Lin S, Wang SY, Yoo C, Singh K, Lin SC. Association Between Serum Ferritin and Glaucoma in the South Korean Population. JAMA Ophthalmol. 2014;132(12):1414–1420. doi:10.1001/jamaophthalmol.2014.2876
Evidence suggests that altered iron metabolism may be associated with oxidative damage to several organ systems, including the eye. Supplementary iron consumption is also associated with greater odds of self-reported glaucoma.
To investigate the association between serum ferritin level and the likelihood of a glaucoma diagnosis in a cross-sectional, population-based study.
Design, Setting, and Participants
Data were collected from 17 476 participants in the first and second years of the Fifth Korea National Health and Nutrition Examination Survey, a cross-sectional study of the South Korean population conducted from January 1, 2010, through December 31, 2011. Data pertaining to the serum ferritin level were aggregated and divided into quartiles. Demographic, comorbidity, and health-related behavior information was obtained via interview.
Main Outcomes and Measures
The presence or absence of glaucoma. The definition of glaucoma was based on criteria established by the International Society of Geographical and Epidemiological Ophthalmology.
Participants whose serum ferritin level was greater than 61 ng/mL (to convert to picomoles per liter, multiply by 2.247) had significantly higher odds of a glaucoma diagnosis when compared with those with a level less than 31 ng/mL, after adjustment for potential confounders (ferritin levels of 31-61 ng/mL: odds ratio [OR], 1.17; 95% CI, 0.84-1.62; ferritin levels of 62-112 ng/mL: OR, 1.60; 95% CI, 1.16-2.20; and ferritin levels of 113-3018 ng/mL: OR, 1.89; 95% CI, 1.32-2.72).
Conclusions and Relevance
Our study reveals that a higher serum ferritin level was associated with greater odds of glaucoma in a representative sample of the South Korean population, even at levels normally observed in the general population. This novel finding may help elucidate the pathogenesis and lead to novel therapeutic approaches for glaucomatous disease.
Glaucoma is an optic neuropathy with retinal ganglion cell loss and structural alteration of the optic nerve head, accompanied by progressive vision loss. Glaucomatous disease is the leading cause of irreversible blindness worldwide1 and a growing public health concern because of an aging worldwide population.2 Elucidation of risk factors associated with disease pathogenesis may help prevent the resulting disability.
Previously identified important risk factors for the development of glaucoma include greater intraocular pressure (IOP), age, and a family history of disease.3 Other concomitant factors that have been postulated to play important roles in disease pathogenesis include high glutamate levels,4 alterations in nitric oxide metabolism,5 vascular alterations,6 and oxidative damage.7,8 An increasing body of evidence suggests that oxidative stress and free radical accumulation may harm the trabecular meshwork (TM),9- 13 resulting in an increase in IOP. Such oxidative stress may also be harmful to retinal ganglion cells, with resultant cell death leading to optic nerve damage characteristic of glaucomatous disease.8
There has been considerable interest in the role of the body’s alteration of iron metabolism as a cause of oxidative damage to several organ systems. Accumulation of iron has been implicated in oncogenesis, aging, and progression of chronic neurodegenerative conditions, such as Alzheimer and Parkinson disease.14- 22 The concentration of serum ferritin, an iron-binding protein that is distributed throughout the body, serves as a surrogate indicator of the body’s iron stores.23 Ferritin is considered to be the preferred marker for the assessment of iron-related oxidative stress.24,25
A previous study26 found an association between self-reported glaucoma and iron ingestion in the US population. Another study27 found that supplementary iron consumption was associated with greater odds of self-reported glaucoma. Such prior work has not assessed the association between body iron stores and glaucomatous disease. The availability of data pertaining to serum ferritin levels in the first and second years of the Fifth Korea National Health and Nutrition Examination Survey (KNHANES V), a large population-based survey conducted from January 1, 2010, through December 31, 2011, in South Korea, has allowed us to perform such an analysis.
All analyses were based on data from KNHANES V. The KNHANES V is a cross-sectional survey conducted under the auspices of the Korea Centers for Disease Control and Prevention with approval by its institutional review board. The KNHANES consists of the health interview, health behavior and nutrition surveys, and a health examination. The survey adhered to the principles outlined in the Declaration of Helsinki for research that involves humans, and all participants provided written informed consent. This nationwide representative study for noninstitutionalized civilians uses a stratified, multistage probability sampling design with a rolling survey sampling model.
Data on demographic characteristics, diet, and health-related variables were collected through personal interview and a self-administered questionnaire. Physical examination and blood and urine sampling were performed at a mobile examination center. Ophthalmologic interview questions and examinations were added in 2008 and thus available for KNHANES V.
In the first (2010) and second years (2011) of the KNHANES V, 21 527 individuals were sampled and 17 476 individuals participated in this survey (8958 of 10 938 in 2010 and 8518 of 10 589 in 2011, for a response rate of 81.2%). Data were available for right eyes of 12 486 participants who were 19 years or older and received an ophthalmologic examination. Participants with retinal abnormalities or a history of stroke were excluded in an effort to minimize the risk of visual field (VF) defects related to these conditions that led to an incorrect diagnosis of glaucomatous disease. Retinal abnormalities that led to exclusion were as follows: retinal detachment, diabetic retinopathy, or any signs of age-related macular degeneration, including drusen, exudation, sensory retinal detachment, retinal pigment epithelium abnormalities, retinal pigment epithelium detachment, subretinal or sub–retinal pigment epithelium hemorrhages, subretinal scars, geographic atrophy, or disciform scars. Participants who lacked data on ferritin levels were also excluded, leaving a total of 9632 individuals for our analyses.
The method of ophthalmologic examination in the KNHANES has been described in a prior publication.28 After an ophthalmology-focused interview, participants underwent visual acuity measurements, automated refraction, slitlamp examination, IOP measurement, fundus photography, and, when deemed appropriate, VF examination. The primary outcome variable was the presence or absence of a glaucoma diagnosis as defined by International Society of Geographical and Epidemiological Ophthalmology (ISGEO) criteria as described in a prior publication.29 Category 1 requires a VF defect consistent with glaucoma and either a vertical cup-disc ratio (VCDR) of 0.7 or greater (97.5th percentile) or asymmetry of VCDR between the right and left eyes of 0.2 or greater (97.5th percentile). Category 2 criteria, used when VF results are not definitive, requires a VCDR of 0.9 or greater (99.5th percentile) or asymmetry of a VCDR of 0.3 or greater (99.5th percentile). Category 3 criteria, used when no information on VF testing or optic disc examination is available, requires a visual acuity less than 3/60 and an IOP greater than the 99.5th percentile for this population (IOP, 21 mm Hg). Testing of visual function was performed with frequency doubling technology (FDT) in those suspected of having glaucoma as defined by the ISGEO criteria, and a defect was deemed present if 2 different test points were abnormal.28The test was performed again if the proportion of fixation errors or false-positive responses was greater than 0.33. If a participant failed the fixation cutoff on 2 VF attempts, the test was considered invalid. Participants who were unable to perform valid VF testing were then considered for a glaucoma diagnosis using ISGEO category 2 or 3 criteria.
The primary predictor variable was serum ferritin level. Blood samples were obtained after a minimum fasting time of 8 hours and were immediately processed, refrigerated, and transported in cold storage to the central testing institute (NeoDin Medical Institute), where analysis was performed within 24 hours of receipt. Serum ferritin was measured by immune radiometric assay using a 1470 WIZARD γ-counter (PerkinElmer).
Potential confounding variables included age; sex; income status; educational level; health-related behavior, such as smoking, alcohol use, exercise, and basal metabolic index; and medical comorbidities, such as refractive status, anemia, renal failure, diabetes mellitus, hypertension, and hyperlipidemia. In addition to assessment of demographic information and medical history, participants underwent examination of blood pressure and testing of blood and urine. Refractive status was divided into 5 groups: emmetropia (−0.99 to 0.99 diopters [D]), mild myopia (−1.00 to −2.99 D), moderate myopia (−3.00 to −5.99 D), severe myopia (<−5.99 D), and hyperopia (>1.00 D).
Complex sample analysis was used for the KNHANES V data for weighting all values following statistical guidance from the Korea Centers for Disease Control and Prevention. Serum ferritin level was sorted into quartiles based on concentrations: 30 ng/mL or less for quartile 1, 31 through 61 ng/mL for quartile 2, 62 through 112 ng/mL for quartile 3, and greater than 112 ng/mL for quartile 4 (to convert to picomoles per liter, multiply by 2.247). The regression model was constructed after identification of potential confounding variables. All risk factors that were identified as being associated with a glaucoma diagnosis by univariate analysis with P < .05 as the cutoff point were then included in the multivariable analysis to assess the possible independent association between quartiles of serum ferritin level and glaucoma status as determined by ISGEO criteria. After ascertainment of such a possible association, 95% CIs of odds ratios (ORs) were identified for each possible association. A test for trends across quartiles was performed by accepting variables for each quartile as being continuous in logistic regression models. Two-sided statistical tests were performed with SPSS statistical software, version 21.0 (SPSS Inc).
After excluding 2854 ineligible participants, 9632 individuals were included in our analysis, of whom 150 met the ISGEO criteria for suspected glaucoma diagnosis, representing 1.1% (SE, 0.1%) of the South Korean population. Table 1 presents information on unadjusted demographic characteristics, comorbidities, health-related behavior, and serum ferritin level between those with and without glaucoma diagnosed using the ISGEO criteria. Participants with glaucoma were older and less educated than those without the disease. There was a predominance of men over women, and health-related behaviors, such as smoking status and exercise, also differed between the 2 groups: those with glaucoma exercised less and smoked more. Glaucoma status was associated with a greater prevalence of severe myopia or hyperopia. Table 2 lists the demographic characteristics of included and excluded study participants, with the latter group being older. There were 47 excluded participants with ISGEO-defined glaucoma, among whom 15 lacked serum ferritin data, 37 had retinal abnormalities, and 6 had a history of stroke.
Among the 12 486 study participants, 88 had self-reported glaucoma, 8536 had no self-reported glaucoma, and 3862 did not respond to this question. Fifty-five participants with self-reported glaucoma had received IOP-lowering glaucoma therapy. Of the participants with self-reported glaucoma, 26 (29.5%) were categorized as having ISGEO-defined glaucoma, and the remaining 62 did not meet such criteria. Among these 62 participants, 6 had VF abnormalities in the absence of documented structural optic nerve damage needed for a category 1 glaucoma diagnosis by ISGEO criteria, and 23 participants had normal VF results and thus did not satisfy ISGEO criteria. Thirty-three participants with self-reported glaucoma did not receive VF testing. Of these participants, only 8 met the criteria for suspecting of having glaucoma, and none met the optic disc criteria for an ISGEO glaucoma diagnosis.
Table 3 lists the demographic characteristics of the 1634 participants suspected of having glaucoma who did not receive VF testing. These participants were, on average, older (mean age, 44.3 vs 42.4 years, respectively) and had relatively higher mean serum ferritin levels (92.3 and 84.7 ng/mL, respectively; P = .054) compared with the remaining 7998 participants. After evaluation for a glaucoma diagnosis based on the ISGEO category 2 and 3 criteria, disease prevalence remained higher among these 1634 participants compared with the other study participants (2.3% and 0.8%, respectively).
Serum ferritin level, which was significantly greater in participants with vs without glaucomatous disease (mean, 116.1 and 85.7 ng/mL, respectively; P = .004), was further categorized into quartiles, and the distribution of quartiles also differed based on glaucoma diagnosis (Table 4) (P = .005). Multivariable logistic regression models were constructed to assess a possible association between the odds of a glaucoma diagnosis among those with various serum ferritin levels. Potential confounding variables were included in the multivariable analysis if the univariate analysis revealed P < .05 for the regression coefficient related to such variables. Confounding factors included in the multivariable analysis were age, sex, educational level, smoking status, exercise, hypertension, diabetes mellitus, and refractive status.
The third and fourth quartile serum ferritin levels (≥62 ng/mL) were associated with significantly higher odds of a glaucoma diagnosis compared with those found with ferritin levels lower than 31 ng/mL, even after adjustment for demographic characteristics, health-related behavior, and comorbidities (ferritin levels of 31-61 ng/mL: OR, 1.17; 95% CI, 0.84-1.62; ferritin levels of 62-112 ng/mL: OR, 1.60; 95% CI, 1.16-2.20; and ferritin levels >112 ng/mL: OR, 1.89; 95% CI, 1.32-2.72; P for trend < .001). Subsequent analyses using serum iron level and serum total iron-binding capacity did not find an association with a glaucoma diagnosis using either measurement (OR, 1.31; 95% CI, 0.81-2.13; and OR, 1.29; 95% CI, 0.81-2.07, respectively) after adjustment for confounders.
We also found that women 50 years or older had significantly higher serum ferritin levels than those younger than 50 years (mean serum ferritin level of 64.8 ng/mL compared with 33.9 ng/mL, respectively; P < .001), and, as expected, there was a greater prevalence (SD) of glaucoma in the former relative to the latter group (2.0% [0.3%] vs 0.3 [0.3%]; P < .001).
Ferritin, an iron-binding protein that is found throughout the body, can store up to 4500 Fe3+ ions with serum concentrations that reflect the body’s iron stores.23 Although exhibiting the potential to protect against oxidative stress by chelating free iron,30,31 increasing ferritin iron content may lead to free iron being more easily released with a potential adverse effect via redox reactions.32,33 Our findings are supportive of earlier work from the United States that suggests that iron metabolism may be important in the pathogenesis of glaucoma.26,27 Serum ferritin level was significantly associated with glaucoma prevalence, and those with the greatest levels of this protein had the highest glaucoma prevalence rates among study participants. In particular, serum ferritin levels in the third and fourth quartiles (≥62 ng/mL) were associated with significantly higher odds of glaucoma diagnosis compared with levels less than 31 ng/mL, even after adjustment for potential confounding variables.
Ferritin has been considered a preferred marker for the assessment of iron-related oxidative stress relative to other iron-related biomarkers.24,25 Our study did not reveal significant association between the diagnosis of glaucoma and any other iron-related biomarkers, such as serum iron level (OR, 1.31; 95% CI, 0.81-2.13) and total iron-binding capacity (OR, 1.29; 95% CI, 0.81-2.07) in a multivariable analysis.
Although iron plays an important role in many cellular functions, excess storage of this oxidant induces DNA damage by generating hydroxyl radicals, which may lead to degenerative organ damage. There is substantial evidence that reactive oxygen species likely play a key role in the pathogenesis of primary open-angle glaucoma. In vitro studies34,35 have led to the hypothesis that iron regulation is important in the pathogenesis of glaucoma via retinal ganglion and TM cells. Lin et al35 proposed potential biological mechanisms for redox active iron that was elevated in chronically stressed TM cells in a hyperoxic model. Changes in expression of iron homeostasis genes also have been observed. Intracellular chelation of iron is protective against apoptosis caused by hydrogen peroxide–induced oxidative stress.
In TM from glaucomatous eyes, there may be additional protection against oxidative stress via the increased expression of endothelial-leukocyte adhesion molecule, a cytokine involved in complex antioxidative pathways.36 Greater IOP and severity of VF defects in patients with glaucoma parallel the degree of oxidative DNA damage that affects the TM.10
In the present study, women 50 years or older had significantly higher serum ferritin levels and greater glaucoma prevalence compared with those younger than 50 years. Prior work24 has similarly found that women older than 50 years have markedly higher ferritin concentrations and 8-hydroxydeoxyguanosine, a reliable biomarker of systemic oxidative DNA damage, than women younger than 50 years, suggesting that ferritin-mediated increased oxidative DNA damage after menopause may be associated with an increased risk of having glaucoma. However, many factors are associated with menopausal status but are difficult to fully assess in this analysis.
In our study, the strong association between a serum ferritin level greater than 61 ng/mL and glaucoma prevalence persisted after adjustment for potential confounding variables. Given that the recommended serum ferritin reference range is 20 to 200 ng/mL and 20 to 300 ng/mL in women and men, respectively, it is concerning that there is profoundly greater glaucoma prevalence observed in those whose serum ferritin levels are within the traditional reference range and far below what would be classified as clinical iron overload based on present criteria.37 Our results support findings of a previous study37 that suggest that the body’s iron levels commonly observed in the general population may be associated with degenerative diseases. Future studies are needed to confirm this association and to determine whether serum ferritin level is important in the pathogenesis of glaucoma. It may be important to discern possible safe ranges for serum ferritin levels in healthy individuals based on risk of potential diseases associated with oxidative damage.
Our study has several limitations, including that even though in the KNHANES V the VF examination was to be administered to all participants who were suspected of having glaucoma, some individuals, for inexplicable reasons, did not receive this test. Among the 9632 participants, 2534 (representing 27.4% of the population) were identified as possibly having glaucoma. There were 1634 individuals suspected of having glaucoma for whom no VF data were available. A further limitation is that the use of the FDT rather than automated perimetric testing to assess VF defects may have resulted in overestimation of the number of individuals with a glaucoma diagnosis. Currently, automated static threshold perimetry is the criterion standard for VF testing.38 Although FDT may not be the ideal choice for VF testing in those suspected of having glaucoma, it offers the benefit of a fast and reliable test that is suited for mass screenings and may predict glaucomatous functional damage earlier than standard perimetry.39
Among the 12 486 participants who were 19 years or older and received the ophthalmology examination portion of the KNHANES V, we found an association between serum ferritin level and glaucoma. In the KNHANES V, there was only 1 FDT result for each participant. Therefore, the 2-2-1 rule for VF interpretation suggested by previous studies26,39 did not apply. In an effort to minimize misdiagnosis, we also excluded participants with retinal disease or history of stroke, thereby reducing errors related to including individuals who may have had VF defects due to nonglaucomatous conditions. Such exclusion, however, may have introduced bias into our findings because 2% of the excluded group was diagnosed as having glaucoma defined by the ISGEO criteria.
Cross-sectional population studies do not allow for the determination of causation, and thus one cannot make any inference from our study regarding whether high serum ferritin levels cause glaucomatous disease. This work simply reports an association between the 2 conditions. Another potential shortcoming of our study is that the biomarker levels were only measured at a single point in time, which may not represent long-term status of these variables. Possible interactions between serum ferritin level and confounders were also not evaluated in the present study.
We found that after adjusting for potential confounding variables, higher serum ferritin level was associated with significantly greater odds of glaucoma diagnosis in a large population-based sample of South Koreans aged 19 years and older. Future prospective studies are needed to confirm this association and more precisely determine the association between ferritin level and the pathogenesis of glaucoma.
Submitted for Publication: February 19, 2014; accepted May 21, 2014.
Corresponding Author: Shan C. Lin, MD, Department of Ophthalmology, University of California, San Francisco, 10 Koret St, Room K301, San Francisco, CA 94143 (LinS@vision.ucsf.edu).
Published Online: August 28, 2014. doi:10.1001/jamaophthalmol.2014.2876.
Author Contributions: Dr S. C. Lin had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: S.-C. Lin, Wang, S. C. Lin.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: S.-C. Lin.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: S.-C. Lin, Singh.
Administrative, technical, or material support: S. C. Lin.
Study supervision: Wang, Singh, S. C. Lin.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by core grant EY002162 from the National Eye Institute (Dr S. C. Lin), by That Man May See Inc (Dr S. C. Lin), and by Research to Prevent Blindness (Dr S. C. Lin).
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.