED TO LIGHT SLEEP

V WILKINSON1, A MALHOTRA2, C NICHOLAS1, C WORSNOP1, A JORDAN2, J BUTLER3, J SABOISKY3, S GANDEVIA3, D WHITE2, J TRINDER1 1 Department of Psychology, University of Melbourne, Parkville, Victoria, Australia, 2Division of Sleep Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts, United States, 3Prince of Wales Medical Research Institute and University of New South Wales, Sydeny, NSW, Australia Introduction: Activity of genioglossus (GG) motor units is a result of a complex pattern of inspiratory, expiratory and tonic drives from the hypoglossal motoneuron pool. We have shown that sleep onset reduces the inspiratory phasic activity of GG motor units. In this paper we report the affect of arousal from sleep on distribution of GG motor units with different discharge patterns. Methods: Sleep and respiratory data were collected for eight subjects (6M). Single motor units were recorded in the GG muscle over q-to-a transitions (arousals) and sorted into motor unit types according to their discharge pattern. Results: 218 motor units were identified, 138 present prior to the arousal from sleep and 80 recruited at or in the three breaths following the arousal. q active units: Most units active during q were tonic (TT, constant activity, no respiratory modulation, 35%), inspiratory tonic (IT, constant activity, inspiratory modulation, 33%) or inspiratory phasic (IP, active during inspiration, 25%) with a small number of expiratory units (expiratory phasic (EP), expiratory tonic (ET)). 67% of motor units retained the same discharge pattern following the arousal, but 49% (17/35) of IP units increased their firing duration to become tonic, primarily IT. When recruited: Of the 80 recruited units, 25% were recruited immediately at the point of q-to-a transition. Recruitment also occurred during the first breath after the arousal (22.5%), the second breath (37.5%), and 15% on the third breath. 63% of arousals in this study occurred during inspiration. Of the 20 immediately recruited units, 11 occurred when the arousal was on an expiratory breath. These differences were non significant. What recruited: IP motor units were most common, and present in all subjects displaying recruited units. The distribution of units were IP (44%), IT (31%), TT (18%) and a small number of expiratory units. For how long: Units were further divided into two categories; a “burst” of activity lasting one or two breaths (55%), or constant activity remaining active past the three breath analysis period (45%). 75% of bursting units were IP/EP, whereas, 78% of constant activity units were tonic (TT/IT/ET). This difference was significant. Conclusion: The GG contains several types of respiratory related motor units which are altered in response to stimuli such as a-to-q or q-to-a transitions. The recruited burst of IP activity is a major component of the increase in activity of GG multiunit recordings at an arousal.

R RATNAVADIVEL, D STADLER, S WINDLER, J BRADLEY, D PAUL, D MCEVOY, P CATCHESIDE 1 Adelaide Institute for Sleep Health, Adelaide, Australia, 2Flinders University of South Australia, Adelaide, Australia Introduction: We have previously shown that sleep disordered breathing significantly improves during slow wave, compared to lighter sleep. The aim of this study was to explore the basis for this finding by examining respiratory and arousal responses to increased inspiratory load and upper airway occlusions during stage 2 compared to slow wave sleep. Methods: 14 patients (11 Male, Mean ⫾ Stdev: Age 57.9 ⫾ 5.2, BMI 32.8 ⫾ 5.4) who had previously demonstrated significant improvements in their posture matched Respiratory Disturbance Index (RDI) in slow wave compared to light sleep, (Overall RDI 49.7 ⫾ 16.5, Slow wave RDI 18.9 ⫾ 14.0). Following 30 secs stable breathing on therapeutic CPAP, in random order, subjects underwent rapid dialdowns to 3 different sub-therapeutic levels and brief airway occlusions. Trials were maintained until arousal or 2 mins and were repeated throughout the night. Posture was matched. Interventions were retrospectively allocated to stage 2 or slow wave sleep, by a technician blinded to intervention details. Mixed model analysis with post hoc testing was performed. Results: Baseline CPAP pressures, Peak Inspiratory flow (PIF) and epiglottic pressures (a measure of ventilatory drive) were matched between stage 2 and slow wave sleep. Following dialdowns, marked pressure dependent reductions in PIF were noted, with a nadir at the 3rd breath. Progressive increases in PIF were noted from this nadir, with partial recovery by the 5th breath, p = 0.047. Marked increases in drive were noted from breath 1–5, p < 0.001. No sleep stage effect was noted for either PIF, p = 0.57 or drive, p = 0.269. Arousals were more frequent during lower dialdown pressures, p < 0.001. With upper airway occlusions, the time to arousal was prolonged during slow wave sleep, 18.0 ⫾ 6.2 vs 23.8 ⫾ 8.4, p = 0.02. The maximum drive achieved on the breath prior to arousal was more negative, –18.2 ⫾ 6.2 vs –27.9 ⫾ 9.9 cmH20, p < 0.001. There was no significant difference in the slope of DpEpi/time. Conclusion: Passive airway function and progressive increases in drive following acute airway challenge appear to be similar between light and slow wave sleep. The probability of arousal is largely dialdown pressure dependent. Maximum drive achievable without arousal is increased during slow wave sleep. Ventilatory chemoresponsiveness is not different. The apparent stability of the upper airway during slow wave sleep may primarily reflect a higher arousal threshold, which allows the development of sustained increased ventilatory drive to translate into higher airflow, without arousals.

P-23 INTER-INDIVIDUAL VARIABILITY IN PHARYNGEAL PATENCY ACROSS THE WAKEFUL RESPIRATORY CYCLE K MADDISON1, J WALSH1, R MCLAUGHLIN2, S BECKER2, J ARMSTRONG2, D SAMPSON2, D HILLMAN1, P EASTWOOD1 1 WASDRI, Perth, WA, Australia, 2OBEL, Perth, WA, Australia Background: Despite reports of variable changes in pharyngeal calibre during breathing it remains widely believed that it is least (and the pharynx is most vulnerable to collapse) at peak inspiratory flow, when

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

A29

Healthy Function of Sleep

intraluminal pressure is at its most negative. Anatomical optical coherence tomography (aOCT) is a new catheter based quantitative imaging technique ideally suited to measure changes in pharyngeal shape and size during the breathing cycle. This study used it to examine these changes and whether they varied between individuals. Methods: 7 awake, healthy, supine subjects (age range 27–45, BMI < 28) breathed via a nasal mask fitted with a pneumotachograph to measure airflow. An aOCT catheter (3 mm OD) was inserted via the nares to the mid-oesophageal level and scans obtained at the level of oro- and velo-pharynx for 3 minutes while breathing at 7.5 breaths.minute–1. Measurements were obtained of airway cross-sectional area (CSA), antero-posterior diameter (A-P) and lateral diameter at end-expiration, mid-inspiration, end-inspiration, mid-expiration and end-expiration. Results: End expiratory oropharyngeal CSA (229.4 ⫾ 113.1 mm2) tended to be larger than velopharyngeal CSA (177.8 ⫾ 66.1 mm2) although this was not statistically significant (p = 0.19). The change in CSA across the respiratory cycle was variable between individuals: relative to end expiration CSA decreased up to 61% in some individuals and increased up to 40% in others. However, mean CSA did not vary significantly across the respiratory cycle for either pharyngeal region (ANOVA p = 0.11) although in the majority of individuals it was least at end expiration and most at mid inspiration. Mean A-P: lateral diameter ratio of the oro and velo-pharynx did not vary significantly across the respiratory cycle. Conclusion: While there was a tendency for minimum CSA to occur at end expiration in both the oro- and velo-pharynx, there was high intersubject variability. These results reflect the complex nature of the forces influencing airway calibre during wakeful breathing with transmural pressure, muscle recruitment and longitudinal tension each having potentially opposing effects on stability of the pharyngeal wall.

P-24 RELATIONSHIP BETWEEN POLYSOMNOGRAPHIC VARIABLES AND HEALTH RELATED QUALITY OF LIFE IN PATIENTS WITH NOCTURNAL HYPOVENTILATION SYNDROMES A PIPER, C HOLLIER, D FLUNT, C MENDADUE Royal Prince Alfred Hospital, Camperdown, Australia Background: Reports of daytime dysfunction such as sleepiness, fatigue and impaired quality of life are common in patients with nocturnal hypoventilation syndromes. However, there is limited information regarding the relationship between physiologic variables and quality of life (QoL) measures in this population. The aim of this study was to explore the relationships between polysomnographic (PSG) indices and daytime arterial blood gases with subjective measures of daytime function including daytime sleepiness, QoL and perceived sleep quality in patients with nocturnal hypoventilation. Methods: Fifty consecutive subjects (29M : 21F) with nocturnal hypoventilation or daytime respiratory failure referred for assessment were recruited. Diagnostic groups: lung disease (n = 13), chest wall restriction (n = 17), obesity-related hypoventilation (n = 20). All subjects underwent nocturnal PSG, awake blood gases and completed questionnaires about sleep and quality of life including Pittsburgh Sleep Quality Index (PSQI), ESS and the SF 36. PSG variables analysed were sleep efficiency (SE), apnea-hypopnea index (AHI), minimum SpO2,%TST < 90%,% REM, and total arousals as well as awake PaCO2. Non-PSG measures chosen to analyse included the total SF36 score and

A30

subcomponents PCS and MCS component scores, SF36-General Health and Vitality domains, ESS and the daytime dysfunction domain of the PSQI. Results: Mean age of the group was 57 ⫾ 14 years with a PaCO2 57 ⫾ 11 mmHg. Overall, subjective sleep quality was poor (PSQI: 10.5 ⫾ 4.8) and daytime sleepiness was moderate (ESS: 12 ⫾ 6).%TST < 90% was correlated positively with total SF36 score (r = 0.36), SF36-Vitality (r = 0.36) as well as the PCS (r = 0.44) and MCS (r = 0.32) and perceived daytime dysfunction (r = –0.32). Total arousals were positively associated with ESS (r = 0.55). ESS was also positively correlated with AHI (r = 0.55), SE (r = 0.34) and negatively correlated with minimum SpO2 (r = –0.33). Discussion: Quality of life was most strongly associated with PSG measures of hypoxic load during sleep, while ESS was associated with sleep disruption and sleep efficiency. No association between daytime CO2 and quality of life was identified in this patient group. The study highlights the importance of measuring quality of life in addition to objective sleep variables in patients with chronic hypoventilation syndromes and identifies the need for QoL measures specific for this population.

P-25 HOME NIV AT THE PRINCE CHARLES HOSPITAL – A HISTORICAL COHORT COMPARISON A FIENE, P ROBINSON, G JORGENSEN, J DOUGLAS The Prince Charles Hospital, Brisbane, QLD, Australia The Prince Charles Hospital Sleep Disorders Centre provides care for outpatients requiring domiciliary non-invasive ventilation (NIV). In this retrospective analysis of our database, we compare our outpatient cohort from 1999 to 2006. Diagnosis, co-morbidities, demographic data, duration of NIV treatment from initiation to death are analysed. Data is compared in 2-yearly cohorts. Results 1999–2000 COPD 16%; Obesity Hypoventilation (OHS) 16%; Motor Neurone Disease (MND) 8%; Kyphoscoliosis (K) 16%; Diaphragm Palsy (DP) 8%; Muscular dystrophy (MD) 0% Neurological diseases other (NO) 16%; Cystic Fibrosis (CF) 0% Respiratory diseases, other (RO) 8%. City 86% Rural 14% PEG insertion 0 2001–2002 COPD 0%; OHS 11%; MND 33%; K 22%; DP 22%; MD 33%; NO 0%; CF 0% City 75% Rural 25% PEG insertion 0 2003–2004 COPD 0%; OHS 25%; MND 25%; K 13%; DP 7%; MD 13%; NO 19%; CF 0% City 62% Rural 38% PEG insertion 5 2005–2006 COPD 4%; OHS 17%; MND 28%; K 0%; DP 7%; MD 25%; NO 45% CF 4% City 72% Rural 28% PEG insertion 9 Increase in service is expressed by number of patients commencing NIV per year: 1999: 7; 2000: 5; 2001: 6; 2002: 8, 2003: 7; 2004: 19; 2005: 24; 2006: 27; 2007: 45 Additionally, the data indicate the distribution of patients’ residence has shifted from mainly Brisbane to rural areas throughout Queensland and northern NSW. No respiratory complications from PEG insertions have occurred. Conclusions: This data demonstrates a shift in the aetiology of respiratory failure. Patients with chronic neuromuscular diseases have replaced those with primary respiratory diseases as our main indication for long term NIV. This has implications on equipment requirements and nursing time to educate carers. A multidisciplinary approach, involving sleep physicians, neurologists, gastroenterologists and palliative care physicians is required. As a result of this approach, PEGs have been safely inserted in patients with significant respiratory impairment.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

The geographic distribution and disability of patients, has required increasing use of telemedicine for follow up.

P-26 6-YEAR REVIEW OF A RESPIRATORY HDU T SATHER, C HUKINS The Princess Alexandra Hospital, Brisbane, Queensland, Australia The Princess Alexandra Hospital Respiratory High Dependency Unit was established in 2001 for the care of high-acuity respiratory patients. Over the past 6 years, it has cared for 826 patients, half of whom underwent Non-Invasive Ventilation. Most patients (75%) underwent Non-Invasive Ventilation for hypercapnoeic respiratory failure. The most common medical diagnoses in patients undergoing NIV were COPD (51%) and pneumonia (16%), followed by sleep-disordered breathing (14%). Non-invasive ventilation was successful in 90% of patients. The most common reasons for failure of NIV were clinical deterioration (46%) and patient intolerance (27%), with planned withdrawal of treatment in a further 18%. Over the 6 years, the annual number of admissions has more than doubled to 216 patients in 2007–8. At the same time, mortality within the unit has decreased from 11.2% to 2.3%, despite no marked change in the demographics or acuity score of patients admitted. Although the reasons for this are not clear, possible contributions may be patient selection & improved level of care with increased experience. These findings show that the Respiratory HDU provides a safe environment for care of patients with predominantly respiratory acute illness requiring high dependency care outside of ICU.

P-27 1-YEAR OUTCOMES IN ACUTE NON-INVASIVE VENTILATION T SATHER, C HUKINS The Princess Alexandra Hospital, Brisbane, Queensland, Australia Non-invasive ventilation (NIV) is offered to many patients for treatment of patients with Type II respiratory failure, despite limited data on longer-term outcomes after therapy(1,2). This retrospective analysis of information collected prospectively aims to identify factors present before commencement of therapy which may help to better-identify patients likely to benefit from NIV, in order to better-inform clinicians and patients in deciding on whether to institute NIV. Baseline data have been collected on patients commenced on NIV at this institution from June 2005–June 2007. Variables at time of commencement recorded were: age, admissions for respiratory illness in the preceding year, co-morbidities, body mass index (BMI), residential status prior to admission, home oxygen use, pre-morbid FEV1 and arterial blood gas parameters at admission. Outcome measures were the number of admissions for any cause & the number of days spent in hospital in the year after discharge. Date of death, if known, was also recorded. 21 patients were identified who were admitted with an exacerbation of COPD requiring NIV. Average age at commencement of NIV was 71 years. Five patients were in residential care & the average Charlson Co-morbidity index was 1.67 (4.19 when age-adjusted) prior to admission. The average BMI was 24.82, ranging from 13.5 to 34.9 kg/m2. Patients had an average of 0.7 admissions in the preceding year & their average FEV1 was 0.9 L. 7 patients were using home oxygen therapy prior to admission. The average arterial pH was 7.22 (7.06–7.39) &

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

average pCO2 was 90 mmHg (35–147) prior to commencement of NIV. The average time until NIV was completely weaned was 95 hours. 8 Patients died during the follow-up, 4 during the index admission. Patients who survived the index admission had an average of 1.1 admissions in the following 12 months, with an average of 8.8 days in hospital (this includes one patient who spent 35 days in rehabilitation). None of the variables examined were statistically significant predictors of outcome, although there was a trend towards poor survival for patients with a BMI of less than 20. These results suggest that patients who receive NIV for acute exacerbations of COPD, including those with a low arterial pH on admission, have fairly infrequent readmissions to hospital & these are not generally prolonged. None of the studied variables were particularly useful in predicting outcome, although this may have been biased by clinical exclusion of patients felt unsuitable to undergo NIV.

Chronobiology and Insomnia P-28 THE EFFECTS OF A 30-MINUTE NAPPING OPPORTUNITY DURING AN ACTUAL NIGHT SHIFT ON PERFORMANCE AND SLEEPINESS IN SHIFT WORKERS M JACKSON1, L RADFORD2, G KENNEDY2, P SWANN3, R PIERCE1, M HOWARD1 1 Institute for Breathing & Sleep, Department of Respiratory & Sleep Medicine, Austin Health, Melbourne, Australia, 2School of Psychology, Victoria University, Melbourne, Australia, 3Department Road Safety, Vic Roads, Melbourne, Australia Introduction: Shiftworkers are particularly susceptible to the effects of sleepiness, both at work and whilst driving home. There is an over representation of shift workers in motor vehicle accidents caused by fatigue [Fell, 1997; Horne, 1995]. A brief nap taken during a night shift or immediately before a night shift may improve vigilance and alleviate sleepiness both at work and on the road after work, and is commonplace in some companies. The current study investigated the effects of a 30-minute napping opportunity before or during a night shift, on simulated driving and vigilance performance. Design: Repeated-measures, crossover design evaluating psychomotor and simulated driving function in three conditions; no nap, a nap prior to or a nap during a first night shift. Participants: Eight healthy nightshift workers. Intervention & measurements: Driving simulation (variation in speed and lane position) (AusEd™), the Psychomotor Vigilance Task (PVT) (lapses and reaction time), the Karolinska Sleepiness Scale (KSS) and a sleepiness symptoms questionnaire were measured at the start end and during an actual night shift. The napping opportunities were provided on different nights at either 20:15 h (before the shift) or at 04:00 h (during the shift). Results: Only three participants slept for an average of 4.88 minutes (⫾8.28 minutes) in the evening, and all participants slept for an average of 23.50 minutes (⫾5.48 minutes) in the morning nap condition. A significant decrement on all measures was observed across the night. There was no effect of nap condition on performance or subjective sleepiness, apart from a reduction in variation in lane position during the final testing period (06:45 h) in the morning nap condition. Conclusion: A napping opportunity, prior to or during a real night shift did not result in consistent performance benefits during or at the end of

A31

Healthy Function of Sleep

the shift. Factors such as the length of nap, time from nap to testing and circadian effects may override the positive benefits reported previously.

rural localities where OC relief is not readily available. Future research should further explore the impact of OC on subsequent work performance, particularly on nights when Doctors are not re-called.

P-29

P-30

DOCTORS’ SLEEP DISTURBANCE WHEN ON-CALL

SHORT WAVELENGTH LIGHT EXPOSURE IN THE AGED: ACUTE AND PHASE SHIFTING RESPONSES

S JAY, M THOMAS, A WEISSENFELD, D DAWSON, S FERGUSON Centre for Sleep Research, University of South Australia, Adelaide, South Australia, Australia Introduction: On-call (OC) work is part of standard rostering practice in most hospitals. Sleep disturbance associated with being re-called to work is clear, often resulting in truncated sleep opportunity and therefore reduced sleep. Quantifying the disturbance when doctors receive only 1–2 short phone calls (but are not recalled) however is less obvious. The aim of these analyses was to investigate sleep during both OC and not-OC nights. Methods: Thirty-eight doctors from 12 facilities participated as part of a larger investigation. Doctors who had at least one valid OC (n = 145) and one valid not-OC (n = 225) sleep were included in analyses. Data was collected for 2–4 weeks using activity monitors and sleep/duty diaries. OC sleep periods were categorised according to level of sleep disturbance; OC-recalled (n = 21), re-called to work once in bed (asleep or attempting sleep); OC-phone-calls (n = 41), answered ⱖ1 calls during once in bed but not re-called; OC-undisturbed (n = 83), not disturbed by OC-related duties after bedtime. Time in bed (TIB) and total sleep time (TST) variables were first compared between the two sleep groups (OC and not-OC) and then between the four groups using mixed effects analysis of variance. Results: Analyses showed that TIB and TST were reduced by 41 min and 39 min respectively when OC. When the OC sleeps were split into categories, analyses revealed that not-OC sleeps were associated with greater TST compared to all OC categories, shown in Fig. 1. TST and TIB on undisturbed OC nights did not differ as compared to OC nights when doctors received ⱖ1 phone-calls.

Total Sleep Time (hrs)

7 6 5 4

Healthy ageing is associated with increased frequency of sleep disturbances and daytime sleepiness. Age-related impairment in the sensitivity to short wavelength light has been demonstrated for light-induced melatonin suppression. The current study examined age-related changes in the response of subjective alertness and circadian phase advancing to short and medium wavelength light exposure. On two separate occasions, young (n = 11, 23.0 ⫾ 2.9 years) and older (n = 15, 65.8 ⫾ 5.0 years) healthy males were exposed to 2 h of intermittent monochromatic light. Short (lmax 456 nm) and medium (lmax 548 nm) wavelength light was matched for photon density (~6 ¥ 1013 photons/ cm2/sec) and timed to begin 8.5 h after dim light melatonin onset (DLMO) determined on a previous visit. Five older subjects participated in only the short wavelength light condition. Subjective alertness was reported on a 9-point scale during and for 5 h following light exposure. The magnitude of phase advance was assessed as the change in plasma melatonin onset (DLMO) and offset (DLMOff) pre- and post-light. The alertness response to short wavelength light was significantly diminished in the older men compared to the younger group (F1,24 = 23.6, p < 0.0001). The alertness response to medium wavelength light was not significantly different between the age groups. For the phase shifting response, larger phase advances were observed in the DLMOff phase marker compared to the DLMO for both light conditions and both age groups. Similar wavelength sensitivity between the age groups was revealed for the phase advancing response (456 nm > 548 nm) but with smaller advances in the older group (non significant). The results demonstrate an age-related impairment in the response to the acute effects of short wavelength light (melatonin suppression, alertness) but not in the phase shifting response. Supported by EU Marie Curie RTN grant (MCRTN-CT-2004-512362) and 6th Framework Project EUCLOCK (018471).

3 2

P-31

1

EFFECTS OF PARTIAL SLEEP RESTRICTION ON SLEEPINESS, NEUROBEHAVIOURAL PERFORMANCE AND COGNITION

0

Not OC

OC recalled

OC phone call(s)

OC undisturbed

Figure 1 TST (mean ⫾ SEM) by On-Call Status Conclusions: Not surprisingly, Doctors obtained less sleep when OC even when sleep was undisturbed. Notably, if recalled to work, Doctors obtained <4 h sleep. In terms of fatigue-risk, this is of significant concern, particularly in facilities where Doctors are frequently OC or in

A32

T SLETTEN2, V REVELL1, B MIDDLETON1, K LEDERLE1, D SKENE1 1 Centre for Chronobiology, University of Surrey, Guildford, Surrey, United Kingdom, 2Sleep and Chronobiology Group, School of Psychology, Psychiatry and Psychological Medicine, Monash University, Melbourne, Victoria, Australia

A SEGAL1, T SLETTEN1, J REDMAN1, S LOCKLEY2, S RAJARATNAM1 1 School of Psychology, Psychiatry and Psychological Medicine, Monash University, Melbourne, Victoria, Australia, 2Division of Sleep Medicine, Brigham and Women’s Hospital, Massachusetts, United States Acute sleep deprivation in humans is associated with increased sleepiness and neurobehavioural performance during wake time. The effects of more moderate, partial sleep restriction that mimics real life

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

conditions are less understood, in particular with respect to complex cognitive tasks. The objective of the current study was to assess the effects of sleep restriction on alertness, performance, mood and cognition in young, healthy adults under controlled laboratory conditions. The study aimed to determine which aspects of neurobehavioural functioning and cognition are most impaired by sleep restriction, and to develop a paradigm to test the efficacy of interventions to reduce excessive daytime sleepiness. Following a screening visit, young, healthy, normal sleepers were asked to maintain a regular sleep-wake schedule of 8:16 h in their home environment for at least 3 weeks prior to the laboratory visit. Participants were randomised to sleep restriction or control conditions. Participants in the sleep restriction condition were sleep restricted for two consecutive nights – 5 h on the last night of the baseline period and 3 h on the night of their laboratory stay. Control subjects continued to maintain an 8 h sleep schedule. Participants attended the sleep laboratory for a period of ~26 h, during which they were placed in a 9 h modified constant posture protocol (starting 2 h after wake time) and underwent at least hourly testing of psychomotor vigilance and subjective sleepiness (Karolinska Sleepiness Scale), in addition to tests of attention (Attentional Blink), mood (Profile of Mood States), learning (Finger Tapping Task) and memory (Paced Auditory Serial Addition Task). To date, 20 participants (12 female, 8 male, mean age 22.9 years) have completed the protocol. Compared to controls, sleepiness scores were significantly higher in the sleep restricted participants (p < 0.005), although the time course differed between the two conditions. Furthermore, significant decrements in performance were demonstrated. Sleep restricted participants exhibited significantly higher mean reaction times (p < 0.005) and increased number of lapses (p < 0.01) on the Psychomotor Vigilance Task. These findings demonstrate that sleep restriction in young healthy adults is associated with significantly higher subjective sleepiness and impaired neurobehavioural performance. The temporal dynamics of sleepiness and performance during the testing period and effects of sleep restriction on cognitive outcomes are being evaluated. Supported by NHMRC (grant # 436758), Compumedics Ltd, Philips Lighting and Faculty Postgraduate Research Scholarship.

P-32 THE EFFECTS OF A BRIEF NAP DURING NIGHT SHIFT N LOVATO1, L LACK1, H WRIGHT1, S FERGUSON2 1 Flinders University, Adelaide, SA, Australia, 2University of South Australia, Adelaide, SA, Australia Introduction: Napping during night shift work has been suggested as a potential countermeasure against the dangerous sleepiness workers face. Although research has consistently demonstrated the alerting benefits of daytime napping, napping during a night shift, especially after a period of prolonged wakefulness (18–24 hours), has not been consistently shown to improve alertness. This study investigated the benefits of a 30 minute nap in a simulated night shift environment, when a prophylactic daytime sleep was allowed prior to the night shift. Methods: The benefits of a 30 minute night nap, relative to no nap, were investigated using a repeated measures, counterbalanced design with 22 subjects. In both conditions subjects obtained a 2-hour sleep in the afternoon from 1500 h–1700 h, which was followed by the night time nap from 0230 h–0300 h in a controlled laboratory environment. Subjective alertness (SSS, KSS, VAS), fatigue and vigour (POMS), cognitive performance (PVT, SDST, LCT) and objective sleepiness (SOL) were measured prior to both the nap and no nap conditions and then several times following both the conditions across the period from 0315 to 0700 h.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Results: The 30 minute nap resulted in some impairment of subjective alertness for a brief period (up to 30 minutes) immediately following the nap, when compared to the no nap condition, probably indicative of sleep inertia. However, after this period, alertness and performance were improved for the rest of the night (0400–0700) following 30 minute nap relative to no nap. Conclusions: When a 2-hour prophylactic sleep is allowed in the afternoon, a 30 minute nap during a night shift is a significant countermeasure against sleepiness.

P-33 EFFECTS OF PARTIAL SLEEP RESTRICTION ON AUTOMATIC COGNITIVE PROCESSING IN HEALTHY YOUNG AND OLDER ADULTS C BAMBRICK, C SWANN, J REDMAN, G YELLAND, S RAJARATNAM Monash University, Clayton, VIC, Australia The effects of sleep restriction on neurobehavioural performance are well established. However, relatively little is understood about the effects of sleep loss on other cognitive functions, such as automatic processing. Furthermore, compared to young adults, there is less research on the effects of restricted sleep on older people, despite evidence of age-related changes to sleep duration, timing and structure. The current study explored the effects of partial sleep restriction on performance on a word recognition task using the masked priming paradigm. The priming effect derived from this task can be used to assess the amount of automatic (or unconscious) processing involved in word recognition. In a controlled laboratory setting, 12 young adults (aged 18–35 years) and 12 older adults (aged 59–72 years) were administered the Karolinska Sleepiness Scale, Psychomotor Vigilance Task and masked priming task. In each trial of the masked priming task, a prime was presented very rapidly (so that the participant was not consciously aware of it) immediately preceding a target word. Both age groups were tested under two conditions in a counterbalanced, repeated-measures design – sleep-restricted and normal sleep. In the sleep restriction condition, the test battery was administered following two consecutive nights of sleep restricted to 60% of each participant’s habitual sleep duration. In the normal sleep condition, participants were required to maintain their habitual sleep-wake pattern. Compliance with the sleepwake schedules was verified by sleep diary, wrist actigraphy and daily call-ins at bedtime and wake time to a time-stamped answering machine. As predicted, sleep restriction resulted in increased subjective sleepiness and psychomotor performance deficits in both age groups. Response times on the masked priming task were not affected by sleep restriction in either group. However, the magnitude of the priming effect significantly increased for young adults (p < 0.05). In contrast, there was no significant increase in the priming effect in older adults. In both age groups, although psychomotor performance and sleepiness were impaired by sleep loss, there was no evidence of impairment to word recognition. It was concluded that the automatic processes supporting word recognition may increase following sleep loss in young adults to maintain baseline performance levels, but that such a compensatory mechanism does not account for the lack of deficits to word recognition in older adults.

A33

Healthy Function of Sleep

P-34 HOW RESTORATIVE IS THE SLEEP OBTAINED BY AIRLINE PILOTS IN ON-BOARD REST FACILITIES? G ROACH, D DARWENT Centre for Sleep Research, University of South Australia, Adelaide, South Australia, Australia Introduction: In some industries, shiftworkers have the opportunity to supplement their normal home sleep with sleep at work. In certain circumstances, particularly in the long-haul transport industries, sleep at work may actually occur in on-board rest facilities. There is some evidence, from research data and anecdotal reports, to suggest that the restorative value of sleep is influenced by the setting in which it is obtained. The purpose of this study was to compare the restorative value of sleep obtained in on-board rest facilities with sleep obtained at home in bed. Methods: Data were extracted from a data set that contained sleep/wake information for over 400 long-haul airline pilots. Participants kept a self-report sleep diary and wore a wrist activity monitor for at least 2 weeks. Participants rated their level of fatigue at the start and end of sleep periods using the 7-point Samn-Perelli Fatigue Checklist (1 = fully alert, 7 = completely exhausted). The restorative value of sleep was calculated as the difference between self-rated fatigue at the start and end of a sleep period. Higher positive scores indicate greater restorative value. Two groups of duration-matched sleeps were selected from the database: on-board sleeps and home-bed sleeps. Results: There were 2,217 on-board sleeps, with an average (⫾st.dev.) time in bed of 2.36 h (⫾1.18 h), total sleep time of 1.72 h (⫾1.11 h), and recovery value of 0.90 (⫾1.13). There were 3,079 home-bed sleeps with an average time in bed of 2.36 h (⫾1.24 h), total sleep time of 1.83 h (⫾1.15 h), and recovery value of 1.29 (⫾1.11). Conclusions: The current analyses indicate that sleep obtained in on-board rest facilities has approximately 70% the restorative value of duration-matched bed sleeps. This difference may be due to psychological factors (e.g. difficultly relaxing the mind while at work) and/or environmental factors (e.g. comfort of the rest facility). This influence of the sleep setting on the restorative value of sleep should be taken into account when planning on-board rest opportunities for airline pilots and other transport workers. Acknowledgements: This research was supported by the Australian Research Council and a consortium that included the University of South Australia, an operator, a regulator and an employee association.

P-35 SLEEP REGULATION IN A DYNAMIC SHIFTWORK ENVIRONMENT: THE INTERACTIONS OF WITHIN- AND BETWEEN-SHIFT SLEEPS M THOMAS, A WEISSENFELD, S JAY, S FERGUSON University of South Australia, Adelaide, SA, Australia Introduction: In safety-critical industries such as healthcare, obtaining adequate restorative sleep whilst working varying shift patterns is essential in minimising fatigue-related risk. Sleep homeostasis is a critical regulatory mechanism for the maintenance of neurobehavioural performance. Whilst homeostatic regulation is well documented in laboratory studies, relatively little is know about sleep regulation in dynamic shiftwork environments such as aviation and healthcare where sleep can be obtained both during and between shifts.

A34

Method: Actigraphic sleep-wake data were collected from a total of six Obstetrics and Gynaecology Registrars working a set pattern of seven consecutive night shifts followed by a sequence of day shifts and days off. Metrics of timing and duration of sleep were calculated for each shift type. Results: Mean Total Sleep Time following day shifts was 6.7 ⫾ 1.4 h (range: 4.3 h–8.2 h), with no evidence of sleep being obtained during shifts. Mean Total Sleep Time following night shifts was 3.5 ⫾ 1.3 h (range: 1.2 h–6.3 h), with Mean Total Sleep Time during night shifts being 3.5 ⫾ 1.6 h (range: 0.0*h–8.5 h). Discussion: The average amount of sleep obtained following night shifts (3.5 h) was more than 3.0 h less than the average amount of sleep obtained following day shifts (6.7 h). Notably, however when the amount of sleep obtained during night shifts is considered, Doctors were obtaining approximately 7.0 h sleep in connection with night shifts. While this 7.0 h is more than the average amount obtained in connection with day shifts (6.7 h), it is obtained in two sleep bouts, which may not be as restorative as a single sleep bout. Further analysis will shed light on the interaction between physiological requirements for sleep and personal adaptive strategy.

P-36 COMPARISON OF INSOMNIA TREATMENT PROVIDERS IN NEW ZEALAND K O’KEEFFE1, P GANDER1, G SCOTT2, H SCOTT3 1 Sleep/Wake Research Centre, Massey University Wellington, Wellington, New Zealand, 2Department of Economics & Finance, Massey University Wellington, Wellington, New Zealand, 3ScottEconomics Ltd, Wellington, New Zealand Introduction: An interview based study was conducted to investigate insomnia treatment in New Zealand. Within the scope of this study, different types of insomnia treatment providers were compared using pre-defined criteria. Methods: After directly approaching insomnia treatment providers known to us, we used a snowball sampling technique to recruit other treatment providers. Eighteen treatment providers consented to participation in a full structured interview. An additional three treatment providers consented to a shortened interview, due to their workload constraints. An interview outline was used to collect information on the profile of patients seen, diagnostic and treatment practices, length of treatment regimes, patient outcomes and the cost of treatment. In order to obtain a reasonable representation of the different insomnia treatments offered in New Zealand, insomnia treatment providers were classified as Specialist Physicians, Psychologists, Pharmacists, General Practitioners (GPs), Health Practitioners and Alternative Health Practitioners. An attempt was made to recruit equal numbers from each group. As part of the interview process, insomnia treatment providers were compared against a number of pre-defined criteria for insomnia and sleep knowledge, and types and validity of treatment. Results: Specialist Physicians, Psychologists and Health Practitioners demonstrated the highest knowledge for types of insomnia and understanding of sleep terminology. Alternative Health Practitioners and Pharmacists demonstrated poor knowledge. Along with GPs, the latter group tended not to use standard sleep terminology, relied solely on interview for diagnosis, rarely used validated treatment approaches and did not assess treatment effectiveness. Health Practitioners were most likely to practise outside their scope of practice and/or competency guidelines and often failed to implement insomnia treatment according to best practice guidelines. Conclusions: There is great disparity between insomnia treatment providers in New Zealand. A significant number of insomnia treatment

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

providers have relatively poor knowledge of insomnia and/or sleep, and implement inadequate insomnia treatment. Given the prevalence of insomnia in New Zealand (Paine et al (2005) Aust N Z J Public Health 29(1); 22–28), this study highlights the critical need for a more systematic approach to insomnia treatment in the mainstream healthcare system.

Paediatric – Sleep Measurement P-37 CHILDREN WITH SLEEP DISORDERED BREATHING HAVE ALTERED MOVEMENT DISTRIBUTION DURING SLEEP S COUSSENS2, M KOHLER2, M BAUMERT2, Y PAMULA1, K LUSHINGTON3, D SAINT2, J MARTIN1, D KENNEDY2 1 CYWHS, North Adelaide SA, Australia, 2Adelaide University, Adelaide SA, Australia, 3UniSA, Adelaide SA, Australia Introduction: Children with sleep disordered breathing (SDB) have been shown to have increased sleep fragmentation and cognitive deficits. However, associations between traditional measures of sleep fragmentation and neurocognitive performance are typically poor. The measurement of movement in sleep has long been poorly defined for clinical applications and often treated simply as artefact. We propose a simple, clearly defined and novel measure for measuring body movement in children with SDB as a measure of sleep fragmentation. In addition we compare the association of this measure and cognitive performance with traditional measures of sleep fragmentation. Methods: Subjects were children (aged 3.0 to 12.9 years) divided into three groups: Primary Snorers (PS: n = 24), Obstructive Sleep Apnoeics (OSA: n = 20) and controls (C: n = 48). We measured movement using standard scoring methods as well as using our novel method. We then compared total numbers of movement events and event distribution during sleep between groups and correlated this with cognitive measures. The distribution of movement events was calculated using survival curve analysis. Results: We found that the total number of body movement per hour of sleep (as measured using the proposed definition and rules) was significantly different between groups (C = 7.1 ⫾ 3.3, PS = 8.2 ⫾ 4.4, OSA = 10.5 ⫾ 5.0. p < 0.05) and was the only movement measure that correlated significantly to certain cognitive performance measures (p < 0.05). We also found a difference in the distribution of movements throughout the night between each of the groups (p < 0.05). Discussion: The results of our study demonstrate that movement, when appropriately analysed, is an important indicator of sleep fragmentation in children with SDB and should be included in normal clinical analysis of PSG results due to its association with daytime function.

P-38 ASSESSMENT OF AN AUTOMATED SLEEP STAGING TOOL BASED ON RECURRENCE PLOT ANALYSIS OF INFANT BREATHING PATTERNS P TERRILL1, S WILSON1, C DAKIN2, S SURESH2, D COOPER2 1 University of Queensland, Brisbane, Queensland, Australia, 2Mater Children’s Hospital, Brisbane, Queensland, Australia It is well known that breathing patterns vary with infant sleep states. However, development of an automated sleep scoring tool based on

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

conventional statistical analysis of breathing patterns has proven problematic. Recent work has demonstrated that non-linear properties of breathing patterns vary with sleep states and that analysis with nonlinear tools such as recurrence plot analysis allows the accurate discrimination of sleep states. In this work we propose using a combination of statistical and non-linear analysis measures to automate the identification of infant sleep states using only breathing dynamics. Full overnight polysomnograms from the CHIME dataset were obtained for 32 healthy infants, aged between 19 and 153 days. These were sleep-staged using 30 second epochs by a human expert using the standard criteria into awake, active and quiet sleep states. For each infant, the respiratory inductive plethysmography channel was extracted, and analysed using a sliding window methodology. For each window, breathing frequency, and the non-linear recurrence plot analysis measure, radius, was calculated at a fixed recurrence of 2%. A “leave one out” training and testing method was used, by which discrimination rules based on these 2 measures were developed using data from 31 infants, and then tested on the remaining 1 infant. This is repeated 32 times such that each infant has been evaluated using rules generated by every other infant. Awake state was able to be discriminated from sleep with an agreement of 86.5% (kappa = 0.52), and quiet sleep was able to be discriminated from a combined active sleep/awake with an agreement of 87.5% (kappa = 0.72). When both discriminators were combined, the stages active sleep, quiet sleep and awake were able to be identified with an average agreement of 76.6% (kappa = 0.60). This tool has the advantage that it uses only a single channel of respiratory data which makes it suitable for use in low cost, minimal channel monitoring environments remote from dedicated sleep laboratories. Possible applications may include use as a diagnostic screening process for infants with suspected sleep disordered breathing; use in epidemiological sleep research to improve on accuracy attained using sleep diaries and actigraphy; and as an adjunct to existing full polysomnograms to reduce clinician time.

P-39 RESPIRATORY VARIABILITY IN TERM AND PRETERM INFANTS: CAN WE DETERMINE WHEN A LONG RESPIRATORY PAUSE BECOMES AN APNOEA? D ELDER, A CAMPBELL, P LARSEN, D GALLETLY University of Otago, Wellington, New Zealand Introduction: There is no agreed definition of apnoea for preterm and term infants. This study aimed to: 1. Document breath-to-breath ventilatory rate variability in term and preterm infants and assess the affect of sleep state, sleep position and age, 2. Determine whether longer respiratory pauses fall within, or outside the tail of the distribution of normal variability. Methods: Nap polysomnography was undertaken on preterm and fullterm infants. Preterm (PT) infants were studied fortnightly where possible and term (T) infants were studied at birth and 3 months. Periods of respiration taken from the nasal pressure trace were downloaded from four epochs: Quiet sleep (QS) and Active Sleep (AS) in both prone and supine. A purpose built programme in Labview was used to determine the inspiratory onset and calculate breath-to-breath intervals (I-I) and other variability measures including standard deviation (SD) and coefficient of variation (CV) measures of ventilatory frequency (VF) and I-Is. For T infants repeated measures analysis in a general linear model was used to look for differences related to sleep state, sleep position and age. For PT infants data were analysed in SAS using the PROC MIXED

A35

Healthy Function of Sleep

procedure for mixed models and the same states compared. For both groups 95th centile and 95% frequency levels of I-Is were calculated for each sleep state as well as two respiratory interval equivalents. Results: All measures of respiratory variability varied significantly with sleep state, with variability more common in AS. Sleep position was not a significant determinant of respiratory variability. For T infants SDVF, CVVF and CVI-I were significantly less at 3 months compared with birth. When histograms of I-Is for both T and PT infants respectively were combined in each sleep state a continuous distribution of the range of I-Is was seen. The 95th centile and 95% cumulative frequency values for two ventilatory intervals in each sleep state, for all groups combined, averaged 5.12 seconds in AS and 4.22 seconds in QS. Discussion: Longer I-I intervals are not clearly differentiated from the range of normal in well term infants and convalescent preterm infants indicating that the term ‘apnoea’ is not useful clinically in this population. Methods of scoring central and hypopnoeic respiratory events in infants under 1 year of age should require a consequence of desaturation or arousal before being counted and not be based on the length of respiratory pause alone. Such events should be thought of as clinically significant respiratory pauses.

P-40 IMPROVING HOME OXIMETRY AS A DIAGNOSTIC TOOL IN CHILDREN WITH SUSPECTED SLEEP DISORDERED BREATHING S YORKSTON, D O’DRISCOLL, R HORNE, G NIXON Ritchie Centre for Baby Health Research, Monash Institute of Medical Research, Melbourne, Victoria, Australia Introduction: Oximetry has been proposed as a screening tool for sleep disordered breathing. However only 20% of snoring children have an abnormal oximetry, with false negative results reported in those with obstructive sleep apnoea (OSA) without desaturation. We hypothesised that pulse rate indices (PRI) and pulse rate standard deviation (PR-SD) would reflect sleep fragmentation and, together with removing periods of wakefulness and poor signal quality, would improve oximetry as a diagnostic tool. Methods: Oximetry (Masimo Radical, 2 s averaging time) recorded during PSG in 5 children (mean age 8.8 y) was analysed using Download 2001TM (v. 2.8.0, Stowood Scientific) software. SpO2 indices included mean SpO2, SpO2 nadir and dips of >3%/h (DI3%). The PRI (PR increases/h) thresholds were chosen based on our previously reported data: heart rate changes from wake to sleep of 10 bpm (PRI10), wake to REM of 8 bpm (PRI-8), and sleep to arousal of 15 bpm (PRI-15). The entire recording was analysed with and without automatic removal of poor signal (full/auto and full). The recording was trimmed by PSG-defined sleep onset and offset, and analysed in the same way (trim/auto and trim). Analysis was also performed after manually excluding poor signal. Rules were developed for estimating sleep period from PR patterns and 4 independent testers trimmed the recordings using these rules. Analysis methods were compared using 1-way ANOVA. Pearson correlations were performed between obstructive apnoea hypopnoea index (OAHI) and total arousal index (ArI) and the PRI and SpO2 results. Results: Trimming the oximetry record led to significantly lower PR-SD (p < 0.05). DI3%, PRI-15 and PR-SD were higher (p < 0.05), and SpO2 nadir was lower (p < 0.05) if poor signal was not removed either automatically or manually. No significant differences were found between the analysis methods for mean SpO2 (98.3 ⫾ 0.1%), PRI-8 (47.3/ h ⫾ 3.9) or PRI-10 (34.9/h ⫾ 4.0). None of the PRI correlated significantly with OAHI or ArI (best values for trim/auto analysis for PRI-15 &

A36

ArI: r = 0.62, p = 0.26). PR-SD was significantly correlated with OAHI (trim/auto: r = 0.97, p < 0.05) and ArI (r = 0.95, p < 0.05). Sleep period after trimming by PR rules was not significantly different from PSG sleep period (mean 3.3% greater than PSG sleep period or 15 min, p = 0.96). Conclusion: More accurate SpO2 and PRI results are achieved by trimming the record and removing periods of poor signal quality. Trimming the record using rules based on PR accurately reflects sleep period. PRI-15 and PR-SD hold the most promise for identifying OSAS and will thus be the focus of further studies.

P-41 THE SLEEPY PRESCHOOLER – BEHAVIOURAL OUTCOMES B MICHELL, S BLUNDEN University of South Australia, Adelaide, SA, Australia There is a large body of research that illustrates the impact of sleep on cognitive and affective functioning and negative behaviours including conduct disorder, aggressive behaviour and ADHD-like symptoms with studies suggesting that if sleep problems are addressed in this age cohort, negative behavioural outcomes problems improve. The trait of empathy, which utilises both affective and cognitive functions, has been linked to negative behaviours such as conduct disorder and aggressivity. This relationship may be mediated by sleep. This study therefore aims to investigate the link between sleep and the affective and cognitive components associated with empathy and sociability in preschool Australian children. The study will be undertaken in 100 children aged 3–4 years old from preschool centres in suburbs of Adelaide SA from July to October 2008. Parents will complete questionnaires to investigate empathy, behaviour and sociability. Sleep will be investigated with subjective (sleep diaries) and objective measures (a 3 day activity monitor). The children will also be interviewed briefly to investigate empathic responding. Relationships between these variables will be explored utilising correlational methods. It is hypothesised that children with reduced sleep duration will display less empathy and sociability as measured by empathy and sociability measures. Implications for childhood development will be discussed.

P-42 CHANGES TO SLEEP ARCHITECTURE AND APNOEA HYPOPNOEA INDEX (AHI) IN CHILDREN EXPOSED TO PASSIVE SMOKING A SALVINI, T GULLIVER John Hunter Childrens Hospital, Paediatric Sleep Unit, Newcastle, NSW, Australia Zang et al in 2006 and 2007 demonstrated a delay in sleep latency and disturbed architecture in adults who smoke. We have collected preliminary PSG data on 114 children aged 8/12 months to 17 years, 40 with at least 1 carer a reported smoker. Results to date demonstrate shorter sleep latencies in children who are exposed to passive smoking. No changes in REM or NREM were observed. Although not statistically significant, trends also suggest that children exposed to passive smoking have higher Apnoea Hypopnoeas Indices (AHI) than those children who are not exposed to passive smoking. Parents/cares were not asked about the degree of their smoking habit as this has proven to be unreliable. Puig et al 2008 reported a decrease

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

in self reporting parental tobacco use, due to greater awareness of the negative effects of passive smoking on children. Further study is required using quantitative methods for analysing smoke exposure in children. Exposed to passive smoking

Sleep Latency (mins) REM%TST NREM%TST AHI (overall) AHI (in REM) OAHI with AHI > 10 Arousal Index

Yes

No

p-values

31.4 18.6 81.1 5.5 9.1 47.6 26

42.5 19.4 80.6 2.4 6.5 16.3 18

0.03 0.28 0.35 0.15 0.22 0.16 0.12

PSG = polysomnography TST = Total Sleep Time OAHI = Obstructive Apnoea Hypopnoea Index

P-43 THE “FIRST NIGHT EFFECT” OF POLYSOMNOGRAPHY ON USUAL SLEEP QUALITY IN HEALTHY INFANTS AT 12 MONTHS OF AGE J LONGLEY, H HEUSSLER, C PARSLEY, C DAKIN Mater Children’s Hospital, Brisbane, Qld, Australia Introduction: There is minimal information comparing usual sleep quality in the home environment with the effect of the sleep laboratory environment and monitoring on usual sleep onset, waking and fragmentation. Most studies on ‘first night effect’ have compared measures on repeated polysomnography (PSG) on consecutive nights. The effect of an unusual sleep environment would be expected to be high in 12 month olds. An understanding of this effect is important for interpretation of PSG. Methodology: Infants were recruited antenatally as part of a larger prospective normal cohort study looking at maturation of sleep and breathing. Infants wore an Actiwatch (AW4 Cambridge Neurotechnology) for a week prior to the PSG night. A minimum of 3 consecutive nights information was required for interpretation. This was compared with full overnight PSG in the sleep laboratory, with concurrent actigraphy. Results: 26 infants were studied at 12 months of age. Of these, full actigraphy information was available on 15 infants. Combined results show a variable daytime nap pattern at 12 months with relatively consolidated sleep at night. There was no significant difference in the time of sleep onset and waking between the PSG night and at home. However, total sleep time was reduced by a mean of 50 minutes on the PSG night and total time asleep 4% less (p = 0.038). This was due to greater sleep fragmentation (p = 0.075), with associated higher mean average activity score (p = 0.015). Conclusions: Surprisingly, the sleep laboratory environment did not appear to affect sleep onset and waking times. The expected later sleep onset on the PSG night was not seen. However, a significant ‘first night effect’ due to sleep fragmentation was demonstrated in this group of 12 month old infants.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

P-44 THE UTILITY OF THE OSA-18 AS A SCREENING QUESTIONNAIRE FOR CHILDREN BEING INVESTIGATED FOR OBSTRUCTIVE SLEEP APNOEA (OSA) M WAWRUSZAK1, N VERGINIS1, M DAVEY1, G NIXON2 1 Melbourne Childrens Sleep Unit, Victoria, Australia, 2Ritchie Centre for Baby Health Research, Monash University, Victoria, Australia Introduction: Increasing demand for polysomnography (PSG) has resulted in increased waiting times. Questionnaires may be useful in prioritising the waiting list. The OSA-18 is an 18 item disease specific questionnaire that has only been validated against daytime nap PSG. It uses a 7-point Likert scale (min 18, max 126) and includes 5 sub-scales (including a “Sleep Disturbance” subscale regarding symptoms of sleep disordered breathing: OSA-18-SD). Aim: To assess the utility of the OSA-18 to prioritise children on a PSG waiting list, using full night PSG. Methods: Parents of children having routine PSG for diagnosis of obstructive sleep apnoea (OSA) from Mar-May 2008 were asked to complete the OSA-18. Linear regression was used to examine the difference in mean OSA-18 score between groups of different OSA severity: primary snoring (PS) (obstructive apnoea-hypopnoea index (OAHI) <1/h), mild (OAHI 1–5/h), moderate (OAHI 5–10/h) or severe OSA (OAHI > 10/h). Multiple linear regression was performed to assess whether age or PSG parameters (OAHI, SpO2 dips <90%/h, SpO2 dips ⱖ4%/h, total arousal index, total sleep time, sleep efficiency, %time in REM & NREM1 sleep) predicted the OSA-18 score. Data is presented as mean ⫾ SD. Results: 224 children had PSG and 126 gave consent. Complete OSA-18 data as available for 101 patients (62 M; age 6.0 ⫾ 3.3 y). OSA-18 scores were significantly different between the OSA severity groups (p = 0.04), with those with severe OSA having a total OSA-18 score on average 16 points higher than those with PS (95%CI 4.2–28.0, p = 0.009). There was also a significant difference in OSA-18-SD scores between patients with varying severities of OSA (p = 0.04), with those with severe OSA having an OSA-18-SD score on average 4 points higher than those with PS (95%CI 1.0–7.9, p = 0.01). However there was too much overlap in OSA-18- and OSA-18-SD scores to be able to use these to differentiate between the OSA severity groups.

OSA-18 total score OSA-18-SD score

Primary snoring

Mild OSA

Moderate OSA

Severe OSA

61.2 ⫾ 17.5 15.0 ⫾ 5.6

60.2 ⫾ 20.3 14.8 ⫾ 5.7

66.4 ⫾ 22.6 17.3 ⫾ 6.1

77.3 ⫾ 11.8 19.5 ⫾ 4.7

Conclusion: While children with severe OSA have a higher OSA-18 score and OSA-18-SD score, the overlap of scores between groups of different severity of OSA precludes its use as a triage tool for prioritising the PSG waiting list.

A37

Healthy Function of Sleep

P-45 A COMPARISON OF NOVEL METHODS FOR MEASURING SLEEP FRAGMENTATION IN CHILDREN A FOSTER1, D O’DRISCOLL1, J YANG1, G NIXON1, M DAVEY2, A WALKER1, V ANDERSON3, J TRINDER4, R HORNE1 1 Ritchie Centre For Baby Health Research, Monash Institute of Medical Research, Victoria, Australia, 2Department of Respiratory and Sleep Medicine, Southern Health, Victoria, Australia, 3Department of Psychology, Royal Children’s Hospital, University of Melbourne, Victoria, Australia, 4 Department of Psychology, University of Melbourne, Victoria, Australia Background: Sleep disordered breathing (SDB) in adults causes frequent arousal from sleep, leading to excessive daytime somnolence and poor cognitive performance. Similarly, SDB in children is associated with neurocognitive deficits despite only 50% of respiratory events terminating with an EEG arousal. Sleep pressure score (SPS) and sleep continuity (SC) are new methods designed to measure sleep fragmentation in children (Tauman, 2004). The aim of this study was to compare the SPS, SC and traditional arousal scoring as measures of sleep fragmentation in children. Methods: 18 children with SDB (7 M/11 F, age 8.8 ⫾ 0.3 y, mean ⫾ SEM) and 17 healthy control (CTRL) children (7 M/10 F, age 9.5 ⫾ 0.5 y) were studied. Routine polysomnography (PSG) was performed. All children in the SDB group had an obstructive apnoea/ hypopnoea index (OAHI) of >5 event/h. PSG confirmed the absence of SDB in the CTRL group. The SPS was calculated using the following arousal indices; total (TAI), respiratory (RAItot, including EEG and non-EEG arousals) and spontaneous (SAI). SC was measured using survival analysis of sleep length (run times) between awakenings. Between groups analyses were conducted using paired Student’s t-tests. The SPS, SC and traditional arousal indices were compared to the OAHI using correlation analysis. Results: TST and SE were reduced in the SDB group compared to the CTRL group (TST: 372 ⫾ 14 min vs. 424 ⫾ 11 min, p < 0.01; SE: 80 ⫾ 2.4% vs. 87 ⫾ 1.8%, p < 0.05). The SDB group had an increased TAI (SDB 25 ⫾ 3 event/h; CTRL 10 ⫾ 1 event/h, p < 0.001) and RAItot (SDB, 14 ⫾ 3 event/h; CTRL, 0.3 ⫾ 0.1 event/h, p < 0.001). SAI did not differ between the two groups (SDB, 7 ⫾ 1 event/h; CTRL, 7 ⫾ 1 event/h). There was a significant correlation between OAHI and SPS (R = 0.88), EEG RAI (R = 0.94), RAItot (R = 0.99) and TAI (R = 0.92, p < 0.001 for all). SC analysis showed a smaller proportion of sleep runs lasting >10 minutes in the SDB group compared to the CTRL group (28% vs. 37%), however this was not statistically significant. Discussion: Our data showed a strong relationship between the SPS and the severity of SDB measured using the OAHI and therefore SPS may provide a useful measure of sleep fragmentation in children. However, standard arousal indices are equally correlated with severity of SDB and are readily available on routine PSG reports.

A38

Sleep and Breathing – Epidemiology P-46 TRENDS IN DEMOGRAPHICS AND SEVERITY OF SLEEP DISORDERED BREATHING OVER TWO DECADES OF DIAGNOSTIC SLEEP STUDIES IN AN ADULT SLEEP LABORATORY S GYULAY, J PRETTO, M HENSLEY John Hunter Hospital, Newcastle, New South Wales, Australia Introduction: Accessibility to diagnostic services, community and physician awareness of sleep disorders, and prevalence of obesity have all increased substantially over the past few decades within Australia. We report trends in demographics and severity of sleep disorders over a twenty-one-year period by retrospective analysis of over 14,000 diagnostic sleep studies with a view to better describing characteristics of diagnostic sleep study subjects. Methods: Demographic and apnoea-hypopnoea index (AHI) data were retrieved from the clinical database of a public metropolitan teaching hospital adult sleep service for the period from 1987–2007. Only those subjects undergoing an initial diagnostic study for potential sleep disorders were included in analysis (n = 14,648). Results: The laboratory increased in size from 2 beds in 1987 to 6 beds in 1994, with the number of new diagnostic studies increasing from 177 in 1987 to 855 in 2007. There was an average increase in body weight from 90 to 102 kg (+13%) for males and from 76 to 89 kg (+17%) for females. The median BMI increased from 29.5 to 31.9 kg/m2 (+8%). The proportion of females increased substantially from 20% to 39%. The rates of morbid obesity (BMI > 40) increased by over 5-fold over the study period (3% to 16% of patients studied). Compared with values from the late 1980s, median AHI fell significantly during the 1990s (p < 0.001), but has now increased (p < 0.001) indicating more severe disease (8.8, 6.6 and 14.3 respectively). There was no significant change in age over this time period (mean of 51 years). Since 1992 there has been a highly significant correlation between annual mean AHI and BMI (r2 = 0.84, p < 0.001) such that for every unit increase in BMI, there has been a corresponding increase of 4 units in AHI. Discussion: There is a continuing trend towards increasing body weight and BMI and significantly higher proportion of females being referred for diagnostic sleep studies. This trend is in line with the increasing rates of obesity observed in the general community in recent times. The substantial increase in body weight over this period has implications in planning and design of diagnostic sleep services in terms of likely subject profile. Our data do not support the hypothesis that increased accessibility to diagnostic services and increased awareness of sleep disorders is resulting in a decline in disease severity. Our data indicates that 84% of the observed variance in AHI since 1992 is attributable to changes in BMI. This is consistent with the premise that worsening severity in sleep-disordered breathing is primarily attributable to increasing obesity.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

P-47 SLEEP APNOEA AS AN INDEPENDENT RISK FACTOR FOR INCIDENT DIABETES AND METABOLIC SYNDROME: THE BUSSELTON HEALTH STUDY N MARSHALL2, K WONG1, C PHILLIPS1, P LIU1, M KNUIMAN3, R GRUNSTEIN1 1 Woolcock Institute for Medical Research, Sydney NSW, Australia, 2 National Health and Medical Research Council Centre for Clinical Research Excellence in Respiratory and Sleep Medicine, Sydney NSW, Australia, 3School of Population Health, University of Western Australia, Perth WA, Australia Background: Previously published cross-sectional population and clinic-based studies have suggested that obstructive sleep apnoea (OSA) is a risk factor for diabetes mellitus, and/or metabolic syndrome (MetS). However, no population-based cohort study has yet confirmed this. Methods: Residents of the Western Australian town of Busselton had their exposure to OSA quantified by the Respiratory Disturbance Index as measured overnight in their own homes (MESAM IV device). Diabetes was defined as either a fasting blood glucose ⱖ7 mmol/l or physician diagnosed diabetes. MetS was three or more of the following: abdominal obesity, hypertrygliceridemia, low HDL choleresterol, hypertension, and/or hyperglyceremia/diabetes. Results: Of the 399 participants 310 did not have diabetes at baseline and had complete data at a mean 4-year follow-up, 10 had moderatesevere OSA (RDI ⱖ 15/hour) and 64 had mild OSA (RDI 5–<15/hour). Moderate-severe OSA was an independent risk factor for incident diabetes (odds ratio OR = 17.81 95% CL 1.73, 183.61) but not mild OSA (OR = 1.75 95%CL 0.32, 9.54). 281 participants did not have metabolic syndrome at baseline. However, neither moderate-severe nor mild OSA were confirmed as significant independent risk factors for incident MetS (Moderate-severe OR = 2.70 95% CL 0.30, 24.4; Mild OR = 2.16 95% CL 0.33, 4.80). Conclusions: These data are consistent with moderate-severe sleep apnoea being an independent risk factor for the development of diabetes. However, the small sample size means that these data should be interpreted with caution. OSA may also be a risk factor for incident metabolic syndrome, although we could not statistically confirm this in our small sample.

P-48 A SIGNIFICANT PROPORTION OF PATIENTS WITH UPPER AIRWAY OBSTRUCTION DURING SLEEP HAVE POSITIONAL OSA HS LIAW2, A VEALE1, S HOMAN1 1 Respiratory Medicine, QEH, Woodville, S.A., Australia, 2University of Adelaide, Adelaide, S.A., Australia Introduction: People with OSA who have their have apnoeas and hypopnoeas predominantly during supine sleep are commonly referred to as having “Positional OSA” (POSA). We report here an inquiry into the prevalence of POSA amongst all people referred to our sleep laboratory for diagnostic PSGs over a 2 month period during 2008. Participants with OSA (AHI ⱖ 5.0) and a supine AHI more than twice lateral AHI, and an AHI in the lateral position <15 and ⱖ5.0 were defined as having “Partial POSA”. Subjects with OSA with a supine AHI more than twice lateral AHI, and an AHI in the lateral position < 5 were defined as having “Strict POSA”. We aimed to 1. Determine the prevalence of Strict and Partial POSA amongst the 110 diagnostic

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

studies performed during the study period. 2. To describe the characteristics of OSA patients with and without POSA. Method: Standard PSG with position validated by video plus weight and ESS. Results: Means and (SD)

All OSA “Non” POSA

Strict POSA

Partial POSA

Supine AHI

% of TST supine

BMI

ESS

N

Age

AHI

Lateral AHI

85

56 (12) 55 (13)

31 (25) 37 (29)

22 (23) 30 (32)

43 (34) 44 (38)

38 (34) 37 (37)

35 (8) 36 (8)

8.9 (6) 9.2 (5)

54 (12)

13 (6)

2 (2)

39 (27)

43 (28)

33 (4)

8.5 (4)

61 (13)

19 (7)

9 (2)

42 (14)

35 (21)

32 (6)

7.8 (5)

60 (71% of OSA) 16 (19% of OSA) 9 (10% of OSA)

Conclusion: Twenty nine% of OSA patients had Partial or Strict POSA. On average POSA patients spent 35 to 43% of their TST on their backs suggesting that measures to encourage them to sleep on their sides could result in a clinically significant improvement in their sleep quality. Patients with POSA were lighter than patients with non-positional OSA and had marginally less symptoms.

P-49 SLEEP-DISORDERED BREATHING IN HEALTHY PREGNANCY: PREVALENCE AND PREDICTION D WILSON1, L ELLETT3, S CROWE2, M JACKSON2, M BARNES1 1 Austin Health, Melbourne, VIC, Australia, 2La Trobe University, Melbourne, VIC, Australia, 3Mercy Hospital for Women, Melbourne, VIC, Australia Introduction: There is mounting evidence that sleep-disordered breathing (SDB) is common during pregnancy. Snoring, symptoms of SDB and a 10–15% risk of developing SDB have all been linked to pregnancy in recent research. However there are limited data on the foetal and maternal consequences of SDB during pregnancy, including maternal hypertension, pre-eclampsia, foetal growth restriction and lower Apgar scores. There has been no study of the prevalence of SDB in healthy pregnancy using objective measures, or the development of a simple screening tool. The aim of this pilot study is to measure the prevalence of SDB in healthy uncomplicated pregnancy and to validate the use of a screening tool for SDB in these women. Methods: Thirteen women in the third trimester (T3) of pregnancy and 10 non-pregnant women of similar age were recruited from a large, public hospital. All underwent overnight polysomnography, and completed the Multivariate Apnea Prediction (MAP) Index Questionnaire (Maislin et al., 1995) at the same time. Results: Initial data analysis shows that 92.3% of the T3 group snored, compared with 50% of controls (p < 0.05).

Supine REM AHI* NonSupine REM AHI* Total AHI

T3

Controls

15.5 (10.1) 16.8 (19.3) 4.9 (7.4)

4.7 (3.6) 4.2 (5.3) 1.5 (1.2)

A39

Healthy Function of Sleep

Upper airway flow limitation was common in the pregnant sample. There was no significant correlation between the MAP index and AHI (p > 0.1), however data collection is ongoing. Conclusions: SDB is more prevalent during the 3rd trimester of pregnancy compared to non-pregnant women, but only during REM sleep. SDB may develop during pregnancy, as most pregnant women did not report any symptoms of SDB (including snoring) prior to pregnancy. Given the negative maternal and foetal outcomes linked to SDB, these findings highlight the importance of developing a reliable prediction method to identify and treat SDB as early as possible in pregnant women.

P-50 SLEEP HEALTH IN TETRAPLEGIA – POLYSOMNOGRAPHIC CHARACTERISTICS OF A COMMUNITY POPULATION J SPONG1, D RILEY1, R PIERCE1, D BROWN1,2, D BERLOWITZ1 1 Institute for Breathing and Sleep, Melbourne, Australia, 2Victorian Spinal Cord Service, Melbourne, Australia Introduction: Sleep disorders in people living with a cervical spinal cord injury (tetraplegia) are common. Despite this, there are few studies reporting subjective and objective characteristics of sleep in this population. This study describes the polysomnographic characteristics of a community population living with tetraplegia. Methods: Full polysomnography were conducted on people living with tetraplegia and compared to values from an age-matched, able-bodied population. Results: Results present means and standard deviations in normal text and the median and interquartile range in italic. 92 people (70 males, 22 females) who had been living with tetraplegia for an average of 11.8 (7.7) years were assessed. On average, the participants were middle aged (45 (12.7) years of age) and overweight (body mass index of 25.7 (5.2)). Compared to able-bodied normal values, the tetraplegia sample had an increased sleep latency (129.3 (83.9) versus 14.2 (14) minutes), an increased REM latency (105.3 (74.5–164) versus 88 (68–120) mins), and more time spent in stages one (11.9% (10.8) versus 5.4% (3.3)) and three sleep (14.8% (9.3) versus 2.9% (2.4)). Discussion: Our data suggest that cervical spinal cord injury may be responsible for the poor sleep reported by people with tetraplegia. This research was supported by the Victorian Neurotrauma Initiative.

P-51 SLEEP HEALTH IN TETRAPLEGIA – SLEEP SYMPTOMS, FUNCTIONAL OUTCOMES AND QUALITY OF LIFE IN A COMMUNITY POPULATION J SPONG1, D RILEY1, R PIERCE1, D BROWN1,2, D BERLOWITZ1 1 Institute for Breathing and Sleep, Melbourne, Australia, 2Victorian Spinal Cord Service, Melbourne, Australia Introduction: Sleep disorders in people living with a cervical spinal cord injury (tetraplegia) are common. Despite this, there are few studies reporting subjective and objective characteristics of sleep in this population. This study describes sleep symptoms, functional outcomes and quality of life in a community population living with tetraplegia. Methods: Postal surveys were distributed to all people living with tetraplegia in Victoria. The postal surveys contained questionnaires

A40

regarding sleep behaviour, daily functioning, quality of life and likelihood of having Obstructive Sleep Apnoea (OSA). Results: Questionnaire data were collected from 163 people (127 males, 36 females) who had been living with tetraplegia for an average of 11.2 (7.5) years. The mean age of the sample was 45.8 years (Standard deviation = 13.6), with the average body mass index indicative of being overweight (25.7 (5)). Compared to published able-bodied normal values for the Basic Nordic Sleepiness Questionnaire, the tetraplegia sample reported more wake time prior to sleep and woke more often during the night. The tetraplegia sample had worse health utility values (0.3 (0.3) versus 0.85 (0.13)), yet had similar daily functioning attributable to sleepiness (Functional Outcomes of Sleep Questionnaire (FOSQ): tetraplegia 17.6 (2.6); able-bodied 17.9 (3.1)). Discussion: People with tetraplegia reported difficulty initiating and maintaining sleep, have very poor health utility values yet report little functional limitations due to sleepiness on the FOSQ. This may suggest that the FOSQ is not a sensitive tool in this population. This research was supported by the Victorian Neurotrauma Initiative.

P-52 THE WESTERN AUSTRALIAN SLEEP HEALTH STUDY (WASHS) J LEE1, A FEDSON1, L SIMPSON1, K WARD1, G LOVE3, M COOPER1, B SINGH2, D HILLMAN2, L PALMER1, S MUKHERJEE2 1 Centre for Genetic Epidemiology & Biostatistics, University of Western Australia, Perth, Western Australia, Australia, 2Department of Pulmonary Physiology, Sir Charles Gardiner Hospital, Perth, Western Australia, Australia, 3Western Australian Sleep Disorders Research Institute, Queen Elizabeth Medical Centre, Perth, Western Australia, Australia Introduction: The Western Australian Sleep Health Study (WASHS) is a population-based clinical, epidemiological, genetic and biospecimen resource that has been developed to investigate the aetiology, natural history and response to therapy of patients with obstructive sleep apnoea (OSA). WASHS is a multi-disciplinary, collaborative project that currently underpins several different projects regarding the development and treatment of OSA. Methods: Participants have been recruited from a sleep clinic in Western Australia since 2006. Data collected from consenting participants includes: (i) a 20-page Sleep Health Questionnaire that covers a range of areas including sleep behaviour, Epworth sleepiness score, medical history, physical activity, family history, smoking and alcohol use and anthropometric measurements; (ii) blood sample for biochemical and genetic analysis; (iii) sleep data from laboratory polysomnography (PSG); and (iv) dual energy x-ray absorptiometry (DEXA) to measure body fat distribution. The development of the WASHS resource will build on the unique WA population health data that has been collected over the last three decades as part of the WA Data Linkage System (WADLS) which includes all births, deaths and hospitalisations. All data collected is stored centrally in the WASHS database, made possible by the WA Genetic Epidemiology Resource (WAGER) which allows the data to be readily accessible by researchers and able to be linked to the WADLS. Results: To date, the WASHS has questionnaire data from 2622 participants, 1977 blood samples, 2747 sleep studies and 116 DEXA scans. Data from 1759 WASHS participants with complete data (men = 1088, 62%; women = 671, 38%) show mean AHI in men is 39.2 (severe) and in women is 27.5 (moderate). Mean BMI in men is 31.5 (SD = 6.52) and in women is 33.7 (SD = 8.62), both classified as obese. Mean Epworth sleepiness score for men is 9.7 (SD = 5.40) and women 9.8 (SD = 8.62). Self-reported doctor diagnosed depression was reported in 32% of men and 51% of women.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

Discussion: The WASHS cohort is predominately male, overweight with moderate to severe sleep apnoea and high rates of doctordiagnosed depression. Data collection for WASHS is ongoing. With the ability to link to the WADLS, the WASHS will be a valuable resource with information on a well-characterised sleep clinic population. By collecting a wide range of information, the WASHS will provide a valuable research tool to assist in understanding factors involved in OSA to better manage, treat and prevent disease.

P-53 A SURVEY OF THE SLEEP HABITS OF VICTORIAN JOCKEYS K GRANT, M HO Monash Medical Centre, Victoria, Australia Jockeys are elite athletes in a high risk occupation. Their work schedules place rigorous demands on their sleep/wake cycles and it is anecdotally accepted that they are sleep deprived due to early rise times and long working hours. They experience high rates of injury. Victorian Workcover Authority figures indicate the injury rates are 3–4 times higher than averages for other occupational groups. Australian Thoroughbred horse racing is valued at $A7.7 billion with an additional $14.5 billion wagered. Considering significant financial investments and the seriousness and frequency of workplace accidents it is important to recognise the consequential effects on the capacity of a jockey to quickly assess and respond to race circumstances. This is the first study to quantify the actual amount of sleep that jockeys have. 210 jockeys (flat, jumps and apprentices) were asked by mail to complete a 14 day sleep diary and an Epworth Sleepiness Scale (ESS). 22 ESS and 9 sleep diaries were returned. The ESS average was 11.7. Total sleep time was 6.9 hours of often fragmented sleep. Average number of days worked over a 14 day period was 12.55. Total sleep time on rest days was 9.62 hours. It is not possible for the study to show a causal link between sleep deprivation and jockey injury rates. However, given the known influence of sleep deprivation on accident risk, cognitive function, reaction times and health there is certainly an imperative to consider sleep debt as a potential threat to jockey safety.

Sleep Neurophysiology 1 P-54 INTERACTIONS BETWEEN SLEEP DEPRIVATION, COGNITIVE LOAD AND POSTURAL CONTROL IN YOUNG AND ELDERLY ADULTS R ROBILLARD1, M BOISSONNEAULT1, N MARTIN1, D FILIPINI1, F PRINCE2, J CARRIER1 1 Centre d’etude du sommeil et des rythmes biologiques, Hopital du Sacre-Coeur de Montreal, Montreal, Quebec, Canada, 2Departement de kinesiologie, Universite de Montreal, Montreal, Quebec, Canada Introduction: Falls increase with age and cause significant injuries in the older population. While a few studies suggest that sleep influences postural control, the impact of age-related changes in sleep mechanisms

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

on postural control is still unknown. This study aimed to determine whether age modulates the interactions between sleep deprivation (SD), attention, and postural control. Methods: Eight young (24.8 ⫾ 1.9 y.) and nine older adults (64.4 ⫾ 4.0 y.) stood still on a force plate in two counterbalanced sleep conditions: after a night of sleep and after a night of total SD (25 h of wakefulness). Two hours after wake time, center of pressure (CoP) displacements were measured in six postural conditions: eyes open (EO) and eyes closed (EC), while doing an interference task, a control task, or no task. Three-way ANOVAs (2 age groups * 2 sleep conditions * 6 postural conditions) were executed on anteroposterior CoP range and speed parameters. Results: SD increased CoP speed in the “EO – no task” condition and in all EC conditions for all subjects (sleep*postural condition interaction p < 0.007; all contrast p < 0.005; Fig. 1). Only older subjects presented an increased CoP range after SD in the “EO – control task” condition and in all EC conditions (age*sleep conditions*postural conditions interaction p < 0.002; all contrasts p < 0.05; Fig. 2). Conclusions: SD increased CoP speed in both age groups. However, in older subjects, the CoP did not only move faster, but also shifted further away, suggesting an even higher risk of crossing postural stability boundaries. Importantly, while the effects of SD on CoP range seemed to depend on tasks and visual input, CoP speed reacted to SD even when visual information was available and no cognitive task was performed, revealing a direct impact of SD. These results suggest that SD may be a significant factor putting older people at higher risk of falling.

P-55 WHAT DO THE TERMS “SLEEPY”, “TIRED”, AND “FATIGUED” MEAN? A SUKOVIC, L LACK, H WRIGHT School of Psychology, Flinders University, Adelaide, SA, Australia Introduction: The terms sleepy, tired, and fatigued are often used interchangeably with potentially untoward effects. For example, a sleepy driver may not stop driving when a roadside billboard warns against being “fatigued”. A common behaviour therapy for insomnia instructs to go to bed only when sleepy but may go to bed prematurely because of feeling fatigued. Recently it has been suggested that questionnaires of daytime fatigue be used in the diagnosis and treatment of chronic insomnia since fatigue, rather than sleepiness, is a prominent daytime symptom. The aim of the present study was to quantify the meaning of the terms sleepy, tired, and fatigued using the Semantic Differential Technique2 and to compare them in good and poor sleepers. Method: From a principal components analysis a set of 13 polar opposite scales (e.g. pleasant-unpleasant, active-passive, light-heavy) was selected on the basis of their high loadings on the three most significant orthogonal components described as evaluative, activity, and potency. The concepts of “sleepy”, “tired”, and “fatigued” were then rated using these 13 scales by 125 young adult university students, half of whom were poor sleepers with a PSQI score >6. Results: Fatigued was judged as more negative, active, and potent than sleepy with tired rated intermediately but closer to fatigued. There were no differences in these ratings between the good and poorer sleepers. The quantified meanings of these terms could also be visualised in 3-Dimensional space and the spatial distance between them calculated. There was a significant distance between sleepy and fatigued in this semantic space. Conclusions: This semantic differential technique showed a considerable difference in the meanings of sleepy and fatigued with tired intermediate but closer to fatigued. In this study we found no difference

A41

Healthy Function of Sleep

between good and poor sleepers in these meanings, however, clinically diagnosed chronic insomniacs should also be studied. In any case these results highlight the need for sleep researchers and clinicians to recognise the distinctions between the terms sleepy, tired, and fatigued. Reference Osgood CE, Suci GJ. Factor analysis of meaning. J. Exper. Psychol. 1955; 50:325–38.

P-56 SLEEP AND SLEEPINESS IN LATE PREGNANCY B SWEENEY, L SIGNAL, P GANDER, L ELLISON-LOSCHMANN Massey University, Wellington, New Zealand Introduction: Sleep changes in pregnancy are often perceived as normal by women and their carers. Poor sleep is associated with increased fatigue, depressive symptoms, and risk of intervention at birth, yet little is known about changes in sleep at this time. The aim of this pilot study was to describe sleep and sleepiness in late pregnancy in a small sample of New Zealand women. Methods: Pregnant women were recruited through their midwife, obstetrician or childbirth education class as part of a larger pilot study investigating links between sleep in late pregnancy and early postpartum, and maternal health. Women completed questionnaires at 35–37 weeks gestation and 12 weeks postpartum. Questions related to sleep quantity and quality, sleepiness, sleep disturbance, sleep disorders, mental and physical health, life stress, support, paid work and care of dependents. Three questions on sleep duration and quality were also asked via telephone at 4 weeks postpartum. Results: 35–37 weeks gestation questionnaire results are reported here. 34 women completed questionnaires (64% response rate). Mean total sleep time (TST) prior to pregnancy was 8.1 hours and women reported getting a good night’s sleep (GNS) on 6 nights/week. At 35–37 weeks TST was 7.2 hours and GNS dropped to 3 nights/week. 81.2% of women were rated as poor sleepers using the General Sleep Disturbance Scale. 94.1% were disturbed during sleep on 3+ nights/week with 76.5% being disturbed every night. 52.9% reported waking too early, and 43.8% were affected by daytime sleepiness on 3+ days/week. 21% of women scored above 10 on the Epworth Sleepiness Scale (ESS) compared to 13.21% in a representative national sample of New Zealand women aged 30–50 years1. The group was split by ESS score (cutoff >10) and comparisons made on self-reported sleep duration and quality. There were no statistically significant differences between the groups, although there was a trend for high scorers to report increased daytime interference from sleepiness and poorer quality sleep. However, this group also reported higher TST, and getting enough sleep/waking refreshed on more nights than those with ESS<10. Nulliparous women reported higher ESS scores (U = 70.50, p = 0.02, r = -0.4). Discussion: Women in late pregnancy report less sleep and increased sleep disturbance than when not pregnant and score higher on the ESS than non-pregnant women. Inconsistencies between TST and sleep quality warrant further investigation, as does the difference in ESS by parity. Reference 1 Harris, R. Personal communication, 2008.

P-57 DECREASED MELATONIN LEVELS IN TRAUMATIC BRAIN INJURY PATIENTS COMPARED TO HEALTHY CONTROLS J SHEKLETON1, D PARCELL1, J REDMAN1, JO PHIPPS-NELSON1, J PONSFORD2, S RAJARATNAM1 1 School of Psychology, Psychiatry and Psychological Medicine, Monash University, Victoria, Australia, 2Monash-Epworth Rehabilitation Research Centre, Epworth Hospital, Victoria, Australia Sleep-wake disturbance and hypersomnia are commonly reported following traumatic brain injury (TBI). Previous studies suggest that the diffuse axonal injuries associated with TBI may damage the structures involved in sleep-wake regulation, altering the timing of sleep and contributing to the emergence of irregular sleep-wake patterns or circadian rhythm disruption. However, the mechanisms that underlie these changes are not well understood. Melatonin, an endogenous sleep promoting hormone, is known to be involved in the circadian regulation of sleep-wakefulness in humans. The onset of the nocturnal rise in circulating melatonin levels (dim light melatonin onset; DLMO) can be used as a marker of the phase of the endogenous circadian pacemaker, located in the hypothalamic suprachiasmatic nuclei. We have previously reported in a smaller sample (n = 10/group) that the timing of DLMO does not differ between TBI individuals and controls. The present study investigated whether circulating levels of melatonin as well as the timing of DLMO are altered following TBI. Twenty-three (17 male, 6 female, 32.5 ⫾ 12.0 years) post-acute TBI patients, and 23 age- and gendermatched control participants (17 male, 6 female, 31.6 ⫾ 11.6 years) were recruited during follow-up outpatient clinic visits (TBI group) and from the general community (control group). The TBI group were studied 429.7 ⫾ 287.6 days post-injury. Post-traumatic amnesia (PTA) duration, a measure of injury severity, was 22.2 ⫾ 17.7 days. All participants attended the sleep laboratory for a single visit during which they remained in modified constant posture conditions in dim light (<10 lux) and provided half-hourly saliva samples from 18:00 h to 00:30 h. Saliva (200 ml) was assayed in duplicate by radioimmunoassay to determine melatonin concentration. Area under the curve (AUC, trapezoid) and DLMO were calculated. AUC was significantly less in TBI participants than in controls (p < 0.05). There was no difference in the timing of DLMO in the two groups. The observed decrease in melatonin levels in TBI participants may contribute to their sleep complaints. Investigation of the efficacy of exogenous melatonin as a treatment for the sleep disturbances in TBI patients is warranted. The study was supported by NHMRC grant # 334002.

P-58 AUDITORY STIMULATION DURING REM SLEEP INHIBITS REMS AND SLEEP DEPENDENT IMPROVEMENT ON THE VISUAL TEXTURE DISCRIMINATION TASK (VDT) S HO, R CONDUIT Monash University, Melbourne, Australia Introduction: Current research suggests that Rapid Eye Movement (REM) sleep plays a fundamental role in the consolidation of memory, particularly in the visual texture discrimination task (VDT). This study aimed to investigate the possible contribution of REM-sleep phasic activity to sleep-dependent improvement in the VDT.

A42

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

Method: 20 participants completed the VDT task before and after a night’s sleep in the sleep laboratory. Sleep was monitored using standard polysomnography (PSG), with additional vertical EOG recordings. A continuous auditory stimulus was used to suppress eye movements (EMs) during REM-sleep in 10 healthy participants. 10 participants receiving no stimulation acted as controls. Results: It was found that EMs could be suppressed during REM sleep without awakening. Participants with EM suppression during REM sleep had poorer post-sleep VDT performance improvement (M = 10.0 msec, SE = 4.47) than control participants (M = 20.0 msec, SE = 6.83; Mann-Whitney U = 21.00, p = 0.02). No significant difference in pre-sleep VDT thresholds between the EM suppression group and control was observed (Mann-Whitney U = 31.00, p = 0.165). Discussion: These results suggest that phasic-REM processes could be a key component in the consolidation of VDT memories and suggest that the neural processes underlying the phasic microstructure of REM sleep are a crucial aspect of memory consolidation in the VDT task.

P-59 AUDITORY INHIBITION OF RAPID EYE MOVEMENTS AND DREAM RECALL FROM REM SLEEP K STUART, R CONDUIT Monash University, Melbourne, Australia Introduction: There is currently debate within the area of sleep and dream research as to whether ponto-genicculo-occipital (PGO) waves or cortical arousal during sleep is the biological substrate underlying dreaming. As eye movements (EMs) are currently considered the best non-invasive indicator of PGO activity in humans, they have been widely used by a number of researchers to phenomenologically study PGO activity. The current study is comprised two experiments. The aim of the first experiment was to investigate the effect of low level repeated auditory stimulation on EMs (and inferred PGO activity) during REM sleep. Based on previous findings, it was predicted that such auditory stimulus presentation during phasic REM sleep would have a suppressive effect on EMs. The aim of the second experiment was to utilize this technique of EM (and inferred PGO) suppression, to examine the effects on subsequent dream reporting on awakening. Method: For the first experiment, 11 participants were monitored in a sleep laboratory using standard polysomnography (PSG), with additional vertical EOG recordings. A continuous auditory stimulus was used to suppress eye movements (EMs) during REM-sleep for 1 minute. In the second experiment, another 13 participants were involved in the same procedures, except they were woken for dream reports after auditory stimulation and control conditions without auditory stimulation. Results: There was a dramatic and significant difference in the frequency and amplitude of eye movements between the pre-stimulation, stimulation and post-stimulation conditions during REM sleep (Friedman’s c2 = 18.558, p < 0.01; Friedman’s c2 = 12.047, p < 0.01). The mean imagery percentage, number of visualisable nouns and imagery ratings of the auditory stimulation condition were significantly lower than that of the no auditory stimulation condition (Wilcoxon’s Z = -2.53, p = 0.01; Wilcoxon’s Z = -2.05, p = 0.04; Wilcoxon’s Z = -2.40, p = 0.02, respectively). Conclusion: The results of this study provide phenomenological support for PGO-based theories of dream reporting on awakening from sleep in humans.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

P-60 SUBJECTIVE PERCEPTION OF SLEEP, BUT NOT OBJECTIVE QUALITY, IS ASSOCIATED WITH POSTPARTUM BLUES IN HEALTHY WOMEN B BEI1, J MILGROM2, J ERICKSEN2, J TRINDER1 1 University of Melbourne, Melbourne, VIC, Australia, 2Austin Health, Melbourne, VIC, Australia Introduction: Existing literature supports a bi-directional linkage between sleep disruption and affective disorders. This study tested whether there was a relationship between Postpartum Blues (PPB) and sleep disruption during the first postpartum week, using both objective and subjective measurements of sleep. Methods: During the third trimester (Time-1), 44 healthy, low-risk women completed questionnaires on mood and sleep (Depression Anxiety Stress Scale, Hospital Anxiety Depression Scale, Positive Negative Affect Schedule, Pittsburgh Sleep Quality Index). 41 of these women wore actigraphy continuously for 7 days. 37 participants completed the same mood and sleep questionnaires at 1 week postpartum (Time-2), and 28 wore actigraphy during the first postpartum week. Attrition was caused by the demands of the research protocol, early deliveries, and hospital procedures. Results: After delivery, average mood improved across all scales, although there was deterioration of mood in 45.95% of the sample. Both objective and subjective night-time sleep significantly worsened after delivery while daytime napping behaviour significantly increased. Multiple regression analyses showed little relationship between Time-1 or Time-2 objective night-time sleep and postpartum mood. However, poorer subjective night-time sleep at both Time-1 and Time-2 was consistently associated with poorer mood reports at Time-2 (ranges .25 to .49). Linear regression analyses showed that more actigraphically measured naps and higher sleep related daytime dysfunction at Time-1 and Time-2 correlated with postpartum mood. (ranges .11 to .28 and .11 to .24 respectively). Both variables were affected by participants’ daytime sleepiness, which depended largely on subjective night-time sleep quality, and a conscious awareness of the need to make up for perceived poor sleep. Conclusion: Compared to actual sleep quality or quantity, the subjective perception of poor sleep, and the conscious awareness of its impact during wake-time, might play a more active role in the occurrence of PPB in a group of healthy, low-risk women.

Sleep & Breathing – Measurements P-61 AN ALTERNATIVE USE FOR PULSE TRANSIT TIME: RELATIONSHIP OF PULSE TRANSIT TIME MEAN AND VARIANCE TO APNOEA-HYPOPNOEA INDEX J BAKKER, A CAMPBELL, A NEILL University of Otago, Wellington, New Zealand Pulse Transit Time (PTT) measurement has been used previously to identify arousals during Obstructive Sleep Apnoea (OSA) as it has been shown to reflect the blood pressure surge following an obstructive event. Another potentially useful way of analysing PTT is to examine the mean and variance in order to determine the overall effects of OSA on the cardiovascular system.

A43

Healthy Function of Sleep

Aim: To report the PTT mean in all sleep stages during five-minute windows and compare to the Apnoea-Hypopnoea Index (AHI; the goldstandard index of OSA severity), and to compare PTT mean and variance in five-minute periods of normal breathing and obstructed breathing in patients with severe OSA. Methods: A sample of 16 patients (mean age 46 years) with varying degrees of OSA (mean AHI 28.9; range 0.4–88.6) were recruited. All underwent overnight polysomnography including PTT monitoring. PTT mean and standard deviation was collected during wakefulness and all sleep stages over ten epochs containing few or no obstructive events, and from Stage 2 of severe OSA patients containing obstructive events at least equivalent to the AHI. Results: Mean PTT correlated negatively with AHI during all sleep stages (Pearson’s correlations all p < 0.05). In patients with severe OSA, mean PTT of diseased and non-diseased Stage 2 sleep was not significantly different, though the standard deviation was higher during periods of obstruction (t(6) = -2.31, p < 0.05; paired samples t-test).

Two trained sleep scientists analysed each NPSG non-consecutively three times in random order using the (1) STD (2) ALT and (3) PRE derivations. Results: 17 subjects (4 F, Age = 44.3 ⫾ 16.8 years, BMI = 31.3 ⫾ 6.8 kg/m2, RDI = 25.4 ⫾ 23.2/hour (mean ⫾ sd)). Averaged Kappa values are summarised below: Interrater agreement

Average Kappa

ALT

PRE

0.75 ⫾ 0.08

0.75 ⫾ 0.10

0.73 ⫾ 0.07

Intrarater agreement

S1 (average ⫾ SD) S2 (average ⫾ SD)

300.0

STD

STD v ALT

STD v PRE

ALT v PRE

0.79 ⫾ 0.08 0.78 ⫾ 0.08

0.75 ⫾ 0.10 0.75 ⫾ 0.10

0.76 ⫾ 0.08 0.76 ⫾ 0.07

Discussion: Substantial agreement was observed between scorers with marginal difference across the different derivations. The lowest average agreement between scorers was observed for the PRE derivation. The results were similar for the intrarater comparisons with substantial agreement observed between the montage types for both scorers. The highest agreement was observed between the STD and ALT montages for both scorers. As there is high agreement obtained between all three EEG derivations it is concluded that the alternative EEG derivations can be adopted in future nocturnal polysomnorgams.

250.0 200.0 150.0 100.0 0

20

40

60

80

100

Figure Correlation between AHI and mean PTT during Stage 2 sleep (Pearson’s r = -0.56, p = 0.05). Conclusion: Further analysis as to the relationship between PTT and AHI is necessary, incorporating blood pressure data and the response to CPAP. The mean PTT during sleep could provide a measure of the chronic effects of OSA on the cardiovascular system allowing comparison between populations with differing disease severity, whereas the variance on PTT reflects acute effects.

P-62 AGREEMENT BETWEEN STANDARD AND ALTERNATIVE EEG ELECTRODE DERIVATIONS FOR MEASURING SLEEP A TURTON, J COPLAND, M HO Monash Sleep Centre, Clayton, Victoria, Australia Introduction: Recent guidelines proposed by the American Academy of Sleep Medicine (AASM) suggest alternative (ALT) EEG derivations for sleep staging nocturnal polysomnography (NPSG). Additionally, an alternative derivation has been suggested by Dyson (1984) using a pre-gelled (PRE) electrode outside the patients hairline. Our aim is to compare sleep stage analysis using the current standard (STD) electrode locations with two alternative derivations. Methods: Each subject had the following electrodes attached: (STD) C4 – A1, C3 – A2, O1 – A2, O2 – A1, (ALT) FZ – CZ, CZ – OZ, FPZ – CZ, (PRE) PreGelled – A1.

A44

P-63 A NOVEL ARTERIAL STIFFNESS PARAMETER, CARDIO-ANKLE VASCULAR INDEX. USEFULNESS AMONG PATIENTS WITH OBSTRUCTIVE SLEEP APNOEA T KUMAGAI2, T KASAI1, M KATO2, K-I MAENO4, S KASAGI1, F KAWANA2, S ISHIWATA3, K NARUI1 1 Sleep Center, Toranomon Hospital, Tokyo, Japan, 2Clinical Physiology and Sleep Center, Toranomon Hospital, Tokyo, Japan, 3Clinical Physiology, Toranomon Hospital, Tokyo, Japan, 4Toranomon Sleep Clinic, Tokyo, Japan Background: Several studies have demonstrated that arterial stiffness is a significant predictor of an increased cardiovascular morbidity and mortality. It has been reported that obstructive sleep apnoea (OSA) was associated with the increased risk of cardiovascular diseases and the progression of arterial stiffness. Recently, a novel stiffness parameter named the cardio-ankle vascular index (CAVI) has been developed. The CAVI is based on the pulse wave velocity (PWV) and compatible with conventional aortic PWV. This novel stiffness parameter has been reported to have the advantage of being easier to measure and being independent of blood pressure levels as compared with the other PWV parameters. However, there are no data about association between OSA and CAVI. Methods: Consecutive subjects who were diagnosed moderate to severe OSA [apnoea-hypopnoea index (AHI) ⱖ20/h] by overnight polysonmography and age-gender matched subject without OSA or with mild OSA (AHI <20/h), recruited from our database were enrolled. In all subjects, CAVI were measured using VaSera VS-1000 (Fukuda denshi, Tokyo, Japan) which monitored electrocardiogram, phonocardiogram and pressures and waveforms of brachial and ankle arteries. PWV from the heart to the ankle was obtained by measuring the length from the

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

aortic valve to the ankle, and then, using following formula, CAVI was automatically calculated and compared between 2 groups. Only subjects whose AHI ⱖ20/h were initiated into continuous positive airway pressure (CPAP) therapy, appropriately and the measurement of CAVI was repeated 1 month later. Results: Finally, subjects whose AHI <20/h (n = 90) and AHI ⱖ20/h (n = 95) were assessed. In subjects with AHI ⱖ20/h, BMI, waist circumference, triglycerides, uric acid and CRP levels were significantly higher than those with <20/h. CAVI in subjects with AHI ⱖ20/h was significantly higher than those with AHI <20/h (7.23 ⫾ 1.24 versus 6.99 ⫾ 1.44, respectively, p = 0.035 in ANCOVA with adjustment of age and blood pressure). In subjects with AHI ⱖ20/h, CAVI was reevaluated at 1 month after initiation of CPAP. Conclusion: Increased CAVI was observed in patients with severer OSA. However, treatment for OSA with CPAP might be able to reduce such increased CAVI.

Conclusion: In a sleep physician assessed group of patients, diagnostic home oximetry was reliable (1% failure rate) and obviated the need for diagnostic PSGs in 75% of patients. This equated to an annual saving of 150 PSGs. References 1. Antic AMJRCCM 2004;169:A543. 2. Antic AMJRCCM 2007;175:A706. 3. Mulgrew. Ann Int Med 2007;146:157–66. 4. AASM/ATS/ACCP consensus. JCSM 2007;3:737–47. 5. Roebuck Sleep& Biol Rhythms 2007;5:A40. Acknowledgements: Drs Belinda Miller, Alan Young, Jeremy Goldin, Simon Frenkel, Andrew Kyoong, Andrew Gillman & Celia Lanteri & Sleep lab staff.

P-65 THE UTILITY OF UNATTENDED HOME TESTING FOR OSA IN HIGH RISK POPULATIONS

P-64

K BURGESS1, A HAVRYK1, WA WHITELAW1, M MIKEL2, G NEWING2, S NEWTON2 1 Peninsula Respiratory Group, Sydney, NSW, Australia, 2Healthy Sleep Solutions, St. Leonards, NSW, Australia

HOME SLEEP APNOEA MONITORING WITH OXIMETRY AUDIT: HIT OR MISS? S HO, T ROEBUCK, E VAN BRAAK, M NAUGHTON Alfred Hospital, Melbourne, Victoria, Australia Background: Although home monitoring for obstructive sleep apnoea (OSA) with oximetry is diagnostically accurate1 and assists management2, it is criticised for selectivity (<5% OSA population)3, small patient numbers4 and low quality assurance5 and abuse in the absence of sleep physician assessment. Aim: To assess a home diagnostic oximetry in a university based teaching hospital with sleep physician assessment (before & after) with PSG backup. Methods: 12 month audit of home diagnostic oximeter monitoring (Novametrix [2 second averaging time; 0.125 Hz sampling rate]5) of patients suspected of OSA attending sleep clinic and assessed by sleep physician before & after oximetry. Outcomes sought included need for diagnostic PSG, provision of CPAP prescription, active “non-CPAP” management (ENT, LAGB, MAS, etc) based upon ⱖ2% oxygen saturation dip rate of more or less than 5 events per hour (ODI2). Service has been in operation 7 years with qualified personnel. Results: In 2007, 200 patients underwent a diagnostic oximetry at home, which represented 35% of all diagnostic (PSG & oximetry) studies. Sleep physician indication for study was “high pre-test probability of OSA” in 80% and “prioritization” in 20%. Inadequate oximetry data occurred in 2 (1%) patients. Of the 200 patients studied: ODI2 Class (eph) Prevalence n (%) Diagnostic PSG n (%) CPAP prescribed# n (%) Active non-CPAP Rx n (%)

ⱕ5 56 (28) 21 (11) 6 (3) 28 (14)

ⱖ5 142 (71) 27 (14) 69 (35) 41 (21)

# At least 48 of 75 patients were known to continue CPAP as of May 2008.

Of the ODI2 ⱖ5 eph group (n = 142), 72 (51%), 37 (26%) and 33 (23%) were mild (5–14 eph), moderate (15–25 eph) and severe (>25 eph) respectively. Of the ODI2 <5 eph group (n = 56), diagnostic PSG was considered necessary in 21 patients in whom PSG-AHI was >5 eph in 14 and AHI >15 eph in 4 patients.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Background: Obstructive sleep apnoea (OSA) is a modern epidemic of great health and economic importance. The major cause is obesity. OSA is thought to make an additive contribution to hypertension, stroke, ischaemic heart disease and glucose intolerance. It is likely that most patients with OSA are currently undiagnosed for a variety of reasons, including lack of symptoms, and lack of sleep laboratory capacity. Therefore home testing of asymptomatic high risk patients may identify patients who would benefit from treatment. The authors have devised an algorithm for use by General Practitioners (GPs) to identify their patients with a high probability of having occult OSA. The ApneaLink is a device that records nasal airflow and pulse oximetry data for over 12 hours. Automatic analysis is available for the respiratory variables commonly used in sleep studies. Aims: To assess the technical adequacy of the results of an ApneaLink® study when used unattended, in the home environment. Also to determine the prevalence of OSA, in a population of patients at high risk of having OSA. Methods: Unattended ApneaLink studies were performed by Healthy Sleep Solutions (HSS) on 683 patients preselected by their GPs according to an algorithm. The algorithm defined high risk patients as those with Hypertension or Diabetes ⫾ Obesity ⫾ symptoms of OSA. Technical adequacy was determined by the reporting physicians. OSA was defined as an AHI ⱖ10. Comorbidities were identified by the GPs. Other values were measured by HSS technicians. Results: 93% of studies were technically adequate. Of those only 13% were “suboptimal”. Mean AHI = 16.3 ⫾ 18.9/hour. 16% patients had AHI ⱖ30/hour. 28% had AHI ⱖ10 but ⱕ30/hour. Mean lowest saturation = 83 ⫾ 7%. Mean BMI = 31.0 ⫾ 6.8. Mean age = 54 ⫾ 8 years. Conclusions: In our hands the unattended ApneaLink study seems technically adequate for the task intended. In the population studied, the algorithm identified a very high prevalence (44%) of previously unsuspected OSA.

A45

Healthy Function of Sleep

P-66 MAN VERSUS MOUSE: THE EFFICACY OF AUTO-SCORING ALGORITHMS K RIXON, L CHENEY, M CHIKAZAZA, P CLARKE, B DUCE, K HANSSEN, C HUKINS, J MILOSAVLJEVIC, J POPOVICH, K SAARINEN, T SATHER, T SHIRLAW Sleep Disorders Centre, Princess Alexandra Hospital, Woolloongabba, Australia Background: Manual scoring of Polysomnography (PSG) is the accepted gold standard for determining the presence of Obstructive Sleep Apnoea (OSA). Most PSG Software Systems available now incorporate some form of Automatic Scoring Algorithm. However, Automatic Scoring is not accepted in clinical practice. Aim/objectives: To assess the agreement between Automatic Scoring algorithms vs Manual Scoring for the diagnosis of suspected OSA. Methods: Twelve Diagnostic Type 1 PSG acquired with Compumedics Profusion 3 software, were scored by 12 observers as part of a Tertiary Sleep Laboratory’s routine Quality Assurance Program. The PSG were also Automatically Scored by the Compumedics Profusion 3 software, including sleep staging (R&K criteria) and scoring of respiratory events (Chicago criteria). Results: Automatically scored results were similar to the mean (⫾SD) of the Manually scored results; Sleep Efficiency 68.2% (Automatic) v 66.7% (Manual) ⫾ 1.2, Arousal Index 9.6 v 9.0 ⫾ 7.5, REM Sleep Time 7.5 v 7.0 ⫾ 2.2, Total Sleep Time 31 v 30.8 ⫾ 1.2. Respiratory Disturbance Index (RDI) of Automatically scored results was 2 SD above the mean of the Manually scored results. Conclusions: Automatic Scoring of PSG (using the Compumedics Profusion 3 Software) is similar to Manual scoring. The study is ongoing.

P-67 IDENTIFYING CENTRAL AND OBSTRUCTIVE SLEEP APNOEA EVENTS IN ECG RECORDINGS A KHANDOKER, J GUBBI, C KARMAKAR, M PALANISWAMI The University of Melbourne, Melbourne, Victoria, Australia Obstructive sleep apnea (OSA) causes a pause in airflow with continuing breathing effort. In contrast, central sleep apnoea (CSA) event is not accompanied with breathing effort. The aim of this study is to differentiate CSA and OSA events from normal breathing events using wavelet based features of ECG signal over 5 second period and two-stage feed forward neural network. Total 19000 clips (each of 5-second duration) from normal breathing events, 2055 clips from 430 CSA and 7379 clips from 1262 OSA events were selected from single lead ECGs (sampling rate = 256 Hz) collected from 39 patients’ overnight sleep studies. Only 5-second ECG (which contains at least one cycle of inspiration and expiration) clips were extracted from pre-scored apnea events. Each ECG clip was decomposed to eight levels of detailed coefficients using mother wavelet Symlets (sym3). From each level of coefficients, Shannon’s entropy, log energy entropy, arithmetic mean and geometric mean were calculated. In total 32 features were extracted. Among them 18 best features were selected by add-remove method with Mahalanobis distance measures. The best feature was found to be Shannon’s entropy value of detailed coefficients at level 8 (0.5~1.0 Hz) (Receiver operating Characteristics (ROC) area = 0.79 between normal events and apnoea (OSA + CSA) clips) and level 6 (2~4 Hz) (ROC area = 0.84 between OSA and CSA clips). At the first stage of classification, apnoea events were classified from normal breathing events and at the second stage, apnoeas were classified into CSA and OSA types. Two-stage Feed

A46

forward Neural network (NN) models with 5 hidden layers was trained using the selected best features. Ten-fold cross validation (randomly 70% for training and remaining 30% for testing events by combining clips) results of the first stage NN model show the average accuracy of 93.0% in recognizing apnoea events from all clips. As for the performance of second stage NN model using the same set of best features, average accuracy of 94.96% in classifying OSA and CSA events was obtained. These results indicate the possibility of noninvasively recognizing OSA/CSA events from normal breathing events based on shorter segments of ECG signals.

P-68 UTILITY OF UNATTENDED APNOEALINK AS A SCREENING TOOL FOR OBSTRUCTIVE SLEEP APNOEA C HUKINS, G KEIR, K RIXON, B DUCE Princess Alexandra Hospital, Woolloongabba, Australia Background: The high prevalence of OSA places demands on the capacity to diagnose this problem with attended laboratory diagnosis. A simplified, unattended screening device would be useful to diagnose OSA without the need for laboratory facilities. Aim: To assess the utility of the ApnoeaLink device (nasal flow and oximetry channels) in the home in screening sleep clinic patients. Methods: 31 patients with suspected OSA underwent home screening with the ApnoeaLink prior to a laboratory diagnostic polysomnography. Concurrent (attended) ApnoeaLink monitoring was performed at the same time as polysomnography. (PSG) ApnoeaLink and laboratory parameters were compared. Results: 17 patients were male and 14 female. Mean ⫾ SD age was 47.1 ⫾ 12.8 years and ESS 16.0 ⫾ 9.0. Unattended oximetry failed in 4 subjects. There was significant correlation between unattended and attended ApnoeaLink AHI (r = 0.72, p < 0.001), mean SpO2 (r = 0.94, p < 0.001) and Oxygen Desaturation Index (4%) (r = 0.78, p < 0.001). There was also significant correlation between unattended ApnoeaLink and PSG AHI (r = 0.80, p < 0.001) and mean SpO2 (r = 0.78, p < 0.001). There was a small bias for a higher AHI in the laboratory polysomnography (PSG-ApnoeaLink AHI 5.4 ⫾ 14.5) but there was much closer agreement in mean SpO2 (PSG-ApnoeaLink SpO2 0.14 ⫾ 1.86%). An unattended ApnoeaLink AHI > 30 was 60% sensitive but 100% specific in detecting laboratory PSG AHI > 30. An unattended ApnoeaLink AHI < 5 was 100% sensitive and 90% specific for a laboratory PSG <5. Conclusions: Home (unattended) ApnoeaLink assessment appears valid in the sleep clinic population for ruling-in severe OSA and possibly ruling out significant OSA but further subject numbers are required. Laboratory confirmation may not be required with a positive ApnoeaLink study.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

P-69 CASE STUDY – DETECTING SITE OF PHARYNGEAL COLLAPSE DURING SLEEP USING ANATOMICAL OPTICAL COHERENCE TOMOGRAPHY: IMPLICATIONS FOR SURGICAL INTERVENTION J WALSH1, K MADDISON1, J ARMSTRONG3, R MCLAUGHLIN3, S BECKER3, D SAMPSON3, D HILLMAN1, P EASTWOOD1 1 West Australian Sleep Disorders Research Institute, Perth, WA, Australia, 2 School of Anatomy & Human Biology, University of Western Australia, Perth, WA, Australia, 3Optical+Biomedical Engineering Laboratory, University of Western Australia, Perth, WA, Australia Background: Identification of pharyngeal regions vulnerable to collapse during sleep is used to guide surgical management of snoring and obstructive sleep apnoea (OSA). Nasendoscopy under sedation (“sleep” nasendoscopy) is the technique most frequently utilised but only allows the airway proximal to collapse to be visualised and is non-quantitative. Further, it is unclear that sedation equates to sleep. Anatomical optical coherence tomography (aOCT) is a novel catheter based quantitative imaging device which can be used in natural sleep to visualise the pharynx at multiple levels. We aimed to examine the suitability of aOCT in identifying the site and length of pharyngeal collapse during sleep and the influence of head posture on collapsing site and length. Methods: One male subject (62 years, 28.7 kg.m-2) with OSA (42.5 events.hour-1) was prepared for standard polysomnography and pneumotachography. An aOCT catheter (3 mm OD) was inserted via the nares to the mid-oesophageal level. A sterno-occipital mandibular immobilisation (SOMI) brace maintained the head in the neutral, extended (10°) and flexed (14°) head postures for a minimum of 45 minutes while the subject slept (stable stage 2) in the supine posture. While the external catheter remained stationary, the internal aOCT probe was ‘pulled back’ from the level of the oesophagus to the nares during times of zero airflow (airway occlusion) in successive apnoeic events. The longitudinal locations of the distal and proximal ends of the collapsing segment were identified to define the site(s) and length of the collapsing segment. Results: Pharyngeal collapse occurred in the retropalatal region in all head postures. The length of the collapsed region was 16.5, 14.1 and 11.7 cm when the head was neutral, extended and flexed, respectively. Retrolingual patency was maintained in all head postures. Conclusion: This case example demonstrated a fixed retroplalatal site of collapse, suggesting potential candidacy for palatal surgery. More generally the case demonstrates a role for aOCT in identifying the collapsible pharyngeal regions during sleep under varying conditions which could include changes in posture or sleep stage or added sedation.

P-70 AUTOMATED VERSUS MANUAL PSG SCORING IN SLEEP DISORDERED BREATHING: EVALUATION OF COMPUMEDICS PROFUSION SOFTWARE P ROCHFORD1, W RUEHLAND1, R PIERCE1, A THORNTON2 1 Institute for Breathing and Sleep, Heidelberg, Victoria, Australia, 2 Royal Adelaide Hospital, Adelaide, South Australia, Australia Algorithms that enable accurate, rapid, automatic scoring of PSGs would represent an important advance for sleep medicine and research. The purpose of this study was to evaluate the accuracy of two versions

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

of automated scoring software (Profusion 2 and Profusion 3, Compumedics) compared with human scorers using PSGs from a clinical population of patients with suspected sleep disordered breathing (SDB). Method: Test PSGs consisted of a segment of 1.5hr duration from each of 17 diagnostic PSGs. 30 experienced scorers from 15 laboratories manually scored these PSGs for sleep stages (R&K), arousals (ASDA) and respiratory events (AASM). The PSGs were also analysed by both versions of Profusion software using the manufacturers recommended settings. Target values and target event sets for each PSG were generated based on a majority score approach utilising custom software. Algorithm performance was evaluated for PSG indices and for discrete events. Results: The test PSGs were representative of mild to moderate SDB with target values that ranged from 19–88 events/hour for apnoeahypopnoea index (AHI) and 16–61 events/hour for arousal index (ArI). The table shows results for these key PSG indices. Values are expressed as average (range) difference from target. P2-Target

P3-Target

Scorers-Target

AHI (events/Hr) +4.5 (-18.2–+17.8) +4.9 (-20.5–+18.3) 0 (-62.9–+56.6) ArI (events/Hr) -25.0 (-55.4–-7.7) -25.1 (-55.4–-8.4) 0 (-39.4–+54.2) TST (Min) +1.6 (-4.7–+9.1) +1.5 (-4.7–+9.1) 0 (-30.1–+57.8)

Conclusions: Slight differences between the two versions of the software were observed. 1. Automated results for AHI and TST fell well within the range of results obtained from a large group of experienced scorers. 2. ArI was consistently underestimated relative to human scorers. 3. Overall, automated scoring of AHI and TST, but not ArI by Profusion 2 and 3 software is close to experienced human scorers and has potential as a scoring aid.

Paediatric Sleep – Treatment and Epidemiology P-71 DEXMEDETOMIDINE PHARMACOKINETICS AND IMPLICATIONS FOR POSTOPERATIVE SLEEP A POTTS, G WARMAN, B ANDERSON University of Auckland, Auckland, New Zealand Postoperative sleep is characteristically of low quality and short duration. In addition, sleep architecture is altered in the postoperative patient, with increased levels of stage-1 and -2 sleep and reduced slow-wave and rapid eye-movement sleep. Many of the drugs commonly used in the intensive care unit, such as opioids and benzodiazepines contribute to sleep disruption (1). Dexmedetomidine (DEX) is an alpha-2 adrenoceptor agonist that has sedative and analgesic effects. One interesting aspect of DEX is that it appears to act through an endogenous sleep-promoting pathway. In contrast to GABAergic agents such as benzodiazepines, DEX produces a sedative response specifically by inhibiting neuronal firing at the locus cereuleus (2). Our study is designed to investigate the effects of DEX on sleep, recovery and mood in postoperative cardiac patients. As a first step in this work we have characterised the pharmacokinetics and pharmacodynamics of DEX postoperatively in a paediatric cardiac population. Fourty-five children who had undergone cardiac surgery received a single bolus dose of DEX (1–2 mg/kg over 10 minutes) on entry to the intensive care unit. Four blood samples were collected over 8 hours and titres of DEX were determined by LC-MS-MS. Non-linear mixed effect modelling of the data indicated a population clearance of 39.2 L/h/70 kg (CV 30.36%),

A47

Healthy Function of Sleep

central volume of distribution of 36.9 L/70 kg (69.49%), intercompartmental clearance of 68.2 L/h/70 kg (37.6%) and a peripheral volume of distribution of 69.9 L/70 kg (48.6%). DEX administration produced a significant reduction in postoperative opioid (control group 345 mg/kg versus DEX group 260 mg/kg, p = 0.002) and benzodiazepine (control group 1 mg/kg versus DEX group 0.4 mg/kg, p = 0.0002) administration and significantly reduced the time to extubation from mechanical ventilation (control group 24 h v DEX group 9 h, p < 0.0001). With a knowledge of the pharmacokinetics of DEX in both adult and paediatric patients we are now in a position to examine the influence of DEX on sleep in the postoperative cardiac intensive care setting. A randomised controlled trial of the efficacy of DEX versus standard treatment (propofol) is underway. References 1. Bourne RS, Mills GH. Sleep disruption in critically ill patients– pharmacological considerations. Anaesthesia. 2004;59(4):374–84. 2. Nelson LE, Lu J, Guo T, Saper CB, Franks NP, Maze M. The alpha2adrenoceptor agonist dexmedetomidine converges on an endogenous sleep-promoting pathway to exert its sedative effects. Anesthesiology. 2003;98(2):428–36.

P-72 FAMILY INTERVENTION FOR SLEEP PROBLEMS IN INFANTS UNDER THREE YEARS S BLUNDEN University of South Australia, Adelaide, South Australia, Australia Background: Night awakening and negative sleep associations are common in infants and toddlers. Treatment regimes include components of graduated extinction frequently leaving a child to cry in distress for long and specific periods of time. Family interventions, such as the Sensible Sleep Solution™ (SSS), that assist parents in limit setting while avoiding distress, are described. Methods: To date, 17 mothers (mean age 32.8.6 y, average education 13.75 y) of 7 girls and 10 boys [aged 8–30 months, mean (SD) age = 34.1 (18.8) mths] have presented at the Sleep Clinic, for behavioural sleep problems averaging 8 months duration. Parents completed a 7 week sleep diary, and sleep and demographic questionnaires pre and post intervention. The psychologically based behavioural intervention was based on stabilisation of sleep/wake rhythm, support of self soothing strategies and parental education for sleep limit setting and was implemented over 4–6 weeks. Changes in total sleep time (TST), sleep onset latency (SOL), family functioning and bedtime behaviours were evaluated. Results: All pre and post diaries were returned. One family has not completed the intervention. The remaining 16 families reported a mean (SD) daily TST increase of 57.0 (0.19) mins and a mean (SD) reduction of SOL from 17.5 (14.7) mins to 16.0 (5.12) mins. Problematic sleep associations were eliminated or replaced in all families. Treatment was achieved without distressed crying and treatment outcomes have been sustained 4–8 weeks after the last visit. All families that completed the intervention reported this method to be relatively stress free with significant improvement in family function and well being. Discussion: Preliminary data suggests that the SSS is an acceptable intervention for families of children under 3 years. It retains the powerful effects of extinction-based procedures with an “anti-distress” component. Early childhood sleep interventions improve family function and have the potential to reduce stress and fatigue in parents.

A48

P-73 DAYTIME COGNITIVE PERFORMANCE IN CHILDREN FOUR YEARS AFTER TREATMENT FOR SLEEP BREATHING DISORDERS C VAN DEN HEUVEL, C WANKLYN, M KOHLER, S BIGGS, D KENNEDY University of Adelaide, Adelaide SA, Australia Introduction: Sleep breathing disorders (SBD) are common in childhood, estimated to affect at least 10% of children in Western countries. They cover a spectrum of severity, but even mild, primary snoring is linked with poor daytime neurocognitive and behavioural performance. However, the long-term efficacy of treatment in ameliorating these deficits is unknown. Methods: Participants were sixteen children (8m : 8f, 9.2 ⫾ 1.9 years) who had previously undergone adenotonsillectomy (AT) for suspected upper airways obstruction during sleep, on average 4.3 ⫾ 0.2 years earlier. We investigated their daytime neurocognition and behaviour in comparison to fifteen healthy controls (6m : 9f, 8.9 ⫾ 2.3 years). Neurocognition and behaviour were examined using the Differential Ability Scales (DAS-II), Conner’s Parent Rating Scale-Revised (CPRS-R) and the Child Behaviour Checklist (CBCL 4–18). Current sleep patterns and the presence/absence of snoring were assessed via home oximetry, actigraphy, the Sleep Disturbance Scale for Children (SDSC) and sleep timing questionnaire (STQ). Results: The AT group did not differ significantly from healthy controls in age, gender, socio-economic status or maternal IQ, which are known confounders of neurocognitive performance. Four years after surgery however, the AT group were still reported by parents to snore more frequently (p < 0.05), have significantly lower non-verbal ability scores (p < 0.05) and more somatic behavioural problems (p < 0.01) relative to controls. Discussion: Prior treatment of suspected SBD in children by adenotonsillectomy appears not to prevent long-term neurocognitive and behavioural morbidity. A prospective study is now underway to confirm whether SBD in childhood causes long-term deficits in daytime performance, despite surgical treatment. Nonetheless, the current data highlights a clear need for earlier detection and improved management of childhood sleep breathing disorders.

P-74 CAN SHORT TERM CPAP PRE-OPERATIVELY HELP PREVENT RESPIRATORY MORBIDITY AND ICU ADMISSION POST-ADENOTONSILLECTOMY IN HIGH RISK CHILDREN WITH SEVERE OSA? G BLECHER, K WATERS, C EVANS 1 The Children’s Hospital at Westmead, Sydney, NSW, Australia, 2University of Sydney, Sydney, NSW, Australia Introduction: Children with severe OSA ⫾ associated risk factors are at increased risk of post–adenotonsillectomy respiratory complications, bleeding and increased length of stay. Our practice has been to observe such patients post-operatively in ICU. We hypothesised that using CPAP during sleep for minimum 2 weeks prior to surgery will help reduce post operative complications and decrease ICU admissions. Methods: This was a retrospective case control study. Sequential patients (allcomers) from January 1, 2006 to Dec 31, 2007 at a children’s tertiary hospital meeting criteria for severe OSA ⫾ risk factors for post-adenotonsillectomy morbidity were trialled on CPAP

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

pre-adenotonsillectomy. Cases were those that tolerated CPAP, controls were those non-compliant. Three or more criteria needed to be met to be offered pre-operative CPAP: AHI >20, minimum SaO2 <80%, TcCO2 rise >15 mmHg, craniofacial abnormality, including Robin sequence, Trisomy, BMI >2SD above/below mean, developmental delay & neuromuscular weakness. Patients established on CPAP were candidates not to be monitored in ICU post-operatively, but left to the surgeons discretion. Medical case notes and PSGs were reviewed. Exclusion criteria: previous adenotonsillectomy, long term CPAP use & insufficient data. Analysis was with SPSS Results: 69 patients were identified, 9 excluded (3 previous adenotonsillectomy, 6 insufficient data). 80% (48/60) had PSGs, the rest were diagnosed clinically, (2% missing). Mean age at diagnostic PSG was 3.95 years, SD 3.20, with minimum age 0.8 years. CPAP was tolerated in 70% (42/60), 27% were non-compliant (16/60; missing data 2/60). 75% (30/41) of those that tolerated CPAP were managed on the general ward. Of the 16 patients non-compliant with CPAP pre-operatively, 7 were managed on the general wards, 2 receiving post-operative CPAP and 1 was an emergency admission to ICU. There was no statistically significant difference in oxygen desaturation or intervention required post-op between cases or controls. Discussion/conclusion: The majority of high-risk children tolerated pre-operative CPAP & were managed on the general ward. Pre-operative CPAP may ameliorate post-adenotonsillectomy morbidity in at-risk children with severe OSA & greatly decrease the need for ICU monitoring. We propose a randomised study with first post-operative night PSG/ partial PSG to examine this further.

P-75 TEENAGE SLEEP AND THE INTRUSION OF DIGITAL TECHNOLOGY INTO THE BEDROOM K LUSHINGTON1, A WILSON1, M NADASAN1, K NGUYEN1, J DOLLMAN1, J MARTIN2, D KENNEDY3 1 University of South Australia, Adelaide SA, Australia, 2Women’s and Children’s Hospital, Adelaide SA, Australia, 3University of Adelaide, Adelaide SA, Australia Recent findings suggest that compared to a decade ago children are spending less time in bed and, hence, less time asleep. This decrease can be attributed to later bed rather than earlier rise times. A contributing factor to later bedtimes may be the intrusion of digital technologies into the bedroom. The intrusion of digital technologies is thought to be especially relevant in older children where sleep regimes are less likely to be regulated by parents. However, knowledge about the use of technology and pre-sleep behaviour in adolescents is not well documented and remains to be expanded. In this study we report findings from 348 adolescents (mean (SD) age = 17.3 (.98), range = 15–20.2 y and female/male = 121/227) from four metropolitan schools (two denominational and two University affiliated Senior Student Colleges). Schools volunteered following a request for participants at the South Australian Science Teacher Association annual conference. Participating students completed an omnibus questionnaire with scales on sleep habits, stress and coping, extracurricular activities, academic expectations, health and physical development. Data were collected in the second week of third term (July). Responses to the digital intrusion question (i.e. “After going to bed how often did you . . . ?”) revealed that girls compared to boys were less likely to report playing electronic games (>5 nights/week 2.5/5.7%), listening to music (18.5/23.9%) and watching TV (6.6/16.0%) but more likely to talk on the phone (19.2/15.0%). When asked “How often do you think you typically get enough sleep on school nights (Sun-Thurs

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

night)?” only 25% of adolescents reported obtaining sufficient sleep [never 12.4/13.2 (male/female); rarely 32.2/30.0; sometimes 27.3/32.6; usually 22.3/22.0; and always 5.0/2.2%]. The present study suggests that up to 50% of Year 11 and 12 adolescents report they sometimes or always play an electronic game, watch TV, listen to music or talk on the phone after going to bed. Given that most adolescents reported insufficient sleep on school nights this suggests that sleep may be unnecessarily shortened in a substantial number of students during a period where alertness and peak academic performance are at a premium. There is a need to educate adolescents, schools, parents and the community about sleep hygiene and the need to minimize the intrusion of digital technology at bedtime.

P-76 A SURVEY OF SLEEPING PATTERNS IN CHILDREN ADMITTED TO HOSPITAL A HERBERT1, J DE LIMA1, C SETON2, D FITZGERALD3, K WATERS2, J COLLINS1 1 Department of Pain Medicine and Palliative Care, The Children’s Hospital at Westmead, Sydney, NSW, Australia, 2Respiratory Support Service, The Children’s Hospital at Westmead, Sydney, NSW, Australia, 3Department of Respiratory Medicine, The Children’s Hospital at Westmead, Sydney, NSW, Australia Introduction: Sleep is important for healing and restoration during illness. The primary aim of this study was to determine the prevalence of poor sleep in children during their admission to a tertiary children’s hospital. A secondary aim was to understand the factors that contributed to poor sleep in these children. Methods: Parents of children, aged between 1 and 18 years, admitted to the hospital were asked to complete a sleep diary during one night of their child’s hospital stay. The diary allowed both the duration and fragmentation of sleep to be quantified. Poor sleep was defined as waking on 3 or more occasions and/or having a sleep efficiency of less than 90% (1). Children older than twelve years were asked to complete a diary independently. The Hospital Ethics and Scientific Review Committee approved the study. Results: Overall, 107 children were surveyed within the hospital. Diagnostic categories for these children included cancer (21 children), acute medical (16 children), chronic medical (24 children), peri-operative (38 children) and psychological (8 children). The overall prevalence of poor sleep was 52%. Unprompted awakenings were defined as spontaneous (17.8%) or due to toileting (17.8%). Factors that were specific to the hospital environment which woke children were nursing cares (25.2%), alarms (12.1%), and pain (12.1%). There was no observed effect of room type on the prevalence of poor sleep (single room compared to a shared ward). Eighteen children (17% of participants) tried some form of intervention to assist with sleep, including ear plugs, behavioural and relaxation strategies as well as medication. This group had a prevalence of poor sleep which was 29% (95% CI: 6 to 52%) less than the other children surveyed. Discussion: Children admitted to hospital have a higher prevalence of poor sleep compared to the prevalence of 18% reported for healthy children in the community (1). Educational interventions may increase parental and hospital staff knowledge of the importance of sleep as well as techniques to modify external factors that disturb sleep, including pain management. Reference (1) Sadeh A, Raviv A, Gruber R. Sleep patterns and sleep disruptions in school-age children. Dev Psychol 2000; (36): 291–301.

A49

Healthy Function of Sleep

P-77 SLEEP APNOEA AND POLYSOMNOGRAPHY CHARACTERISTICS IN SINGAPORE CHILDREN – EXPERIENCE OF A PAEDIATRIC SLEEP DISORDERS CENTRE J TANG, A GOH, OH TEOH, OM CHAY, CH TAN KK Women’s and Children’s Hospital, Singapore, Singapore Introduction: Obstructive sleep apnea hypopnea syndrome (OSAHS) is not uncommon in Singapore with a prevalence of up to 55% in the morbidly obese children. Polysomnographic characteristics of a local referred cohort have not been previously discussed. Objective: To describe the polysomnographic characteristics and patient profile of patients referred to the Sleep Disorders Centre (SDC) KKH for evaluation of possible OSAHS. Method: Review of polysomnography data of patients admitted for possible OSAHS during the period Jan 2000 to Dec 2007. Results: A total of 1442 children were evaluated. Mean age was 8.5 years (1 mth–19 years), 70% were males, 55.5% were overweight and racial distribution was 2.6: 1.0: 0.3: 0.1 for Chinese: Malays: Indians: Others respectively. More obese Malays were studied. Overall prevalence of OSAHS in the referred cohort was 45% with a higher prevalence in overweight children (55% vs 34%). In children > 6 years, overweight status and male gender were significantly associated with presence of OSAHS (p < 0.05) while not significant in those ⱕ6 years of age. In both groups older age was significantly associated with OSAHS (p < 0.01) while race was not (p > 0.05). Significantly disrupted sleep architecture with reduced sleep efficiency, deep sleep and increased arousals with abnormal gaseous exchange was demonstrated in children with OSAHS. OAHI correlated significantly with%idea body weight for height (r = 0.31), SaO2 nadir (r = -0.75) and peak CO2 (r = 0.16). Conclusion: Prevalence of OSAHS in our referred cohort was high with increased incidence in older children Overweight boys were more likely to have OSAHS in children >6 years old. Disturbed sleep and abnormal gaseous exchange is of concern in these children.

(63.8%), they woke on average once per night and spent a mean of 32 minutes awake (SD 41). 2% of parents believed their child to have a significant sleep problem and 15% believed their child to have a small sleep problem. Past research identified 3 factors as indicators of poor infant sleepers (Sadeh, 2004); waking 3 or more times per night, having less than 9 hours night-time sleep and spending more than 1 hour in the night awake. In this sample, 15.3% had less than 9 hours sleep, 11.5% were awake for more than 1 hour and 7.3% were waking 3 or more times. 25.4% of the sample experienced at least of one of these symptoms of poor sleep. They also took significantly longer to be put to bed and went to sleep later (p < 0.01), and were more likely to have noisy breathing or snoring episodes (p < 0.05). Children spending more than an hour awake at night more often slept in a room with adults (p < 0.05). Multiple Logistic Regression revealed no particular behaviours (for example, method of going to sleep or place of sleep) to be significantly correlated with symptoms of poor sleep. Discussion: The results from this survey support past findings. Children with identified sleep problems are more likely to share a bedroom with adults. This could be due to parents being more aware of problem behaviours, or could be the outcome of reverse causality e.g. sharing the room due to repeated disturbances. These findings support the need for further research using a larger sample, as well as objective monitoring to identify behaviours that may contribute to infant sleep problems, and to eradicate any confounding due to parental report.

Sleep and Breathing – Clinical 2 and Periodic Leg Movements P-79 CARDIORESPIRATORY COORDINATION DECREASES WITH THE DEGREE OF SEVERITY OF OBSTRUCTIVE SLEEP APNOEA M KABIR1, H DIMITRI2, P SANDERS2, D ABBOTT1, M BAUMERT1 1 University of Adelaide, Adelaide, South Australia, Australia, 2Royal Adelaide Hospital, Adelaide, South Australia, Australia

P-78 A SURVEY OF THE SLEEP OF THREE-MONTH-OLD INFANTS R GIBSON1, P GANDER1, S CHENG2, S-B BROGAN2, H DUCKETT2, J DOUWES2 1 Sleep/Wake Research Centre, Massey University, Wellington, New Zealand, 2Centre For Public Health Research, Massey University, Wellington, New Zealand Introduction: There is limited data concerning the sleep of infants across Australasia. Method: This study describes survey data on the sleep of 3 month old infants (N = 260) from a prospective asthma cohort covering rural and urban New Zealand. The aims of this study are to provide descriptive information on infants’ sleep, to identify factors associated with problem sleep, and to examine potential associations with health outcomes such as obesity and asthma. Reported findings are a summary to June 2008. Results: 60% of infants slept in a cot in their own room, 29% in a cot in their parents’ room. 82% slept on their back and 3% on their front. Only 1.5% of infants were sleeping in their parent’s bed. 35% had soft toys on their bed. Three month olds were having a mean of 9.8 (SD1.5) hours sleep per night and 4.7 hours per day (SD 1.7). They took approximately 26.7 minutes to put to bed (SD 33.4) at an average bedtime of 20:22. The majority of infants were left to go to sleep alone

A50

Introduction: Obstructive sleep apnea (OSA) adversely affects autonomic cardiac control. Previous studies found changes in heart rate variability, but little is known about the underlying mechanisms. In this study we aimed to determine the effect of OSA on the coordination between heart rhythm and respiration as a possible cause of abnormal HRV. Methods: Overnight sleep studies were analyzed in 248 patients (age: 49.4 ⫾ 12.3 years, BMI: 34.1 ⫾ 8.14 kg/m2, gender: 157 males, 91 females) with suspected OSA. The cohort was trichotomized based on the severity of OSA: 0 ⱕ AHI ⱕ 15, 15 < AHI < 30, and AHI ⱖ 30. Cardiorespiratory coordination was assessed by means of phase locking between heart rate and respiratory rate, using Hilbert transform. Heart rate was computed from the R-R intervals of body surface ECG while the respiratory rate was calculated from abdominal sensors. The effects of AHI, age, gender, BMI and sleep stage were assessed with 2-way ANOVA and Tukey’s multiple comparison tests. Results: OSA caused a decrease in the amount of phase locking between heart rate and respiration (see Table 1). Post-hoc analysis revealed a significant reduction of cardiorespiratory coordination in patients with severe sleep apnea (AHI ⱖ 30) compared to subjects without OSA (0 ⱕ AHI ⱕ 15). REM sleep caused a significant decrease in cardiorespiratory coordination compared to SW sleep, as determined by post-hoc analysis. Cardiorespiratory coordination was as not significantly affected by age, gender or BMI.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

Table 1. Percentage of sleep time revealing phase locking between heart rate and respiration

AHI 0–15 16–29 ⱖ30

Awake

S1 Sleep

S2 Sleep

SW Sleep

REM Sleep

8.6 ⫾ 6.1* 8.5 ⫾ 6.7 6.2 ⫾ 3.4*

8.4 ⫾ 4.3* 7.9 ⫾ 3.8 5.5 ⫾ 4.1*

10.1 ⫾ 4.2* 8.6 ⫾ 3.3 6.1 ⫾ 3.4*

12.6 ⫾ 5.7* 11.6 ⫾ 4.6 9.9 ⫾ 5.3*

8.3 ⫾ 4.0* 7.2 ⫾ 4.0 5.0 ⫾ 2.2*

Significant differences in the post-hoc comparison between AHI groups are asterisked; *P < 0.05.

Discussion: Cardiorespiratory coordination during sleep is affected by sleep stage and reduced in patients with severe OSA. It appears that OSA perturbs the interaction between the cardiac and respiratory oscillators, which causes a reduction in cardiorespiratory coordination. The assessment of cardiorespiratory coordination might provide an ECG based screening tool for OSA.

P-80 PREDICTORS OF SLEEP LATENCY IN A POPULATION PRESENTING TO THE SLEEP LABORATORY A ROY, A THORNTON, R ANTIC Royal Adelaide Hospital, Adelaide, SA, Australia Aim: To determine the predictors of Sleep Latency (SL) in a population of patients presenting to the Sleep Laboratory for Polysomnography (PSG). Methods: Patients presenting for PSG over a period of 2 months in 2008 were included in the study. Patient demographics, the Epworth Sleepiness Score (ESS), Body Mass Index (BMI) and history of depression were ascertained at presentation. The Stanford Sleepiness Score (SSS) in a scale of 1 to 7 was assessed just prior to the lights being turned off in preparation for sleep. After PSG, predictors of SL were sought. Statistical analysis: Unlike absolute values of SL, the natural logarithm of SL was normally distributed and thus, was more appropriate for fitting statistical models. To investigate the influence of SSS, ESS, BMI, the age of subjects and the Respiratory Disturbance Index (RDI) on SL, ordinary linear regression models were fitted to the data with log SL as the outcome. An independent samples t-test was used to analyse the influence of gender and depression on log SL. Results: Results of 129 patients were available for analysis. The median age of patients was 50 years (range 19 to 80); there were 72 males (56%). BMI ranged from 21.8 to 65.1 kg/m2 (median 32). Median SL was 17 minutes (range 1 to 234). 79 (61%) had moderate or severe sleep apnoea (RDI >15/min). 44 (34%) had history of depression. The median SL was significantly longer in patients with depression than in those without (24 minutes vs. 15 minutes, p = 0.009). There was a significant relationship between log SL and SSS, with unit increase in SSS associated with 12% reduction in median SL (p = 0.019). There was also a significant relationship between log SL and age, with unit increase in age associated with a 1.65% increase in median SL (p = 0.014). The estimated median sleep latency for a 40 year-old was 15 minutes, compared to 20 minutes for a 60 year-old and 28 minutes for an 80 year-old. ESS, BMI, RDI and gender had no significant influence on SL. Depression, SSS and age remained significant independent predictors of log SL when entered into a multivariate model. Conclusions: Our results suggest that depressive illness, low SSS and increasing age are the main predictors of prolonged SL in a population

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

of patients presenting to the Sleep Laboratory for PSG. The ESS, RDI, BMI and the gender of patients have no significant influence on SL in this population.

P-81 SMS TEXTING IMPROVES ATTENDANCE OF PATIENTS AT A SLEEP CLINIC J DOUGLAS, T YOUNG The Prince Charles Hospital, Brisbane, Queensland, Australia Background: With the obesity epidemic and increased awareness of sleep disorders there has been an exponential rise in patients accessing sleep clinics. Given the limited resources, maximising patient attendance at clinics is vital. There is no published data on attendance rates in sleep clinics. Patients with sleep disorders may have accompanying cognitive deficits, memory & concentration problems, and mood disturbances. These factors may contribute to lower sleep clinic attendance rates. Methods: A retrospective audit of attendance rates at a private practice sleep clinic based at a major teaching hospital was performed. The control period was from February to April 2006 and the intervention period was the same months in 2007. One administration officer text messaged all patients with a mobile phone number via Groupwise (a Queensland Health email system). Patients with mobile phones were sent a reminder text 1 week prior to their appointment. Patients without a mobile phone were telephoned to remind them of their forthcoming appointment. SPSS was used for statistical analysis. Results: 62% of patients had a mobile phone. There was an increased attendance rate in 2007 compared to 2006 (1035 (86.8%) compared to 825 (84.1%) p < 0.03) There was an increase in attendance rates in the group receiving a reminder SMS message compared to those not receiving a text message (646 (87.9%) vrs 389 (85.1%) p < 0.031) in the 2007 cohort. Discussion: We have demonstrated that a simple process of text messaging patients to remind them of forthcoming appointments is efficient in improving attendance rates at a sleep clinic. Given the majority of patients have mobile phones, this is an effective communication strategy for this patient cohort.

P-82 THE EFFECT OF SURGERY ON OSA SEVERITY IN HIP REPLACEMENT PATIENTS WITH UNDIAGNOSED OSA M ANSON, P ROCHFORD, N BERGMAN, R PIERCE, F O’DONOGHUE 1 Austin Hospital, Heidelberg, Melbourne Victoria, Australia, 2Warringal Hospital, Heidelberg, Melbourne, Australia Introduction: Obstructive Sleep Apnoea (OSA) may be worse in the postoperative period due to effects of analgesia, REM rebound following sleep disruption and restriction to supine posture. This may lead to an increase in postoperative morbidity. We present preliminary results examining changes in severity of undiagnosed OSA following hip replacement surgery. Methods: Patients undergoing hip arthroplasty and without a previous diagnosis of OSA were recruited from surgeons’ rooms and outpatients’ clinics. Epworth Sleepiness Scale and Multivariate Apnoea Prediction Questionnaire (MAP) were administered. Those with MAP scores ⱖ0.55

A51

Healthy Function of Sleep

were invited to have a baseline PSG. Patients with OSA (AHI>5) underwent a second PSG either postoperatively (surgery group) on the third postoperative night or a week after initial PSG but before planned surgery (control group). Groups were assigned by alternating allocation. Average Interval between PSGs was 5.7 (surgery) and 6.7 days (control). Results: 18 males and 10 females were recruited. No females and 13 males had MAP scores ⱖ0.55. 8 of 13 males consented to PSG and all demonstrated OSA. Table 1 shows results for 7 subjects. Compared with baseline, postoperative AHI, AHI (supine) and hypoxaemia were worse. A rebound in REM sleep was not evident. The average supine AHI at baseline was higher than the overall AHI and partly accounts for the observed postoperative changes.

Discussion: This electrical pattern has not to our knowledge been described previously in the sleep laboratory setting. Enquiry as to the use of DBS should now form part of a sleep history. The proximity of the subthalamic nucleus to circadian control centres in the hypothalamus raises the possibility that DBS is causally linked to this patient’s sleep disturbance, and that DBS should be added to the already long list of sleep disorders associated with parkinsonism.

Table 1. %SpO2 <90 =%sleep time at O2 sat <90. All figures mean ⫾ SD

AHI (supine)

AHI

surgery n=4 control n=3

%SpO2<90

REM%TST

age

BMI

Pre

Post

Pre

Post

Pre

Post

Pre

Post

P-84

64.8 ⫾8.7 60.0 ⫾1.7

31 ⫾3.1 31.8 ⫾3.2

21.4 ⫾17.2 32.4 ⫾7.2

80.3 ⫾26.2 28.8 ⫾5.2

53.1 ⫾10.6 51.9 ⫾26.4

80.3 ⫾26.2 45.1 ⫾23.3

7.4 ⫾7.6 6.3 ⫾5.0

30.4 ⫾36.4 4.7 ⫾3.7

11.5 ⫾8.3 15.1 ⫾8.4

3.3 ⫾3.5 17.5 ⫾9.0

COMPARISONS BETWEEN THE NEW AASM AND THE CURRENT ASDA LIMB MOVEMENT CRITERIA

Conclusion: Preliminary results suggest that OSA is more severe in the immediate postoperative period, partly due to the requirement to remain supine, but also due to the effect of surgery. These results require confirmation with greater numbers of subjects. Recruitment is continuing.

P-83 POLYSOMNOGRAPHY IN A PATIENT WITH A DEEP BRAIN PACEMAKER P TEUWEN, S MORRISON, R ALLEN, P SILBURN Respiratory and Sleep Specialists, Brisbane, QLD, Australia Introduction: Deep Brain Stimulation (DBS) is a surgical treatment for disorders including parkinsonism, epilepsy and clinical depression. DBS involves the implantation of an electrode into the subthalamic nucleus, wired to a neurotransmitter or pacemaker implanted subcutaneously below the clavicle or in the abdomen. We describe a case study of a 65 year old patient referred for polysomnography (PSG) who had had a deep brain pacemaker in situ for the treatment of severe parkinsonism since May 2006. Neurological symptoms included difficulty walking, weakness of all four limbs and speech impairment. The movement related symptoms improved dramatically when the pacemaker was active, though speech impairment worsened. Sleep symptoms included snoring and daytime somnolence with nocturnal insomnia, i.e. an inverted circadian rhythm, with onset several months after commencement of DBS. Method: An attended full PSG was performed using Compumedics Profusion 2 according to the American Academy of Sleep Medicine (2007) guidelines. The DBS pacemaker was active throughout the study. Results: Throughout the study an artefact of 40Hz was recorded. This rendered sleep staging impossible. The artefact was visible through all EEG leads, EOG leads and ECG traces. A regular cyclic pattern was noted over a 30 min timeframe, with double peaks in amplitude occurring approximately every 2 min:

A52

B DUCE, L CHENEY, K SAARINEN, S MCWHAE, C HUKINS Sleep Disorders Centre, Princess Alexandra Hospital, Brisbane, QLD, Australia The definition and scoring of Periodic Limb Movements in Sleep (PLMS) have been defined according to the American Sleep Disorders Association criteria (ASDA), published in 1993. Recently the American Academy of Sleep Medicine (AASM), the new incarnation of ASDA, has proposed a new set of definitions and scoring rules. These new criteria differ particularly with respect to the amplitude and the duration required in defining a limb movement. The significance of these changes to limb movement definition has yet to be determined. Fifty-three consecutive patients referred to a tertiary sleep unit for a full-night diagnostic polysomnography were initially included in the study. Each diagnostic study was scored according to usual criteria (R&K, Chicago, ASDA). The limb movement portion of the study was subsequently rescored using the AASM criteria by a different scorer. Included patients tended to be middle-aged (Median age 56 ⫾ 9 years), predominantly male (M : F 37:16), and obese (BMI 35 ⫾ 6 kg/m2). Their sleep was characterised by slightly longer than normal sleep latencies (27 ⫾ 8 min) with reduced sleep efficiency (67 ⫾ 10%). The selected patients displayed a Respiratory Disturbance Index corresponding to moderate OSA (RDI 24 ⫾ 11) with mild SpO2 desaturations (Nadir SpO2 86 ⫾ 4%). Scored limb movements that were not associated with respiratory events increased with the use of the AASM movement rules (LMI 7.8 ⫾ 16.3/hour to 27.8 ⫾ 14.6/hour p < 0.001). These limb movements were distributed evenly between NREM and REM sleep for both rule sets. Scored PLM’s increased with the AASM movement criteria (0.2 ⫾ 11.4/hour to 11.2 ⫾ 13 p < 0.01) with a concomitant increase in the number of PLM episodes (0.1 ⫾ 1.1/hour to 0.8 ⫾ 1.5/ hour p < 0.01). Very few arousals appeared to be associated with these PLMS (1.4 ⫾ 12/hour) and this did not change with the AASM movement criteria. In summary, utilisation of the new AASM movement criteria results in enhanced limb movement and PLM indices. These results may suggest that these index changes may not have any clinical significance, however an exploration of “subcortical” or autonomic arousals may reveal otherwise. Moreover the determination of new threshold indices for abnormal results may be necessary. Further research is required to determine the clinical significance of these new criteria.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

P-85 PIEZO ELECTRIC SENSORS AND LEG EMG DO NOT DETECT LIMB MOVEMENTS EQUALLY DURING A POLYSOMNOGRAM A ROSE, S WINDLER, P CATCHESIDE, L HARMER, A YEO, M JURISEVIC, J MERCER, D MCEVOY, N ANTIC Adelaide Institute of Sleep Health, Repatriation General Hospital, South Australia, Australia Periodic limb movements of sleep (PLMS) are a common finding in a polysomnogram (PSG). They have been reported in up to 45% of elderly patients who were not specifically questioned for Restless Legs Syndrome. Periodic limb movements have been traditionally detected using an electromyogram (EMG) and this is the recommended method (AASM 2007 Manual for the scoring of sleep and associated events). However, Piezo electric sensors (PES) have been widely used as an alternate means of recording limb movements in a PSG as they are technically easier to use However this device has never been compared to the EMG. Methods: This study compared PLMS index and limb movement related arousal index between the two sensors – EMG and Piezo Electric Sensor (Compumedics – leg sensor). They were used simultaneously in 20 consecutive patients who underwent laboratory PSG (Compumedics profusion online version 2) primarily for the assessment of sleep disordered breathing. Limb movements were scored using the AASM 2007 scoring criteria. Results: Eleven patients had PLMS as defined by an EMG-PLMS index of >15/hour (a threshold of limb movement activity considered “significant” in accordance with the ICSD 2nd edition of the AASM). The ROC curve for PES detection of significant PLMS revealed an AUC = 0.674 ⫾ 0.122[SEM] (p = 0.169). The sensitivity and specificity of detection of limb movements using the PES was 63.6% and 77.8% respectively. The PLMS index detected by the EMG was [mean ⫾ SEM] 33.1 ⫾ 9.1 and the same index detected by PES was 19.6 ⫾ 4.3/hour. The correlation of the PLMS index using the 2 devices was poor – R2 = 0.065, p = 0.144. There was better agreement between the two devices when considering only those limb movements associated with EEG arousal, (limb movement related arousal index detected by EMG 2.6 ⫾ 0.7 and PES 2.9 ⫾ 0.6/hour, R2 = 0.548, p < 0.001). Conclusion: Piezo Electric Sensor is likely to be an unreliable means of detecting PLMS compared to EMG. However, they show better agreement with EMG-detected limb movements when considering only those limb movements associated with sleep arousal. Since limb movement related sleep fragmentation is possibly more clinically relevant than limb movements alone Piezo Electric Sensors might still be useful in the clinical assessment of limb movements in sleep. Further study with a large sample and standardisation of criteria for application of PES and interpretation of data derived is needed.

Sleep Neurophysiology 2 P-86 A SYSTEMATIC ANALYSIS OF WAKING EEG DATA TO DETERMINE THE TIME COURSE AND MAGNITUDE OF SLEEP INERTIA D MULLER, M VAN DEN BERG, L SIGNAL, A GARDEN, P GANDER Massey University, Wellington, New Zealand Introduction: Sleep inertia, the transient confusion and reduced performance seen on sudden awakening, is considered to be one of the

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

most serious contraindications to the use of napping in the workplace. Physiologically it is unclear what underpins this, however one hypothesis is that it is a gradual transition between sleep and wakefulness requires time to reach completion. Disparities in methodology in previous studies have resulted in inconsistent findings, highlighting the need for a controlled, systematic approach. This study used waking EEG spectral power analysis as a key measure to detect changes in neurophysiological alertness. To ensure results are not confounded, artefact (e.g. eye movements, lead movement) must be removed. However there are presently no standardised methods for these processes. Methods: 12 participants completed the protocol on 4 occasions (20, 40 and 60 minute nap, +no-nap control). Each protocol began with a full night’s sleep recording followed by 20 hours of continuous wakefulness then a nap opportunity. A 6-minute test battery was administered at frequent intervals using the Karolinska Sleepiness Scale (KSS), N-back working memory task and 90 seconds of controlled EEG recording (Fz, Pz and Oz). Waking EEG artefact was removed using a 2 step process: 1) Visual screening for non-physiological artefact using standard criteria developed to guide decision making and 2) Independent Component Analysis (ICA) for removal of vertical and horizontal eye movements. Up to 3 components were visually identified, that reduced eye movements in the EEG data without altering spectral power. Spectral output files were then created for each channel. Results: 2 researchers scored a total of 1296 recordings and double scored 10%. 26 recordings (2%) were deemed unusable due to technical recording problems/poor quality. 1237 (97%) of remaining files had non-physiological artefact rejected prior to running ICA. 3630 spectral output files for Fz, Pz and Oz were created – 95% of potential outputs for the ‘usable’ recordings and 93% of total recording events. Across all channels 80% of data was identified as suitable for analysis, which is significantly higher than previous studies. Discussion: Having standard processes is critical to enable comparisons between sleep inertia studies. Spectral analysis of waking EEG provides valuable data, however inappropriate artefact inclusion can alter results. The data collection methods and analysis used in this study enabled us to retain a significant amount of data using a controlled and comparable process.

P-87 SLEEP ONSET ESTIMATOR D CVETKOVIC, I COSIC RMIT University, Melbourne, Victoria, Australia Introduction: The electroencephalographic (EEG) alterations during the human sleep onset have been evaluated by several studies in the past. However, the analysis part has been limited due to standard signal processing methods. Methods: This paper has attempted to evaluate a number of advanced parameters for improved sleep onset estimation, such as EEG nonparametric coherence, power frequency and spectral band power. The EEG inter-hemispheric coherence was computed to investigate the functional connectivity between the left and right brain hemispheres. A well known non-parametric spectral estimation algorithm is the Capon’s approach, known as Minimum Variance Distortion-less Response (MVDR) was applied for coherence computation. The power frequency or the weighted mean frequency of the EEG spectrum has been defined as the ‘brain-rate’, as a preliminary EEG frequency indicator of wakesleep transition. The EEG band powers were performed using shorttime Fourier transform (STFT) by extracting absolute and relative (band/total) EEG frequency bands.

A53

Healthy Function of Sleep

Results: The analysis structure of sleep depth estimation was performed. The results indicated a sharp suppression in coherence at 4 Hz (narrow band) from wake (W) to stage 1 (S1), followed by the recovery period where coherence is fairly consistent around. The ‘brain-rate’ at W was 7–7.3 Hz and during stage 1 (S1) to stage 2 (S2) it was decreased to 5.5 Hz. In the middle of S2, it covered back to 7 Hz and progressively decreased to 4 Hz during S3 and S4 until it reached the next sleep stage/cycle. The EEG relative and absolute band power results revealed a clear suppression in alpha band from S1 onset and remained fairly constant (absolute) and continuously decreasing (relative) during S2-S4. In terms of relative delta band, there is an increase in delta relative power from W to S1-4 transition. Discussion: It is possible to objectively estimate the wake-to-sleep transition, or drowsiness period from the EEG coherence, brain-rate and spectral delta/alpha band parameters. There have been many studies, reporting on what is the ‘right’ moment of sleep onset. We assume that this moment is from the onset of stage 1 (S1). All of our parameters evaluated have shown a potential as the solid indicators in sleep onset detection. However, further testing is needed on multiple human subjects, on-line algorithm design in neuro-feedback applications, artefact removal techniques and investigation on EEG spatial and temporal dynamic changes during sleep onset.

P-88 HOMEOSTATIC AND CIRCADIAN MODULATION OF BEST AND WORST PERFORMANCE DURING TWO COGNITIVE TASKS X ZHOU, G ROACH, R MATTHEWS, N CROUCHER, T WENG, L WILLIAMS, C SARGENT Centre for Sleep Research, the University of South Australia, Adelaide, South Australia, Australia Introduction: Under elevated sleep pressure, fluctuations in attention make it difficult to sustain performance during the Psychomotor Vigilance Task (PVT). While micro-sleeps result in lapses and response errors during PVT, compensatory mechanisms act to preserve normal reaction time (RT). Consequently, best performance (fastest RTs) is better maintained or less vulnerable to elevated sleep pressure than worst performance (slowest RTs). In most studies however, these two performance domains were either not considered or were examined solely on PVT at a single circadian phase. Therefore, the current study examined best and worst performance from two tasks requiring sustained attention, PVT and Serial Addition Subtraction task (SAS), across circadian phases under various levels of homeostatic sleep pressure. Method: Nine healthy males (18–35 years) lived in a time isolation laboratory for 11 days during a forced desynchrony protocol. Following an adaptation and baseline night sleep, participants were scheduled to seven consecutive 28h ‘days’ (a : r = 1:2), rendering desynchrony between the sleep-wake cycle and near-24h circadian rhythms. Given that each 28-h day began 4h later relative to the circadian system, each time point of a day coincided with a circadian phase 4h later compared with the same point of the previous day. Every 2.5h during wake, a performance battery was administrated, including a 10 min PVT and 5 min SAS. Results: Data are yet to be analysed. Six repeated-measures ANOVAs (Wakefulness x Circadian Phase) will be conducted on best and worst performance, including mean 10% fastest and slowest RT (PVT and SAS) and corresponding mean accuracy rate (SAS). Main Wakefulness effect would be significant for worst performance but not best. If the circadian force has a role, Circadian main effect or and Wakefulness x Circadian interaction would be significant.

A54

Discussion: Findings would contribute to understanding performance variability under the influence of homeostatic and circadian processes.

P-89 SLEEP BEHAVIOUR IN AUSTRALIAN OLYMPIC SWIMMERS DURING HIGH-INTENSITY TRAINING C SARGENT1, S HALSON2, G ROACH1 1 Centre For Sleep Research, University of South Australia, Adelaide, SA, Australia, 2Department of Physiology, Australian Institute of Sport, Canberra, ACT, Australia Introduction: Although sleep is important for exercise performance and recovery, few studies have examined the sleep behaviour of elite athletes. Anecdotally, coaches and athletes suggest that when training volume and/or intensity is high, the level of sleep disturbance is also high. To address this issue, we examined the sleep behaviour of members of the Australian Olympic swimming team during a period of high-intensity training. Methods: Twelve members of the Australian Olympic swimming team, including four women and eight men (age 22 ⫾ 4 years; mean ⫾ SD), underwent 14 days of high-intensity training at the Australian Institute of Sport. Sleep and wake patterns were assessed using wrist activity monitors and sleep diaries. For each training day, athletes recorded their bed/wake times and rated their subjective sleep quality and post-sleep fatigue levels. Results: During the 14 days of high-intensity training, athletes spent 8.7h ⫾ 42 min (mean ⫾ SD) in bed between 20:56 h ⫾ 49 min and 05:53 h ⫾ 72 min. Athletes obtained 6.2h ⫾ 4 min of sleep, with a sleep onset latency of 1.3h ⫾ 24 min and a sleep efficiency of 69 ⫾ 5%. Athletes reported their sleep quality as ‘good’ and post-sleep fatigue levels as ‘a little tired, less than fresh’. Discussion: Despite sufficient sleep opportunity, the amount of sleep obtained by this group of athletes (~6h) during 14 days of high-intensity training was considerably less than the recommended daily target of 8h. This disparity may have several explanations. First, the poor sleep efficiency observed in this group of athletes may be characteristic of Olympic swimmers and independent of the level of training undertaken. This issue may have been resolved if baseline sleep levels were assessed. Second, high-intensity training may increase muscle soreness at night, which in turn, could disrupt sleep. The athletes’ subjective reports of sleep quality, however, do not support this explanation. Finally, the time-of-day at which high-intensity training is performed may affect sleep efficiency. The 6am start times employed in swimming are likely to truncate athletes’ sleep in the morning. While earlier bedtimes may compensate for these awakenings, the ‘forbidden zone’ in the evening makes it is difficult to get to sleep earlier than normal. This is supported by the increased sleep onset latency (>1h) observed in this group of athletes. These findings highlight the importance of scheduling training times that optimise sleep opportunity. Later training start times that extend the sleep period may result in improved training performance.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

P-90 SLEEP INERTIA AND ALCOHOL IMPAIRMENT IN YOUNG ADULTS: NEUROCOGNITIVE EFFECTS AND INTERACTIONS M TOKLEY, M BALL, D BRUCK Victoria University, Melbourne, Australia Introduction: This study investigated the combined decrements of alcohol consumption and sleep inertia in young adults. Cognitions investigated included those required in emergency situations; mental tracking, visual scanning, working memory, and sustained, selective, and divided attention. Methods: The study involved 24 subjects (18–26 years). During the alcohol administration night, 10-minute testing blocks occurred under (i) baseline sober and (ii) baseline.05 blood alcohol content (BAC) conditions. Subsequently, subjects were awoken from stage 4 and assessed in two consecutive 10-minute blocks (iii) and (iv). Subjective measures of sleepiness and clear-headedness were also taken. The same procedure was used during the sober night (with condition (ii) excluded). Results: Mental tracking, visual scanning and attentional functions showed decrements between sober baseline (i) and conditions of alcohol (ii) and further decrements under conditions of sleep inertia (iii) and (iv). When the sober and alcohol nights were compared, no further decrements were found with combined alcohol impairment and sleep inertia. Performance speed on a complex working memory task showed a similar pattern, whilst performance accuracy on this task was adversely affected by alcohol only, with no further decrements with sleep inertia, or combined alcohol impairment and sleep inertia. Hence, conditions of sleep inertia caused a speed-accuracy trade-off on the working memory task. Subjective sleepiness increased whilst subjective clear-headedness decreased with alcohol only and sleep inertia only. However, only subjective sleepiness was further affected by the combination of alcohol and sleep inertia, revealing a synergistic effect. Discussion: Moderate alcohol impairment and sleep inertia do not interact synergistically on neurocognitive functioning to produce further decrements in performance than those caused by the effects of alcohol or sleep inertia alone. Hence, when awoken abruptly in an emergency situation, prior alcohol consumption to.05 BAC will not further impede cognitive functioning that is already compromised by a state of sleep inertia. Alcohol impairment and sleep inertia in combination produce more subjective sleepiness than that reported under conditions of alcohol or sleep inertia alone.

Sleep and Breathing – Treatment P-91 ADVANCEMENT SPLINTS AND UPPER AIRWAY ANATOMY IN OBSTRUCTIVE SLEEP APNOEA A CHAN1,2,3, K SUTHERLAND1,2, R SCHWAB4, B ZENG1,2,3, R LEE1,2,3, MA DARENDELILER5, P CISTULLI1,2,3 1 Centre for Sleep Health and Research, Department of Respiratory Medicine, Royal North Shore Hospital, St Leonards, NSW, Australia, 2 Woolcock Institute of Medical Research, University of Sydney, NSW, Australia, 3Department of Respiratory and Sleep Medicine, St George Hospital, Kogarah, NSW, Australia, 4University of Pennsylvania, Philadelphia, PA, United States, 5Sydney Dental Hospital, University of Sydney, NSW, Australia Introduction: The mechanisms by which mandibular advancement splints (MAS) improve obstructive sleep apnoea (OSA) are not well understood. This study aimed to evaluate their mechanism of action by assessing the effect of MAS on upper airway anatomy in patients with OSA. Methods: Patients with OSA were prospectively recruited for MAS treatment. Magnetic resonance imaging (MRI) of the upper airway was performed in awake supine patients, with and without the MAS. Segmentation of the upper airway lumen and soft tissue structures (tongue, soft palate, fat pads and lateral pharyngeal walls) was performed. Results: MRI was performed in 11 responders (reduction of apnoeahypopnoea index [AHI] by 50% or greater) and 9 non-responders (less than 50% reduction of AHI). At baseline, there were no significant differences in the upper airway volume of responders and nonresponders; and no significant differences in the volume of surrounding soft tissue structures (tongue, soft palate, fat pads and lateral pharyngeal walls). A significant increase in velopharyngeal volume (6.3 ⫾ 0.7 cm3 vs 8.5 ⫾ 0.8 cm3; p < 0.001) and hypopharyngeal volume (8.6 ⫾ 0.7 cm3 vs 9.9 ⫾ 0.7 cm3; p < 0.01), and a reduction in fat pad volume (3.2 ⫾ 0.3 cm3 vs 2.6 ⫾ 0.3 cm3; p < 0.05) occurred in responders following mandibular advancement, but not in nonresponders. An increase in lateral pharyngeal wall volume occurred in both responders (19.3 ⫾ 1.7 cm3 vs 21.4 ⫾ 1.6 cm3; p < 0.05) and non-responders (17.2 ⫾ 1.5 cm3 vs 18.7 ⫾ 1.1 cm3; p < 0.05) following mandibular advancement. Conclusions: These preliminary findings suggest that the effect of mandibular advancement on upper airway volume is not the same in all patients. Mandibular advancement increased the volume of the velopharynx and hypopharynx in responders, but not in non-responders. There were no significant differences in the upper airway anatomy of responders and non-responders at baseline. The significance of the changes observed in some soft tissue structures (fat pads and lateral pharyngeal walls) is unclear and requires further investigation.

P-92 A COMPARISON OF CPAP INTERFACE IN THE CONTROL OF LEAK: NASAL MASK AND CHIN STRAP VERSUS FULL FACE MASK S ROWLAND, C HENNESSY, P CATCHESIDE Adelaide Institute for Sleep Health, Repatriation General Hospital, South Australia, Australia Continuous positive airway pressure (CPAP) is the treatment of choice for obstructive sleep apnoea. Mouth leaks result in a loss of pressure,

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

A55

Healthy Function of Sleep

reducing the therapeutic effect and can cause patient discomfort and potentially reduced compliance. Options for managing leaks are the addition of a chinstrap (CS) if using a nasal system, or use of a full face mask (FFM). Chin straps are a cheaper option than a FFM but there is little evidence to validate their effectiveness in the control of mouth leaks. The aims of this study were to determine if patient compliance, mask/mouth leaks, and patient comfort were improved with the use of a nasal mask with CS, compared to a full face mask. Method: The study was a prospective randomised crossover trial. Randomly selected patients newly referred to a tertiary based sleep unit to commence CPAP were invited to participate in the trial of 3 different mask configurations (nasal mask only; nasal mask and CS; FFM). Each mask type was selected for “best comfortable seal” from a range in the clinic and used for a period of 4 weeks, with the order of use fully randomised. Patients were reviewed and data including compliance, minutes spent with a large leak (>2x expected leak), and a patient comfort questionnaire were collected at 4, 8, and 12 weeks. CPAP compliance and mask leak data were collated for the last 2 weeks of use allowing a 2 week period of acclimatisation to each new mask set up. Patient preference was also recorded. Results: To date 26 patients (17 males, 9 females, mean ⫾ SEM age 57.7 ⫾ 1.9 years, BMI 35.3 ⫾ 1.4 kg/m2) with moderate to severe OSA (AHI 56.9 ⫾ 4.3) have completed the study. An additional 7 withdrew before completion. There were no significant differences in hours of CPAP use or minutes of large leak (ANOVA p > 0.05) but patient satisfaction with the mask and quality of sleep were significantly higher with the nasal mask and nasal mask with CS, than with the FFM (all p < 0.05). Patients also reported fewer problems fitting and keeping the nasal mask in place, and with leaks (Chi2, p < 0.05). Patients preferred the nasal mask over the chin strap and FFM options (N = 16, 6 and 4 respectively, p = 0.008). Discussion: Preliminary data show no significant differences in gross mask leaks or patient compliance between the mask options. However, patients perceive some subjective benefits and are more satisfied with a nasal mask. When significant mouth leaks are suspected, it may be more appropriate and more cost effective to trial a nasal mask with chin strap before a FFM.

P-93 OUTCOMES WITH CPAP AT 2 MONTHS COMPARING OPTIMAL PRESSURE DETERMINED BY ATTENDED LABORATORY POLYSOMNOGRAPHY AND UNATTENDED APAP IN THE HOME C HUKINS, T SATHER, K RIXON, K HANSSEN, G KEIR, T TANG Sleep Disorders Centre, Princess Alexandra Hospital, Woolloongabba, Australia Background: The large burden of disease with obstructive sleep apnoea requires simplified strategies for treatment. Unattended auto-titrating CPAP (APAP) is one strategy to allow for determination in the home of optimal settings for fixed pressure CPAP. Aim: To compare outcomes between laboratory-determined (LabCPAP) to APAP determined treatment settings. Methods: Patients with OSA recommended treatment with CPAP following diagnostic sleep study were randomised to either laboratory or home APAP titration. Epworth Sleepiness Scale (ESS), SF-36 and FOSQ quality of life measures and neuro-cognitive function (Cantab Battery) were measured at baseline and after 2 months of treatment with CPAP. Results: 11 subjects (aged 51.6 ⫾ 12.7 years with RDI 36.6 ⫾ 27.6) have completed the protocol to date. There was no difference in average

A56

CPAP usage at 2 months (Lab-CPAP 6.60 ⫾ 1.17, APAP 4.96 ⫾ 2.68 hours, p = 0.24). ESS improved in each group (Lab-CPAP 12.7 ⫾ 5.65 to 6.71 ⫾ 2.61, p < 0.001 and APAP 13.6 ⫾ 5.53, P < 0.001) but there was no difference between groups. There were improvements in the Total score of the FOSQ in both arms but no difference between treatments (Lab-CPAP 62.3 ⫾ 13.0 to 77.6 ⫾ 9.0, p < 0.05 and APAP 66.6 ⫾ 14.8 to 86.3 ⫾ 14.2, p < 0.05). There were improvements in the SF-36 domains of Vitality (Lab-CPAP 35.1 ⫾ 12.5 to 49.6 ⫾ 13.2 and APAP 43.0 ⫾ 12.4 to 58.2 ⫾ 6.6, both p < 0.05 but no difference between treatment arms) and General Health (Lab-CPAP 35.1 ⫾ 9.9 to 46.7 ⫾ 10.4 and APAP 45.7 ⫾ 15.3 to 55.7 ⫾ 7.4, both p < 0.05 but no difference between treatment arms). There was a trend for improvements in Spatial Working Memory after 8 weeks treatment in the LabCPAP cohort only (reduced Total Errors p = 0.07). Conclusions: CPAP usage and improvements in subjective sleepiness and Quality of Life are similar after 2 months of CPAP whether optimal pressure is determined in the laboratory or unattended with APAP. There is a trend for improvement in Spatial Working Memory in the laboratory titrated cohort only. The use of unattended APAP to determine optimal CPAP settings appears to be valid.

P-94 PREDICTING EFFECTIVE CPAP PRESSURE IN SINGAPOREAN POPULATION WITH OSA USING AN EXISTING PREDICTION FORMULA P SONG, SA RAHIM, TH ONG Singapore General Hospital, Singapore Purpose: To test the validity of an existing prediction formula for predicting effective Continuous Positive Airway Pressure (CPAP) in a Singapore population of patients with Obstructive Sleep Apnea (OSA). Background: A group from Taiwan previously proposed a prediction formula for the minimum effective CPAP pressure needed to treat patients with OSA. As the ethnic Chinese makeup of the Taiwanese population is rather similar to that which we have in Singapore, we hypothesized that this equation would be applicable in our local population also. Methods: A retrospective analysis of polysomnographic data from attended, full-night studies or the titration portion of split night studies carried out at the sleep disorders unit of a tertiary hospital in Singapore. The data collected were applied into a previously validated formula to determine the effective CPAP pressure. [P(eff) = 0.52 + 0.174 ¥ BMI + 0.042 ¥ AHI] Results: 1106 polysomnographic studies were performed in 2007, 203 patients between age 21–90 years underwent a titration study (either full titration study or split-night study) to determine the optimal pressure needed to abolish sleep disordered breathing (SDB). 158 patients (121 males, 37 females) went through manual CPAP titration. The Pearson correlation coefficient value is 0.445 between the formulated and titrated CPAP pressure using the data generated from the 158 patients. We repeated the correlation after excluding non-Chinese patients (n = 131, 69%) and the Pearson correlation coefficient value is 0.438. Conclusions: Correlation between the predicted CPAP pressure generated by the above-mentioned formula and the laboratory-titrated pressure was poor even after excluding the non-Chinese patients in the Singaporean population. Clinical implications: CPAP pressure formula extracted from the Taiwanese population could not be applied to an ethnically similar

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

Abstracts

Singaporean population of OSA patients. Use of prediction formulas for deciding on CPAP pressures must be carefully validated within the intended patient population.

P-96

P-95

M WOODS1, N MCLEAN1, K GODDE3, P EASTWOOD2, D HILLMAN3, J KIRKNESS2 1 School of Psychology, The University of Western Australia, Perth, WA, Australia, 2School of Anatomy and Human Biology, The University of Western Australia, Perth, WA, Australia, 3Western Australia Sleep Disorders Research Institute, Sir Charles Gairder Hospital, Perth, WA, Australia

MILD TO MODERATE OBSTRUCTIVE SLEEP APNOEA AND SYMPTOMS THEREOF ARE IMPROVED BY SIGNIFICANT WEIGHT LOSS, A PROSPECTIVE, RANDOMISED CONTROLLED TRIAL IN PROGRESS S HOMAN1, A THORNTON3, M NOAKES2, AM SOUTHCOTT1, B SMITH1, A VEALE1 1 Queen Elizabeth Hospital, Woodville, S.A., Australia, 2CSIRO Human Nutrition, Adelaide, S.A., Australia, 3Royal Adelaide Hospital, Adelaide, S.A., Australia Introduction: Weight loss is predicted to reduce the AHI of patients with OSAHS. In the mild to moderate range of OSA weight loss may also have a beneficial effect on symptoms of sleepiness. Methods: A randomised trial comparing the effect of weight loss on severity and symptoms of OSA compared to the change in symptoms using CPAP. Overweight patients having undergone a diagnostic PSG study that demonstrated mild to moderate OSA, with significant sleepiness without co morbidity affecting ability to diet were recruited. Subjects were randomised to treatment arms; Control (C), CPAP (CP), Diet (D) and D&CP, for 8 weeks then followed by repeat PSG, questionnaires, blood screen and weight and girth measurement. CP used Autoset (Resmed), weight loss was achieved using partial meal replacements (Kicstart). Results: The 34 patients; C 10, CP 9, D&CP 7, D 8, that have completed the study were well matched for BMI. Significant weight loss was achieved in the D group (6%) and the D&CP group (3%). ESS decreased in the CP (-5, P = 0.005), D&CP (-9, P = 0.004) and D (-4, P = 0.048). AHI decreased only in the D&CP (17/hour→ 10/hour, P = 0.024). Combining the results of all groups, change in AHI with weight correlated poorly (r2 = 0.12) however when only those with a substantial weight change (>⫾5%) were considered the relationship was strong (r2 = 0.43). Discussion: There are small numbers in each group therefore caution is required in drawing conclusions from these results; the data of individuals may distort results significantly. Change in weight (ⱖ5%) was associated with AHI change. Improvement in ESS was not associated with AHI outcome but with treatment group. With the small number of patients recruited to date, the additional benefit on ESS seen in the D&CP group was not proven to be an additive effect of weight loss and CPAP prescription and it may be surmised that individuals’ preferences for D or CP led to greater compliance benefit with one aspect of the treatments offered.

A NOVEL THERAPY FOR SLEEP DISORDERED BREATHING AND THE RELATIONSHIP OF SELF-EFFICACY AND OUTCOME EXPECTATIONS WITH PERCEIVED SLEEP QUALITY AND TREATMENT ACCEPTABILITY

Continuous positive airway pressure (CPAP) is the standard treatment for obstructive sleep apnoea. High Nasal Flow therapy (HNF) may improve treatment adherence due to its lighter, less intrusive design. This study explores perception of sleep quality and treatment acceptability, and their association with self-efficacy and outcome expectations. Method: Nine patients (7 male, 2 female) with mild-to-moderate obstructive sleep apnoea (apnoea-hypopnoea index, AHI = 15 ⫾ 4 events/hour), attended the sleep laboratory for two overnight sleep studies. On one night they received CPAP (Autoset 6–16 cm H2O) and on the other night, HNF therapy (20–24 L/min) in random order. In addition to polysomnographic recordings, a morning questionnaire, treatment acceptability questionnaire, and scales of self-efficacy and outcome expectations were administered following each treatment study. A score for the perception of sleep quality was created from five questions in the morning questionnaire. Multivariate regression analysis was used to compare measurements of combined data obtained on the two nights as well as stratified by treatment. P < 0.05 was considered significant. Results: Measures of sleep quality outcomes (self-efficacy, outcome expectations, perception of sleep quality and treatment acceptability) were similar following HNF and CPAP. In multivariate analysis both self-efficacy (R2 = 0.79, p < 0.001) and outcome expectations (R2 = 0.72, p = 0.002) were associated with the combination of perception of sleep quality and treatment acceptability in the final model. Self-efficacy and outcome expectations were more strongly associated with perception of sleep quality and treatment acceptability using high nasal flow (R2 = 0.90, p < 0.01 & R2 = 0.87, p < 0.01, respectively) than continuous positive airway pressure (R2 = 0.65, p = 0.21 & R2 = 0.76, p = 0.12, respectively). Conclusion: Patients who perceive good sleep and positive experience with a treatment will be more confident in their ability to use the treatment and in its effectiveness. Further investigation of the effect of high nasal flow therapy on sleep quality may provide additional insight into factors that influence treatment adherence in obstructive sleep apnoea.

P-97 THERAPEUTIC STRATEGY IN ATTENUATING OBSTRUCTIVE SLEEP APNOEA BY SENSORY NEUROLOGICAL STIMULATION MODALITIES M DAULATZAI, A KHANDOKER The University of Melbourne, Melbourne, Victoria, Australia Obstructive sleep apnoea (OSA) is a disorder in which anatomic and physiologic perturbation is correlated with sleep-disordered breathing.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research

A57

Healthy Function of Sleep

Muscles of the oropharynx that normally keep the upper airway (UA) patent during wakefulness decrease their tone, and supine body position makes it easier for the tongue to fall backward and further narrow the upper airway. Since waking drive to breathe is lost, the chemoreceptor response is reduced. Patients with OSA have alterations in the mechanical properties of the UA and a compensatory neuromuscular responses. Thus what is paramount here is a need to control UA patency. The ability to ventilate is dependent on the integrity of medullary respiratory neurons, chest wall muscle contraction, and indeed the patency of the UA. Of the structures that comprise the upper respiratory tract, the “pharynx” is the only structure susceptible to collapse. Pharyngeal dilator muscles maintain UA patency during wakefulness and sleep. Continuous neuromuscular stimulation from medullary cranial motor neurons sustain pharyngeal airway patency during wakefulness. Conversely, during sleep in OSA, there is a reduction in the motoneuron output to the UA muscles, resulting in UA narrowing. Unlike normal, in OSA subjects, the balance of pressure within, across, and external to the airway is altered, thus altering the anatomy and physiology, and producing the airway obstruction. A remarkable feature across clinical

A58

studies was the observation that coexisting neuropathy increases the prevalence of sleep apnea, CCHS is characterized by diminished nocturnal drive to breathe. They do not alter respiratory rate to a forehead cold pressor challenge and show diminished respiratory-related heart rate variation. This reflects “nonchemoreceptor” respiratory reflex deficit in the syndrome, and indicates failure of integration of respiratory motor output with afferent input. There is an analogous situation in OSA. Consistently, topical UA anesthesia by tetracaine led to an increase in apnoea duration. This study utilizes different sensory modalities including air puff, cold and warm water stimulation to enhance nerve regeneration, stimulate pharyngeal muscle’s response to stimuli, and hence control OSA. Further, this strategy is simple and non-invasive and does not include chemoreceptor components. Thus sensory stimulation of the UA is a viable strategy in OSA. OSA appears to have diminution in sensory function; UA mucosa which has a rich sensory innervation is amenable to stimulation. The present hypothesis is that stimulation of UA sensory receptors during obstructed inspiratory efforts contributes to apnoea termination.

© 2008 The Authors Journal compilation © 2008 Japanese Society of Sleep Research