EPIRETINAL MEMBRANE, ALSO REFERRED AS PRIMARY

Transcription

1 Prevalence and Associations of Epiretinal Membranes in the Visual Impairment Project DANIEL J. MCCARTY, PHD, BICKOL N. MUKESH, PHD, VATSAL CHIKANI, MPH, JIE J. WANG, MMED, PHD, PAUL MITCHELL, MD, PHD, HUGH R. TAYLOR, MD, AND CATHERINE A. MCCARTY, PHD, MPH PURPOSE: To determine the prevalence and factors associated with epiretinal membranes in a random sample of the population aged 40 years and older in Victoria, Australia. DESIGN: Population-based cross-sectional study. METHODS: Detailed eye examinations, including retinal photographs, were conducted in 1992 and 1997 in 3271 people (83% of the eligible) in Melbourne and 1473 (92% of the eligible) in rural Victoria. Eyes present with either cellophane macular reflex (CMR) or preretinal macular fibrosis (PMF) were classified as having epiretinal membranes. Eyes with both CMR and PMF present were classified as having PMF. Age-standardized prevalence rates and 95% confidence limits were calculated by the direct methods using Segi s world population. RESULTS: Epiretinal membranes were observed in 253 of 4313 participants (6.0%; 95% confidence interval [CI] 5.2 to 6.7), bilaterally in 19%. Prevalence increased significantly by age group (0.5% for 40 to 49 years, 2.6% for 50 to 59 years, 9.4% for 60 to 69 years, 15.1% for 70 to 79 years, and 11.3% for 80 years and older). Prevalence was similar in males and females after adjusting for age. The overall age- and gender-standardized prevalence of CMR was 4.8% (95% CI 4.0 to 5.6) and PMF was 1.7% (95% CI 1.2 to 2.3). A decrease in visual acuity (<6/6) was significantly associated with idiopathic Accepted for publication Mar 10, From the Marshfield Clinic Research Foundation, Marshfield, Wisconsin (D.J.M., B.N.M., V.C., C.A.M.); Centre for Eye Research Australia, University of Melbourne, East Melbourne, Australia (D.J.M., B.N.M., H.R.T., C.A.M.); Department of Ophthalmology, University of Sydney (Centre for Vision Research, Westmead Millennium Institute), Sydney, Australia (J.J.W., P.M.); and Vision Cooperative Research Centres, Australia (J.J.W., P.M., H.R.T.). The Melbourne VIP was sponsored in part by grants from the National Health and Medical Research Council, the Victorian Health Promotion Foundation, the Ansell Ophthalmology Foundation, the estate of the late Dorothy Edols, the Jack Brockhoff Foundation, the Eye, Ear, Nose, and Throat Research Institute, the Ian Potter Foundation, the Felton Bequest, the Apple Family Bequest, and the Hugh D. Williamson Trust. Catherine McCarty was the recipient of the Wagstaff Fellowship in Ophthalmology from the Royal Victorian Eye and Ear Hospital. Inquiries to Daniel J. McCarty, PhD, Marshfield Clinic Research Foundation, 1000 North Oak Avenue (ML2), Marshfield, WI 54449; fax: (715) ; mccarty.daniel@mcrf.mfldclin.edu PMF (odds ratio [OR] 1.9; 95% CI 1.0 to 3.6) and CMR (OR 1.5; 95% CI 1.1 to 2.0) after adjusting for age. CONCLUSIONS: The prevalence of epiretinal membranes was similar to that reported in other populationbased studies. Population shifts in the age distribution to older ages could lead to an increase in mild visual impairment caused by epiretinal membranes. (Am J Ophthalmol 2005;140: by Elsevier Inc. All rights reserved.) EPIRETINAL MEMBRANE, ALSO REFERRED AS PRIMARY retinal folds, silent central retinal vein obstruction, 1 wrinkling of the internal retinal surface, 2 preretinal gliosis, preretinal macular fibrosis, 3 cellophane maculopathy, and macular pucker, 4 is a condition characterized by proliferation of abnormal tissues on the surface of the macula or central retina of the eye. 5 Epiretinal membranes are associated with a variety of ocular diseases, 6 such as vascular occlusion, diabetic retinopathy, cataract, retinal detachment surgery, 7,8 and posterior vitreous detachment, 9,10 and also occur after photocoagulation, inflammation, or prior trauma. 6,11 Epiretinal membranes occurring without any antecedent ocular condition or surgical procedure are termed idiopathic 6 ; this definition includes posterior vitreous detachment. Pathogenic theories of idiopathic epiretinal membranes include migration of retinal glial cells through defects in the internal limiting membrane, such as retinal pits or holes, and from the optic nerve head in association with posterior vitreous detachment. 8,12 17 These cells engage in proliferative activity that results in the fibrocellular tissue termed epiretinal membranes. Other cell types have also been implicated in the proliferative process, including fibrocytes, myofibroblasts, macrophages, inflammatory cells, hyalocytes, and vascular elements. Clinically, epiretinal membranes start as a mild glinting, water-silk, shifting light reflex on ophthalmoscopic examination 3,11 known as cellophane macular reflex (CMR). At this time, superficial retinal vessels cast visible shadows on the pigment epithelium with the oblique slit-lamp illumination. As the membrane thickens and contracts, superfi- 288.e BY ELSEVIER INC. ALL RIGHTS RESERVED /05/$30.00 doi: /j.ajo

2 cial retinal folds or traction lines appear and a glinting reflex becomes opaque and gray, and is seen as clumps or bands extending over the retinal surface known as preretinal macular fibrosis (PMF). As the retinal traction lines develop, the small vessels become increasingly tortuous and macular edema may develop. This change is often apparent on fluorescein angiography. Treatment for epiretinal membranes is uncommonly performed and consists of vitrectomy with membrane peeling. A small percentage of patients may undergo spontaneous retraction of the membrane. Machemer 18 first described the surgical removal of epiretinal membranes using pars plana vitrectomy techniques. Since then, many case series have reported pars plana vitrectomy combined with epiretinal membrane tearing as an effective method of treating epiretinal membranes The majority of patients who develop epiretinal membranes are over age 50 years. 11,12,23,24 Occasionally, epiretinal membranes can develop in children and young adults. 25 In early stages, epiretinal membranes may be asymptomatic or may create a mild reduction in visual acuity, which seldom progresses below 20/200. 3,12,26,27 Metamorphopsia, central blurring, and distortion of the Amsler grid pattern develop once the foveal center is involved. Absolute scotomas are rare. In some cases, membrane contraction may exert tangential traction on the macular retina causing severe vision loss. 24 Roth and Foos 2 found a 5.4% incidence of surface wrinkling retinopathy in a series of 500 autopsy eyes; 29.6% of them were bilateral. Although none of these individuals had undergone ocular surgery, a number of them had ocular abnormalities such as diabetic retinopathy, retinal tears, and vaso-occlusive disease. There are only a few published population-based studies of epiretinal membranes The purpose of this study is to determine the prevalence and factors associated with epiretinal membranes in a random sample of the population aged 40 years and older in Victoria, Australia. METHODS VISUAL IMPAIRMENT PROJECT (VIP) STUDY METHODS HAVE been published previously. 33 The VIP is a populationbased study of eye diseases in the 40 and over age group. Using 1986 Census Collector Districts information, nine pairs of adjacent Census Collector Districts in Melbourne and four pairs of Census Collector Districts in rural Victoria were selected by cluster random sampling. A household census was conducted by field interviewers to identify eligible residents. The eligibility criteria for participation in the study were: (1) 40 years of age or older and (2) resident at the target address for 6 months or more. The eligible residents were interviewed to collect basic demographic information and elicit data on their use of glasses or contact lenses and eye health care services. They were then invited to attend the eye examination at the local screening center. Modified home visits were offered for residents who were not mobile enough to attend the screening center. The protocol was approved by the Human Research and Ethics Committee at the Royal Victorian Eye and Ear Hospital, and written, informed consent was obtained from all participants. A standardized ophthalmologic examination was performed on all eligible participants. Participants current optical correction was measured using the Humphrey automatic lens analyzer model 330. Distance visual acuity was measured using the logarithm of minimal angle of resolution (logmar) Early Treatment Diabetic Retinopathy Study 4-m chart. The standard ambient visual acuity room illumination was 100 lux. Near visual acuity was measured using the logmar word reading card at the participant s preferred reading distance. A directional E card on the logmar word reading card was used for illiterate or non-english-speaking people. Participants underwent a detailed clinical examination of each eye. Intraocular pressure was measured with handheld applanation tonometer after pupil dilation. 33 One drop of tropicamide (0.5%) and one drop of phenylephrine hydrochloride (10%) were instilled in both eyes to obtain a minimum pupil size of 6 mm. The fundus was examined by both indirect biomicroscopy at the slit lamp (using a 90-diopter convex lens) and by indirect ophthalmoscopy. Color stereoscopic photographs, centered on both the optic disk and fovea of each eye (Diabetic Retinopathy Study fields 1 and 2), were taken using a Topcon EFT camera (Topcon Corporation, Tokyo, Japan) with Kodachrome 64 slide film (Kodak, Australia). The 35 mm slides were mounted in clear plastic sheets (Bardes Products, Milwaukee, Wisconsin) that permitted close apposition of stereoscopic pairs for grading at the Sydney grading center. The photographs were graded by two ophthalmologists with retinal subspecialty training (P.M., J.J.W.). 34 Epiretinal membranes were identified and graded using methods described by Klein and associates. 28 Epiretinal membranes were classified into an early form without retinal folds (CMR) or a later stage with retinal folds (PMF). Eyes with both CMR and PMF present were classified as having PMF. Epiretinal membranes were classified as secondary if related to past cataract surgery, retinal vascular disease, and retinal detachment. If they were present in people without a secondary cause, they were considered idiopathic. The presence of macular edema or previous retinal laser treatment was also recorded. Macular edema was defined as thickening of the retina within 1 disk diameter of the center of the macula. The personal interview was conducted by trained interviewers. Information regarding ocular history, previous diagnosis of ocular disorders, diabetes, gout, arthritis, hypertension, current and past use of medications, alcohol intake, smoking history, and self-reported height and VOL. 140, NO. 2 EPIRETINAL MEMBRANES IN THE VISUAL IMPAIRMENT PROJECT 288.e2

3 TABLE 1. Weighted Prevalence of Epiretinal Membrane (ERM) and 95% Confidence Interval in Either Eye by Age and Gender Variable Total (n) PMF CMR Any ERM Sex* Male ( ) 5.0 ( ) 6.0 ( ) Female ( ) 4.6 ( ) 5.9 ( ) Age (y) ( ) 0.4 ( ) 0.5 ( ) ( ) 1.9 ( ) 2.6 ( ) ( ) 7.9 ( ) 9.4 ( ) ( ) 11.6 ( ) 15.1 ( ) ( ) 8.3 ( ) 11.3 ( ) Total* ( ) 4.8 ( ) 6.0 ( ) CMR cellophane macular reflex; ERM epiretinal membrane; PMF preretinal macular fibrosis. *Age-standardized prevalence. TABLE 2. Age-Standardized Prevalence (95% CI) of Idiopathic and Secondary Epiretinal Membranes by Gender Epiretinal Membrane Classification Male Female Total Idiopathic (number at risk 3916) PMF 1.2 ( ) 1.6 ( ) 1.4 ( ) CMR 4.5 ( ) 4.2 ( ) 4.4 ( ) Any ERM 5.4 ( ) 5.3 ( ) 5.4 ( ) Secondary (number at risk 397) PMF 7.2 ( ) 5.7 ( ) 6.5 ( ) CMR 10.6 ( ) 11.9 ( ) 11.2 ( ) Any ERM 15.5 ( ) 17.5 ( ) 16.4 ( ) CI confidence interval; CMR cellophane macular reflex; ERM epiretinal membrane; PMF preretinal macular fibrosis. weight was collected during interview. Interview data were entered directly into a PARADOX (Corel, Ottawa, Canada) database that included consistency checks both within and between time points. Data from the vision, eye examinations, and photoscopic gradings were entered twice into the computer, and any discrepancies were corrected. Statistical analyses were performed with SAS (SAS Institute, Cary, North Carolina, USA), and P.05 was considered statistically significant. Age-standardized prevalence rates and 95% confidence limits 35 were calculated by the direct methods using Segi s world population. To measure the overall prevalence of epiretinal membranes, weighting of Melbourne and rural Victoria data was carried out based on the information of the 1996 census. 2 analyses were employed to assess the significance of univariate associations, and multiple logistic regression was employed to determine the independence of potential risk factors. Epiretinal membrane cases were defined when epiretinal membranes were present in either eye. Subject-specific analyses were performed. RESULTS A TOTAL OF 4744 PARTICIPANTS WERE IN THE STUDY: 3271 (84% of eligible residents) from urban and 1473 (92% of eligible residents) from rural cohort. Of these, gradable fundus photographs for epiretinal membranes were available for 4313 participants (91%). Among the excluded were 401 participants (8%) who had ungradable or no fundus photographs taken and 30 participants with signs of late age-related maculopathy. The mean age of participants was 60.1 years (SD 12.8), and 53% were women. Epiretinal membranes were observed in 253 of 4313 participants (6.0%; 95% confidence interval [CI] 5.2 to 6.7). The prevalence of CMR was 4.8% (95% CI 4.0 to 5.6), and PMF was 1.7% (95% CI 1.2 to 2.3) in this study population. Age- and gender-specific prevalence rates of epiretinal membranes are listed in Table 1. The overall prevalence of epiretinal membranes was not significantly different by gender (6.0% in male; 95% CI 4.7 to 7.3, and 5.9% in women; 95% CI 5.5 to 6.3). 288.e3 AMERICAN JOURNAL OF OPHTHALMOLOGY AUGUST 2005

4 TABLE 3. Prevalence of Epiretinal Membranes by Secondary Cause Secondary Cause N PMF CMR Any ERM Retinal detachment (10.3%) 10 (4.5%) 33 (14.7%) Cataract surgery (8.0%) 20 (14.5%) 31 (22.5%) Diabetic retinopathy 57 2 (3.5%) 2 (3.5%) 4 (7.0%) Vein occulusion 23 3 (13.0%) 1 (4.3%) 4 (17.4%) CMR cellophane macular reflex; ERM Epiretinal membrane; PMF preretinal macular fibrosis. FIGURE. Age-adjusted visual acuity by presence or absence of idiopathic and secondary epiretinal membranes. Prevalence of epiretinal membranes increased significantly by age (P.001): 0.5% for persons 40 to 49 years of age, 2.6% for persons 50 to 59 years of age, 9.4% for persons 60 to 69 years of age, 15.1% for persons 70 to 79 years of age, and 11.3% for persons aged 80 years or older. Participants aged 70 years and older were four times (95% CI 3.1 to 5.2) more likely to have epiretinal membranes when compared with participants less than 70 years of age. The prevalence of CMR was higher for each age group compared with the prevalence of PMF. There were 250 participants with epiretinal membranes with gradable photographs of both eyes. Of these, 48 (19%) were bilateral. The prevalence of epiretinal membranes without known predisposing causes was considered idiopathic, and with other potential ocular pathology was considered secondary. The prevalence rates for these groups are shown in Table 2. The overall prevalence of idiopathic epiretinal membranes in 3916 subjects was 5.4% (95% CI 4.6 to 6.1), with significantly higher prevalence of CMR (4.4%; 95% CI 3.5 to 5.2) than PMF (1.4%; 95% CI 0.9 to 1.9). The overall prevalence was similar in men (5.4%) and women (5.3%). The prevalence of secondary epiretinal membranes in 397 subjects without known ocular pathologies was 16.4% (95% CI 12.4 to 20.5) with nonsignificantly higher prevalence of CMR (11.2%; 95% CI 7.6 to 14.8) than PMF (6.5%; 95% CI 3.5 to 9.5). Women (17.5%; 95% CI 13.7 to 21.4) were slightly more affected with secondary epiretinal membranes than men (15.5%; 95% CI 8.6 to 22.4). The prevalence of secondary epiretinal membranes by each cause is listed in Table 3. The most common cause of secondary epiretinal membranes was retinal detachment (10.3%) and retinal vein occlusion (13%). The highest prevalence of secondary CMR was seen in participants who had undergone cataract surgery (14.5%). The prevalence rates of secondary epiretinal membranes were significantly higher for cataract surgery (P.001), retinal detachment (P.001), and retinal vein occlusion (P.02, Fisher s exact test) when compared with the idiopathic group. For diabetic retinopathy the rate was slightly higher, but the difference was not statistically significant (P.36, Fisher s exact test). LogMAR visual acuity was assessed in the worst eye in subjects with idiopathic and secondary epiretinal membranes and in subjects without epiretinal membranes (Figure). Compared with eyes without epiretinal membranes, eyes with idiopathic epiretinal membranes had a decrease in visual acuity after adjusting for age. The average number of letters read in idiopathic epiretinal membrane was 56.5 (SD 6.8) compared with 54.2 (SD 7.4) letters in the group with no epiretinal membrane (P.001). No statistical significant difference was observed in secondary epiretinal membrane after adjusting for age (P.18). Logistic regression analysis was performed to identify factors significantly associated with idiopathic epiretinal membrane, after excluding subjects with any known secondary cause for epiretinal membrane (397) and all cases with diabetic retinopathy (Table 4). After adjusting for age and sex, the results indicated an association of PMF with urban population (odds ratio [OR] 3.1; 95% CI 1.3 to 7.0), myopia (OR 2.2; 95% CI 1.0 to 4.6), and visual acuity 6/6 (OR 1.9; 95% CI 1.0 to 3.6). CMR was significantly associated with visual acuity 6/6 (OR 1.5; 95% CI 1.1 to 2.0), as was the overall prevalence of epiretinal membranes (OR 1.6; 95% CI 1.2 to 2.2). A moderate protective association was found between smoking (OR 0.6; 95% CI 0.5 to 0.9) and epiretinal membranes. Glaucoma, history of cardiovascular disease, and self-reported diabetes were not significantly associated with idiopathic epiretinal membranes. VOL. 140, NO. 2 EPIRETINAL MEMBRANES IN THE VISUAL IMPAIRMENT PROJECT 288.e4

5 TABLE 4. Relationship of Idiopathic Epiretinal Membrane (ERM) to Selected Characteristics After Excluding Cases With Diabetic Retinopathy and Other Known Secondary Causes Age- and Sex-adjusted Odds Ratios (95% CI) Characteristic PMF CMR Any ERM Urban 3.1 ( ) 0.9 ( ) 1.1 ( ) Smoking 0.5 ( ) 0.7 ( ) 0.6 ( ) Alcohol consumption 0.9 ( ) 1.1 ( ) 1.0 ( ) Self-reported high blood pressure 0.7 ( ) 0.9 ( ) 0.9 ( ) Self-reported diabetes 1.3 ( ) 1.3 ( ) 1.4 ( ) Early ARM 0.5 ( ) 1.0 ( ) 0.9 ( ) Myopia 2.2 ( ) 0.8 ( ) 1.1 ( ) Hyperopia 0.7 ( ) 0.8 ( ) 0.8 ( ) Glaucoma 2.3 ( ) 1.0 ( ) 1.1 ( ) History of cardiovascular disease 1.2 ( ) 1.0 ( ) 1.0 ( ) BMI 1.0 ( ) 1.0 ( ) 1.0 ( ) Visual acuity 6/6 1.9 ( ) 1.5 ( ) 1.6 ( ) ARM Age-related maculopathy; BMI Body mass index; CI confidence interval; CMR cellophane macular reflex; PMF preretinal macular fibrosis. TABLE 5. Comparison of Age-Standardized* Prevalence Rates of Epiretinal Membranes in Five Population-Based Studies (95% CI) Study Country Date Conducted Population at Risk Age-Standardized Rate (95% CI) Beaver Dam Eye Study* USA % ( ) Blue Mountains Study Australia % ( ) Hisayama Study Japan % ( ) Latino Eye Study USA % ( ) Melbourne VIP Study Australia % ( ) CI confidence interval. *Prevalence of idiopathic epiretinal membrane. Age-standardized prevalence rates were calculated by the direct method using Segi s world population. DISCUSSION FEW POPULATION-BASED STUDIES OF PREVALENCE OF epiretinal membrane have been reported previously Our results are similar to those reported in the Blue Mountains Eye Study. 29 The methods of the two surveys were similar except that the VIP used two-field photographs and the Blue Mountains Eye Study used seven fields. Table 5 compares the age-standardized prevalence of epiretinal membrane in the Melbourne VIP study with the other four population-based studies: Beaver Dam Eye Study, 28 Blue Mountains Eye Study, 29,31 Hisayama Study, 30 and Latino Eye Study. 32 The age-standardized rates for each study were calculated using Segi s world population 35 to make them comparable. The age-standardized rate of epiretinal membrane in the VIP (5.1%; 95% CI 3.8 to 6.4) was similar to the Blue Mountains Eye Study (5.4%; 95% CI 4.1 to 6.8), and higher than the Hisayama Study (2.8%; 95% CI 1.3 to 4.4). The Hisayama Study 30 used nonstereoscopic 45-degree fundus photographs to grade epiretinal membranes, which may be a plausible reason for the low prevalence in this study compared with other studies. The Latino Eye Study 32 and the Beaver Dam Eye Study 28 reported a higher prevalence of epiretinal membrane. It is possible that finer grading system of epiretinal membranes with a seven-step scale (1/1, 2/ 2, 2/2, 3/ 3, 3/3, 4/ 4, 4/4) could have accounted for higher prevalence of epiretinal membrane in the Latino Eye Study and the Beaver Dam Eye Study, 28 or it could be a true difference that is currently unexplained. Fraser-Bell and associates 32 reported that the higher prevalence of epiretinal membrane in the Latino Eye Study may have been due to the higher prevalence of diabetes in their population. Age-standardized prevalence of epiretinal membrane was higher for secondary (16.4%) compared with idiopathic (5.4%), reflecting the presence of underlying ocular pathology associated with aging. 288.e5 AMERICAN JOURNAL OF OPHTHALMOLOGY AUGUST 2005

6 We found that a decrease in visual acuity ( 6/6) was significantly associated with PMF (OR 1.9) and CMR (OR 1.5) (Table 4). The Beaver Dam 28 and Blue Mountains 29 studies found a significant association of decrease in visual acuity with PMF, but not with CMR. The Latino Eye Study reported a significant reduction in visual acuity in eyes with central PMF after adjusting for age. 32 We also found a significant association between PMF and urban residence (OR 3.1; 95% CI 1.3 to 7.0). The urban population of Melbourne may have a higher prevalence of undiagnosed diabetes, which is one of the factors associated with epiretinal membranes. This may account for higher prevalence of epiretinal membrane in urban population compared with rural population. Refractive error, such as increasing myopia, was significantly associated with PMF. This result was also reported by Appiah and Hirose, 6 who found in 324 cases of idiopathic premacular gliosis that 31.6% of eyes had myopia. However, contrary to this result, the Blue Mountains Eye Study 29 found a significant association of CMR with hyperopia, but not with myopia. The Blue Mountains Eye Study 29 reported that diabetes is a risk factor for PMF, after excluding any subjects with diabetic retinopathy. Similar to other studies, 30,32 we did not find any association between diabetes and epiretinal membranes. Consistent with other studies, 28,29,32 no other systemic associations were found, including the lack of any association with vascular risk factors (hypertension and cardiovascular heart diseases). We found a moderate protective association with smoking (OR 0.6; 95% CI 0.5 to 0.9). All the other studies 28 30,32 also found a nonsignificant protective association between smoking and epiretinal membranes after adjusting for age and gender. This finding may be an artifact reflecting a relative decreased survival among smokers in the older age groups. Several factors limit the interpretation of the results of this study. First, it is a cross-sectional study and further prospective studies will help to clarify the relationship between ocular factors and the development and progression of epiretinal membranes. Second, grading of epiretinal membranes may be affected by the presence of media opacities, such as cataract, resulting in poorly focused retinal photographs, which may lead to underestimation of the prevalence of epiretinal membranes. This could explain the lower prevalence of epiretinal membranes in people 80 years of age or older, in whom more severe cataract is present. In conclusion, the prevalence of epiretinal membranes in this population was similar to that reported in the Blue Mountains 29 and Hisayama 30 population-based studies. Idiopathic PMF was significantly associated with urban population, myopia, and visual acuity 6/6 after adjusting for age and gender. The population shifts in the age distribution to the older ages and trends for increasing myopia prevalence could lead to an increase in mild visual impairment due to epiretinal membranes. ACKNOWLEDGMENTS We are grateful for the help provided by the VIP project team and I. V. Nayak. REFERENCES 1. Wise GN. Macular changes after venous obstruction. Arch Ophthalmol 1957;58: Roth AM, Foos RY. Surface wrinkling retinopathy in eyes enucleated at autopsy. Trans Am Acad Ophthalmol Otolaryngol 1971;75: Wise GN. Preretinal macular fibrosis (an analysis of 90 cases). Trans Ophthalmol Soc U K 1972;92: Tanenbaum HL, Schepens CL, Elzeneiny I, Freeman HM. Macular pucker following retinal detachment surgery. Arch Ophthalmol 1970;83: Yanoff M, Fine BS. Ocular pathology, 4th ed. New York: Mosby-Wolfe, 1996; Appiah AP, Hirose T. Secondary causes of premacular fibrosis. Ophthalmology 1989;96: Hagler WS, Aturaliya U. Macular puckers after retinal detachment surgery. Br J Ophthalmol 1971;55: Lobes LA Jr, Burton TC. The incidence of macular pucker after retinal detachment surgery. Am J Ophthalmol 1978;85: Wiznia RA. Posterior vitreous detachment and idiopathic preretinal macular gliosis. Am J Ophthalmol 1986;102: Hirokawa H, Jalkh AE, Takahashi M, et al. Role of the vitreous in idiopathic preretinal macular fibrosis. Am J Ophthalmol 1986;101: Wise GN. Clinical features of idiopathic preretinal macular fibrosis. Schoenberg Lecture. Am J Ophthalmol 1975;79: Fine SL. Idiopathic preretinal macular fibrosis. Int Ophthalmol Clin 1977;17: Green WR, Kenyon KR, Michels RG, et al. Ultrastructure of epiretinal membranes causing macular pucker after retinal re-attachment surgery. Trans Ophthalmol Soc U K 1979;99: Clarkson JG, Green WR, Darcy Massof BA. A histopathologic review of 168 cases of preretinal membrane. Am J Ophthalmol 1977;84: Toshihiro K, Toshihiko K, Hajime I. Epiretinal membrane formation: light and electron microscopic study in an experimental rabbit model. Arch Ophthalmol 1995;113: Bellhorn MB, Friedman AH, Wise GN, Henkind P. Ultrastructure and clinicopathologic correlation of idiopathic preretinal macular fibrosis. Am J Ophthalmol 1975;79: Smiddy WE, Maguire AM, Green WR, et al. Idiopathic epiretinal membranes. Ophthalmology 1989;96: Machemer RA. A new concept of vitreous surgery. 7. Two instrument techniques in pars plana vitrectomy. Arch Ophthalmol 1974;92: McDonald HR, Verre WP, Aaberg TM. Surgical management of idiopathic epiretinal membranes. Ophthalmology 1986;93: VOL. 140, NO. 2 EPIRETINAL MEMBRANES IN THE VISUAL IMPAIRMENT PROJECT 288.e6

7 20. Bustros SD, Thompson JT, Michels RG, et al. Vitrectomy for idiopathic epiretinal membranes causing macular pucker. Br J Ophthalmol 1988;72: Margherio RR, Cox MS, Trese MT, et al. Removal of epimacular membranes. Ophthalmol 1985;92: Grewing R, Mester U. Results of surgery for epiretinal membranes and their recurrences. Br J Ophthalmol 1996;80: Allen WA, Gass JD. Contraction of a perifoveal epiretinal membrane simulating a macular hole. Am J Ophthalmol 1976;82: Sidd RJ, Fine SL, Owens SL, Patz A. Idiopathic preretinal gliosis. Am J Ophthalmol 1982;94: Smiddy WE, Michels RG, Gilbert HD, Green WR. Clinicopathologic study of idiopathic macular pucker in children and young adults. Retina 1992;12: Mills PV. Preretinal macular fibrosis. Trans Ophthalmol Soc U K 1979;99: Pearlstone AD. The incidence of idiopathic preretinal macular gliosis. Ann Ophthalmol 1985;17: Klein R, Klein BEK, Wang Q, Moss SE. The epidemiology of epiretinal membranes. Trans Am Ophthalmol Soc 1994;92: Mitchell P, Smith W, Chey T, et al. Prevalence and associations of epiretinal membranes. The Blue Mountains Eye Study, Australia. Ophthalmology 1997;104: Miyazaki M, Nakamura H, Kubo M, et al. Prevalence and risk factors for epiretinal membranes in a Japanese population: the Hisayama study. Graefes Arch Clin Exp Ophthalmol 2003;241: Fraser-Bell S, Guzowski M, Rochtchina E, et al. Five-year cumulative incidence and progression of epiretinal membranes. The Blue Mountains Eye Study. Ophthalmology 2003;110: Fraser-Bell S, Ying-Lai M, Klein R, Varma R, the Los Angeles Latino Eye Study Group. Prevalence and association of epiretinal membranes in Latinos: the Los Angeles Latino Eye Study. Invest Ophthalmol Vis Sci 2004;45: Livingston PM, Carson CA, Stainislavsky YL, et al. Methods for a population-based study of eye disease: the Melbourne Visual Impairment Project. Ophthalmic Epidemiol 1994;1: McCarty DJ, Fu CL, Harper CA, et al. Five-year incidence of diabetic retinopathy in the Melbourne Visual Impairment Project. Clin Experiment Ophthalmol 2003;31: Breslow NE, Day NE. Statistical methods in cancer research. Volume II the design and analysis of cohort studies. In: Heseltine E, ed. Lyon, France: International Agency for Research on Cancer, Oxford, UK: Oxford University Press, 1987: e7 AMERICAN JOURNAL OF OPHTHALMOLOGY AUGUST 2005

8 Biosketch Dr. McCarty is an epidemiologist at the Marshfield Clinic Research Foundation. His primary research interests include diabetes, obesity and ophthalmic epidemiology. Prior to coming to the Marshfield Clinic, Dr. McCarty was Clinical Research Unit Head at the University of Melbourne s Department of Ophthalmology and was an Australian National Health and Medical Research Council Postdoctoral Fellow at the International Diabetes Institute in Melbourne, Australia. VOL. 140, NO. 2 EPIRETINAL MEMBRANES IN THE VISUAL IMPAIRMENT PROJECT 288.e8