Inflammatory Markers in Age-Related Maculopathy: Title and subTitle BreakCross-sectional Analysis From the Muenster Aging and Retina Study FREE

Age-related maculopathy (ARM) and its late stage, age-related macular degeneration (AMD), are degenerative disorders of the central part of the retina and are the leading cause of visual impairment and legal blindness among older persons in industrialized countries. Although the natural history of ARM has been studied and described in several epidemiological investigations,^{1 - 3} its pathogenesis and progression are insufficiently understood. Because of this lack of understanding, treatment for ARM remains limited.^{4 - 6} With the growing population of older persons, it is imperative to identify factors associated with the onset and progression of ARM.

As in other degenerative diseases associated with aging,^{7 - 10} it has been hypothesized that chronic inflammation may underlie ARM.^{11 - 12} Immunohistochemical investigations revealed that drusen contain numerous proteins common to the inflammatory process.^{7 - 8} Histopathological changes in exudative ARM also suggest an immune-mediated reaction.¹³

Fibrinogen and C-reactive protein (CRP) are established systemic inflammatory biomarkers linked to cardiovascular diseases.^{14 - 17} The investigation of inflammatory biomarkers in ARM may be complicated by the observation that cardiovascular risk factors and disease tend to coincide with the occurrence of ARM.^{18 - 21} In fact, epidemiological studies^{22 - 26} have shown inconsistent associations between inflammatory markers and ARM. More recently, however, Seddon et al²⁷ reported a relationship between high-sensitivity CRP and ARM. In a case-control setting, serum CRP levels were significantly higher among participants with advanced ARM than among those without ARM. After adjustment for age, sex, and other variables, including smoking history and body mass index, CRP concentrations were significantly associated with the presence of intermediate and advanced stages of AMD. To further evaluate the inflammatory hypothesis, we studied the association between fibrinogen and CRP levels and the risk of ARM in the baseline examination of the Muenster Aging and Retina Study (MARS).

The MARS is a longitudinal study designed to identify prognostic medical, environmental, and genetic factors for the progression of ARM. From June 2001 to October 2003, a cohort of 1060 subjects was assembled. Ophthalmologists in Muenster and the surrounding county of Westphalia were asked to refer eligible patients to the MARS center. In addition, subjects were identified by reviewing patient records from a retina specialty clinic (St Franziskus Hospital, Muenster, Germany). Eligibility criteria constituted morphologic evidence of ARM in at least 1 eye, nonexudative disease in at least 1 eye, and age at diagnosis between 60 and 80 years. However, 8 persons not meeting the age criteria were also included in the study, so the age range of the cohort was 59 to 82 years. Persons without ARM, primarily volunteers and family members or friends of the study participants, were recruited as control subjects for future genetic investigations. Persons having a diagnosis of narrow-angle glaucoma were excluded because of a contraindication to pupil dilatation for mydriatic fundus photography. The recruitment and research protocols were reviewed and approved by the Institutional Review Board of the University of Muenster, and written informed consent was obtained from all study participants.

At the time of enrollment, all subjects were interviewed by a trained interviewer using a standardized risk factor questionnaire. Information was obtained about demographic characteristics, smoking history, lifestyle, nutrition, and current use of medications, micronutrients, and supplements. Subjects were asked to report any prior diagnosis of medical conditions, including eye disease or surgery, hypertension, dyslipidemia, diabetes mellitus, and cardiovascular events (myocardial infarction or stroke). Baseline examinations included measuring of height, weight, pulse rate, and oscillometric blood pressure. Hypertension was defined as a systolic blood pressure of 140 mm Hg or higher, a diastolic blood pressure of 90 mm Hg or higher, or the use of antihypertensive medications. Dyslipidemia was defined as a total cholesterol–high-density lipoprotein cholesterol ratio of 5 or higher. A history of cardiovascular events was assumed if the participant responded affirmatively to questions regarding past myocardial infarction or stroke. Diabetes mellitus was defined as the current use of hypoglycemic medication or a hemoglobin A_1c level greater than 6.5%. Body mass index, calculated as weight in kilograms divided by the square of height in meters, was divided into 3 categories (<25, 25-29.9, or ≥30).

All subjects underwent a comprehensive ophthalmologic examination using slitlamp microscopy. Best-corrected visual acuity in all eyes was measured with Early Treatment of Diabetic Retinopathy Study charts. Scotoma and metamorphopsia were assessed with the Amsler chart. In addition, subjective visual function was evaluated using the validated German version of the 25-item National Eye Institute Visual Function Questionnaire.²⁸

After pharmacologic mydriasis, the fundus was examined through the dilated pupil with the slitlamp for a clinical assessment of the fundus morphologic characteristics. Subsequently, 30° stereoscopic digital color photographs were taken from both eyes, centered on the fovea (FF 450 fundus camera; Carl Zeiss, Oberkochen, Germany). For digital images, a Kodak Professional DCS camera (Eastman Kodak Company, Rochester, NY) with approximately 1200 × 1152-pixel resolution and Visupac picture analysis software (Hiko, Pirmasens, Germany) were used. All digital images were stored on a magnetic optical disk. Photographs were evaluated on a 22-inch computer screen. Total magnification was approximately 38-fold.

The presence and severity of retinal lesions were graded according to the protocol of the International ARM Epidemiological Study Group.²⁹ We used a standard grid with central, middle, and outer circles, the radii of which were 500, 1500, and 3000 μm, respectively. Spokes split the middle and outer circles into upper, lower, temporal, and nasal zones, resulting in 9 subfields.

In collaboration with staff from the Rotterdam Study,² we trained staff members in taking and grading digitalized fundus photographs. The interobserver agreement of color fundus photographs was assessed by comparing 25 randomly selected digitalized photographs from the 3 MARS graders with 75 photographs from the Rotterdam Study graders. There was substantial, consistent agreement between the MARS and Rotterdam Study gradings (weighted κ value range, 0.60-0.80) for various classifications (early stage of ARM and atrophic or neovascular AMD), and there was very good agreement among the 3 MARS observers (weighted κ value, >0.80).

Fundus signs within the standard grid around the fovea included the following items: the number of drusen (0, <10, 10-19, or ≥20) in 4 drusen size categories (<63, 63 to <125, and ≥125 μm), the largest drusen size (<63, 63 to <125, 125 to <175, or ≥175 μm), the most frequent drusen size (<63, 63 to <125, 125 to <175, or ≥175 μm), and the confluence of drusen (none, <10%, <50%, or ≥50%). To assess the phenotype of ARM, we separately graded the following for each of the 9 subfields: the most severe drusen type (hard, soft distinct <125 μm, soft distinct ≥125 μm, soft indistinct, or reticular), the extent of the grid area occupied by the drusen (<1%, <10%, <25%, <50%, or ≥50%), and the grid area occupied by atrophic or neovascular AMD (none, <25%, <50%, or ≥50%). Retinal pigment epithelium changes were determined by the presence of increased pigmentation or hypopigmentation (none, <125, 125 to <175, or ≥175 μm) within the central, middle, and outer circles. Atrophic AMD was defined as any sharply delineated area 175 μm or larger of hypopigmentation or depigmentation or the apparent absence of the retinal pigment epithelium, with visible choroidal vessels. Neovascular ARM was characterized by serous or hemorrhagic retinal pigment epithelium detachment, subretinal neovascular membrane, subretinal hemorrhage, or periretinal fibrous scar.

We used the Rotterdam Study² classification system for the definition of ARM stages. Therefore, fundus signs were stratified into the following 5 levels of severity: stage 0, no sign of ARM or hard drusen (<63 μm) only; stage 1, soft distinct drusen (≥63 μm) only, or pigment epithelium changes only and no soft drusen (≥63 μm); stage 2, soft indistinct drusen (≥125 μm) or reticular drusen only, or soft distinct drusen (≥63 μm) with pigment epithelium changes; stage 3, soft indistinct drusen (≥125 μm) or reticular drusen with pigment epithelium changes; and stage 4, atrophic or neovascular AMD. A fundus without ARM signs was considered stage 0. Early ARM corresponded to stages 1 through 3, and late ARM (AMD) was defined as stage 4.

Digital fundus images were not taken in some patients for the following reasons: refusal (2 patients), technical failure of the retina camera (5 right eyes and 15 left eyes), severe opacities of the lens or cornea (6 right eyes and 5 left eyes), and poor dilatation (2 right eyes and 5 left eyes). Because of other adverse eye conditions (retinal vein occlusion, diabetic retinopathy, epiretinal gliosis, macular foramen, and laser coagulation scar), the ARM grade was not assessed in 44 right eyes and 40 left eyes. Overall, gradable digital fundus photographs were available in at least 1 eye of 1003 subjects (95%) and in both eyes of 974 subjects (92%).

Fibrinogen and CRP assays were performed on thawed blood samples. Clotting fibrinogen levels were measured in sodium citrate plasma according to the method of Clauss³⁰ on a BCS coagulation analyzer using Multifibern U (Dade Behring, Marburg, Germany). C-reactive protein levels were measured in serum with highly sensitive (detection limit, 0.2 mg/L) particle-enhanced immunonephelometry on a BN II nephelometer (N High Sensitivity CRP, Dade Behring). Both tests were carried out according to the instructions of the manufacturer. Interassay coefficients and intra-assay coefficients of variation were 5.2% and 3.5%, respectively, for fibrinogen measurements and 8.0% and 4.7%, respectively, for CRP measurements.

Of the 1060 subjects examined at baseline, 873 (82%) had bilateral gradable fundus photographs and complete data on fibrinogen, CRP, and potential confounders. Study participants were classified as having a normal fundus, early ARM, or late ARM according to the worst stage in either eye. Continuous variables and categorical variables were compared among the groups using the Kruskal-Wallis test and the χ² test, respectively. We compared the distribution of fibrinogen and CRP levels among the groups by means of medians and quartiles using the Mann-Whitney test. In addition, cumulative distribution functions were calculated.

Logistic regression was used to model the association of fibrinogen and CRP levels with the risk of ARM. For these analyses, the ARM stages of individual eyes were used as outcome variables. Early and late stages of ARM were combined into a single, binary response variable of ARM. We used a marginal model accounting for correlations between the outcomes of pairs of eyes with an additional variable for the log odds ratio (OR) between left and right eyes.³¹ Fibrinogen and CRP levels were categorized as quartiles based on levels among controls with normal fundus in both eyes. We calculated ORs and 95% confidence intervals (CIs) for each fibrinogen and CRP quartile, with the lowest quartile as the reference category. Multivariate logistic regression analyses were performed, adjusting for age and sex (model 1) and for current smoking, hypertension, dyslipidemia, overweight, diabetes mellitus, and history of cardiovascular disease (model 2). All analyses were performed using SAS version 8 (SAS Institute Inc, Cary, NC).

Table 1 gives the baseline characteristics of the study population according to their ARM status. Of the 873 participants, 181 (21%) had normal results on bilateral fundus examination, 422 (48%) had unilateral or bilateral early ARM, and 270 (31%) had unilateral or bilateral late ARM. The mean age of the participants was 70.8 years; 59% were women. Compared with controls, patients with early or late ARM were older, more frequently were smokers, and had a higher prevalence of hypertension, overweight, diabetes mellitus, and history of cardiovascular disease.

The median levels for both inflammatory markers increased with maculopathy stage (Table 2). Among patients with late ARM, the fibrinogen and CRP levels were significantly higher than those among participants without ARM. By contrast, while the fibrinogen levels were also significantly higher among patients with early ARM, their CRP levels were nonsignificantly higher (P = .07).

The cumulative frequency plots (Figure 1 and Figure 2) illustrate that the distributions of both inflammatory markers were shifted to slightly higher values for patients with early and late stages of ARM compared with controls. However, the distinction was not pronounced.

Table 3 and Table 4 give the ORs for the binary ARM outcome, that is, the combination of early and late stages of ARM based on single-eye observations, according to the quartiles of fibrinogen or CRP levels. In unadjusted analyses, the risk of ARM increased with higher concentrations of fibrinogen and CRP. For fibrinogen, persons in the highest quartile had a nearly 2-fold increased risk of ARM compared with those in the lowest quartile (OR, 1.90; 95% CI, 1.29-2.80; P<.001 for trend) (Table 3). Furthermore, a positive but not statistically significant association between CRP levels and ARM was found (OR, 1.43; 95% CI, 0.97-2.10; P = .06 for trend) (Table 4). Adjustment for age and sex resulted in attenuated associations, which were markedly affected by additional adjustments for classic cardiovascular risk factors. The multivariate-adjusted ORs for the highest vs the lowest quartile were 1.37 (95% CI, 0.91-2.06; P = .04 for trend) for fibrinogen (Table 3) and 1.12 (95% CI, 0.73-1.73; P = .67 for trend) for CRP (Table 4).

Histopathological and immunohistochemical findings suggest that immune-mediated processes may play a causal role in the pathogenesis of ARM.^{13 ,32} These immune reactions are similar to chronic inflammatory reactions in other age-related degenerative diseases, such as atherosclerosis and Alzheimer disease.^{7 - 10}

Evidence of inflammatory cell involvement in the later stages of ARM includes the presence of giant multinucleated cells and other leukocytes in the choroid of AMD eyes,^{33 - 35} and it is likely that these cells represent a source of potent lytic enzymes, necessary for neovascularization and degradation of the Bruch membrane. In the early stages of ARM, alterations of retinal microglia capable of acting as antigen-presenting cells are detectable.³⁶ In addition, immunohistochemical analyses have shown that drusen contain different immune reactants, including complement, immunoglobulins, and acute-phase proteins.^{7 - 9 ,37 - 39} A possible trigger of inflammation may be an injured retinal pigment epithelium, caused by hypoxia, oxidative insults, lipofuscin accumulation, drusen constituents, or other abnormal changes in the extracellular environment in the vicinity of the Bruch membrane.

Fibrinogen and CRP are hepatic acute-phase proteins whose blood concentrations vary in different pathologic processes, including infections and tissue damage. Both proteins are well-established biomarkers for acute and chronic inflammation and are linked to cardiovascular diseases.^{14 - 17}

In a crude analysis, our study showed an association between raised concentrations of fibrinogen and CRP and a moderate risk of early or late ARM. The relationship was somewhat stronger with fibrinogen than with CRP. Overall, the results revealed no statistical significance after considering concomitant cardiovascular risk factors. Therefore, our findings do not support a clear independent role of inflammatory markers in ARM beyond what is explained by concomitant cardiovascular risk factors and disease. This may be interpreted as a lack of any contribution to ARM. However, the concurrence of ARM with cardiovascular disease in the same individuals makes it methodologically difficult to separately quantify the contribution of mechanisms and their respective markers that are common to both diseases. Therefore, another interpretation compatible with our findings is that chronically raised levels of inflammation (eg, due to disseminated atherosclerosis) confer an increased susceptibility and response to cellular damage in other organs (eg, the retina). Furthermore, our findings support the hypothesis by Friedman⁴⁰ that the pathogenesis of ARM is largely based on ocular atherosclerotic mechanisms impairing choroidal perfusion.

Evidence from other epidemiological studies is not consistent regarding the association of inflammatory markers with ARM. In the Beaver Dam Eye Study,²⁶ high white blood cell counts were related to an increased 10-year risk of development of large drusen, depigmentation of retinal pigment epithelium, and progression of ARM. In the same study, hypoalbuminemia, another inflammatory sign, had an association with the 10-year incidence of choroidal neovascularization. In the Blue Mountains Eye Study,²² elevated plasma fibrinogen concentrations were associated with late ARM but not with early ARM. However, this finding was not confirmed in the long-term follow-up phase of that study.⁴¹ Furthermore, plasma fibrinogen levels were unrelated to ARM in the Cardiovascular Health Study,²⁴ the Eye Disease Case-Control Study,²³ or the Rotterdam Study.²⁵

Recently, Seddon et al²⁷ examined the relationship between high-sensitivity CRP levels and ARM in 930 individuals from the multicenter Age-Related Eye Disease Study (AREDS). After adjustment for confounding variables, CRP levels were significantly associated with the presence of intermediate and advanced stages of ARM. The OR for the highest vs the lowest quartile of CRP was 1.65 (95% CI, 1.07-2.55). In contrast with our findings, controlling for cardiovascular and other confounders in the study by Seddon et al resulted in a strengthening rather than an attenuation of the crude and age- and sex-adjusted ORs. This occurred despite the association of these factors with maculopathy stages that were similar to those found in the MARS. Furthermore, median CRP levels in the AREDS were 3.4 mg/L among AMD cases and 2.7 mg/L among disease-free controls. These concentrations were much higher than those observed in our study, and the reasons are unclear. Differences in biochemical determinations could not have affected strengths of associations unless they were differential at varying concentration levels, which appears unlikely. We used the Rotterdam Study classification system to categorize the ARM stages, while Seddon et al used the AREDS classification system. In addition, the AREDS and MARS cohorts comprised populations from different countries. The study populations also differed in mean ± SD age (69.0 ± 5.1 years in the AREDS vs 70.8 ± 5.5 years in the MARS), race (97% vs 100% white), and the presence of hypertension as defined in the AREDS (35% vs 61%).

There are potential limitations of our study. First, our analyses were cross-sectional, which may be problematic when analyzing risk relations in older persons. If ARM patients have a higher mortality compared with controls, this may introduce a selective survival bias. A higher frequency of fatal cardiovascular disease in ARM patients could result in such a selection. However, if present, the selection would lead to lower levels of inflammatory markers in ARM patients who survive owing to less or absent concurrent cardiovascular disease. The risk associated with raised fibrinogen or CRP levels would be underestimated rather than overestimated. Second, comparisons between the characteristics of our controls and the age-specific data from the German National Health Interview and Examination Survey 1998^{42 - 44} indicate a slightly better general health status, particularly a lower distribution of cardiovascular risk factors, among our controls. This might falsely inflate the differences observed in the ARM patients. However, comparing the CRP levels in the MARS with those from another population-based survey in Germany,⁴⁵ we found no substantial difference in CRP concentrations. Therefore, a bias due to low inflammation levels among our controls appears unlikely. Third, levels of fibrinogen and CRP are systemic biomarkers and might not reflect inflammation levels in the ocular region.

After adjustment for cardiovascular risk factors, the baseline examinations of the MARS cohort did not demonstrate that elevated fibrinogen and CRP concentrations were significantly associated with ARM. Our results do not indicate a role of systemic inflammation in ARM beyond what is already present owing to concurrent cardiovascular disease.

Correspondence: Hans-Werner Hense, MD, PhD, Institute of Epidemiology and Social Medicine, University of Muenster, Domagkstrasse 3, 48149 Muenster, Germany (hense@uni-muenster.de).

Submitted for Publication: June 21, 2004; final revision received January 25, 2005; accepted February 17, 2005.

Financial Disclosure: None.

Funding/Support: The study was supported by grant HE 2293/5-1 from Deutsche Forschungsgemeinschaft, by the ProRetina Deutschland, and by the Intramurale Förderüng fund of the University of Muenster.

Acknowledgment: We thank Ada Hooghart and Corina Brussee from the Rotterdam Study team for their contributions to fundus grading.