Prevalence and Associations of Epiretinal Membranes in Latinos: The Los Angeles Latino Eye Study

Abstract

purpose. To determine age- and gender-specific prevalence and associations of epiretinal membranes (ERMs) in adult Latinos.

methods. The Los Angeles Latino Eye Study (LALES) is a population-based study of eye disease among Latinos aged 40 or more years. Complete ophthalmic examinations included stereoscopic fundus photography. Masked photographic grading was used to identify and classify ERMs as cellophane macular reflex (CMR) without retinal folds or preretinal macular fibrosis (PMF) with folds.

results. Of the 6142 persons examined at the clinic, 5982 (97%) had gradable retinal photographs. The mean age of the participants was 54.7 ± 10.7 years; 58% were women. ERMs were present in 18.5% (95% confidence interval [CI]: 17.5%–19.5%) of the participants. Of the participants with ERMs, 19.9% had bilateral ERMs. The prevalence of ERMs increased from 10.1% in persons 40 to 49 years of age to 35.7% in those aged 70 to 79 years and was 22.5% in persons aged 80 years or more. The prevalence was similar in men and women. CMR was present in 16.3% (95% CI: 15.3%–17.2%) and PMF in 2.2% (95% CI: 1.9%–2.6%). Retinal folds involved the fovea in 11% of PMF cases. On average, eyes with central PMF had poorer visual acuity than did eyes without (P < 0.0002). Epiretinal membranes (ERMs) were present in 71% of eyes with macular holes. ERMs were also more common in individuals who had undergone cataract surgery (39.9%), those with proliferative diabetic retinopathy (25.7%), and those with any retinal disease (27.5%).

conclusions. ERMs occur frequently in Latinos, often bilaterally. The associations of ERMs with proliferative retinopathy, retinal lesions, macular holes, and cataract surgery were confirmed. Central PMF is associated with reduced visual acuity.

Epiretinal membranes are caused by the migration of cells through small focal defects in the internal limiting membrane after posterior vitreous detachment, ^{1 2 3} or by retinal breaks and detachments. ^{4 5 6 7} These cells proliferate and create a thin veil of tissue at the retinal–vitreous interface. Ultrastructural examination of ERMs has revealed glial cells, retinal pigment epithelial cells, fibrous astrocytes, fibrocytes, and myofibroblasts. ^{8 9 10} Purely glial cells are reported to occur in the earlier form of ERM. Whereas a prominent fibrous, nonglial component has been reported in membranes causing traction. ¹¹  

The early form of ERM, sometimes termed cellophane macular reflex (CMR), ^{12 13} is usually asymptomatic. However, the more severe form, known as preretinal macular fibrosis (PMF) ^{1 3 6 7 13 14} or macular pucker, ^{5 8 15} can cause significant loss of visual acuity and visual symptoms such as distortion and metamorphopsia ¹⁶ by tangential traction in the macula area. Abnormal retinal vascular tortuosity may be seen at the posterior pole with fluorescein angiography. ^{17 18} In very severe cases, macular edema and retinal detachment have been known to occur. ^{2 19 20}  

Known associations with ERMs include intraocular surgery and various retinal disorders such as retinal detachment and breaks, retinal vein occlusions and diabetic retinopathy. ^{4 6 7 12 13 21} They have been termed idiopathic ^{1 2 3 12 13 14} when they occur after posterior vitreous detachment without the presence of other known secondary associations. 

ERMs are common in older whites. ^{12 13} However, there are only two previous population-based studies documenting the prevalence of ERMs in whites, in which photographic masked grading was used. ^{12 13} Further, to our knowledge, there have been no reports of their prevalence in other racial groups. The Los Angeles Latino Eye Study (LALES) was designed to provide population-based prevalence estimates of eye disorders in Latinos. Data were collected between 2000 and 2003, including stereoscopic retinal photographs, which provided the opportunity to obtain prevalence estimates of ERMs among Latinos and their associated risk factors. 

Methods

Study Cohort

The LALES is a population-based survey of vision and common eye diseases of an urban population of older Latinos. Details of the study design, sampling plan, and baseline data are reported elsewhere. ²² In summary, a door-to-door census of permanent residents was conducted in six census tracts in La Puente, Los Angeles County, California. The study cohort consisted entirely of self-identified Latinos, aged 40 years or more. The Institutional Review Board at the University of Southern California approved the study protocol. All study procedures adhered to the principles outlined in the Declaration of Helsinki for research involving human subjects. Participants provided written, informed consent to participate in the study at the time of the census. 

Interview and Examination Procedures

A detailed home interview was conducted to determine demographic factors and information regarding ocular and medical histories, various risk factors, and access to medical and ocular care. All eligible Latinos were invited to visit a local eye examination center for a comprehensive ocular examination by a trained ophthalmologist and technicians using standardized protocols. Visual acuity for each LALES participant was measured using standard Early Treatment Diabetic Retinopathy Study (ETDRS) protocols with a modified ETDRS distance chart ²³ transilluminated with a chart illuminator (Precision Vision, La Salle, IL). Automated refraction was performed (Humphrey Autorefractor Model 599; Carl Zeiss Meditec, Dublin, CA), followed by subjective refraction, using standard protocols. Lens opacification was graded at the slit lamp by a trained ophthalmologist according to the Lens Opacities Classification System II (LOCS II). ²⁴ Pseudophakia and aphakia were also documented during the slit lamp examination. The ophthalmic examination included 30° stereoscopic color retinal photographs (Topcon TRC 50EX; Topcon Corp. of America, Paramus, NJ) of Diabetic Retinopathy Study fields ²⁵ 1 (optic disc) and 2 (macula) and a modified field 3 (temporal to macula) in all participants. Slide film was used (Ektachrome 100; Eastman Kodak, Rochester, NY). 

Grading

ERMs were diagnosed with masked photographic grading by trained graders. The site of the ERM was classified by using a grid with an outer radius of 3000 μm, corresponding to the anatomic macula, and an inner radius of 500 μm, which was placed over one of the stereo pair when grading. 

Definitions

The presence of diabetes was defined by a self-reported history of diabetes, and/or hemoglobin A1c greater than 7.0%, and/or random glucose greater than 200 mg/dL. Hypertension was defined as having a history of hypertension or elevated blood pressure when measured in the clinic (diastolic blood pressure greater than 85 mmHg or systolic blood pressure greater 140 mmHg). Refractive error was defined by spherical equivalent (sphere + cylinder/2) in diopters. Applanation intraocular pressure (Goldmann) was measured. After agreement by two of three glaucoma specialists, glaucoma was diagnosed by the presence of a characteristic glaucomatous optic disc (seen on stereophotographs) and by visual field changes. 

Classification of ERMs

Two types of ERMs were identified adopted from Klein et al. ¹² The first or earlier stage consisted of a CMR only, described as a patch or patches of irregular increased reflection from the inner surface of the retina. The second stage, PMF, a more severe form, occurs as the membrane thickens and contracts, becoming opaque and gray with the appearance of superficial retinal folds or traction lines. ¹ Subjects with both CMR and PMF were allocated to the PMF group. ERMs were evaluated to determine whether there was involvement of the center point of the fovea. 

ERMs were further classified as primary (or idiopathic) and secondary. Primary or idiopathic ERMs were those present in subjects without a secondary cause for the development of ERMs. Secondary ERMs were defined as those occurring in eyes in the presence of such conditions as retinal vascular disease, retinal detachment, and cataract surgery. Retinal vein occlusion was present in 58 subjects, proliferative diabetic retinopathy (PDR) in 74, and a history or signs of retinal detachment in 36. 

Data Handling and Statistical Methods

Data were entered into an automated database (Access-98; Microsoft, Redmond, WA) with internal automated quality control checks. Commercial statistical analysis software (Statistical Analysis System, ver. 8; SAS Institute Inc., Cary, NC) was used for tabulations and statistical analyses. Statistical analysis of the data included frequency distributions, χ² analyses, t-tests, and logistic regression analyses. In the logistic regression, age, intraocular pressure, and spherical equivalent were used as continuous variables, whereas glaucoma, early age-related maculopathy, hypertension, smoking, and diabetes were used as dichotomous variables. Odds ratios (OR) and 95% confidence intervals (CI) are reported. 

Results

Of the 7789 eligible Latinos, 6870 (88%) permitted and completed an in-home interview, and 6142 (78%) subjects were examined in the clinic. Those who completed an in-home interview were, on average, 2 years younger (54.9 ± 11 years) and more likely to be women (58%) when compared with those who did not complete an interview (n = 919; 56.8 ± 11 years, female 47%; P < 0.0001). Participants who completed the in-home interview, but refused to participate in the clinical examination were older (55.5 years, P < 0.0001); had higher income, education, and acculturation levels; and were more likely to have health insurance. 

Of the 6142 in-clinic–examined participants, 5982 (97%) had photographs in at least one eye that were considered gradable for ERMs. Among the 160 participants excluded were 104 individuals who had no photographs taken, 31 whose photographs were considered ungradable because of poor photograph quality or dense media opacities, and 25 with late age-related macular degeneration (exudative or geographic atrophy) in one or both eyes, which was thought to confound the grading of ERMs. The mean age of participants included in the analyses was 54.7 ± 10.7 years and 58% were women (Fig. 1) . 

Prevalence of ERMs

Epiretinal membranes were present in 1105 participants (18.5%; 95% CI: 17.5%–19.5%). The earlier, less severe form without retinal folds, CMR, was present in 16.3%, (95% CI: 15.3%–17.2%) and the later stage with retinal folds, PMF, in 2.2% (95% CI: 1.9%–2.6%) of the study group (Table 1) . The overall prevalence of any ERMs was higher in those aged 70 to 79 years (35.7%) compared with those aged 40 to 49 years (10.1%). This age-related increase was significant (P for trend < 0.0001). Prevalence of PMF also was higher in Latinos aged 70 to79 years than in those who were younger (Fig. 2) . 

Women were more likely to have an ERM than men, although the difference was small (18.7% vs. 18.2%; P = 0.01; Fig. 2 ). This reflects the slightly higher prevalence of the less severe form, CMR, among women than among men (16.4% vs. 16.1%; P = 0.01). There was no significant difference in the overall prevalence rate of PMF between men and women P = 0.86, after age-adjustment (Table 2) . 

Bilaterality of ERMs

Bilaterality of ERMs was assessed in participants with ERMs with gradable photographs of both eyes (n = 1088). ERMs were bilateral in 217 (19.9%) of the 1088 participants. CMR was bilateral in 204 (21.3%) of participants with any CMR, including those with CMR in one eye and PMF in the other. Of all participants with PMF and gradable photographs of both eyes, 13 (11.2%) of 116 had bilateral PMF and 40 (34.5%) of 116 had mixed CMR/PMF. The prevalence of any bilateral ERMs increased with age (P for trend < 0.0001; Fig. 3 ). ERMs were more likely to be bilateral in women than in men after age adjustment (P = 0.01). 

ERMs and Visual Acuity

After adjustment for age and lens opacities, the best corrected visual acuity in eyes with central PMF (11% of all eyes with PMF) was significantly lower than eyes without central PMF (logarithm of minimum angle of resolution [logMAR] score 0.29 vs. 0.11, P < 0.0002). Furthermore, after adjustment for age and lens opacities, eyes with PMF had borderline lower visual acuity than eyes without PMF (logMAR score 0.14 vs. 0.11, P = 0.07). However, after adjustment for age and lens opacities, eyes with CMR had visual acuity similar to that of eyes without CMR (logMAR score 0.10 vs. 0.11, P = 0.31). 

Primary and Secondary ERMs

The age-specific prevalences of primary (idiopathic) and secondary ERMs are presented in Table 3 . When stratified by the presence or absence of lens opacities, primary (idiopathic) ERMs were less common in those with lens opacities. Secondary ERMs were more prevalent than primary (idiopathic) ERMs in all age groups. The prevalence of secondary ERMs were further stratified by cause, including any retinal disease, PDR, and cataract surgery (Table 4) . The highest prevalence of CMR was found in participants who had undergone cataract surgery (32.3%). The highest prevalence of PMF was seen in those with PDR (13.5%). After age-adjustment, PMF was more frequent in persons with retinal disease (OR: 2.7; 95% CI: 1.2–5.7), PDR (OR: 5.2; 95% CI: 2.6–10.7), and post cataract surgery (OR: 1.9; 95% CI: 1.1–3.4). Full-thickness macular holes were seen in seven eyes from seven persons. In five persons (71%), they were associated with an ERM in the same eye (four with PMF and three with CMR). 

To identify risk factors of idiopathic ERMs, various characteristics and their association with ERMs were assessed. These included presence of early age-related maculopathy, elevated intraocular pressure, diabetes, hypertension, history of smoking, and refractive error. After adjustment for age, none of these previously suggested risk factors were significantly associated with ERMs in our study (OR: 0.6–1.9, P > 0.05). 

Discussion

This report documents the prevalence and associations of a common ocular condition, ERM, and its relationship with visual acuity. The LALES provides unique data on the presence and type of ERM among Latinos, using standardized protocols for grading these lesions from stereoscopic fundus photographs. The LALES population has a socioeconomic demographic profile similar to all U.S. Latinos. ²⁶ The number of eligible subjects and the response rate achieved is comparable to previous population-based prevalence studies ^{12 13} of ERMs. 

The grading of ERMs in our study was similar to that in the Beaver Dam Eye Study and Blue Mountains Eye Study, with the same definitions of early and late lesions used. Further, the LALES fundus photographs were graded by masked graders at the Ocular Epidemiology Grading Center at the University of Wisconsin (Madison, WI), who were trained by graders from the Beaver Dam Eye Study. Thus, due to similar methods we are able to compare our data with those in the Beaver Dam and Blue Mountains studies. ^{12 13} Similar to these studies, the prevalence of ERMs was higher in older than in younger Latinos and most of these ERMs were of the early form (CMR). The prevalence of ERMs in Latinos (18.5%) was higher than rates found in whites in the Beaver Dam Eye Study (11.8%) and the Blue Mountains Eye Study (7%). One explanation for this may have been the higher prevalence of diabetes in our population (because a significant association has been reported ¹³ ). However, although we found a relationship with PDR, no significant association with history of diabetes or diagnosis of diabetes at the examination was present after age-adjustment. Although this higher prevalence of ERMs may represent a true difference, we cannot exclude the possibility that the CMR is more readily visible against the slightly more pigmented fundus of Latinos and that this difference leads to systematic differences in grading. Further evidence of a grading difference is the similarity in the prevalence of later ERMs with retinal folds (PMF), being 2.2% in our group compared with 2.8% in the Beaver Dam ¹² and 2.2% in the Blue Mountains ¹³ eye studies. The diagnosis of PMF is less likely to be subject to grading differences because of the presence of retinal folds. 

The bilateral tendency of ERMs has been described. ^{12 13} ERMs commonly occur after vitreous detachment, which is frequently bilateral. Incident ERMs were more than twice as frequent in second eyes in the Blue Mountains than those without the lesion at baseline. ¹⁴ Our rate of bilaterality of 19.9% was similar to that found in both the Beaver Dam and Blue Mountains eye studies (19.5% and 31%, respectively). CMR was more than seven times more prevalent than PMF and was twice as likely to occur bilaterally than PMF (21.3% vs. 11.1%). 

Our study confirmed the association of ERMs with previously reported risk factors, such as cataract surgery, PDR, and retinal lesions. Cataract surgery was associated with ERMs, especially the earlier form (CMR). ^{7 8 13 14 21} It is possible that ERMs are more commonly visible after cataract surgery, because lens opacification may obscure their detection before surgery. Further, the timing of the incidence of ERMs relative to the cataract surgery is not known in our study. Similarly, the lower prevalence of ERMs in the oldest age group (80 years and older) may be from grading difficulties or represent a survivor effect. This same reduction in prevalence in older individuals was reported in the Beaver Dam and Blue Mountains eye studies. 

The association of the more severe form of ERM (PMF) with PDR and other retinal lesions was confirmed in our study, supporting other reports. ^{5 6 7 12 13} ERMs are reported to coexist commonly with macular holes. We found the prevalence of macular holes to be 0.1% in our population of Latinos. Most macular holes were associated with an ERM (71%). We cannot comment on whether ERMs preceded these lesions. No other significant systemic or ocular associations were found with PMF or with CMR after age adjustment. In particular, we found no association with diabetes in contrast to the Blue Mountains Eye Study. 

Only PMF was associated with a small, but significant reduction in visual acuity after age adjustment and only if it involved the center of the fovea. This was consistent with the Beaver Dam and Blue Mountains studies. ^{12 13} Central PMF occurred in only 11% of all PMF cases, equating to an overall prevalence of 0.2%. Therefore, there did not appear to be a significant effect of ERMs on visual acuity in our population. A limitation of our study is that there was no information available on the proportion of people with ERMs with metamorphopsia (neither on history nor on Amsler grid testing). 

In summary, as in whites, ERMs are common retinal lesions among Latinos. The earlier form, CMR, may be more prevalent in Latinos than in whites, but the more severe form, PMF is equally prevalent. ERMs were associated with cataract surgery, and the more severe form, PMF, occurred more frequently with retinal lesions, such as diabetic retinopathy and retinal detachment. They also were found in most eyes with macular holes. On average, central PMF had only a minor effect on vision, while noncentral PMF and CMR had no effect on visual acuity. 

Appendix

The Los Angeles Latino Eye Study Group

University of Southern California, Los Angeles, CA (study site): Rohit Varma, MD, MPH; Sylvia H. Paz, MS; LaVina Abbott; Stanley P. Azen, PhD; Lupe Cisneros, COA; Elizabeth Corona; Carolina Cuestas, OD; Denise R. Globe, PhD; Sora Hahn, MD; Mei Lai, MS; George Martinez; Susan Preston-Martin, PhD; Ronald E. Smith MD; Mina Torres MS; Natalia Uribe OD; Jennifer Wong, MPH, Joanne Wu, MPH; and Myrna Zuniga. 

Battelle Survey Research Center, St. Louis, MO: Sonia Chico, BS; Lisa John, MSW; Michael Preciado, BA; and Karen Tucker, MA 

Ocular Epidemiology Grading Center, University of Wisconsin, Madison, WI: Ronald Klein, MD, MPH; Stacy M. Meuer; Michael Neider; Tiffany Jan; Sandy C. Tomany MS; Michael Knudtson MA; Karl Jensen; Carol Hoyer; Andy Ewen; Maria Swift; Anne Mosher; and Ellen Hall. 

⁵  See Appendix for members of the Los Angeles Latino Eye Study Group.

 Supported by National Institutes of Health Grants U10-EY-11753 and EY-03040 and an unrestricted grant from Research to Prevent Blindness, New York, New York. RV is a Research to Prevent Blindness Sybil B. Harrington Scholar.

 Submitted for publication November 26, 2003; revised January 8, 2004; accepted January 15, 2004.

 Disclosure: S. Fraser-Bell, None; M. Ying-Lai, None; R. Klein, None; R. Varma, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Age and gender distribution of examined participants in the LALES.

Table 1.

 

Prevalence of Epiretinal Membranes in the LALES

Age Group (y)	PMF n (%)	CMR Only n (%)	Any n (%)
40–49	20 (0.9)	215 (9.3)	235 (10.1)
50–59	19 (1.1)	255 (14.0)	274 (15.1)
60–69	51 (4.4)	321 (27.4)	372 (31.7)
70–79	36 (6.5)	161 (29.2)	197 (35.7)
80+	7 (5.8)	20 (16.7)	27 (22.5)
Total	133 (2.2)	972 (16.3)	1105 (18.5)

 Data are expressed as the number of subjects, with the percentage of the total group (n = 5982) in parentheses.

Figure 2.

 

Prevalence of epiretinal membranes by age and gender in the LALES.

Table 2.

 

Age- and Gender-Specific Prevalence of Epiretinal Membranes in the LALES

Age Group (y)	PMF			CMR
		Males n (%)	Females n (%)	Males n (%)	Females n (%)
40–49	10 (1.0)	10 (0.7)	102 (10.5)	113 (8.4)
50–59	7 (0.9)	12 (1.1)	83 (11.0)	172 (16.1)
60–69	19 (4.0)	32 (4.6)	134 (28.1)	187 (26.9)
70–79	16 (6.5)	20 (6.5)	73 (29.7)	88 (28.8)
80+	2 (4.2)	5 (6.9)	8 (16.7)	12 (16.7)
Total	54 (2.2)	79 (2.3)	400 (16.1)	572 (16.4)

 Data are as described in Table 1 .

Figure 3.

 

Bilaterality of CMR and PMF.

Table 3.

 

Age-Specific Prevalence of Idiopathic and Secondary Epiretinal Membranes in the LALES

Age Group (y)	Idiopathic Epiretinal Membranes (Number at Risk = 5637)					Secondary Epiretinal Membranes* (Number at Risk = 327)
		PMF n (%)	CMR Only n (%)	Any n (%)	PMF n (%)	CMR Only n (%)	Any n (%)
40–49	19 (0.8)	211 (9.2)	230 (10.1)	1 (3.0)	4 (12.1)	5 (15.2)
50–59	15 (0.9)	246 (14.0)	261 (14.8)	4 (7.4)	9 (16.7)	13 (24.1)
60–69	42 (3.9)	291 (27.2)	333 (31.1)	9 (8.7)	30 (29.1)	39 (37.9)
70–79	27 (6.1)	123 (27.8)	150 (33.9)	9 (8.3)	38 (34.9)	47 (43.1)
80+	4 (5.4)	8 (10.8)	12 (16.2)	3 (6.5)	12 (26.1)	15 (32.6)
Total	107 (1.9)	879 (15.6)	986 (17.5)	26 (7.5)	93 (27.0)	119 (34.5)

*  Epiretinal membranes occurring in participants with retinal disease (retinal vascular disease, retinal detachment, surgery for retinal disease, proliferative diabetic retinopathy) or after cataract surgery.

Table 4.

 

Age-Specific Prevalence of Secondary Epiretinal Membranes by Cause Among Latinos in the LALES

Age Group (y)	Retinal Disease* n = 91, †					PDR n = 74, †					Cataract Surgery n = 220, †
		PMF	CMR	Any	PMF	CMR	Any	PMF	CMR	Any
40–49	10.0	10.0	20.0	8.3	0.0	8.3	0.0	21.4	21.4
50–59	9.5	14.3	23.8	19.1	9.5	28.6	5.9	23.5	29.4
60–69	13.5	16.2	29.7	11.5	26.9	38.5	11.1	35.2	46.3
70–79	0.0	35.3	35.3	14.3	0.0	14.3	8.7	37.0	45.7
80+	0.0	16.7	16.7	0.0	0.0	0.0	6.9	25.6	32.6
Total	8.8	18.7	27.5	13.5	12.2	25.7	8.2	32.3	39.9

*  Retinal disease refers to retinal vascular disease, retinal detachment, and PDR.

†  The denominator for any secondary cause is n = 327 in Table 3 . In Table 4 , the sum of the three categories; retinal disease, PDR, and cataract surgery is 385, because in 56 participants, there are overlapping causes (PDR and cataract surgery or retinal disease and cataract surgery.

The authors thank the LALES External Advisory Committee for their advice and contributions: Roy Beck, MD, PhD (Chairman); Natalie Kurinij, PhD; Leon Ellewein, PhD; Helen Hazuda, PhD; Eve Higginbotham, MD; Lee Jampol, MD; M. Cristina Leske, MD; Donald Patrick, PhD; and James M. Tielsch, PhD.