- Email: [email protected]
Obstructive Sleep Apnea and Increased Risk of Glaucoma
Obstructive Sleep Apnea and Increased Risk of Glaucoma A Population-Based Matched-Cohort Study Ching-Chun Lin, MA,1 Chao-Chien Hu, MD,2,3,4,5 Jau-Der Ho, MD, PhD,2,3 Hung-Wen Chiu, PhD,1 Herng-Ching Lin, PhD6,7 Purpose: Previous studies had reported an increased prevalence of glaucoma in patients with obstructive sleep apnea (OSA). However, the risk of open-angle glaucoma (OAG) among patients with OSA remains unclear. Using a nationwide, population-based dataset in Taiwan, this study aimed to examine the prevalence and risk of OAG among patients with OSA during a 5-year follow-up period after a diagnosis of OSA. Design: A retrospective, matched-cohort study. Participants and Controls: This study used data sourced from the Longitudinal Health Insurance Database 2000. We included 1012 subjects with OSA in the study cohort and randomly selected 6072 subjects in the comparison group. Methods: Each subject in this study was individually traced for a 5-year period to identify those subjects who subsequently received a diagnosis of OAG. Cox proportional hazards regression was performed to calculate the 5-year risk of OAG between the study and comparison cohorts. Main Outcome Measures: The incidence and risk of OAG between the study and comparison groups. Results: During the 5-year follow-up period, the incidence rate per 1000 person-years was 11.26 (95% confidence interval [CI], 8.61e14.49) and 6.76 (95% CI, 5.80e7.83) for subjects with and without OSA, respectively. After adjusting for monthly income, geographic region, diabetes, hypertension, coronary heart disease, obesity, hyperlipidemia, renal disease, hypothyroidism, and the number of outpatient visits for ophthalmologic care during the follow-up period, stratified Cox proportional hazards regression revealed that the hazard ratio for OAG within the 5-year period for subjects with OSA was 1.67 (95% CI, 1.30e2.17; P<0.001) that of comparison subjects. Conclusions: Our results suggest that OSA is associated with an increased risk of subsequent OAG diagnosis during a 5-year follow-up period. Financial Disclosures(s): The authors have no proprietary or commercial interest in any of the materials discussed in this article. Ophthalmology 2013;120:1559-1564 ª 2013 by the American Academy of Ophthalmology.
Obstructive sleep apnea (OSA) is a condition characterized by intermittent upper airway obstruction during sleep.1 The obstructive respiratory disturbances may last up to 2 minutes, leading to hypoxia, hypercapnea, sleep disruption, negative intrathoracic pressure, and sympathetic activation, which can trigger arousal from sleep by increasing ventilatory drive.1-4 Epidemiologic studies have indicated that the prevalence ranges between 2% and 20%.5,6 Furthermore, 24% of men and 9% of women in Wisconsin have been demonstrated to suffer from OSA by the time they reach middle age.7 Often, OSA leads to severe hypoxemia and increases vascular resistance, which may influence the development of ganglion cell loss.8-10 Previous studies have demonstrated OSA to be associated with several eye disorders, such as floppy eyelid syndrome, keratoconus, papilledema, optic neuropathy, filamentary or infectious keratitis, papillary conjunctivitis, and glaucoma.2,11-15 Furthermore, several studies have showed an
2013 by the American Academy of Ophthalmology Published by Elsevier Inc.
increased prevalence of glaucoma in patients with OSA.2,3,5,16 One cross-sectional study by Mojon et al2 investigating 69 patients with OSA in Switzerland found that all of the 3 glaucoma patients with OSA and the respiratory disturbance index correlated positively with a diagnosis of OAG.2 Another case-control study by Girkin et al16 selected 667 patients with glaucoma and nonglaucomatous controls in Birmingham, Alabama. The results of their study indicated that patients with glaucoma were more likely to have a previous diagnosis of OSA relative to control subjects.16 Karakucuk et al5 also showed that patients with OSA had a high prevalence of glaucoma.5 Moreover, a study by Lin et al3 reported that a high prevalence of glaucoma was found in patients with OSA, especially in patients with moderate and severe OSA. However, although previous studies have shown an increased prevalence of glaucoma among patients with OSA, no study has explored the risk of glaucoma after an OSA diagnosis. Therefore, using a nationwide,
ISSN 0161-6420/13/$ - see front matter http://dx.doi.org/10.1016/j.ophtha.2013.01.006
1559
Ophthalmology Volume 120, Number 8, August 2013 population-based dataset in Taiwan, this study examined the prevalence and risk of OAG diagnosis among patients with OSA over a 5-year follow-up period after a diagnosis of OSA, compared with patients without OSA during the same period, adjusting for sociodemographic characteristics.
Methods Database The data used in this study were sourced from the Longitudinal Health Insurance Database 2000 (LHID2000), which is derived from the Taiwan National Health Insurance (NHI) program and maintained by the Taiwan National Health Research Institute, includes the registration files and original claims data for the reimbursement of 1 000 000 beneficiaries under the NHI program. These 1 000 000 beneficiaries were randomly selected from the year 2000 Registry of Beneficiaries (n ¼ 23.72 million) of the NHI. The completeness and accuracy of the claims data of NHI research database were appropriately monitored,17,18 with hundreds of studies based on these data having been published in peer-reviewed journals. The LHID2000 consists of deidentified secondary data released to the public for research purposes and was therefore exempted from full review after consultation with the Taipei Medical University’s Institutional Review Board.
Study Sample This retrospective, cohort study included study and comparison cohorts. We first selected the study cohort by identifying those subjects who had received their first diagnosis of OSA (International Classification of Diseases, 9th edition, Clinical Modification [ICD-9-CM] codes 327.23, 780.51, 780.53, or 780.57) after undergoing polysomnography in ambulatory care visits from January 1, 2001, to December 31, 2004 (n ¼ 2279). In Taiwan, if a patient has a confirmed OSA diagnosis after polysomnography, he or she would receive another OSA diagnosis at their next ambulatory care visit. Therefore, this study only included patients who had been given OSA diagnoses after undergoing polysomnography to increase the validity of the OSA diagnoses sourced from the dataset. We assigned the date of the first ambulatory care visit in which OSA was diagnosed after undergoing polysomnography as the index date for the study cohort. In addition, we excluded subjects aged <40 years (n ¼ 813) because the prevalence of OAG was low for this age group. Furthermore, we excluded study subjects who had never visited an ophthalmologist during the 4 years preceding the index date (n ¼ 422). By only including study subjects who had visited an ophthalmologist, we sought to decrease the possibility that any of the study subjects were suffering from OAG asymptomatically before the index date. We further excluded subjects who had received a diagnosis of OAG (ICD-9-CM codes 365.1e365.11) before their index use of healthcare services (n ¼ 32). Because the Taiwan NHI program began in 1995, the LHID2000 only allows us to trace use of medical services as far back as 1996; therefore, we cannot rule out patients who had a diagnosis of OAG before 1996. However, OAG is a chronic disease and requires patients to receive medication regularly. Therefore, we believe we have excluded all cases who received an OAG diagnosis before index date. As a result, 1012 subjects with OSA were included in the study cohort. We also selected the comparison cohort from the LHID2000. We first excluded all subjects who had ever received a diagnosis of OSA since the initiation of the NHI program in 1995. We then
1560
randomly selected 5060 subjects (5 for every patient in the study cohort) matched with the study cohort in terms of gender, age (40e44, 45e49, 50e54, 55e59, 60e64, 65e69, and >69 years), urbanization level (5 levels; from 1 [most urbanized] to 5 [least urbanized]), and the year of index date using the SAS program proc SurveySelect (SAS System for Windows, Version 8.2, Cary, NC). Although, for study subjects, the year of index date was the year in which the study subjects received their first diagnosis of OSA, for comparison subjects the year of index date was simply a matched year in which comparison subjects had a medical utilization. We assigned the date of their first use of medical services occurring during that matched year as the index date for the comparison subject. Similarly, to reduce the possibility of surveillance bias, in this study we only included comparison subjects who had visited an ophthalmologist at least once during the 4 years preceding the index date. All the towns/cities of Taiwan were stratified into 5 different urbanization levels based on a prior study.19 This study matched the comparison cohort with the study cohort on urbanization level to help control for unmeasured neighborhood socioeconomic characteristics between the study and comparison cohort. As for the comparison cohort, we assigned their first use of medical care occurring during the year of index as their index date. We further ensured that none of the selected comparison subjects had ever received a diagnosis of OAG before their index date. Thereafter, each subject in this study (n ¼ 6072) was individually traced for a 5-year period from their index date to identify those subjects who subsequently received a diagnosis of OAG (ICD-9-CM codes 365.1e365.11). To increase the validity of glaucoma diagnosis we only included glaucoma cases who underwent treatment with topical antiglaucoma medication. In this way, we can avoid patients who had received a diagnosis of OAG but only get an eye examination to check OAG from an ophthalmologist.
Statistical Analysis We used the SAS statistical package (SAS System for Windows, Version 8.2) to carry out all the statistical analyses in this study. We used the KaplaneMeier method to calculate the 5-year OAG-free survival rates, with the log-rank test also being used to examine differences in OAG-free survival rates between the study and comparison cohorts. In addition, we used stratified Cox proportional hazard regressions (stratified on gender, age group, urbanization level, and year of index date) to calculate the hazard ratio (HR) and corresponding 95% confidence interval (CI) for the relationship between OSA and subsequent risk of OAG diagnosis after adjusting for monthly income, selected comorbidities (hypertension, diabetes, coronary heart disease, hyperlipidemia, obesity, renal disease, migraine, and hypothyroidism) and number of outpatient visits for ophthalmologic care during the follow-up period, censoring those subjects who died during the 5-year follow-up period. Of the sampled patients, 749 died; this included 141 from the study cohort (9.8% of study cohort) and 638 from the comparison cohort (8.9% of the comparison cohort). In this study, we have also examined the proportional hazards assumption and found this assumption to be satisfied because the survival curves for both strata (patients in the study cohort and comparison cohort) had hazard functions that were proportional over time. A 2-sided P<0.05 was considered significant in this study.
Results Table 1 shows the distribution of demographic characteristics and comorbidities between the study cohort and the comparison cohort. Of the total of 6072 subjects, the mean age was 56.0 years
Lin et al
Glaucoma and Obstructive Sleep Apnea
Table 1. Demographic Characteristics for the Sampled Taiwanese Patients Stratified by the Presence/Absence of Obstructive Sleep Apnea (OSA), 2001e2004 (n ¼ 6072) Patients with OSA (n [ 1012) Variables Gender Male Female Age (yrs) 40e44 45e49 50e54 55e59 60e64 65e69 70 Urbanization level 1 (most urbanized) 2 3 4 5 (least urbanized) Number of outpatient visits for ophthalmologic care (mean SD) Monthly income (NT$) 1e15 840 15 841e25 000 25 001 Diabetes Hypertension Coronary heart disease Obesity Hyperlipidemia Renal disease Hypothyroidism Migraine
Comparison patients (n [ 5060)
n
%
n
%
609 403
60.2 39.8
3045 2015
60.2 39.8
173 202 184 118 94 70 171
17.1 20.0 18.2 11.6 9.3 6.9 16.9
865 1010 920 590 470 350 855
17.1 20.0 18.2 11.6 9.3 6.9 16.9
368 294 179 100 71
36.4 29.0 17.7 9.9 7.0
1,840 1,470 895 500 355
36.4 29.0 17.7 9.9 7.0
P >0.999 >0.999
>0.999
5.938.64
5.539.04
0.126 0.582
286 409 317 226 511 278 33 342 83 8 55
28.3 40.4 31.3 22.3 50.5 27.5 3.3 33.8 8.2 0.8 5.4
1396 1994 1670 657 1519 636 21 857 188 27 137
27.6 39.4 33.0 13.0 30.0 12.6 0.4 16.9 3.7 0.5 2.7
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.324 <0.001
NT$ ¼ New Taiwan Dollar; SD ¼ standard deviation.
(standard deviation, 11.9) and 60% were male. After matching for gender, age group, and urbanization level, subjects with OSA had a higher prevalence of the following comorbidities than comparison subjects: hypertension (P<0.001), diabetes (P<0.001), coronary heart disease (P<0.001), hyperlipidemia (P<0.001), obesity (P<0.001), renal disease (P<0.001), and migraine (P<0.001). However, there was no difference in the prevalence of hypothyroidism (P ¼ 0.324) between the study and comparison cohorts. Table 2 presents the incidence of OAG diagnosis during the 5year follow-up period after the index date of the sampled subjects. The incidence rate per 1000 person-years was 11.26 (95% CI, 8.61e14.49) and 6.76 (95% CI, 5.80e7.83) for subjects with and without OSA, respectively. The log-rank test revealed that subjects with OSA had significantly lower 5-year OAG-free survival rates compared with comparison subjects (chi square value, 2163; P<0.001). The figure for the KaplaneMeier curves is displayed in Figure 1. Table 2 also shows the crude and adjusted HR of OAG diagnosis between the cohorts. After adjusting for monthly income, geographic region, diabetes, hypertension, coronary heart disease, obesity, hyperlipidemia, renal disease, hypothyroidism, and number of outpatient visits for ophthalmologic care during the follow-up period, stratified Cox proportional hazard regressions (stratified on
gender, age group, urbanization level, and year of index date) showed that the HR for OAG diagnosis within the 5-year period for subjects with OSA was 1.67 (95% CI, 1.30e2.17; P<0.001) that of comparison subjects. Table 3 reveals results from further analyses stratified by gender. Among females, the adjusted hazard of OAG diagnosis during the 5year follow-up period was 1.55 (95% CI, 1.04e2.31) times greater for those with OSA than for those without. However, among males, the adjusted HR for OAG diagnosis within the 5-year period for subjects with OSA was only 1.45 (95% CI, 1.02e2.16) that of comparison subjects.
Discussion We found that patients with OSA were independently associated with a 1.67 times increased risk of OAG diagnosis within the first 5 years after their diagnosis, after adjusting for monthly income, diabetes, hypertension, coronary heart disease, obesity, hyperlipidemia, renal disease, hypothyroidism, and the number of outpatient visits for ophthalmologic care during the follow-up period. Several previous studies exploring the link between OSA and OAG were cross-sectional or retrospective in nature and
1561
Ophthalmology Volume 120, Number 8, August 2013 Table 2. Crude and Adjusted Hazard Ratios (HR) of Glaucoma among the Sample Patients during the 5-year Follow-up Periods (n ¼ 6072) Total (n [ 6072) Development of Glaucoma 5-year follow-up period Yes Incidence rate per 1000 person-years (95% CI) Crude HR (95 % CI) Adjusted HR (95 % CI)y
n
%
228 3.8 7.51 (6.58e8.53) e e
Patients with Obstructive Sleep Apnea (n [ 1012) n
%
57 5.6 11.26 (8.61e14.49) 1.71* (1.26e2.32) 1.67* (1.30e2.17)
Comparison Patients (n [ 5060) n
%
171 3.4 6.76 (5.80e7.83) 1.00 1.00
CI ¼ confidence interval. The HR was calculated by stratified Cox proportional hazards regressions (stratified on gender, age group, urbanization level, and the year of index date). *P<0.001. y Adjustments are made for the number of outpatient visits for ophthalmologic care during the follow-up period, monthly income, diabetes, hypertension, coronary heart disease, obesity, hyperlipidemia, renal disease, hypothyroidism, and migraine.
only showed the prevalence of OAG among patients with OSA. One cross-sectional study in Switzerland indicated that patients with OSA are a high-risk population for glaucoma and that respiratory disturbance index correlated positively with intraocular pressure.2 Another case-control study by Girkin et al16 found that patients with glaucoma were more likely to have a previous diagnosis of OSA relative to control subjects. However, this finding was of borderline significance at an alpha level of 0.05 (P ¼ 0.06).16 Furthermore, 2 studies conducted on institutional databases in Turkey and Taiwan reported a high prevalence of glaucoma among patients with OSA.3,5 Nevertheless, methodologic variation (e.g., different populations, study designs, and disease definitions) or limits (e.g., inappropriate
Figure 1. Open-angle glaucoma-free survival rates for patients with obstructive sleep apnea and comparison group in Taiwan
1562
consideration of confounding effects) render more specific comparisons between investigations difficult. Both a mechanical theory and a vascular theory have been proposed to underpin the association between OSA and OAG. The mechanical theory suggests that increased intraocular pressure and stretching of the lamina cribrosa is the reason for glaucoma.20,21 The intermittent upper airway obstruction experienced in OSA patients during sleep would cause hypoxia with a subsequent decrease in PaO2 and increase in PaCO2. This episodic vascular insufficiency may compromise optic nerve perfusion and oxygenation, and subsequently cause optic neuropathy.3,21 Moreover, OSA is often, although not invariably, associated with an increased body mass index. The positive association between obesity and intraocular pressure is proposed to be related to direct damage of the optic nerve.22 This intraocular pressure elevation may be because of the overproduction of aqueous fluid or impaired aqueous outflow.19,23 The intracranial pressure of OSA patients has been found to be markedly elevated in an episodic fashion, paralleling apneic episodes. It is hypothesized that hypercapnea leads to cerebral venous dilatation.24 The vascular theory holds that repetitive or prolonged episodes of hypoxia cause direct damage to the optic nerve. The vascular theory proposes that the association between OSA and OAG can be explained by 3 mechanisms: Impaired optic nerve head blood flow autoregulation secondary to repetitive prolonged apneas, optic nerve vascular dysregulation secondary to arteriosclerosis, and arterial blood pressure variations or imbalance between nitric oxide (a vasodilator) and endothelin (a vasoconstrictor).17,25,26 The vascular theory postulates that there is insufficient blood supply to nourish the nerve fiber layer and/or optic nerve. This arises on account of the recurrent apnea/hypopnea characteristic of OSA which leads to transient hypoxemia and increased vascular resistance which in turn compromises optic nerve head perfusion.3 The main strength of our study lies in its longitudinal and large population database, which allowed us to consider the relationship between OSA and OAG. Furthermore, this database tracks medical utilizations in Taiwan, where >98% of the population are of Han Chinese ethnicity. This homogenous population may avoid potential confounding
Lin et al
Glaucoma and Obstructive Sleep Apnea
Table 3. Crude and Adjusted Hazard Ratios (HR) of Glaucoma among the Sample Patients During the 5-year Follow-up Periods, by Gender Group Gender Group Male Development of Glaucoma Five-year follow-up period Yes Incidence rate per 1000 person-years (95% CI) Crude HR (95 % CI) Adjusted HR (95 % CI)y
Patients with OSA (n, %) 29 9.52 1.51* 1.45z
(4.8) (6.50e13.50) (1.05e2.17) (1.02e2.16)
Female Comparison Patients (n, %) 97 (3.2) 6.43 (5.24e7.80) 1.00 1.00
Patients with OSA (n, %) 28 13.90 1.99* 1.55z
(7.0) (9.42e19.81) (1.27e3.11) (1.04e2.31)
Comparison Patients (n, %) 74 (3.6) 7.25 (5.74e9.05) 1.00 1.00
CI ¼ confidence interval; OSA ¼ obstructive sleep apnea. The HR was calculated by stratified Cox proportional hazards regressions (stratified on gender, age group, urbanization level, and the year of index date). *P<0.01. y Adjustments are made for the number of outpatient visits for ophthalmologic care during the follow-up period, monthly income, diabetes, hypertension, coronary heart disease, obesity, hyperlipidemia, renal disease, hypothyroidism, and migraine. z P<0.05.
by race. Last, clinical diagnostic criteria were used for the identification of OSA and OAG, with the latter further requiring the confirmatory results of a polysomnography to ensure diagnostic validity. Nevertheless, the findings of this study need to be considered in light of several limitations. One of the limitations of our study is that the LHID2000 database represents patients who had sought treatment for OSA, but lacks information on the severity of OSA, such as apneaehypopnea index (AHI) scores or respiratory disturbance index scores. Therefore, we were unable to evaluate whether patients with more severe OSA had a higher risk of OAG than those with mild OSA. Another limitation is that the only patients who seek treatment are those who have some symptoms of suspected OSA, such as snoring, persistent daytime sleepiness or drowsiness while driving, and hypersomnolence. Because the NHI database only included patients who sought treatment, it is possible that some patients may have had OSA but did not receive a diagnosis because they did not seek or receive medical care for this. These patients may have been selected and recruited in the comparison cohort, which would have worked to bias our results toward the null. Moreover, patients with OSA have significantly higher risks of all considered comorbidities, and this might imply that patients with OSA have generally poorer health than those without OSA. OSA may simply be a marker for poor (vascular) health, and not necessarily an independent risk factor for OAG. However, after adjusting for socioeconomic characteristics and medical comorbidities, we found that patients with OSA had an increased risk of OAG diagnosis within the first 5 years after their diagnosis. Therefore, OSA is an independent risk factor for OAG. Finally, this study only included OAG patients who received treatment with topical antiglaucoma medication. Those (n ¼ 8) who received an OAG diagnosis but were untreated during the follow-up period were not counted in this study. However, we reanalyzed the data and found the
results remained even if those 8 cases were included in the OAG cases. In conclusion, our results suggest that OSA is associated with an increased risk of subsequent OAG diagnosis during the first 5 years after OSA diagnosis. We found that the hazard of receiving an OAG diagnosis during the 5-year follow-up period was 1.67 times greater in patients with OSA than in gender- and age-matched comparison subjects, after adjusting for socioeconomic characteristics and medical comorbidities. The authors hope that this study encourages clinicians to alert OSA patients of the association of between OSA and OAG as a means of raising the issue and encouraging treatment of those who need it. Further study is needed to clarify our findings and to explore the mechanisms underlying this association.
References 1. Bendel RE, Kaplan J, Heckman M, et al. Prevalence of glaucoma in patients with obstructive sleep apnoeaea crosssectional case-series. Eye (Lond) 2008;22:1105–9. 2. Mojon DS, Hess CW, Goldblum D, et al. High prevalence of glaucoma in patients with sleep apnea syndrome. Ophthalmology 1999;106:1009–12. 3. Lin PW, Friedman M, Lin HC, et al. Normal tension glaucoma in patients with obstructive sleep apnea/hypopnea syndrome. J Glaucoma 2011;20:553–8. 4. Ayas NT, Brown R, Shea SA. Hypercapnia can induce arousal from sleep in the absence of altered respiratory mechanoreception. Am J Respir Crit Care Med 2000;162:1004–8. 5. Karakucuk S, Goktas S, Aksu M, et al. Ocular blood flow in patients with obstructive sleep apnea syndrome (OSAS). Graefes Arch Clin Exp Ophthalmol 2008;246:129–34. 6. Onen SH, Mouriaux F, Berramdane L, et al. High prevalence of sleep-disordered breathing in patients with primary openangle glaucoma. Acta Ophthalmol Scand 2000;78:638–41. 7. Young T, Palta M, Dempsey J, et al. The occurrence of sleepdisordered breathing among middle-aged adults. N Engl J Med 1993;328:1230–5.
1563
Ophthalmology Volume 120, Number 8, August 2013 8. Chervin RD, Guilleminault C. Obstructive sleep apnea and related disorders. Neurol Clin 1996;14:583–609. 9. Geyer O, Cohen N, Segev E, et al. The prevalence of glaucoma in patients with sleep apnea syndrome: same as in the general population. Am J Ophthalmol 2003;136:1093–6. 10. Marcus DM, Costarides AP, Gokhale P, et al. Sleep disorders: a risk factor for normal-tension glaucoma? J Glaucoma 2001;10:177–83. 11. McNab AA. The eye and sleep. Clin Experiment Ophthalmol 2005;33:117–25. 12. Purvin VA, Kawasaki A, Yee RD. Papilledema and obstructive sleep apnea syndrome. Arch Ophthalmol 2000;118:1626–30. 13. Bucci FA Jr, Krohel GB. Optic nerve swelling secondary to the obstructive sleep apnea syndrome. Am J Ophthalmol 1988;105:428–30. 14. Mojon DS, Mathis J, Zulauf M, et al. Optic neuropathy associated with sleep apnea syndrome. Ophthalmology 1998;105:874–7. 15. Robert PY, Adenis JP, Tapie P, Melloni B. Eyelid hyperlaxity and obstructive sleep apnea (O.S.A.) syndrome. Eur J Ophthalmol 1997;7:211–5. 16. Girkin CA, McGwin G Jr, McNeal SF, Owsley C. Is there an association between pre-existing sleep apnoea and the development of glaucoma? Br J Ophthalmol 2006;90:679–81. 17. Cheng CL, Kao YH, Lin SJ, et al. Validation of the National Health Insurance Research Database with ischemic stroke
18.
19. 20. 21. 22. 23. 24.
25. 26.
cases in Taiwan. Pharmacoepidemiol Drug Saf 2011;20: 236–42. Lin HC, Xirasagar S, Chen CH, Hwang YT. Physician’s case volume of intensive care unit pneumonia admissions and inhospital mortality. Am J Respir Crit Care Med 2008;177: 989–94. Lin HC, Chao PZ, Lee HC. Sudden sensorineural hearing loss increases the risk of stroke: a 5-year follow-up study. Stroke 2008;39:2744–8. Abdal H, Pizzimenti JJ, Purvis CC. The eye in sleep apnea syndrome. Sleep Med 2006;7:107–15. Parish JM, Shepard JW Jr. Cardiovascular effects of sleep disorders. Chest 1990;97:1220–6. Fingeret M, Lewis TL. Primary Care of the Glaucomas. Norwalk, CT: Appleton & Lange; 1993:3-5. Cheung N, Wong TY. Obesity and eye diseases. Surv Ophthalmol 2007;52:180–95. Sugita Y, Iijima S, Teshima Y, et al. Marked episodic elevation of cerebrospinal fluid pressure during nocturnal sleep in patients with sleep apnea hypersomnia syndrome. Electroencephalogr Clin Neurophysiol 1985;60:214–9. Hayreh SS. Acute ischemic disorders of the optic nerve. Pathogenesis, clinical manifestations, and management. Ophthalmol Clin North Am 1996;9:407–42. Lüscher TF. The endothelium and cardiovascular diseaseea complex relation. N Engl J Med 1994;330:1081–3.
Footnotes and Financial Disclosures Originally received: September 17, 2012. Final revision: December 18, 2012. Accepted: January 3, 2013. Available online: April 17, 2013.
6 School of Medical Laboratory Sciences and Biotechnology, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan.
Manuscript no. 2012-1417.
1
Graduate Institute of Biomedical Informatics, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan. 2
Department of Ophthalmology, Taipei Medical University, Taipei, Taiwan.
3 Department of Ophthalmology, Taipei Medical University Hospital, Taipei, Taiwan. 4
Department of Ophthalmology, Shin Kong Wu-Ho-Su Memorial Hospital, Taipei, Taiwan. 5
School of Medicine, Fu-Jen Catholic University, Hsingchuang, Taiwan.
1564
7 Sleep Research Center, Taipei Medical University Hospital, Taipei, Taiwan.
Financial Disclosures: The authors have no proprietary or commercial interest in any of the materials discussed in this article. Correspondence: Herng-Ching Lin, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan, 250 Wu-Hsing St., Taipei 110, Taiwan. E-mail: [email protected].
Obstructive sleep apnea (OSA) is a condition characterized by intermittent upper airway obstruction during sleep.1 The obstructive respiratory disturbances may last up to 2 minutes, leading to hypoxia, hypercapnea, sleep disruption, negative intrathoracic pressure, and sympathetic activation, which can trigger arousal from sleep by increasing ventilatory drive.1-4 Epidemiologic studies have indicated that the prevalence ranges between 2% and 20%.5,6 Furthermore, 24% of men and 9% of women in Wisconsin have been demonstrated to suffer from OSA by the time they reach middle age.7 Often, OSA leads to severe hypoxemia and increases vascular resistance, which may influence the development of ganglion cell loss.8-10 Previous studies have demonstrated OSA to be associated with several eye disorders, such as floppy eyelid syndrome, keratoconus, papilledema, optic neuropathy, filamentary or infectious keratitis, papillary conjunctivitis, and glaucoma.2,11-15 Furthermore, several studies have showed an
2013 by the American Academy of Ophthalmology Published by Elsevier Inc.
increased prevalence of glaucoma in patients with OSA.2,3,5,16 One cross-sectional study by Mojon et al2 investigating 69 patients with OSA in Switzerland found that all of the 3 glaucoma patients with OSA and the respiratory disturbance index correlated positively with a diagnosis of OAG.2 Another case-control study by Girkin et al16 selected 667 patients with glaucoma and nonglaucomatous controls in Birmingham, Alabama. The results of their study indicated that patients with glaucoma were more likely to have a previous diagnosis of OSA relative to control subjects.16 Karakucuk et al5 also showed that patients with OSA had a high prevalence of glaucoma.5 Moreover, a study by Lin et al3 reported that a high prevalence of glaucoma was found in patients with OSA, especially in patients with moderate and severe OSA. However, although previous studies have shown an increased prevalence of glaucoma among patients with OSA, no study has explored the risk of glaucoma after an OSA diagnosis. Therefore, using a nationwide,
ISSN 0161-6420/13/$ - see front matter http://dx.doi.org/10.1016/j.ophtha.2013.01.006
1559
Ophthalmology Volume 120, Number 8, August 2013 population-based dataset in Taiwan, this study examined the prevalence and risk of OAG diagnosis among patients with OSA over a 5-year follow-up period after a diagnosis of OSA, compared with patients without OSA during the same period, adjusting for sociodemographic characteristics.
Methods Database The data used in this study were sourced from the Longitudinal Health Insurance Database 2000 (LHID2000), which is derived from the Taiwan National Health Insurance (NHI) program and maintained by the Taiwan National Health Research Institute, includes the registration files and original claims data for the reimbursement of 1 000 000 beneficiaries under the NHI program. These 1 000 000 beneficiaries were randomly selected from the year 2000 Registry of Beneficiaries (n ¼ 23.72 million) of the NHI. The completeness and accuracy of the claims data of NHI research database were appropriately monitored,17,18 with hundreds of studies based on these data having been published in peer-reviewed journals. The LHID2000 consists of deidentified secondary data released to the public for research purposes and was therefore exempted from full review after consultation with the Taipei Medical University’s Institutional Review Board.
Study Sample This retrospective, cohort study included study and comparison cohorts. We first selected the study cohort by identifying those subjects who had received their first diagnosis of OSA (International Classification of Diseases, 9th edition, Clinical Modification [ICD-9-CM] codes 327.23, 780.51, 780.53, or 780.57) after undergoing polysomnography in ambulatory care visits from January 1, 2001, to December 31, 2004 (n ¼ 2279). In Taiwan, if a patient has a confirmed OSA diagnosis after polysomnography, he or she would receive another OSA diagnosis at their next ambulatory care visit. Therefore, this study only included patients who had been given OSA diagnoses after undergoing polysomnography to increase the validity of the OSA diagnoses sourced from the dataset. We assigned the date of the first ambulatory care visit in which OSA was diagnosed after undergoing polysomnography as the index date for the study cohort. In addition, we excluded subjects aged <40 years (n ¼ 813) because the prevalence of OAG was low for this age group. Furthermore, we excluded study subjects who had never visited an ophthalmologist during the 4 years preceding the index date (n ¼ 422). By only including study subjects who had visited an ophthalmologist, we sought to decrease the possibility that any of the study subjects were suffering from OAG asymptomatically before the index date. We further excluded subjects who had received a diagnosis of OAG (ICD-9-CM codes 365.1e365.11) before their index use of healthcare services (n ¼ 32). Because the Taiwan NHI program began in 1995, the LHID2000 only allows us to trace use of medical services as far back as 1996; therefore, we cannot rule out patients who had a diagnosis of OAG before 1996. However, OAG is a chronic disease and requires patients to receive medication regularly. Therefore, we believe we have excluded all cases who received an OAG diagnosis before index date. As a result, 1012 subjects with OSA were included in the study cohort. We also selected the comparison cohort from the LHID2000. We first excluded all subjects who had ever received a diagnosis of OSA since the initiation of the NHI program in 1995. We then
1560
randomly selected 5060 subjects (5 for every patient in the study cohort) matched with the study cohort in terms of gender, age (40e44, 45e49, 50e54, 55e59, 60e64, 65e69, and >69 years), urbanization level (5 levels; from 1 [most urbanized] to 5 [least urbanized]), and the year of index date using the SAS program proc SurveySelect (SAS System for Windows, Version 8.2, Cary, NC). Although, for study subjects, the year of index date was the year in which the study subjects received their first diagnosis of OSA, for comparison subjects the year of index date was simply a matched year in which comparison subjects had a medical utilization. We assigned the date of their first use of medical services occurring during that matched year as the index date for the comparison subject. Similarly, to reduce the possibility of surveillance bias, in this study we only included comparison subjects who had visited an ophthalmologist at least once during the 4 years preceding the index date. All the towns/cities of Taiwan were stratified into 5 different urbanization levels based on a prior study.19 This study matched the comparison cohort with the study cohort on urbanization level to help control for unmeasured neighborhood socioeconomic characteristics between the study and comparison cohort. As for the comparison cohort, we assigned their first use of medical care occurring during the year of index as their index date. We further ensured that none of the selected comparison subjects had ever received a diagnosis of OAG before their index date. Thereafter, each subject in this study (n ¼ 6072) was individually traced for a 5-year period from their index date to identify those subjects who subsequently received a diagnosis of OAG (ICD-9-CM codes 365.1e365.11). To increase the validity of glaucoma diagnosis we only included glaucoma cases who underwent treatment with topical antiglaucoma medication. In this way, we can avoid patients who had received a diagnosis of OAG but only get an eye examination to check OAG from an ophthalmologist.
Statistical Analysis We used the SAS statistical package (SAS System for Windows, Version 8.2) to carry out all the statistical analyses in this study. We used the KaplaneMeier method to calculate the 5-year OAG-free survival rates, with the log-rank test also being used to examine differences in OAG-free survival rates between the study and comparison cohorts. In addition, we used stratified Cox proportional hazard regressions (stratified on gender, age group, urbanization level, and year of index date) to calculate the hazard ratio (HR) and corresponding 95% confidence interval (CI) for the relationship between OSA and subsequent risk of OAG diagnosis after adjusting for monthly income, selected comorbidities (hypertension, diabetes, coronary heart disease, hyperlipidemia, obesity, renal disease, migraine, and hypothyroidism) and number of outpatient visits for ophthalmologic care during the follow-up period, censoring those subjects who died during the 5-year follow-up period. Of the sampled patients, 749 died; this included 141 from the study cohort (9.8% of study cohort) and 638 from the comparison cohort (8.9% of the comparison cohort). In this study, we have also examined the proportional hazards assumption and found this assumption to be satisfied because the survival curves for both strata (patients in the study cohort and comparison cohort) had hazard functions that were proportional over time. A 2-sided P<0.05 was considered significant in this study.
Results Table 1 shows the distribution of demographic characteristics and comorbidities between the study cohort and the comparison cohort. Of the total of 6072 subjects, the mean age was 56.0 years
Lin et al
Glaucoma and Obstructive Sleep Apnea
Table 1. Demographic Characteristics for the Sampled Taiwanese Patients Stratified by the Presence/Absence of Obstructive Sleep Apnea (OSA), 2001e2004 (n ¼ 6072) Patients with OSA (n [ 1012) Variables Gender Male Female Age (yrs) 40e44 45e49 50e54 55e59 60e64 65e69 70 Urbanization level 1 (most urbanized) 2 3 4 5 (least urbanized) Number of outpatient visits for ophthalmologic care (mean SD) Monthly income (NT$) 1e15 840 15 841e25 000 25 001 Diabetes Hypertension Coronary heart disease Obesity Hyperlipidemia Renal disease Hypothyroidism Migraine
Comparison patients (n [ 5060)
n
%
n
%
609 403
60.2 39.8
3045 2015
60.2 39.8
173 202 184 118 94 70 171
17.1 20.0 18.2 11.6 9.3 6.9 16.9
865 1010 920 590 470 350 855
17.1 20.0 18.2 11.6 9.3 6.9 16.9
368 294 179 100 71
36.4 29.0 17.7 9.9 7.0
1,840 1,470 895 500 355
36.4 29.0 17.7 9.9 7.0
P >0.999 >0.999
>0.999
5.938.64
5.539.04
0.126 0.582
286 409 317 226 511 278 33 342 83 8 55
28.3 40.4 31.3 22.3 50.5 27.5 3.3 33.8 8.2 0.8 5.4
1396 1994 1670 657 1519 636 21 857 188 27 137
27.6 39.4 33.0 13.0 30.0 12.6 0.4 16.9 3.7 0.5 2.7
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.324 <0.001
NT$ ¼ New Taiwan Dollar; SD ¼ standard deviation.
(standard deviation, 11.9) and 60% were male. After matching for gender, age group, and urbanization level, subjects with OSA had a higher prevalence of the following comorbidities than comparison subjects: hypertension (P<0.001), diabetes (P<0.001), coronary heart disease (P<0.001), hyperlipidemia (P<0.001), obesity (P<0.001), renal disease (P<0.001), and migraine (P<0.001). However, there was no difference in the prevalence of hypothyroidism (P ¼ 0.324) between the study and comparison cohorts. Table 2 presents the incidence of OAG diagnosis during the 5year follow-up period after the index date of the sampled subjects. The incidence rate per 1000 person-years was 11.26 (95% CI, 8.61e14.49) and 6.76 (95% CI, 5.80e7.83) for subjects with and without OSA, respectively. The log-rank test revealed that subjects with OSA had significantly lower 5-year OAG-free survival rates compared with comparison subjects (chi square value, 2163; P<0.001). The figure for the KaplaneMeier curves is displayed in Figure 1. Table 2 also shows the crude and adjusted HR of OAG diagnosis between the cohorts. After adjusting for monthly income, geographic region, diabetes, hypertension, coronary heart disease, obesity, hyperlipidemia, renal disease, hypothyroidism, and number of outpatient visits for ophthalmologic care during the follow-up period, stratified Cox proportional hazard regressions (stratified on
gender, age group, urbanization level, and year of index date) showed that the HR for OAG diagnosis within the 5-year period for subjects with OSA was 1.67 (95% CI, 1.30e2.17; P<0.001) that of comparison subjects. Table 3 reveals results from further analyses stratified by gender. Among females, the adjusted hazard of OAG diagnosis during the 5year follow-up period was 1.55 (95% CI, 1.04e2.31) times greater for those with OSA than for those without. However, among males, the adjusted HR for OAG diagnosis within the 5-year period for subjects with OSA was only 1.45 (95% CI, 1.02e2.16) that of comparison subjects.
Discussion We found that patients with OSA were independently associated with a 1.67 times increased risk of OAG diagnosis within the first 5 years after their diagnosis, after adjusting for monthly income, diabetes, hypertension, coronary heart disease, obesity, hyperlipidemia, renal disease, hypothyroidism, and the number of outpatient visits for ophthalmologic care during the follow-up period. Several previous studies exploring the link between OSA and OAG were cross-sectional or retrospective in nature and
1561
Ophthalmology Volume 120, Number 8, August 2013 Table 2. Crude and Adjusted Hazard Ratios (HR) of Glaucoma among the Sample Patients during the 5-year Follow-up Periods (n ¼ 6072) Total (n [ 6072) Development of Glaucoma 5-year follow-up period Yes Incidence rate per 1000 person-years (95% CI) Crude HR (95 % CI) Adjusted HR (95 % CI)y
n
%
228 3.8 7.51 (6.58e8.53) e e
Patients with Obstructive Sleep Apnea (n [ 1012) n
%
57 5.6 11.26 (8.61e14.49) 1.71* (1.26e2.32) 1.67* (1.30e2.17)
Comparison Patients (n [ 5060) n
%
171 3.4 6.76 (5.80e7.83) 1.00 1.00
CI ¼ confidence interval. The HR was calculated by stratified Cox proportional hazards regressions (stratified on gender, age group, urbanization level, and the year of index date). *P<0.001. y Adjustments are made for the number of outpatient visits for ophthalmologic care during the follow-up period, monthly income, diabetes, hypertension, coronary heart disease, obesity, hyperlipidemia, renal disease, hypothyroidism, and migraine.
only showed the prevalence of OAG among patients with OSA. One cross-sectional study in Switzerland indicated that patients with OSA are a high-risk population for glaucoma and that respiratory disturbance index correlated positively with intraocular pressure.2 Another case-control study by Girkin et al16 found that patients with glaucoma were more likely to have a previous diagnosis of OSA relative to control subjects. However, this finding was of borderline significance at an alpha level of 0.05 (P ¼ 0.06).16 Furthermore, 2 studies conducted on institutional databases in Turkey and Taiwan reported a high prevalence of glaucoma among patients with OSA.3,5 Nevertheless, methodologic variation (e.g., different populations, study designs, and disease definitions) or limits (e.g., inappropriate
Figure 1. Open-angle glaucoma-free survival rates for patients with obstructive sleep apnea and comparison group in Taiwan
1562
consideration of confounding effects) render more specific comparisons between investigations difficult. Both a mechanical theory and a vascular theory have been proposed to underpin the association between OSA and OAG. The mechanical theory suggests that increased intraocular pressure and stretching of the lamina cribrosa is the reason for glaucoma.20,21 The intermittent upper airway obstruction experienced in OSA patients during sleep would cause hypoxia with a subsequent decrease in PaO2 and increase in PaCO2. This episodic vascular insufficiency may compromise optic nerve perfusion and oxygenation, and subsequently cause optic neuropathy.3,21 Moreover, OSA is often, although not invariably, associated with an increased body mass index. The positive association between obesity and intraocular pressure is proposed to be related to direct damage of the optic nerve.22 This intraocular pressure elevation may be because of the overproduction of aqueous fluid or impaired aqueous outflow.19,23 The intracranial pressure of OSA patients has been found to be markedly elevated in an episodic fashion, paralleling apneic episodes. It is hypothesized that hypercapnea leads to cerebral venous dilatation.24 The vascular theory holds that repetitive or prolonged episodes of hypoxia cause direct damage to the optic nerve. The vascular theory proposes that the association between OSA and OAG can be explained by 3 mechanisms: Impaired optic nerve head blood flow autoregulation secondary to repetitive prolonged apneas, optic nerve vascular dysregulation secondary to arteriosclerosis, and arterial blood pressure variations or imbalance between nitric oxide (a vasodilator) and endothelin (a vasoconstrictor).17,25,26 The vascular theory postulates that there is insufficient blood supply to nourish the nerve fiber layer and/or optic nerve. This arises on account of the recurrent apnea/hypopnea characteristic of OSA which leads to transient hypoxemia and increased vascular resistance which in turn compromises optic nerve head perfusion.3 The main strength of our study lies in its longitudinal and large population database, which allowed us to consider the relationship between OSA and OAG. Furthermore, this database tracks medical utilizations in Taiwan, where >98% of the population are of Han Chinese ethnicity. This homogenous population may avoid potential confounding
Lin et al
Glaucoma and Obstructive Sleep Apnea
Table 3. Crude and Adjusted Hazard Ratios (HR) of Glaucoma among the Sample Patients During the 5-year Follow-up Periods, by Gender Group Gender Group Male Development of Glaucoma Five-year follow-up period Yes Incidence rate per 1000 person-years (95% CI) Crude HR (95 % CI) Adjusted HR (95 % CI)y
Patients with OSA (n, %) 29 9.52 1.51* 1.45z
(4.8) (6.50e13.50) (1.05e2.17) (1.02e2.16)
Female Comparison Patients (n, %) 97 (3.2) 6.43 (5.24e7.80) 1.00 1.00
Patients with OSA (n, %) 28 13.90 1.99* 1.55z
(7.0) (9.42e19.81) (1.27e3.11) (1.04e2.31)
Comparison Patients (n, %) 74 (3.6) 7.25 (5.74e9.05) 1.00 1.00
CI ¼ confidence interval; OSA ¼ obstructive sleep apnea. The HR was calculated by stratified Cox proportional hazards regressions (stratified on gender, age group, urbanization level, and the year of index date). *P<0.01. y Adjustments are made for the number of outpatient visits for ophthalmologic care during the follow-up period, monthly income, diabetes, hypertension, coronary heart disease, obesity, hyperlipidemia, renal disease, hypothyroidism, and migraine. z P<0.05.
by race. Last, clinical diagnostic criteria were used for the identification of OSA and OAG, with the latter further requiring the confirmatory results of a polysomnography to ensure diagnostic validity. Nevertheless, the findings of this study need to be considered in light of several limitations. One of the limitations of our study is that the LHID2000 database represents patients who had sought treatment for OSA, but lacks information on the severity of OSA, such as apneaehypopnea index (AHI) scores or respiratory disturbance index scores. Therefore, we were unable to evaluate whether patients with more severe OSA had a higher risk of OAG than those with mild OSA. Another limitation is that the only patients who seek treatment are those who have some symptoms of suspected OSA, such as snoring, persistent daytime sleepiness or drowsiness while driving, and hypersomnolence. Because the NHI database only included patients who sought treatment, it is possible that some patients may have had OSA but did not receive a diagnosis because they did not seek or receive medical care for this. These patients may have been selected and recruited in the comparison cohort, which would have worked to bias our results toward the null. Moreover, patients with OSA have significantly higher risks of all considered comorbidities, and this might imply that patients with OSA have generally poorer health than those without OSA. OSA may simply be a marker for poor (vascular) health, and not necessarily an independent risk factor for OAG. However, after adjusting for socioeconomic characteristics and medical comorbidities, we found that patients with OSA had an increased risk of OAG diagnosis within the first 5 years after their diagnosis. Therefore, OSA is an independent risk factor for OAG. Finally, this study only included OAG patients who received treatment with topical antiglaucoma medication. Those (n ¼ 8) who received an OAG diagnosis but were untreated during the follow-up period were not counted in this study. However, we reanalyzed the data and found the
results remained even if those 8 cases were included in the OAG cases. In conclusion, our results suggest that OSA is associated with an increased risk of subsequent OAG diagnosis during the first 5 years after OSA diagnosis. We found that the hazard of receiving an OAG diagnosis during the 5-year follow-up period was 1.67 times greater in patients with OSA than in gender- and age-matched comparison subjects, after adjusting for socioeconomic characteristics and medical comorbidities. The authors hope that this study encourages clinicians to alert OSA patients of the association of between OSA and OAG as a means of raising the issue and encouraging treatment of those who need it. Further study is needed to clarify our findings and to explore the mechanisms underlying this association.
References 1. Bendel RE, Kaplan J, Heckman M, et al. Prevalence of glaucoma in patients with obstructive sleep apnoeaea crosssectional case-series. Eye (Lond) 2008;22:1105–9. 2. Mojon DS, Hess CW, Goldblum D, et al. High prevalence of glaucoma in patients with sleep apnea syndrome. Ophthalmology 1999;106:1009–12. 3. Lin PW, Friedman M, Lin HC, et al. Normal tension glaucoma in patients with obstructive sleep apnea/hypopnea syndrome. J Glaucoma 2011;20:553–8. 4. Ayas NT, Brown R, Shea SA. Hypercapnia can induce arousal from sleep in the absence of altered respiratory mechanoreception. Am J Respir Crit Care Med 2000;162:1004–8. 5. Karakucuk S, Goktas S, Aksu M, et al. Ocular blood flow in patients with obstructive sleep apnea syndrome (OSAS). Graefes Arch Clin Exp Ophthalmol 2008;246:129–34. 6. Onen SH, Mouriaux F, Berramdane L, et al. High prevalence of sleep-disordered breathing in patients with primary openangle glaucoma. Acta Ophthalmol Scand 2000;78:638–41. 7. Young T, Palta M, Dempsey J, et al. The occurrence of sleepdisordered breathing among middle-aged adults. N Engl J Med 1993;328:1230–5.
1563
Ophthalmology Volume 120, Number 8, August 2013 8. Chervin RD, Guilleminault C. Obstructive sleep apnea and related disorders. Neurol Clin 1996;14:583–609. 9. Geyer O, Cohen N, Segev E, et al. The prevalence of glaucoma in patients with sleep apnea syndrome: same as in the general population. Am J Ophthalmol 2003;136:1093–6. 10. Marcus DM, Costarides AP, Gokhale P, et al. Sleep disorders: a risk factor for normal-tension glaucoma? J Glaucoma 2001;10:177–83. 11. McNab AA. The eye and sleep. Clin Experiment Ophthalmol 2005;33:117–25. 12. Purvin VA, Kawasaki A, Yee RD. Papilledema and obstructive sleep apnea syndrome. Arch Ophthalmol 2000;118:1626–30. 13. Bucci FA Jr, Krohel GB. Optic nerve swelling secondary to the obstructive sleep apnea syndrome. Am J Ophthalmol 1988;105:428–30. 14. Mojon DS, Mathis J, Zulauf M, et al. Optic neuropathy associated with sleep apnea syndrome. Ophthalmology 1998;105:874–7. 15. Robert PY, Adenis JP, Tapie P, Melloni B. Eyelid hyperlaxity and obstructive sleep apnea (O.S.A.) syndrome. Eur J Ophthalmol 1997;7:211–5. 16. Girkin CA, McGwin G Jr, McNeal SF, Owsley C. Is there an association between pre-existing sleep apnoea and the development of glaucoma? Br J Ophthalmol 2006;90:679–81. 17. Cheng CL, Kao YH, Lin SJ, et al. Validation of the National Health Insurance Research Database with ischemic stroke
18.
19. 20. 21. 22. 23. 24.
25. 26.
cases in Taiwan. Pharmacoepidemiol Drug Saf 2011;20: 236–42. Lin HC, Xirasagar S, Chen CH, Hwang YT. Physician’s case volume of intensive care unit pneumonia admissions and inhospital mortality. Am J Respir Crit Care Med 2008;177: 989–94. Lin HC, Chao PZ, Lee HC. Sudden sensorineural hearing loss increases the risk of stroke: a 5-year follow-up study. Stroke 2008;39:2744–8. Abdal H, Pizzimenti JJ, Purvis CC. The eye in sleep apnea syndrome. Sleep Med 2006;7:107–15. Parish JM, Shepard JW Jr. Cardiovascular effects of sleep disorders. Chest 1990;97:1220–6. Fingeret M, Lewis TL. Primary Care of the Glaucomas. Norwalk, CT: Appleton & Lange; 1993:3-5. Cheung N, Wong TY. Obesity and eye diseases. Surv Ophthalmol 2007;52:180–95. Sugita Y, Iijima S, Teshima Y, et al. Marked episodic elevation of cerebrospinal fluid pressure during nocturnal sleep in patients with sleep apnea hypersomnia syndrome. Electroencephalogr Clin Neurophysiol 1985;60:214–9. Hayreh SS. Acute ischemic disorders of the optic nerve. Pathogenesis, clinical manifestations, and management. Ophthalmol Clin North Am 1996;9:407–42. Lüscher TF. The endothelium and cardiovascular diseaseea complex relation. N Engl J Med 1994;330:1081–3.
Footnotes and Financial Disclosures Originally received: September 17, 2012. Final revision: December 18, 2012. Accepted: January 3, 2013. Available online: April 17, 2013.
6 School of Medical Laboratory Sciences and Biotechnology, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan.
Manuscript no. 2012-1417.
1
Graduate Institute of Biomedical Informatics, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan. 2
Department of Ophthalmology, Taipei Medical University, Taipei, Taiwan.
3 Department of Ophthalmology, Taipei Medical University Hospital, Taipei, Taiwan. 4
Department of Ophthalmology, Shin Kong Wu-Ho-Su Memorial Hospital, Taipei, Taiwan. 5
School of Medicine, Fu-Jen Catholic University, Hsingchuang, Taiwan.
1564
7 Sleep Research Center, Taipei Medical University Hospital, Taipei, Taiwan.
Financial Disclosures: The authors have no proprietary or commercial interest in any of the materials discussed in this article. Correspondence: Herng-Ching Lin, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan, 250 Wu-Hsing St., Taipei 110, Taiwan. E-mail: [email protected].