Progression rates according to C-reactive protein (A) and interleukin 6 (B) levels.
Seddon JM, George S, Rosner B, Rifai N. Progression of Age-Related Macular DegenerationProspective Assessment of C-Reactive Protein, Interleukin 6, and Other Cardiovascular Biomarkers. Arch Ophthalmol. 2005;123(6):774-782. doi:10.1001/archopht.123.6.774
Age-related macular degeneration (AMD) and cardiovascular disease share common risk factors. Inflammatory biomarkers, including C-reactive protein (CRP), interleukin 6 (IL-6), soluble tumor necrosis factor alpha receptor 2, soluble intercellular and vascular adhesion molecules (intercellular adhesion molecule 1 and vascular cell adhesion molecule 1), and lipid biomarkers, including lipoprotein(a) and apolipoprotein B, have all been associated with cardiovascular disease. We previously found an association between AMD and CRP in a cross-sectional analysis, but the prospective relationships between AMD, CRP, and the other cardiovascular disease markers are unknown.
To test the hypothesis that baseline cardiovascular disease biomarkers are associated with subsequent increased risk for progression of AMD.
Design, Setting, and Participants
This prospective cohort study involved 251 participants aged 60 years and older who had some sign of nonexudative AMD and visual acuity of 20/200 or better in at least one eye at baseline. The AMD status was assessed by standardized grading of fundus photographs, and stored fasting blood specimens obtained at baseline were analyzed for levels of the various biomarkers. The average follow-up time was 4.6 years.
Main Outcome Measures
Relationship between biomarkers and incidence rates of progression of AMD.
Comparing the highest quartile with the lowest quartile, CRP was associated with progression of AMD, with a multivariate adjusted relative risk (RR) of 2.10 (95% confidence interval [CI], 1.06-4.18; P for trend, .046) controlling for body mass index, smoking, and other cardiovascular variables and a multivariate adjusted RR of 2.02 (95% CI, 1.00-4.04; P for trend, .06) controlling additionally for antioxidant nutrients. Interleukin 6 was also related to progression of AMD, with a multivariate adjusted RR of 1.81 (95% CI, 0.97-3.36; P for trend, .03). Comparing the highest quartile with the lowest quartile, the effect estimates for vascular cell adhesion molecule 1 (multivariate adjusted RR, 1.94) and apolipoprotein B (adjusted RR, 1.39) were in the positive direction but were not statistically significant (P for trend, .08 and .24, respectively). The CRP and IL-6 levels were both significantly related to higher body mass index and current smoking.
Higher levels of the systemic inflammatory markers CRP and IL-6 are independently associated with progression of AMD.
Age-related macular degeneration (AMD) is the leading cause of irreversible visual impairment and blindness among persons aged 60 years and older.1 With the elderly population steadily growing, the burden related to this loss of visual function will increase. With limited treatment options available, prevention remains the best approach for addressing this public health concern.
Over the past decade, knowledge has grown considerably about modifiable factors related to AMD, most notably cigarette smoking2,3 and nutritional factors4- 9 as well as obesity10,11 and lipid levels.12 Many of these factors associated with AMD are also related to cardiovascular disease (CVD). We have hypothesized that cardiovascular disorders and AMD share common antecedents and proposed that novel biomarkers associated with CVD be evaluated for their potential relationship with AMD.13 These factors include C-reactive protein (CRP), a physiological marker of systemic inflammation that we have recently shown to be related to advanced AMD in a large case-control study within a multicenter randomized trial of antioxidant supplements, the Age-Related Eye Disease Study.14 Although CRP, an acute-phase reactant, is the inflammatory marker most consistently associated with CVD,15,16 other indicators of inflammation have also been shown to be related to CVD over the past decade, including interleukin 6 (IL-6),17,18 tumor necrosis factor alpha receptor 2 (TNF-α-R2),19- 21 soluble intercellular adhesion molecule 1 (ICAM-1), and soluble vascular cell adhesion molecule 1 (VCAM-1).21,25 Additionally, lipid biomarkers, including lipoprotein(a) (Lp[a]) and apolipoprotein B (ApoB), have been found to be associated with CVD.26- 28
Inflammation is associated with angiogenesis and neovascularization, which can occur in inflammatory eye diseases,29 similar to the most advanced and debilitating neovascular form of AMD. In addition to CVD, stroke and Alzheimer disease may also have an inflammatory component.30- 32 It is possible that AMD may represent another chronic, age-related inflammatory disease. The presence, biogenesis, and composition of drusen, yellowish deposits in the macula that are the hallmark of AMD, have been studied in relation to inflammation. It has been hypothesized that drusen, especially large, soft drusen, which confer a higher risk of progression,7,10 may be an indicator of immune-mediated and local inflammatory events in the eye.33- 37 Furthermore, ApoB has been reported to be one of the many lipid components in the molecular composition of drusen.34,37
Thus far, one case-control study of AMD has implicated CRP,14 but no prospective studies have been reported, and the relationships between the other CVD biomarkers and AMD are unknown. We therefore conducted a prospective study to explore the relationship between baseline levels of these CVD biomarkers and incidence rates of progression of AMD.
The longitudinal study design of the Progression of Age-Related Macular Degeneration Study has been described previously.8,10 The study aim is to evaluate multiple risk factors for the onset and progression of AMD. The study population consists of patients with nonexudative AMD and best corrected visual acuity of 20/200 or better in at least one eye aged 60 years and older at baseline. Other inclusion criteria included willingness to participate in a long-term study that involved annual dilated eye examinations and fundus photography. Patients were excluded if they were unable to speak English or had decreased hearing or cognitive function. All patients were examined by J.M.S. at the Massachusetts Eye and Ear Infirmary, Boston.
Of the 397 persons who were eligible for enrollment between July 1989 and May 1998, 366 (92%) were enrolled. The Human Subjects Committee at the Massachusetts Eye and Ear Infirmary approved the study, and all subjects signed a consent form to participate. Of the 366 participants enrolled, 36 were not considered for analyses because of inability to complete the initial study examination (n = 5), lack of follow-up data (n = 17), or lack of one or more primary independent variables (n = 14). To conduct the analyses, we excluded men and women who reported, at baseline, ever having a diagnosis of cancer (n = 61) (except nonmelanoma skin cancer), which could influence the dietary and other variables assessed in these analyses. Finally, to maintain the quality of data for these analyses, we excluded 8 individuals whose dietary questionnaires had inadequate or missing answers (n = 6) or extreme values (n = 2). Participants whose responses were not included in the analyses (n = 8) did not differ significantly by age, sex, or education from those who were included. Of the 261 individuals, 251 had biomarkers assessed and were included in these analyses.
Information was collected from various sources, including a fasting blood specimen, a standardized risk factor questionnaire that was administered over the telephone by a trained interviewer, a clinical interview, a validated food frequency questionnaire,38 and measurements of height, weight, and blood pressure. Body mass index during the initial examination was calculated as weight in kilograms divided by height in meters squared. At baseline and at each annual visit, a complete dilated eye examination was performed, including a refraction, assessment of best corrected visual acuity,39 lens evaluation, and slitlamp examination of the macula. Stereoscopic color fundus photographs of the macula were also obtained. We used our 5-grade classification scale of AMD, the Clinical Age-Related Maculopathy Staging System,8,10,40- 43 which was modified from the Age-Related Eye Disease Study7 and evaluated for reliability.42 Macular characteristics were graded within a 3000-μm radius centered on the foveal center. Eyes with extensive small drusen (≥15 small drusen, <63 μm), nonextensive intermediate drusen (<20 drusen, ≥63 μm but <125 μm), or pigment abnormalities associated with AMD were assigned a grade of 2. Eyes with extensive intermediate or large (≥125 μm) drusen were assigned a grade of 3. Eyes with any geographic atrophy (≥350 μm) received a grade of 4. If there was evidence of retinal pigment epithelial detachment or choroidal neovascular membrane, a grade of 5 was assigned. Eyes received a grade of 1 if none of these signs was present. Advanced AMD was defined as grade 4 or 5.
Fasting blood specimens were all drawn in the morning at the baseline visit and were processed immediately and then frozen in liquid nitrogen freezers until analysis. All assays but the cytokine and adhesion molecule assays were measured using a Hitachi 911 analyzer (Roche Diagnostics, Indianapolis, Ind). C-reactive protein levels were measured using a high-sensitivity assay from Denka Seiken (Niigata, Japan). Interleukin 6, TNF-α-R2, ICAM-1, and VCAM-1 levels were measured by enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, Minn). Lipoprotein(a) concentrations were measured using an assay from Denka Seiken that was not affected by the Kringle type 2 repeats.44 We used TNF-α-R2 to measure the presence of inflammation because it has been shown to be more stable and require less blood than TNF-α, thus making it more reliable as an indicator of disease. The ApoB assay was performed by an immunoturbidimetric technique using reagents from Wako Chemicals USA (Richmond, Va).
Progression to advanced AMD was defined either as one eye progressing from a grade of less than 4 to grade 4 or 5 or progressing from grade 4 to grade 5 at any follow-up visit. Although both grades 4 and 5 are classified as advanced disease, eyes with geographic atrophy (grade 4) can progress to neovascular disease (grade 5). Sunness et al45 reported 4-year conversion rates of 11% for individuals with bilateral atrophy and 34% for those with unilateral atrophy and choroidal neovascular membrane in the fellow eye, and a 19% rate of progression overall. They also noted that the development of choroidal neovascular membrane had a deleterious effect on the degree of visual acuity loss. Therefore, we also included progression from atrophy to choroidal neovascular membrane (grade 4 to 5) as an outcome.
Regression was not considered in the analyses. Each subject was considered to have progressed only once during the follow-up period, counting the first eye that progressed. The rationale for this definition of progression was that AMD is a progressive disease and regression at this advanced stage is uncommon. In our data, among 107 people who progressed from baseline, only 2 individuals regressed at a subsequent visit; both of these subjects then subsequently progressed to advanced AMD.
Our principal method of analysis was the Cox proportional hazards model. We computed an adjusted relative risk (RR) of progression for sex-specific quartiles 2 to 4 of CRP vs sex-specific quartile 1 after controlling for age-sex group (men aged 60-69 years, men aged 70-79 years, men aged ≥80 years, women aged 60-69 years, women aged 70-79 years, and women aged ≥80 years), log total energy intake (calories), and protein intake (quartiles). In addition, the full multivariate model included the number of years of education (<12 or ≥12), smoking status (current, past, or never), body mass index (<25, 25-29, or ≥30), systolic blood pressure (analyzed continuously), cardiovascular disease, energy-adjusted log beta carotene intake (continuous), self-reported alcohol intake (grams per day as a continuous variable), physical activity (number of times per week of vigorous physical activity as a continuous variable), and initial AMD grade in worse eye (1-5, categorical). Tests for trend were also conducted corresponding to the adjusted and multivariate RRs by substituting a single trend variable of CRP coded as 1 to 4 and interpreted as a continuous variable. Ninety-five–percent confidence intervals (CIs) were computed for the multivariate RRs, and 2-sided P values were computed for all models. A similar approach was used to model the relationship between progression of AMD and the other biomarkers: IL-6, TNF-α-R2, ICAM-1, VCAM-1, Lp (a), and ApoB. Furthermore, the analyses concerning biomarkers were repeated, adding adjustment for the individual nutrients reported to be beneficial in the Age-Related Eye Disease Study: quartiles of total intake of zinc and vitamins C and E from food and supplements.7
Finally, to better understand the effect of other AMD risk factors on CRP and IL-6, linear regression analyses were performed for log CRP and log IL-6 on all other covariates in Table 1. Furthermore, rates of progression to advanced AMD were calculated for various levels of CRP and IL-6 (Figure). All analyses were performed using SAS software, version 8.0 (SAS Institute Inc, Cary, NC).
Two hundred fifty-one participants (96 men, 155 women) with a mean age of 72 years were included in the final analyses. Almost all subjects (99.6%) were white. The average follow-up time was 4.6 years. Of the 251 participants with AMD, 96 progressed to advanced stages of disease.
Table 1 displays the relationships between baseline characteristics and the various biomarkers unadjusted for other variables. Physical activity was inversely related to CRP, IL-6, and TNF-α-R2. Current smoking had a positive association with all biomarkers but VCAM-1 and Lp(a). Systolic blood pressure was positively related to CRP, IL-6, TNF-α-R2, ICAM-1, VCAM-1, and ApoB. Cardiovascular disease was positively related to CRP, IL-6, TNF-α-R2, ICAM-1, VCAM-1, and Lp(a). Body mass index was positively related to CRP, TNF-α-R2, and VCAM-1. Alcohol intake was inversely related to TNF-α-R2 and VCAM. Energy intake showed a positive association with Lp(a). Beta carotene intake was inversely related to CRP, IL-6, VCAM-1, and Lp(a). Zinc intake was inversely related to IL-6 and ApoB. Vitamin C intake was intake inversely related to CRP and IL-6. Vitamin E intake was inversely related to IL-6 and ApoB.
Relative risks for progression to advanced AMD by quartiles of inflammatory biomarkers are shown in Table 2. The highest quartile of CRP was positively related to progression of AMD, with a twofold greater risk compared with the lowest quartile of CRP after controlling for covariates (RR, 2.10; 95% CI, 1.06-4.18). After further adjustment for antioxidant nutrients, the effect was essentially unchanged, with an RR of 2.02 (95% CI, 1.06-4.18). The P values for trend for an increasing risk of progression with higher levels of CRP were .046 and .06, respectively. Similarly, the highest quartile of IL-6 was related to progression of AMD compared with the lowest quartile (multivariate RR after controlling for covariates, 1.81; 95% CI, 0.97-3.36; multivariate RR after further adjustment for antioxidant nutrients, 1.96; 95% CI, 1.04-3.71). The trend for a higher risk of progression of AMD with higher levels of IL-6 was statistically significant for both models (P values for trend, .03 and .02, respectively). For both CRP and IL-6, the attenuation of the effect compared with the simple adjusted model indicates the presence of positive confounding factors, most likely smoking and body mass index. Both of these variables are positively related to AMD2,3,10,11 and positively correlated with CRP and IL-6, as described below. Although TNF-α-R2 was weakly associated with progression in the age- and sex-adjusted analyses, the effect became nonsignificant after multivariate adjustment for known risk factors for AMD. No association was seen with ICAM-1, but VCAM-1 showed a slight, nonsignificant trend for a positive association with progression of AMD comparing the highest quartile with the lowest quartile (multivariate RR after controlling for covariates, 1.94; 95% CI, 0.99-3.80; multivariate RR after further adjustment for antioxidant nutrients, 1.76; 95% CI, 0.89-3.47). The RRs for progression to advanced AMD according to quartiles of the lipid biomarkers are shown in Table 3. There was no association with Lp(a) and a weak positive association with ApoB that was not significant (RR = 1.40 for both multivariate models).
After we reviewed the crude relationship between CRP, IL-6, and progression to advanced AMD, it appeared there were cutpoints that defined low-, middle-, and high-risk groups. As seen in the Figure, CRP levels less than 0.5 mg/L conferred the lowest risk, within the range from 0.5 to 10.0 mg/L there was little variation in risk of AMD, and levels of 10 mg/L or higher were associated with the highest risk of progression of AMD. Interleukin 6 levels of 6.0 pg/mL or higher were associated with increased risk for progression of AMD.
Table 4 displays the results of linear regression analyses of log CRP and log IL-6 on all covariates in Table 1, including known risk factors for AMD. Both CRP and IL-6 were significantly positively associated with current smoking as well as intermediate and high levels of body mass index. In addition, log CRP was inversely associated with vitamin C intake. As previously reported, the relationship between fish intake and AMD is modified by levels of linoleic acid intake.5,8 In the present study, we also found that the association between fish intake and CRP was modified by linoleic acid intake, in that fish intake was inversely related to CRP only in the low linoleic acid group (quartiles 1 and 2) (β = −0.94 ± 0.30, P = .003 for ≥2 servings of fish per week vs ≤1 serving of fish per week). There was no relationship between fish intake and CRP in the high linoleic acid group (quartiles 3 and 4).
In this prospective longitudinal study, we examined several biomarkers for CVD to test the hypothesis that these biomarkers are associated with an increased risk for progression to advanced AMD. We found that both CRP and IL-6, markers of systemic inflammation, were significantly and independently related to AMD after adjustment for known and potential confounding factors. The highest quartile of CRP was significantly associated with progression of AMD, with a twofold greater risk compared with the lowest quartile of CRP. Similarly, the highest quartile of IL-6 was significantly related to progression of AMD, with an almost twofold greater risk compared with the lowest quartile. For CRP, the estimates of risk were increased above the first quartile, and the trend was marginally significant. The IL-6 values in the third and fourth quartiles were associated with increased risk, and the trend for increasing risk of AMD with increasing levels of IL-6 was statistically significant. These results strengthen and expand the previously reported finding of a cross-sectional association between CRP and advanced stages of AMD.14
Smoking and body mass index are known risk factors associated with AMD,2,3,10,11 and we found that current smoking as well as intermediate to high levels of body mass index had a significant positive association with both CRP and IL-6. Thus, these two variables meet the definition of a confounder. However, even after adjustment for these and other factors, CRP and IL-6 were related to progression of AMD. Although some unmeasured and therefore uncontrolled factors might still be confounding these relationships, they would have to be both highly associated with CRP and IL-6 and strong risk factors for AMD to substantially alter the relationships seen.
To the best of our knowledge, the results in this article are the first in a prospective cohort study to show an association between both CRP and IL-6 levels and progression from early or intermediate disease to advanced stages of AMD. Our findings lend further support to the hypothesis that AMD may be partially mediated through inflammatory and immune-related mechanisms. Basic research has suggested that AMD shares biological pathways similar to those found in other inflammatory diseases, such as Alzheimer disease and atherosclerosis33- 37; AMD may have “a common response pattern,” since drusen resemble plaque and deposit compositions seen in these other diseases.33- 37 Drusen constituents have been implicated in inflammatory and immune-mediated processes, including dendritic cells, suggesting that drusen biogenesis may be a result of the eye’s inability to return “to tolerance” from a state of chronic localized inflammation.35 Leukocytes have been found in the choroid of eyes with AMD, and autoantibodies directed against drusen, RPE, and retina components have been detected in serum samples from patients with AMD.46 There are several mechanisms that may be involved in the relationship between inflammation and AMD, one of which is oxidative stress. Age, smoking,2,3 obesity,10,11 dietary fat,5,6,8 and insufficient antioxidants in the diet4,7 are some of the known risk factors for AMD that can lead to oxidative stress. Oxidative stress may cause injury to the RPE, Bruch membrane, and choroid, layers in the eye involved in the pathophysiology of AMD, and the eye may then react to this injury with an inflammatory response.33- 37
Reducing levels of CRP and IL-6 might have a favorable impact on the progression rate of AMD. One treatment to reduce the levels would be statin medications, since they have been shown not only to reduce low-density lipoprotein cholesterol levels but also to have anti-inflammatory as well as antioxidant properties.47,48 There have been a few conflicting reports regarding the potential beneficial effect of statin therapy in reducing the risk of AMD.48,49 Furthermore, some favorable results have been reported from case series of patients with neovascular AMD treated with triamcinolone,50 which has anti-inflammatory properties.
In this article, TNF-α-R2, ICAM-1, VCAM-1, Lp(a), and ApoB were not found to be significantly associated with AMD. However, VCAM-1 did show a nonsignificant trend for a positive association with progression of AMD. This finding corresponds to the report that elevated VCAM-1 levels were associated with late stages of coronary artery disease.24 Perhaps VCAM-1 may be an indicator of the late stages of disease. Additionally, ApoB is a marker for atherosclerosis,27,28 and in this study ApoB was found to have a nonsignificant but weakly positive association with AMD. It is not known why these biomarkers had different results, but it is possible that some of the relationships that were not statistically significant in these analyses may become statistically significant with a larger sample size.
There are several strengths of this study, including the well-characterized study population and prospective design, providing a unique opportunity to evaluate our hypotheses. By using a prospective longitudinal study design we were able to evaluate the temporal relationships between events that occur prior to the progression of the disease. Further strengths of the study include standardized data-collection instruments, including interviews and direct measurements of height, weight, and blood pressure; assessment of AMD end points by standardized ophthalmologic examination and fundus photography for all subjects; and a low rate of loss to follow-up. Measures of the various biomarkers were taken from single fasting blood specimens that were stored in a repository at −140°C. These are standard methods in use in several large-scale epidemiologic studies throughout the country. In fact, these are the same methods used in the studies that established CRP and IL-6 as markers for cardiovascular diseases.18 The medians and ranges of the biomarkers in the various quartiles in our study are similar to those in other published studies of biomarkers and cardiovascular diseases.15- 28 Since a single measurement was taken, we cannot evaluate the effects of changes in the levels of these biomarkers over time. Misclassification was unlikely because biomarker values were quantified by using objective laboratory methods without knowledge of the participants’ maculopathy status, and AMD grade was assigned without knowledge of the biomarker levels.
This study population is comparable to other populations of similar age range and sex distribution for parameters including smoking status at baseline and rates of progression.7 Also, the biological effects of these biomarkers are not likely to differ in major ways among various populations of patients with AMD. Therefore, the results might be applicable to the 8 million people in the United States, as well as similar populations elsewhere, who have early or intermediate stages of AMD and are at high risk for future progression to advanced AMD and visual loss.51
In summary, to our knowledge, this is the first prospective study to report a positive association between the systemic inflammatory markers CRP and IL-6 and the rate of progression to advanced AMD. Smoking and obesity were significantly related to both CRP and IL-6 levels. Higher values of CRP and IL-6 were found to be significantly related to AMD independent of these and other established risk factors. These results, together with the cross-sectional report of an association between AMD and CRP14 and previous basic research on this topic,33- 37,46 may shed light on the mechanisms and pathogenesis of AMD development and prognosis. Moreover, levels of inflammatory factors may add clinically relevant, predictive information concerning risk of AMD and response to treatment, in addition to known, more well-established risk factors. Anti-inflammatory agents may have a role in preventing AMD, and inflammatory biomarkers may provide a method of identifying people for whom these agents would be more or less effective. These results and hypotheses require confirmation by additional prospective studies and possibly randomized trials.
Correspondence: Johanna M. Seddon, MD, ScM, Epidemiology Unit, Massachusetts Eye and Ear Infirmary, 243 Charles St, Boston, MA 02114 (Johanna_Seddon@meei.harvard.edu).
Submitted for Publication: December 8, 2004; accepted: December 13, 2004.
Author Contributions:Study concept and design: Seddon. Acquisition of data: Seddon, George, Rifai. Analysis and interpretation of data: Seddon, George, Rosner, Rifai. Drafting of the manuscript: Seddon, George, Rosner. Critical revision of the manuscript for important intellectual content: Seddon, George, Rosner, Rifai. Statistical analysis: Seddon, Rosner. Obtained funding: Seddon. Administrative, technical, and material support: Seddon, George, Rifai. Study supervision: Seddon, Rosner.
Financial Disclosure: MEEI has pending US and international patent applications that include related subject matter. In the event MEEI receives proceeds related to subject matter described in this article, all proceeds will be distributed per the institutional policies.
Funding/Support: This study was supported by the Foundation Fighting Blindness Inc, Owing Mills, Md; the Massachusetts Lions Eye Research Fund Inc, Northboro; the Epidemiology Unit Research Fund, Massachusetts Eye and Ear Infirmary, Boston; DSM Inc, Parsippany, NJ; in part by grant EY013982 from the National Eye Institute, Bethesda, Md; and by a Lew R. Wasserman Merit Award from Research to Prevent Blindness Inc, New York, NY. Some of these markers are also included in EY013834 (D. Schaumberg).
Role of the Sponsor: The funding agencies did not influence the design and conduct of the study; the collection, analysis, and interpretation of the data; or the preparation or approval of the manuscript.
Acknowledgment: We thank Marion McPhee, BEd, from the Harvard University Channing Laboratory, Boston, Mass, for her SAS programming assistance.