DIABETICMedicine DOI: 10.1111/j.1464-5491.2010.02962.

Original Article: Complications The prevalence of retinopathy in men with Type 2 diabetes and obstructive sleep apnoea
S. D. West, D. C. Groves, H. J. Lipinski, D. J. Nicoll, R. H. Mason, P. H. Scanlon and J. R. Stradling
Oxford Centre for Respiratory Medicine, Churchill Hospital and Oxford Eye Hospital, John Radcliffe Hospital, Oxford, UK Accepted 4 January 2010

Abstract
Aims

To clarify the relationship between obstructive sleep apnoea (OSA) and diabetic retinopathy.

Research design and methods

A cohort of 240 men from primary and secondary care previously participated in a study on the prevalence of OSA in Type 2 diabetes and provided anthropometric information, details of their diabetes, had glycated haemoglobin (HbA1c) measured and overnight oximetry performed. They were re-contacted for permission to review their routine screening clinical retinal photographs, which were then scored by a trained grader, providing detailed retinopathy, maculopathy and photocoagulation scores.

One hundred and eighteen men both consented and had retinal photographs available to review. Of these, 24% had OSA, with mean sd 4% oxygen saturation (SaO2) dips h of 20.9 16.6 vs. 2.8 2.1 in the non-OSA group. As expected, the OSA group had a signicantly higher mean body mass index of 31.9 5.2 vs. 28.5 5.1 kg m2 and neck size 44.5 3.6 vs. 41.9 2.5 cm, but the two groups did not differ signicantly in age, diabetes duration, diabetes treatment, HbA1c, smoking history or proportion with known hypertension. Retinopathy and maculopathy scores were signicantly worse in the OSA group (P < 0.0001). Multiple regression analysis showed only OSA (R2 = 0.19, P < 0.0001) and HbA1c (R2 = 0.04, P = 0.03) to be signicant independent predictors of retinopathy. OSA was the only independent signicant predictor of the total microaneurysm score (R2 = 0.21, P = 0.004), a detailed retinopathy subclassication. OSA was the only independent signicant predictor of maculopathy (R2 = 0.3, P < 0.001).
Results Conclusion

In men with Type 2 diabetes, there is a strong association between retinopathy and OSA, independent of conventional retinopathy risk factors.

Diabet. Med. 27, 423430 (2010)

Keywords Embletta, obstructive sleep apnoea, oximetry, retinopathy, Type 2 diabetes Abbreviations AHI, apnoeahypopnoea index; BMI, body mass index; CPAP, continuous positive airway pressure;

ESS, Epworth Sleepiness Score; HbA1c, glycated haemoglobin; OSA, obstructive sleep apnoea; SaO2, oxygen saturation; sd, standard deviation

Introduction
Obstructive sleep apnoea (OSA) is a common condition, affecting an estimated 4% of men and 2% of women [1]. It is mainly caused by central obesity, with a strong correlation between neck circumference and severity of OSA [2]. OSA is characterized by recurrent episodes of upper airway obstruction

Correspondence to: Sophie D. West, Cardiothoracic Centre, Freeman Hospital, Newcastle upon Tyne NE7 7DN, UK. E-mail: Sophie.West@nuth.nhs.uk

during sleep, apnoea, hypoxia, increased inspiratory effort and subsequent arousal to terminate the apnoea, leading to fragmented sleep and daytime sleepiness. Each arousal is associated with a burst of sympathetic activity, which causes immediate transient rises in pulse and blood pressure (Fig. 1). OSA is associated with diurnal hypertension, insulin resistance, Type 2 diabetes, the metabolic syndrome, and therefore an increased cardiovascular risk, all to some extent independent of obesity [37]. Retinopathy is a microvascular complication of diabetes, which is the commonest cause of blindness in the working age

2010 The Authors. Journal compilation 2010 Diabetes UK. Diabetic Medicine, 27, 423430

423

DIABETICMedicine

Diabetic retinopathy and obstructive sleep apnoea S. D. West et al.

FIGURE 1 Tracing (left to right) of blood pressure (Finapres, tracing equivalent to intra-arterial line), pulse rate, oxygen saturation and air ow, over a 5-min period in a person with severe obstructive sleep apnoea (OSA). Obstructive apnoeas are shown by periods of air ow cessation, pulsus paradoxus on the blood pressure tracing and subsequent oxygen desaturation; arousals required to terminate the apnoea are indicated by normalization of oxygen saturation, pulse rate rises and transient blood pressure increases. The process is repetitive, occurring hundreds of times per night.

population in the UK [8]. The UK Prospective Diabetes Study (UKPDS) found that improved glucose and blood pressure control both reduced the risk of retinopathy progression [9,10]. Other potentially modiable risk factors are elevated serum lipids [11] and smoking, although the role of smoking in Type 2 diabetes is controversial [10]. Non-modiable risk factors include duration of diabetes [12], a complex relationship with age at diagnosis [1013], a genetic predisposition in some families [14] and an association with ethnicity [15,16]. Routine annual screening is performed in the UK (http://www. retinalscreening.nhs.uk), to identify diabetic retinopathy early and thus allow prompt referral and treatment of those with sightthreatening features. Our previous work has shown a high prevalence of OSA in men with Type 2 diabetes, with an estimated 23% having OSA, based on questionnaire responses and overnight sleep studies [17]. Another study has suggested rates of up to 86% in obese people with Type 2 diabetes [18]. The relationship of OSA with retinopathy has not, however, been previously well established. We hypothesized that the frequent and transient episodes of both intermittent hypoxia and blood pressure surges seen in OSA could cause retinal damage and worsening of diabetic retinopathy.

Subjects and methods

Subjects

invited them to complete the Berlin questionnaire, which predicted whether they were at high or low risk of OSA [19]. It included the question Do you have high blood pressure? and asked for their height and weight in order to calculate body mass index (BMI). We also asked participants to measure their neck size. Further questions regarding the duration of diabetes and type of treatment were asked (insulin oral glucose-lowering agents insulin plus oral glucose-lowering agents diet alone). The most recent glycated haemoglobin (HbA1c) result, within 3 months of the sleep study, was recorded for all questionnaire respondents. We performed home overnight oximetry sleep studies in 140 subjects from the high risk for OSA group and 100 subjects from the low risk group (out of a total 938 subjects who returned questionnaires). Subjects from each group were selected on the basis of their postcodes, with those within 20 miles of the base hospital being preferred, for ease of equipment delivery. A small portable battery-operated, wrist-worn monitor with attached nger probe was used (Pulsox-3i; Konica Minolta, Tokyo, Japan). This continuously measures the oxygen saturation and the pulse rate, which can be downloaded via a computer and analysed (Stowood Scientic Instruments, Oxford, UK)., A systematic literature review of home diagnosis of sleep apnoea, found that a normal overnight oximetry tracing substantially reduced the probability of nding OSA on a more comprehensive sleep study in the vast majority of patients [20]. Subjects also completed the Epworth Sleepiness Score (ESS).

In a previous study conducted in 20042005, we wrote to all men with Type 2 diabetes, aged 1875 years, on the database of the Oxford Centre for Diabetes, Endocrinology and Metabolism (a secondary care, hospital-based centre) and also all men with Type 2 diabetes on ve local primary care general practice databases (total n = 1676) (Fig. 2) [17]. We

Diagnosis of obstructive sleep apnoea

Those subjects who had more than 10, > 4% oxygen saturation (SaO2) dips h on oximetry, consistent with a diagnosis of signicant OSA, or those who had fewer than

424

2010 The Authors. Journal compilation 2010 Diabetes UK. Diabetic Medicine, 27, 423430

Original article

DIABETICMedicine

Screening questionnaires sent to 1676 men with type 2 diabetes from local hospital and selected general practice databases

938 Replies received (56%)

528 (56%) high risk of OSA

362 (39%) low risk of OSA

39 (4%) Known OSA

9 (1%) Incomplete

Screening oximetry performed in 124 high risk men

Screening oximetry performed in 116 low risk men Current study 240 men who had screening oximetry in 2004-5 written to in 2008 to seek consent to view routine screening retinal photographs and correlate results with previous sleep studies

38 High risk men had OSA on oximetry (31%)

15 Low risk men had OSA on oximetry (13%)

More detailed home sleep study performed in those with OSA

36 Had study: 33 confirmed OSA

11 Had study: 11 confirmed OSA

118 Consented

All men with OSA offered opportunity to be reviewed in Sleep Clinic

90 No OSA (76%) 31% with retinopathy 28 OSA (24%) 54% with retinopathy

16 Men commenced CPAP

1 man commenced CPAP

FIGURE 2 Flow diagram of study. CPAP, continuous positive airway pressure; OSA, obstructive sleep apnoea.

10 > 4% SaO2 dips h, but had an oxygen saturation tracing consistent with a diagnosis of OSA when viewed by an expert (JRS, blinded to retinopathy status), were contacted in order to have an unattended portable multichannel monitor home sleep study to verify the diagnosis of OSA. The sleep study equipment measured body position, body movement, nasal pressure via nasal cannula, oximetry, pulse rate, plus respiratory effort via thoracic and abdominal bands (Embletta PDS 3.0; Flaga Medical, Reykjavik, Iceland) [21]. From these measurements, a validated apnoeahypopnoea index (AHI) and oxygen desaturation events h were automatically calculated, with manual editing of areas of artefact. More than one trace was used in order to validate the AHI, to eliminate any possible false positive results for intermittent apnoeas caused by people intermittently mouth breathing. All of those subjects who were found to have OSA on sleep studies were offered a clinic appointment to discuss the results and determine whether they required treatment.

Retinopathy grading

For the current study, we wrote again in 2008 to the cohort of 240 men who had been previously studied. We sought written consent to look at their digital retinal photographs, taken for annual retinopathy screening by the Oxfordshire Diabetic Eye Screening Service as part of their diabetes care. This service offers

digital photographic screening to all people with diabetes, unless they are already known to be under the Oxford Eye Hospital for monitoring or treatment of diabetic retinopathy; these Eye Hospital clinical retinal images use a different technique and were therefore not suitable for inclusion in this study. The screening images were all taken on 45 non-mydriatic cameras. All patients had their eyes dilated with 1% tropicamide prior to photography and two images of each eye were taken. The rst image was centred on the macula and the second centred on the optic disc. Non-responders were contacted by telephone to encourage participation. In those men who gave consent, a single trained ophthalmology grader (DCG) reviewed their digital images, blind to OSA status, and graded them according to the criteria shown in Table 1. The images were regraded by a second trained individual, also blind to OSA status and the original grading scores, to ensure accuracy and validity. Where there were discrepancies between the graders, a third trained grader (HJL) acted as arbitrator. The grading structure involves using the overall grades recommended by the English National Screening Programme of no diabetic retinopathy (R0), background (R1), pre-proliferative (R2) and proliferative retinopathy (R3), but subcategorizing these more closely in areas where the national grading form leaves it more open to individual interpretation [22]. Maculopathy and prior photocoagulation were graded as absent (M0, P0) or present (M1, P1). Sight-threatening diabetic

2010 The Authors. Journal compilation 2010 Diabetes UK. Diabetic Medicine, 27, 423430

425

DIABETICMedicine

Diabetic retinopathy and obstructive sleep apnoea S. D. West et al.

Table 1 Detailed method of grading diabetic retinopathy, including subcategories, performed on each eye, in the English National Screening Programme

DD, disc diameter; IRMA, intra-retinal microvascular abnormalities; M, maculopathy; P, photocoagulation; R, retinopathy.

retinopathy is dened as pre-proliferative retinopathy (R2) or proliferative retinopathy (R3) or sight-threatening maculopathy (M1). The presence of photocoagulation scars was recorded, but the retinopathy and maculopathy level was determined by the retinopathy lesions seen and was not altered by the presence of photocoagulation scars. Visual acuity was measured with a Snellen chart. Grading and acuity from the worst eye determined the overall score. The most recent HbA1c of responders was documented.

logistic regression analyses were performed. A P value of < 0.05 was considered to be statistically signicant. Both the initial study and this study obtained local ethics committee approval (OCREC 03.091 and H0603 13).

Results
Of the 240 men who previously had overnight sleep studies, 155 gave written consent to have their digital retinal photographs reviewed by the study investigators. Five men refused consent, 51 men did not reply despite reminders, 10 had moved and were not contactable and 19 had died since the original study. There was no statistically signicant difference between the responders and the non-responders in BMI, neck size, duration of diabetes, 4% SaO2 dip rate on sleep study or ESS. Of the 155 who gave consent, 118 had digital photographs available to view. There were patients who had not had screening photographs taken because they were either known to attend the Eye Hospital (n = 12) or because they had not attended for screening photographs (n = 25). For the 118 participants with appropriate photography, their previous sleep study data were accessed and they were divided into two groups: those with signicant OSA on a sleep study (24%) and those without (76%). The characteristics of these two groups are shown in Table 2. Of those with signicant OSA, 11 (39%) went on to commence continuous positive airway pressure (CPAP) treatment following

Analysis

Analysis was by spss version 14.0 (SPSS Inc., Chicago, IL, USA). All data are expressed as mean standard deviation (sd). Kappa values were calculated to measure the inter-rater agreement between retinopathy graders. Differences between the non-OSA and OSA groups were assessed with unpaired t-tests and chisquare tests with Fishers exact test when comparing proportions. Ninety-ve per cent condence intervals of the differences between the groups were determined. The full microaneurysm score for both eyes summed together (07 for each eye) was also used in the results, with a division of people into either 010 total or 1114 total for 2 2 chi-square testing. Pearsons correlation coefcient was used to determine the relationship of retinopathy with different variables. In order to determine the independent predictors of retinopathy, stepwise forward multiple linear and

426

2010 The Authors. Journal compilation 2010 Diabetes UK. Diabetic Medicine, 27, 423430

Original article

DIABETICMedicine

Table 2 Characteristics of patients with digital photographs, grouped according to whether obstructive sleep apnoea was present on the sleep study or not Non-OSA n = 90 Age (years) BMI (kg m2) Neck size (cm) Duration of diabetes (years) Diabetes treatment (%) Diet only Oral glucose-lowering agents Insulin and oral agents Insulin only Known hypertension (%) HbA1c % Smoking history Never smoked (%) Ex-smoker (%) Current smoker (%) 4% SaO2 dips h Mean nocturnal SaO2% ESS 66.3 10.2 28.5 5.1 41.9 2.5 9.7 7.0 10 57 14 19 58 7.7 1.2 36 61 3 2.8 2.1 94.9 1.4 7.5 3.8 OSA n = 28 67.2 31.9 44.5 10.4 8.8 5.2 3.6 8.2

P value 0.7 0.003 < 0.0001 0.7 0.2

7 43 32 18 71 8.3 1.8 32 57 11 20.9 16.6 91.9 4.4 10.0 4.9

0.2 0.6 0.3

< 0.0001 < 0.0001 0.004

The two groups were compared with unpaired t-tests or chi-square tests when comparing proportions. Mean sd or %. BMI, body mass index; ESS, Epworth Sleepiness Score; HbA1c, glycated haemoglobin; OSA, obstructive sleep apnoea; SaO2, oxygen saturation; sd, standard deviation.

their original sleep study and clinical review if they were found to have signicant daytime sleepiness necessitating treatment.

Patient characteristics

venous beading, new vessels at the disc or elsewhere, brous proliferation and pre-retinal or vitreous haemorrhage. There was no signicant difference between the groups regarding visual acuity. There was good inter-rater agreement between retinopathy graders (weighted kappa = 0.74).

As expected, the OSA group had a signicantly higher BMI and neck size than the non-OSA group. There were no signicant differences between the groups in the duration of diabetes, the type of diabetes treatment, HbA1c (either at the time of the original questionnaire study or at the time of the more recent study, P = 0.5), the proportion with known hypertension and the smoking history. As expected, the OSA group had signicantly more > 4% SaO2 dips h on their overnight sleep study, signicantly lower mean overnight SaO2, and signicantly worse ESS.

Multiple regression analysis

Retinopathy

The OSA group had signicantly more retinopathy, maculopathy and microaneurysms than the non-OSA group (Table 3). Of the more detailed grading subcategories, based on the mean of both eyes, the OSA group had a signicantly higher percentage of people with the following abnormalities: haemorrhages inside and outside the arcades, haemorrhages less than one disc diameter from the fovea, hard exudates inside the arcades and cotton wool spots inside the arcades. There was no signicant difference between the groups in: hard exudates outside the arcades, cotton wool spots outside the arcades, intraretinal microvascular abnormalities inside or outside the arcades,

Univariate amalysis showed signicant associations between retinopathy and the following: HbA1c (r = 0.3, P < 0.0001), duration of diabetes (r = 0.2, P = 0.04), presence or absence of OSA (r = 0.2, P = 0.04) and 4% SaO2 dip rate (r = 0.3, P < 0.0001). There was no signicant correlation with BMI or known hypertension. In multiple regression analyses, OSA remained a signicant predictor for all measures of retinopathy, except prior photocoagulation. Multiple linear regression for retinopathy scoring showed 19% of the variance was explained by OSA and 4% by HbA1c (R2 = 0.19 and 0.04, respectively, P < 0.0001 and P = 0.03). Age, BMI, duration of diabetes and known hypertension were not statistically signicant independent variables. Multiple logistic regression for maculopathy showed OSA accounted for 30% of the variance (R2 = 0.30, P < 0.0001). There were no other signicant independent variables. Multiple logistic regression analysis of total microaneurysm score showed 21% of the variance was accounted for by OSA, which was the only statistically signicant predictor (R2 = 0.21, P = 0.004). Having OSA more than tripled the likelihood of having a combined microaneurysm score greater than 10.

2010 The Authors. Journal compilation 2010 Diabetes UK. Diabetic Medicine, 27, 423430

427

DIABETICMedicine

Diabetic retinopathy and obstructive sleep apnoea S. D. West et al.

Table 3 Grading of diabetic retinopathy in non-OSA and OSA groups Non-OSA n = 90% of people with lesion Retinopathy R1 R2 R3 Maculopathy M1 Photocoagulation P1 Microaneurysm score (both eyes combined) 1114 Visual acuity Haemorrhages inside the arcades Haemorrhages outside the arcades Haemorrhages less than one disc diameter from the fovea Hard exudates inside the arcades Cotton wool spots inside the arcades OSA n = 28% of people with lesion

95% condence interval of the difference

P value

27 4 0 4 2 8 6 9.5 (9.5) 12 16 4 3 4

18 36 0 29 11 26 6 9.6 (5.6) 42 42 27 31 23

< 0.0001

7 to 41 )3 to 20 1 to 37

< 0.0001 0.09 0.02 > 0.99 0.001 0.006 0.001 < 0.0001 0.008

10 to 50 5 to 46 5 to 40 10 to 50 2 to 35

Results are expressed as a percentage of men in each group with levels of retinopathy (R1R3), maculopathy (M1), photocoagulation (P1) and total microaneurysm score (both eyes combined) of greater than 10. Visual acuity is shown as 6 x, with x being the mean (sd). Detailed retinopathy grading subcategories (using the mean of both eyes) are also shown. The two groups were compared with chi-square tests (with Fishers exact test) to compare proportions and an unpaired t-test for visual acuity. 95% condence intervals of the differences between groups are shown. OSA, obstructive sleep apnoea; sd, standard deviation.

Discussion
This study suggests a signicant association between obstructive sleep apnoea and diabetic retinopathy in men with Type 2 diabetes. The men with OSA had mean body mass indices and neck circumferences which were larger than those of the non-OSA group, but the relationship between OSA and retinopathy was independent of the potential confounders of BMI, age, duration of diabetes, known hypertension and HbA1c. Men with more severe retinopathy requiring hospital treatment were excluded as their digital retinal photographs were not the same as those in the retinopathy screening programme, with the macular image being non-compatible; they represent an important group to include in further studies of this nature. The presence of photocoagulation scars did not alter the retinopathy score in this study; these were present in three men in the OSA group (11%) and two of the non-OSA group (2%). As photocoagulation could potentially lower the retinopathy score by treating the affected retinal areas, these results were reviewed. Of the ve people with photocoagulation scars, three had sight-threatening maculopathy with R2 M1 scores. One person in each group had R1 scores, a score indicating potentially non-sight threatening retinopathy in the presence of previous treatment for more severe lesions; if the scoring had been different, this may have increased the

association with OSA. Women were not included in this or the original study; the sex differences in body fat distribution mean OSA is less common in women and therefore they were not studied originally. This was a cross-sectional study, but with a time lag between sleep studies and retinal photography examination. The retinal photographs were graded separately by two trained graders, blind of OSA status, giving considerable validity to the results. As this was a cross-sectional study, neither initiation of retinopathy nor its acceleration by OSA has been proved, but the results allow hypothesis generation about reasons for this association. Aspects of the pathophysiology of OSA may affect the vulnerable retina, provoking or accelerating retinopathy in people with Type 2 diabetes and OSA. Alternatively, OSA may not be the direct cause, but may be a marker of some other predictive factor, which has not been directly measured. It may be, for example, that the particular central obesity distribution leading to OSA also predisposes to retinopathy in some way. To partially investigate this possibility, we entered neck circumference, our only measure of upper body obesity, into the regression analyses. This, however, produced no decrement in the effect of OSA on retinopathy scores. Randomized controlled trials of OSA treatment on retinopathy are required to investigate this association further. The fact that this study did not nd known hypertension or duration of diabetes to be signicant independent variables relating to retinopathy or maculopathy,

428

2010 The Authors. Journal compilation 2010 Diabetes UK. Diabetic Medicine, 27, 423430

Original article

DIABETICMedicine

when these have been well-established causative factors in other studies, was perhaps surprising, particularly as it is well established that OSA is associated with hypertension. This may, however. relate to the fact in this study these measures were based on patient recall, rather than medical notes. This may have potentially led to greater noise in this data which could account for its non-signicance. As the sample size was relatively small, this study needs to be repeated in a larger population. Blood pressure should be measured and physician recorded history of hypertension noted. Waist circumference, lipid levels and medication use, including brates and angiotensin-converting enzyme inhibitors, should all be recorded. The relationship of OSA with retinopathy has not been previously well established. In one study, 44 people with Type 2 diabetes underwent overnight oximetry [23]. They were divided into sight-threatening (pre-proliferative and or proliferative retinopathy) or non-sight-threatening (background changes or normal) retinopathy groups based on their digital retinal photographs. Those in the sight-threatening group (n = 21) had signicantly more sleep-disordered breathing than those in the non-sight-threatening group [mean (sd) 4% oxygen desaturation index 13.6 16.5 vs. 3.8 3.1, P = 0.03]. Another previous study performed sleep studies in 118 people with proliferative and 48 with non-proliferative diabetic retinopathy and found 48% and 29%, respectively, had evidence of sleep-disordered breathing [24]. Potential mechanisms for the relationship between OSA and retinopathy have been little studied. The recurrent episodes of intermittent hypoxia, reoxygenation, sympathetic surges and blood pressure rises (both intermittent and sustained) might all affect the retina adversely. For example, there might be increased vascular damage from the blood pressure changes (both the multiple transient rises at night and the increased diurnal values) and increased new vessel growth from the recurrent hypoxia. It is known that tight glycaemic control can delay progression of retinopathy [13]. It is not known whether the treatment of OSA would also delay the progression of retinopathy. The mainstay of treatment for symptomatic OSA is CPAP, where a positive pressure supplied by a pump is applied to the upper airway via a mask, to splint the airway open. CPAP is a well-established treatment and has been found in randomized controlled trials to improve sleep quality, daytime sleepiness and self-reported health status [25,26]. It also decreases 24-h blood pressure, urine catecholamine excretion, arterial stiffness, total cholesterol and improves baroreceptor reex sensitivity; in an uncontrolled observational study, it also decreased fatal and non-fatal cardiovascular events [7,27,28]. Its effects in diabetes are less clear; small uncontrolled studies show it can improve insulin resistance and glycaemic control in people with Type 2 diabetes and OSA, but our own randomized controlled study showed no effect [29,30]. It could be hypothesized that alleviation of the intermittent hypoxia and blood pressure surges could improve the environment of the retina, leading to stabilization, or possibly improvement, of diabetic retinopathy. Future randomized controlled trials of CPAP treatment for the OSA will need to

address this, using both short-term markers and long-term progression of retinopathy. One small study has measured retinaspecic mRNAs in 35 patients with Type 2 diabetes. The degree of overnight hypoxia was signicantly associated with these possible markers of retinal damage [31]. This study has shown an association between retinopathy and obstructive sleep apnoea in men with Type 2 diabetes. If these results are conrmed by larger studies of this nature, it will emphasize the importance of identifying men with Type 2 diabetes who also have OSA. Patients with symptoms of OSA, such as loud snoring, witnessed apnoeas and excessive daytime sleepiness, should certainly be referred for sleep studies, if only for symptom relief.

Competing interests
Nothing to declare.

Acknowledgements

The original study was supported by research grants from Diabetes UK and the Grand Charity. Funding for the retinopathy component came from internal charitable funds. The funding bodies had no role in the design or analysis of the studies. We are grateful to Arifa Rahman for regrading the retinal photographs and to Irene Statton for statistical help with the inter-observer agreements.

References
1 Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med 1993; 328: 12301235. 2 Davies RJ, Ali NJ, Stradling JR. Neck circumference and other clinical features in the diagnosis of the obstructive sleep apnoea syndrome. Thorax 1992; 47: 101105. 3 Nieto FJ, Young TB, Lind BK, Shahar E, Samet JM, Redline S et al. Association of sleep-disordered breathing, sleep apnoea, and hypertension in a large community-based study. Sleep heart health study. JAMA 2000; 283: 18291836. 4 Punjabi NM, Sorkin JD, Katzel LI, Goldberg AP, Schwartz AR, Smith PL. Sleep-disordered breathing and insulin resistance in middle-aged and overweight men. Am J Respir Crit Care Med 2002; 165: 677682. 5 Meslier N, Gagnadoux F, Giraud P, Person C, Ouksel H, Urban T et al. Impaired glucose-insulin metabolism in males with obstructive sleep apnoea syndrome. Eur Respir J 2003; 22: 156160. 6 Coughlin SR, Mawdsley L, Mugarza JA, Calverley PM, Wilding JP. Obstructive sleep apnoea is independently associated with an increased prevalence of metabolic syndrome. Eur Heart J 2004; 25: 735741. 7 Marin JM, Carrizo SJ, Vicente E, Agusti AG. Long-term cardiovascular outcomes in men with obstructive sleep apnoea hypopnoea with or without treatment with continuous positive airway pressure: an observational study. Lancet 2005; 365: 1046 1153. 8 Bunce C, Wormald R. Leading causes of certication for blindness and partial sight in England and Wales. BMC Public Health 2006; 6: 58.

2010 The Authors. Journal compilation 2010 Diabetes UK. Diabetic Medicine, 27, 423430

429

DIABETICMedicine

Diabetic retinopathy and obstructive sleep apnoea S. D. West et al.

9 Stratton IM, Adler AI, Neil HA, Matthews DR, Manley SE, Cull CA et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. Br Med J 2000; 321: 405412. 10 Stratton IM, Kohner EM, Aldington SJ, Turner RC, Holman RR, Manley SE et al. UKPDS 50: risk factors for incidence and progression of retinopathy in Type II diabetes over 6 years from diagnosis. Diabetologia 2001; 44: 156163. 11 Chew EY, Klein ML, Ferris FL III, Remaley NA, Murphy RP, Chantry K et al. Association of elevated serum lipid levels with retinal hard exudate in diabetic retinopathy. Early Treatment Diabetic Retinopathy Study (ETDRS) Report 22. Arch Ophthalmol 1996; 114: 10791084. 12 Klein R, Klein BE, Moss SE, Davis MD, DeMets DL. The wisconsin epidemiologic study of diabetic retinopathy X. Four-year incidence and progression of diabetic retinopathy when age at diagnosis is 30 years or more. Arch Ophthalmol 1989; 107: 244249. 13 Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologic Study of diabetic retinopathy XIV. Ten-year incidence and progression of diabetic retinopathy. Arch Ophthalmol 1994; 112: 12171228. 14 Leslie RD, Pyke DA. Diabetic retinopathy in identical twins. Diabetes 1982; 31: 1921. 15 Emanuele N, Sacks J, Klein R, Reda D, Anderson R, Duckworth W et al. Ethnicity, race, and baseline retinopathy correlates in the veterans affairs diabetes trial. Diabetes Care 2005; 28: 19541958. 16 Simmons D, Clover G, Hope C. Ethnic differences in diabetic retinopathy. Diabet Med 2007; 24: 10931098. 17 West SD, Nicoll DJ, Stradling JR. Prevalence of obstructive sleep apnoea in men with type 2 diabetes. Thorax 2006; 61: 945950. 18 Foster GD, Sanders MH, Millman R, Zammit G, Borradaile KE, Newman AB et al.; Sleep AHEAD Research Group Obstructive sleep apnea among obese patients with type 2 diabetes. Diabetes Care 2009; 32: 10171019. 19 Netzer NC, Stoohs RA, Netzer CM, Clark K, Strohl KP. Using the Berlin Questionnaire to identify patients at risk for the sleep apnoea syndrome. Ann Intern Med 1999; 131: 485491. 20 Flemons WW, Littner MR, Rowley JA, Gay P, Anderson WM, Hudgel DW et al. Home diagnosis of sleep apnoea: a systematic review of the literature: an evidence review cosponsored by the American Academy of Sleep Medicine, the American College of Chest Physicians, and the American Thoracic Society. Chest 2003; 124: 15431579.

21 Dingli K, Coleman EL, Vennelle M, Finch SP, Wraith PK, Mackay TW et al. Evaluation of a portable device for diagnosing the sleep apnoea hypopnoea syndrome. Eur Respir J 2003; 21: 253259. 22 Harding S, Greenwood R, Aldington S, Gibson J, Owens D, Taylor R et al. Grading and disease management in national screening for diabetic retinopathy in England and Wales. Diabet Med 2003; 20: 965971. 23 Merritt SP, Moxham J, Wong A, Steier J, Carroll P, Williams AJ. Sleep disordered breathing, a causative factor in the development of severe diabetic retinopathy? Thorax 2007; 61: S59. 24 Shiba T, Sato Y, Takahashi M. Relationship between diabetic retinopathy and sleep-disordered breathing. Am J Ophthalmol 2009; 147: 10171021. 25 Jenkinson C, Davies RJ, Mullins R, Stradling JR. Comparison of therapeutic and subtherapeutic nasal continuous positive airway pressure for obstructive sleep apnoea: a randomised prospective parallel trial. Lancet 1999; 353: 21002105. 26 Ballester E, Badia JR, Hernandez L, Carrasco E, de Pablo J, Fornas C et al. Evidence of the effectiveness of continuous positive airway pressure in the treatment of sleep apnoea hypopnea syndrome. Am J Respir Crit Care Med 1999; 159: 495501. 27 Pepperell JC, Ramdassingh-Dow S, Crosthwaite N, Mullins R, Jenkinson C, Stradling JR et al. Ambulatory blood pressure after therapeutic and subtherapeutic nasal continuous positive airway pressure for obstructive sleep apnoea: a randomised parallel trial. Lancet 2002; 359: 204210. 28 Kohler M, Pepperell JC, Casadei B, Craig S, Crosthwaite N, Stradling JR et al. Continuous positive airway pressure and measures of cardiovascular risk in men with OSAS. Eur Respir J 2008; 32: 14881496. 29 Harsch IA, Schahin SP, Bruckner K, Radespiel-Troger M, Fuchs FS, Hahn EG et al. The effect of continuous positive airway pressure treatment on insulin sensitivity in patients with obstructive sleep apnoea syndrome and type 2 diabetes. Respiration 2004; 71: 252 259. 30 West SD, Nicoll DJ, Wallace TM, Matthews DR, Stradling JR. Effect of CPAP on insulin resistance and HbA1c in men with obstructive sleep apnoea and type 2 diabetes. Thorax 2007; 62: 969974. 31 Wong A, Merritt S, Butt AN, Williams A, Swaminathan R. Effect of hypoxia on circulating levels of retina-specic messenger RNA in type 2 diabetes mellitus. Ann N Y Acad Sci 2008; 1137: 243 252.

430

2010 The Authors. Journal compilation 2010 Diabetes UK. Diabetic Medicine, 27, 423430

Copyright of Diabetic Medicine is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.