Retinal Vein Occlusion and Vascular Mortality

Transcription

1 Retinal Vein Occlusion and Vascular Mortality Pooled Data Analysis of 2 Population-Based Cohorts Sudha Cugati, MS, 1 Jie Jin Wang, MMed, PhD, 1 Michael D. Knudtson, MS, 2 Elena Rochtchina, MApplStat, 1 Ronald Klein, MD, MPH, 2 Barbara E. K. Klein, MD, MPH, 2 Tien Yin Wong, MPH, PhD, 3,4 Paul Mitchell, MD, PhD 1 Purpose: To assess the association of retinal vein occlusion (RVO) with cardiovascular and cerebrovascular mortality. Design: Pooled data from 2 population-based cohort studies. Participants: At baseline, the Beaver Dam Eye Study (BDES) examined 4926 persons aged 43 to 86 years (from ) and the Blue Mountains Eye Study (BMES) examined 3654 persons aged 49 to 97 years (from 1992 to 1994). Methods: Retinal vein occlusion was assessed from retinal photographs. Vascular deaths were determined using either death certificates (BDES) or the Australian National Death Index (BMES). Cox regression analysis was used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). Main Outcome Measure: Vascular (cardiovascular and cerebrovascular) mortality was determined. Results: Of 8384 baseline participants, 96 (1.14%) had RVO at baseline (BDES, n 38; BMES, n 58). Over 12 years, 1312 (15.7%) died of cardiovascular-related conditions and 341 (4.1%) died of cerebrovascular-related conditions. Age-standardized vascular mortality rates were 26.0% and 5.3%, respectively, in persons with RVO and 17.1% and 4.5%, respectively, in those without RVO. After adjusting for age, gender, body mass index, hypertension, diabetes, smoking, glaucoma, and study site, RVO was not associated with cardiovascular-related mortality (HR, 1.2; 95% CI, ) or cerebrovascular-related mortality (HR, 0.9; 95% CI, ) among participants of all ages. However, in persons aged less than 70 years, baseline RVO was associated with higher cardiovascular mortality (combined BDES and BMES: HR, 2.5; 95% CI, ; BDES: HR, 2.5; 95% CI, ; BMES: HR, 2.1; 95% CI, ). Conclusions: Retinal vein occlusion in persons aged 43 to 69 years may signal a doubling of the risk of cardiovascular mortality. Ophthalmology 2007;114: by the American Academy of Ophthalmology. Originally received: March 2, Accepted: June 22, Manuscript no Centre for Vision Research, Department of Ophthalmology, Westmead Millennium Institute, University of Sydney, Sydney, Australia. 2 Department of Ophthalmology & Visual Sciences, University of Wisconsin Medical School, Madison, Wisconsin. 3 Centre for Eye Research Australia, University of Melbourne, Melbourne, Australia. 4 Singapore Eye Research Institute, National University of Singapore, Singapore. Supported by the Australian National Health & Medical Research Council, Canberra, Australia (grant nos.: , , , ); National Institutes of Health, Bethesda, Maryland (grant no.: EY [RK, BEKK]); and Research to Prevent Blindness, Inc., New York, New York (RK, BEKK). Correspondence to Jie Jin Wang, MMed, PhD, Centre for Vision Research, Department of Ophthalmology, Westmead Millennium Institute, University of Sydney, Westmead Hospital, Hawkesbury Road, Westmead, New South Wales, Australia jiejin_wang@wmi.usyd.edu.au. Although retinal vein occlusion (RVO) is relatively uncommon in older populations, with a prevalence of 0.3% to 1.6%, 1 3 a 5-year incidence of 0.8%, 3 and a 10-year incidence of 1.7%, 4 it is an important cause of moderate to severe visual loss in older persons. 5 Retinal vein occlusion frequently is associated with concomitant systemic vascular conditions. However, associations of RVO and cardiovascular or cerebrovascular diseases have not been demonstrated consistently, 3,6 11 with most reports using clinical samples. Only the Beaver Dam Eye Study (BDES) has reported mortality associated with RVO from a populationbased sample. 3 Participants with branch RVO at baseline did not have an increased 8-year risk of dying of ischemic heart disease. 3 There have been no population-based studies of cerebrovascular mortality and RVO. In the current study, we aimed to assess the long-term (10 12 years) vascular mortality associated with RVO using pooled data from 2 representative populations: the BDES and the Blue Mountains Eye Study (BMES) cohorts. Patients and Methods The BDES and BMES are 2 population-based cohort studies of vision and common eye diseases in predominantly white pop by the American Academy of Ophthalmology ISSN /07/$ see front matter Published by Elsevier Inc. doi: /j.ophtha

2 Cugati et al Retinal Vein Occlusion and Vascular Mortality ulations examined using similar protocols. Both were approved by the human research ethics committees in their respective institutions and were conducted adhering to tenets of the Declaration of Helsinki. The BDES examined 4926 of 5924 eligible residents aged 43 to 86 years (83.1% response) from 1988 through 1990 in Beaver Dam, Wisconsin. The BMES examined 3654 of 4433 eligible residents aged 49 years or older (82.4% response) from 1992 through 1994 in the Blue Mountains region west of Sydney, Australia. At baseline, all participants underwent a comprehensive eye examination after pupil dilatation. Stereoscopic retinal photographs (30 ) of the macula and other retinal fields of both eyes were taken using a Zeiss FF3 camera (Carl Zeiss, Oberkochen, Germany). 2,3 Retinal photographs were available for at least 1 eye in 98% of participants (BDES, 4829/4926; BMES, 3583/3654), and complete data were available for 8378 persons (98% of 8580 seen at baseline examinations). Signs of central or branch RVO were assessed from photographs, and all cases were adjudicated by retinal specialists (RK, PM). 2,3 Details relating to deaths were obtained from either death certificates for the BDES or the Australian National Death Index for the BMES. Causes of death were collected from death certificates and recorded using International Classification of Diseases (ICD) codes. Codes for cardiovascular death included ICD 9 ( , , , ) and ICD 10 (I01.0 I09.9, I11.0 I11.9, I13.0 I13.9, and I20.0 I51.9), and cerebrovascular death included ICD 9 ( ) and ICD 10 (I60.0 I69.9). The validity of death certificates 12 and Australian National Death Index data 13,14 were reported previously. At baseline, systolic and diastolic blood pressures were recorded. Hypertension was defined as systolic blood pressure of 140 mmhg or more, diastolic blood pressure of 90 mmhg or more, or current self-reported use of antihypertensive medications. Current smokers were defined using interviewer-administrated questionnaires. Diabetes was defined as a previous history of diabetes or hyperglycemia, or a glycosylated hemoglobin more than 2 standard deviations above the mean for appropriate age and gender groups, or casual blood glucose level 200 mg/dl (11.1 mmol/l) or more in the BDES or fasting blood glucose 7.0 mmol/l or more in the BMES. Serum total cholesterol and high-density lipoprotein (HDL) were measured from casual (BDES) or fasting (BMES) blood specimens using standard procedures. In the BMES, primary open-angle glaucoma was diagnosed when typical glaucomatous visual field defects on the Humphrey 30-2 test (Humphrey Instruments, San Leandro, CA) were present, matching optic disc rim thinning, including an enlarged cup-to-disc ratio ( 0.7) or cup-todisc ratio asymmetry ( 0.3) between the 2 eyes. In the BDES, primary open-angle glaucoma was diagnosed based on 2 of the following 3 criteria: (1) visual field defect compatible with diagnosis of glaucoma; (2) cup-to-disc ratio of 0.8 or more, or difference in the ratio of 0.2 or more between the 2 eyes; 3) intraocular pressure of 22 mmhg or more. Statistical Analysis System software (version 9; SAS Institute, Cary, NC) was used for analyses. Age-standardized mortality rates were obtained after direct age standardization of patients with RVO to the entire study population. Cox regression models estimated hazard ratios (HRs) and 95% confidence intervals (CIs) for cardiovascular and cerebrovascular death after adjusting for age; gender; body mass index; current smoking; presence of hypertension, diabetes, and glaucoma; total cholesterol and HDL levels (per 10 mmol/l); and study site. Subgroup analyses were conducted stratified by gender and age group ( 70 years and 70 years). Results Table 1 shows age- and gender-specific prevalence of RVO at baseline, plus baseline characteristics by RVO status in the 2 populations. Age-specific RVO prevalence was nearly double in the BMES compared with the BDES. Participants in both studies with baseline RVO were significantly older and were more likely to have hypertension. In the BDES, participants with RVO also were more likely to have diabetes. In the BMES, participants with RVO were more likely to report a history of angina or stroke. Prevalent glaucoma was more common in those with RVO compared with those without RVO (Table 1). Of 8378 participants, RVO was detected in 96 participants (1.1%) at baseline (BDES, n 38; BMES, n 58), of whom 71 had branch RVO (BDES, n 31; BMES, n 40). Over 10 to 12 years, 1312 participants (15.7%) died of cardiovascular or cardiovascular-related conditions. The cumulative cardiovascular mortality was significantly higher among participants with than without RVO at baseline (30.2% vs. 15.5%). After age standardization, mortality rates were 26.0% and 17.1% per 1000 personyears among persons with and without RVO, respectively. In participants of all ages, there was no significant difference in this mortality risk between persons with and without RVO, after adjusting for age and gender (HR, 1.2; 95% CI, ) or after further adjustment for body mass index, hypertension, diabetes, glaucoma, current smoking, and study site (HR, 1.2; 95% CI, ). In participants aged 43 to 69 years, however, RVO was associated with a 2-fold higher risk of cardiovascular mortality in a multivariable model (HR, 2.5; 95% CI, ; Table 2). After replacing the covariable hypertension by systolic blood pressure in the multivariate regression model, the association was similar for participants aged 43 to 70 years (HR, 2.4; 95% CI, ) and persisted after further adjustment for baseline history of stroke or acute myocardial infarction (HR, 2.5; 95% CI, ). This association was not found among participants aged 70 years or older or in either gender subgroup of all ages (Table 2). Similar findings were found for the individual study data, although these findings did not reach statistical significance (Table 2). Baseline total cholesterol was not significant in the multivariate model in either age group, whereas HDL cholesterol was significant in the model for participants aged 43 to 69 years. After further adjustment for HDL in the multivariate model, the association between RVO and mortality remained significant in participants aged 43 to 69 years (HR, 2.6; 95% CI, ). The associations of branch RVO and central RVO with cardiovascular mortality in this age group were similar to those observed for any RVO (branch RVO: adjusted HR, 2.5; 95% CI, ; central RVO: HR, 2.2; 95% CI, ). Over the 10 to 12 years, 341 participants (4.1%) died of cerebrovascular-related conditions. The cumulative cerebrovascular mortality was higher in participants with than without RVO at baseline (7.3% vs. 4.0%). After age standardization, mortality rates were 5.3% versus 4.5% per 1000 person-years among persons with and without RVO, respectively. There was no significant difference overall in the mortality of persons with and without RVO (multivariate adjusted HR, 1.1; 95% CI, ; Table 3). In analyses stratified by gender, however, men with RVO at baseline had a nonsignificant, 2-fold higher risk of cerebrovascular mortality than men without RVO (HR, 2.3; 95% CI, ; Table 3). This association was attenuated after further adjustment for history of stroke (HR, 1.8; 95% CI, ). No significant association was observed in women, or in men and women combined for those 70 years of age or older (Table 3). There were no cerebrovascular events among persons in the younger group (aged years) 521

3 Ophthalmology Volume 114, Number 3, March 2007 Table 1. Age- and Gender-Specific Prevalence (%) of Retinal Vein Occlusion and Baseline Characteristics by Retinal Vein Occlusion Status and Study Site Retinal Vein Occlusion Not Present Beaver Dam Eye Study Retinal Vein Occlusion Present P Value Retinal Vein Occlusion Not Present Blue Mountains Eye Study Retinal Vein Occlusion Present Retinal vein occlusion prevalence by age and gender (prevalence) Age group 55 yrs (0.3%) (0.8%) yrs (0.4%) (0.7%) yrs (1.2%) (1.6%) 75 yrs (1.9%) (3.7%) Total (0.8%) (1.6%) P value for trend Men (0.8%) (1.4%) Women (0.8%) (1.8%) P value Baseline characteristics Age (yrs), mean (95% CI) 61.8 ( ) 70.1 ( ) 65.9 ( ) 71.6 ( ) Body mass index, mean (95% CI) 28.8 ( ) 29.2 ( ) 26.2 ( ) 25.5 ( ) Hypertension (%) Diabetes (%) Current smoker (%) Body mass index category (%) History of angina (%) History of acute myocardial infarction (%) History of stroke (%) Presence of primary open-angle glaucoma (%) P Value CI confidence interval. with RVO. We were not able to analyze data by individual study because of small numbers. Discussion In this pooled data analysis of 2 white population-based cohorts, we found that in persons aged 43 to 69 years, the presence of RVO was associated with a 2-fold higher risk of cardiovascular mortality. Also in men, presence of RVO was associated with a nonsignificant, 2-fold higher risk of cerebrovascular mortality for all ages. These findings support an association between RVO and vascular mortality, suggested by previous clinical studies. 7 9 In our study population, glaucoma patients receiving timolol treatment had an increased risk of vascular mortality. 15 We therefore included glaucoma as a covariable in our multivariate adjusted models. The association of RVO with cardiovascular mortality in the younger subgroup seemed to be independent of other vascular risk factors and glaucoma. The long-term prognosis of RVO was documented in the early 1970s, 10 and no overall decrease in life expectancy among patients with RVO was found, although the number of deaths resulting from vascular causes in these patients was double the national rate. A subsequent study in the 1990s also documented an association between RVO and vascular mortality. 8 A recent study using the Framingham algorithm 7 estimated higher mortality resulting from cardiovascular disease among patients with RVO. Our study provides evidence from population-based samples to support these earlier clinic-based findings. The pathogenesis of branch and central RVO remains uncertain, but is likely multifactorial. Two different mechanisms have been postulated, that is, thrombosis in the vein resulting from compression by atherosclerotic changes in the adjacent artery and a local alteration of the blood flow resulting from unfavorable physiologic factors. The most common risk factors are increasing age, arterial hypertension, and primary open-angle glaucoma. 2,4 In a small proportion of patients with branch RVO, retinal vasculitis may be present as a result of various relatively rare conditions such as Behçet disease, lupus, Eales disease, retinal periphlebitis, viral phlebitis, frosted branch angiitis, Takayasu arteritis, lyme borreliosis, human immunodeficiency virus, and toxoplasmosis. 16 Other more frequent systemic associations include diabetes and hypercoagulable states. 11,17 Because the central retinal vein narrows as it traverses the lamina cribrosa, any condition that contributes to further narrowing or increases the turbulence of flow through this partial barrier may predispose a patient to thrombosis of the 522

4 Table 2. Long-term Cardiovascular Mortality by Presence of Retinal Vein Occlusion at Baseline and by Study Site Baseline Retinal Vein Occlusion Status Cugati et al Retinal Vein Occlusion and Vascular Mortality Hazard Ratio (95% Confidence Interval) Combined Beaver Dam and Blue Mountains Eye Studies n (%) Age and Gender Adjusted Adjusted* Beaver Dam Eye Study Adjusted Blue Mountains Eye Study Adjusted All ages and both genders Absent 1283/8282 (15.5) Present 29/96 (30.2) 1.2 ( ) 1.2 ( ) 1.1 ( ) 1.2 ( ) 70 yrs Absent 443/5743 (7.7) Present 8/37 (21.6) 2.7 ( ) 2.5 ( ) 2.5 ( ) 2.1 ( ) 70 yrs or more Absent 840/2539 (33.1) Present 21/59 (35.6) 1.0 ( ) 1.0 ( ) 0.9 ( ) 1.0 ( ) Men Absent 668/3640 (18.4) Present 14/38 (36.8) 1.4 ( ) 1.4 ( ) 1.2 ( ) 1.6 ( ) Women Absent 615/4642 (13.3) Present 15/58 (25.9) 1.0 ( ) 1.1 ( ) 1.0 ( ) 1.0 ( ) *Adjusted for age, gender, body mass index, current smoking, presence of hypertension, diabetes, glaucoma, and study site. Adjusted for age, gender, body mass index, current smoking, and presence of hypertension, diabetes, and glaucoma. central retinal vein. Thus, glaucoma is very frequent, particularly for central RVO. 2 The association of RVO and cardiovascular mortality among persons aged 43 to 69 years could be explained in part by the pathogenesis of RVO. Retinal vasculitis is an important cause of RVO in young adults, 18 and systemic inflammation has been found associated with all-cause or cardiovascular mortality in the elderly The association of RVO and cerebrovascular mortality found in men but not in women could be explained in part by the higher risk of stroke-related mortality among men than women. 22 It is possible that the small number of patients with RVO who died of cerebrovascular causes could have led to a chance finding or a biased estimate. In these pooled population-based data, there was only 1 woman with RVO who had a stroke-related death; hence, our findings should be treated with caution. Reasons for the higher prevalence of RVO in the BMES compared with the BDES are not entirely clear. Differences in the photographic grading of some questionable cases between the 2 studies may be responsible in part. 3 We carefully regraded all BMES cases and excluded 8 questionable cases, but the difference remained (Table 1). The older age range (49 97 years in BMES vs years in Table 3. Long-term Cerebrovascular Mortality by Presence of Retinal Vein Occlusion at Baseline and by Study Site Baseline Retinal Vein Occlusion Status, Combined Beaver Dam and Blue Mountains Eye Study n (%) Hazard Ratio (95% Confidence Interval) Age and Gender Adjusted Adjusted* All ages Absent 334/8282 (4.0) Present 7/96 (7.3) 0.9 ( ) 0.9 ( ) 70 yrs Absent 87/5743 (1.5) Present 0/37 (0) N/A 70 yrs or more Absent 247/2539 (9.7) Present 7/59 (11.9) 1.0 ( ) 1.0 ( ) Men Absent 153/3640 (4.2) Present 6/38 (15.8) 2.5 ( ) 2.3 ( ) Women Absent 181/4642 (3.9) Present 1/58 (1.7) 0.2 ( ) 0.2 ( ) *Adjusted for age, gender, body mass index, current smoking, presence of hypertension, diabetes, glaucoma, and study site. 523

5 Ophthalmology Volume 114, Number 3, March 2007 BDES) and higher prevalence of hypertension in the BMES population are possible reasons. The age- and gender-standardized prevalence of hypertension among BMES participants was 48.7% compared with 43.6% in the BDES (P 0.01). 23 We cannot exclude possible differences in environmental and genetic factors between the 2 study populations. The cross-sectional associations of RVO and baseline cardiovascular risk factors in the 2 populations were similar. The population-based nature of our samples tends to eliminate potential selection bias, whereas use of similar grading procedures in photographic documentation of RVO tends to eliminate misclassification. Although these study strengths enhance the quality of the evidence provided, small numbers in subgroup analyses limits study power, so possible chance findings cannot be excluded. Our study may have potential implications for clinical management of patients with RVO. Although there are previous clinical reports on the systemic prognosis of RVO, available evidence regarding the risk of cardiovascular disease has been conflicting. Our study suggests that some patients with RVO may benefit from a careful cardiovascular risk assessment. However, we should emphasize that our study was not designed to address this question directly. Independent confirmation of the cardiovascular associations found in our study in larger case series is useful to determine whether routine cardiovascular risk assessments should be performed for patients with RVO. References 1. Wong TY, Larsen EK, Klein R, et al. Cardiovascular risk factors for retinal vein occlusion and arteriolar emboli. The Atherosclerosis Risk in Communities & Cardiovascular Health studies. Ophthalmology 2005;112: Mitchell P, Smith W, Chang A. Prevalence and associations of retinal vein occlusion in Australia: the Blue Mountains Eye Study. Arch Ophthalmol 1996;114: Klein R, Klein BE, Moss SE, Meuer SM. The epidemiology of retinal vein occlusion: the Beaver Dam Eye Study. Trans Am Ophthalmol Soc 2000;98: Cugati S, Wang JJ, Rochtchina E, Mitchell P. Ten-year incidence of retinal vein occlusion in an older population: the Blue Mountains Eye Study. Arch Ophthalmol 2006;124: Klein R, Wang Q, Klein BE, et al. The relationship of agerelated maculopathy, cataract, and glaucoma to visual acuity. Invest Ophthalmol Vis Sci 1995;36: Mansour AM, Walsh JB, Henkind P. Mortality and morbidity in patients with central retinal vein occlusion. Ophthalmologica 1992;204: Martin SC, Butcher A, Martin N, et al. Cardiovascular risk assessment in patients with retinal vein occlusion. Br J Ophthalmol 2002;86: Tsaloumas MD, Kirwan J, Vinall H, et al. Nine year follow-up study of morbidity and mortality in retinal vein occlusion. Eye 2000;14: Priluck IA, Robertson DM, Hollenhorst RW. Long-term follow-up of occlusion of the central retinal vein in young adults. Am J Ophthalmol 1980;90: Rubinstein K, Jones EB. Retinal vein occlusion: long-term prospects: 10 years follow-up of 143 patients. Br J Ophthalmol 1976;60: Elman MJ, Bhatt AK, Quinlan PM, Enger C. The risk for systemic vascular diseases and mortality in patients with central retinal vein occlusion. Ophthalmology 1990;97: Iso H, Jacobs DR Jr, Goldman L. Accuracy of death certificate diagnosis of intracranial hemorrhage and nonhemorrhagic stroke: the Minnesota Heart Survey. Am J Epidemiol 1990; 132: Powers J, Ball J, Adamson L, Dobson A. Effectiveness of the National Death Index for establishing the vital status of older women in the Australian Longitudinal Study on Women s Health. Aust N Z J Public Health 2000;24: Magliano D, Liew D, Pater H, et al. Accuracy of the Australian National Death Index: comparison with adjudicated fatal outcomes among Australian participants in the Long-term Intervention with Pravastatin in Ischaemic Disease (LIPID) study. Aust N Z J Public Health 2003;27: Lee AJ, Wang JJ, Kifley A, Mitchell P. Open-angle glaucoma and cardiovascular mortality: the Blue Mountains Eye Study. Ophthalmology 2006;113: Lang GE, Spraul CW. Risk factors for retinal occlusive diseases [in German]. Klin Monatsbl Augenheilkd 1997;211: Appiah AP, Trempe CL. Risk factors associated with branch vs. central retinal vein occlusion. Ann Ophthalmol 1989;21: 153 5, Fong AC, Schatz H. Central retinal vein occlusion in young adults. Surv Ophthalmol 1993;37: De Martinis M, Franceschi C, Monti D, Ginaldi L. Inflammation markers predicting frailty and mortality in the elderly. Exp Mol Pathol 2006;80: Margolis KL, Manson JE, Greenland P, et al. Leukocyte count as a predictor of cardiovascular events and mortality in postmenopausal women: the Women s Health Initiative Observational Study. Arch Intern Med 2005;165: Shankar A, Mitchell P, Rochtchina E, Wang JJ. The association between circulating white blood cell count, triglyceride level and cardiovascular and all-cause mortality: populationbased cohort study. Atherosclerosis. In press. 22. Lotufo PA, Bensenor IM. Stroke mortality in Sao Paulo ( ): a description using the Tenth Revision of the International Classification of Diseases. Arq Neuropsiquiatr 2004;62: Klein BE, Klein R, Jensen SC, Linton KL. Hypertension and lens opacities from the Beaver Dam Eye Study. Am J Ophthalmol 1995;119: Answers for CME credit: 1. C; 2. B; 3. D; 4. B; 5. A 524