Article Text

Download PDFPDF

Sleep disorders in practice

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


1R. H. Mason, 2S. D. West, 3V. Chong, 1J. R. Stradling. 1The Churchill Hospital, Oxford, UK, 2Freeman Hospital, Newcastle upon Tyne, UK, 3The Eye Hospital, Oxford, UK

Introduction and Objectives Diabetic retinopathy is an important cause of visual loss worldwide. We and others recently showed that retinopathy was more common in a group of patients with diabetes if they also had obstructive sleep apnoea (OSA).1 2 Although the numbers were small, maculopathy was also related to the presence of OSA.2 This study was designed to see if OSA was also related to the presence of diabetic macular oedema (DME), which is at the more severe end of the diabetic retinopathy spectrum, poorly responsive to treatment and the most common cause of blindness in patients with diabetes.

Methods 62 patients who had received laser therapy for DME provided anthropometric data, filled in the Epworth Sleepiness Score (ESS) and were screened with a home sleep study (ApnoeaLink, ResMed) to identify the presence of OSA (oxygen desaturation index (ODI), apnoea–hypopnoea index (AHI) and snoring). These results were compared with available prevalence figures from relevant control populations.

Results 58 patients (27 males) had >2 h of valid overnight data. Average (SD) age 62.9 (10.6), neck circumference 40.6 cm (5.3), body mass index (BMI) 30.9 (6.2), HbA1c 7.7% (1.4), ESS 7.5 (4.7). The prevalence of an ODI ⩾10 was 53%, that of AHI ⩾15 was 47%, and 57% had either or both. These prevalences are considerably higher than in any of the control data, 6% in a randomly selected population,3 and 17% in randomly selected patients with type 2 diabetes.4 Those with OSA were not significantly sleepier than those without, but they were older and more obese.

Conclusions Individuals with DME show a much higher prevalence of OSA (defined via ODI or AHI) than both the general population and unselected patients with type 2 diabetes. This hypothesis-generating study suggests that OSA might contribute to DME, but a continuous positive airways pressure (CPAP) intervention study will be required to prove cause and effect.



B. Jayaraman, P. Farrow. Kent & Canterbury Hospital, Canterbury, UK

Background Abnormal thyroid function is recognised to be associated with obstructive sleep apnoea syndrome (OSAS), but both the SIGN (Scottish Intercollegiate Guidelines Network) guidelines and NICE (National Institute for Health and Clinical Excellence) technical appraisal on OSAS recommend that thyroid function tests (TFTs) may be considered rather than mandatory in such patients. Studies supporting the frequency of abnormal TFTs in patients with OSAS are limited in the literature; therefore, we have performed a review of our experience in investigating OSAS patients from a single institution over a decade.

Method The study was performed in a single sleep-breathing unit, between 1998 and 2008, in investigating patients with OSAS, who were recommended to undergo a trial of continuous positive airways pressure (CPAP) by the multidisciplinary team. A total of 679 patients were found to be eligible and, in addition to their demographics, the presence or absence of normal or abnormal TFTs were collected. 62% of the patients (422/679) included in the study had their TFTs measured and the results were obtained from this cohort of patients.

Results 6.6% (22) were diagnosed to have biochemical hypothyroidism and were already on thyroxine replacement, and 1.4% (6) patients were picked up on screening. 3.1% (13) were diagnosed to have biochemical hyperthyroidism. Male sex (83%; mean age 57 years) was found to be an another significant factor in this study with abnormal TFTs.

Discussion Abnormal (hypothyroidism (6.6%) and hyperthyroidism (3.1%)) TFT results were considerably more common (9.7%) in our patients than for reported spontaneous background incidence rates (1–2% and 0.5–2%, respectively) in the UK. The disparity is more marked especially as these overall reported spontaneous background rates are 10 times higher in females, whereas in our study the affected patients were predominantly male. Previous reports have shown the association of abnormal thyroid function and OSAS, but have not led to definitive testing recommendations. We conclude that the substantial excess rates of thyroid disease within our patients with OSAS are sufficient to warrant mandatory thyroid function testing and that it needs to be included in the national guidelines.

Abstract S19 Table


T. Aung, S. Bianchi. Royal Hallamshire Hospital, Sheffield, UK

Diagnosis of obstructive sleep apnoea requires clinical assessment and objective demonstration of sleep disordered breathing (SDB). The polysomnographic definition of apnoea and hypopnoea has recently been updated by the American Academy of Sleep Medicine (AASM). We were interested in how this change affects our clinical practice in terms of diagnosis and treatment decisions.

Objectives To compare apnoea–hypopnoea indices (AHIs) derived using two apnoea/hypopnoea definitions published by the AASM (AHIold and AHInew) and to examine the impact of recent definition change on the prevalence and severity of SDB.

Method Retrospective review of 38 limited polysomnograms with re-scoring using modified AASM guidelines.

Definitions Previous AASM scoring criteria (AHIold) define hypopnoea as a >50% airflow reduction from baseline or a lesser airflow reduction with associated >3% oxygen desaturation or arousal. The updated AASM Manual for Scoring of Sleep and Associated Events (AHInew) defines hypopnoea as a ⩾50% airflow reduction from baseline and ⩾3% desaturation or arousal.

Results The AHIold was 17.02±5.19. The AHInew was 7.29±3.48, representing a 57% reduction. 15/38 (39%) patients previously classified as positive for SDB lowered their AHI to <5 (normal). All patients (23/38) previously diagnosed with moderate SDB were reclassified as mild. Pearson correlation coefficient was used to assess the correlation between AHIs derived from different hypopnoea definitions, and the correlation value (r) is 0.64 and the p value is 0.0001, which is statistically significant. Based on AHIold, all patients classified as moderate SDB were offered continuous positive airway pressure (CPAP) therapy. Only 16% of this group were compliant with CPAP treatment at 12 months.

Conclusions This study demonstrates that the new AASM definitions used to calculate AHI have led to marked change in diagnosis and clinical practice. The compliance data suggest that the new scoring guidelines may be more appropriate in severity scoring and choice of first-line therapy.

Abstract S20 Table



1R. H. Mason, 1M. Hardinge, 2J. M. T. Perkins, 3M. Kohler, 1J. R. Stradling. 1The Churchill Hospital, Oxford, UK, 2John Radcliffe Hospital, Oxford, UK, 3University Hospital of Zurich, Zurich, Switzerland

Introduction and Objectives Abdominal aortic aneurysms (AAAs) are associated with life-threatening complications such as thrombosis and rupture. Apart from smoking, the reasons for a more rapid rate of expansion and subsequent rupture are largely unknown. Therefore any potentially modifiable risk factors affecting rate of expansion and improving mortality and morbidity are of great importance. We have shown previously that there is an association between thoracic aortic root size in individuals with Marfan syndrome and degree of obstructive sleep apnoea (OSA).1 We therefore wished to identify whether the same relationship might exist between AAA and OSA.

Methods 86 patients (aged 18–75 years), under surveillance for an AAA (⩾3 cm), had at least two ultrasound measurements of aortic diameter over a minimum time interval of 6 months. Rates of expansion were calculated. In addition, anthropometric data, Epworth Sleepiness Score (ESS), HbA1c and cholesterol were recorded. The degree of OSA ((oxygen desaturation index (ODI), apnoea–hypopnoea index (AHI) and snoring)) was quantified with a home sleep study (ApnoeaLink, ResMed); these results were compared with available prevalence figures from relevant control populations.

Results 76 patients (70 males) had >2 h of valid overnight data. Average (SD): age 68.7 (5.7), neck circumference 41.5 cm (4.0), body mass index (BMI) 28.7 (4.9), ESS 7.1 (3.7), HbA1c 5.9 (0.9), cholesterol 4.2 (1.0), current aortic diameter 4.3 cm (0.9), rate of expansion 2.2 mm/year (2.3). The prevalence of an ODI ⩾10 was 36%, that of AHI ⩾15 was 36%, and 44% had either or both. These prevalences are higher than those identified in similar age-matched control groups, estimated to be ∼20%.2 Neither current diameter nor rate of expansion of the AAA was significantly correlated with the degree of OSA (ODI or AHI).

Conclusions About twice as many individuals with AAAs had evidence of OSA compared with appropriately matched control populations. Despite this higher prevalence, there was no suggestion that OSA influenced the rate of expansion. Presumably similar risk factors influence the prevalence of OSA and AAA.


A. C. Hutchings, J. C. T. Pepperell, H. Potter. Musgrove Park Hospital, Taunton, UK

Introduction and Objectives High levels of noise in hospitals cause sleep deprivation and increase stress response in patients. Noise increases error rate and makes communication more difficult amongst staff. The World Health Organization suggests that patients should not be exposed to noise levels >35 dB. We measured noise levels on a general medical ward, and aimed to reduce noise through simple interventions. We surveyed patients’ perception of noise levels before and after these interventions.

Methods Baseline noise levels were measured as 30 s epochs over 2 weeks through day and night periods, and again postinterventions. The interventions were: behaviour modification through a staff and visitor education programme, quiet-closing bins, closing the doors to the bays at night and restricting visitors to designated times. Patients completed a questionnaire on the monitored ward area before and after interventions. Patients scored their perception of noise across multiple situations. Thematic analysis identified the most disruptive sources of noise. Decibel levels were analysed before and after interventions using Student t tests. Patients’ perceptions of noise were analysed using Mann–Whitney U test. Results are presented as mean (SD); p<0.05 was taken as statistically significant.

Results At baseline noise levels exceeded the recommended guidelines by >25 dB (mean noise in day 65.0 dB (SD ±1.31), mean noise at night 62.5 dB (SD ±0.9)). Following interventions, mean noise levels at night fell significantly to 61.6 dB (SD ±1.1, p = 0.04). We significantly reduced the amount of time that noise levels exceeded 60 dB at any time from 82% to 76% (p = 0.04), with the greatest impact at night (75% to 59% p = 0.02). Patients reported a strong trend towards less disruption from visitors and a trend towards improved sleep quality following the intervention. Thematic analysis identified noise from mobile telephones as a particular issue.

Conclusions We determined that noise levels on a medical ward exceed recommended guidelines and that this has a negative impact on the patient experience. We have shown that simple, cost-effective measures result in quieter wards, especially at night.

View Abstract