eTable 1. Distribution of Demographics, Comorbidities, and Health-Related Behaviors by Quintiles of Self-reported Calcium Supplement Consumption
eTable 2. McFadden Pseudo-R2 Values for Individuals Variables and for the Full Model
Kakigi CLM, Singh K, Wang SY, Enanoria WT, Lin SC. Self-reported Calcium Supplementation and Age-Related Macular Degeneration . JAMA Ophthalmol. 2015;133(7):746-754. doi:10.1001/jamaophthalmol.2015.0514
Despite widespread use of calcium supplementation among elderly people, little is known about the association between such consumption and the prevalence of age-related macular degeneration (AMD) in the United States.
To investigate the association between self-reported supplementary calcium consumption and the prevalence of AMD in a representative US sample.
Design, Setting, and Participants
This cross-sectional study included 3191 participants 40 years and older in the 2007-2008 National Health and Nutrition Examination Survey (NHANES) who were evaluated for the presence or absence of AMD by fundus photography. Participants were interviewed regarding use of dietary supplements and antacids during the 30-day period preceding enrollment. Self-reported supplementary intake of calcium was aggregated and divided into quintiles. Fundus photographs were graded for the presence or absence of AMD. Information regarding demographics, comorbidities, and health-related behaviors was obtained via interview. Multivariable logistic regression models were created to determine the odds of an AMD diagnosis among participants in each quintile of self-reported calcium vs participants not self-reporting supplementary calcium consumption after adjusting for confounders.
Self-reported use of calcium supplements.
Main Outcomes and Measures
Presence or absence of AMD by fundus photography.
A total of 248 participants (7.8%) were diagnosed with AMD. Mean ages were 67.2 years for those with AMD and 55.8 for those without AMD. After adjustment for potential confounding variables, study participants who self-reported consumption of more than 800 mg/d of supplementary calcium were found to have higher odds of an AMD diagnosis based on fundus photography evaluation compared with those not self-reporting supplementary calcium consumption (odds ratio, 1.85; 95% CI, 1.25-2.75). The association between self-reported supplementary calcium intake and AMD was stronger in older than younger individuals (odds ratio, 2.63; 95% CI, 1.52-4.54). A clear dose-response association between the quintiles of self-reported supplementary calcium intake and AMD was not established.
Conclusions and Relevance
Self-reported supplementary calcium consumption is associated with increased prevalence of AMD, with the findings suggesting a threshold rather than a dose-response relationship. The stronger association in older individuals may be due to relatively longer duration of calcium supplementation in older individuals.
Age-related macular degeneration (AMD) is the third leading cause of blindness and visual impairment worldwide and the leading cause of blindness in developed countries.1 In 1990 and 2010, the worldwide prevalence of AMD was estimated to be 5% and 7%, respectively.2 The aging population is projected to result in a prevalence of 196 million AMD cases by 2020 and 288 million AMD cases by 2040.1 In 2011, the prevalence of AMD within the United States alone was estimated to total 7.2 million, including 5.2 million individuals 60 years and older.3
Age-related macular degeneration is a multifactorial disease. Proposed risk factors include cigarette smoking; obesity; low dietary intake of vitamins A, C, and E, zinc, lutein, and ω3 fatty acids; health-related behaviors related to cardiovascular risk factors; and certain genetic markers.4 Several mechanisms have been proposed for the pathogenesis of AMD, including senescence, choroidal ischemia, oxidative damage, and variations in vascular endothelial growth factor expression leading to choroidal neovascularization.4 Current treatment options, including vascular endothelial growth factor inhibitors, laser photocoagulation, and photodynamic therapy, although helpful in limiting vision loss, remain less than ideal for long-term therapy. Prevention of AMD has proven more challenging. The Age-Related Eye Disease Study found that consumption of supplemental antioxidants and minerals delays the progression of AMD.5 Other preventive measures include smoking cessation, physical activity, and avoidance of excessive sunlight exposure.
Calcium is widely used by older men and women in the United States, with an estimated 43% of the US population and approximately 70% of older women reporting supplemental calcium consumption.6,7 Known adverse effects of calcium supplementation include increased risk of kidney stones,8 gastrointestinal symptoms,9 and cardiovascular events.10 Disruption of calcium homeostasis is associated with permanent neuronal damage, which may play an important role in the pathogenesis of neurologic conditions, such as Alzheimer, Huntington, and Parkinson diseases.11 Calcium dysregulation has also been implicated in the pathogenesis of neurodegenerative eye diseases, such as glaucoma.12 In prior population-based studies, we found that there is a threshold dose above which self-reported calcium supplementation is associated with increased odds of self-reported glaucoma after controlling for potential confounders, further supporting the idea that calcium may damage nerve cells and contribute to the development of neurodegenerative eye disease.13,14 Calcifications can be seen in the drusen that characterize AMD.15- 17 A detailed analysis of the subretinal pigment epithelial deposit compositions of a donor eye with known AMD revealed an unexpectedly high concentration of calcium.18
Although previous studies15-18 have suggested that calcium plays an important role in the development of AMD, to the best of our knowledge, no prior studies have examined directly the association between self-reported supplemental calcium intake and AMD. The National Health and Nutrition Examination Survey (NHANES) is an annual national population study run by the Centers for Disease Control and Prevention (CDC) designed to assess the health status of the US population. It includes an extensive interview questionnaire, which collects data on numerous health conditions, medication use, and nutritional intake, as well as a physical examination component. In addition, the 2007-2008 data release of NHANES included a calculated intake of a variety of nutrients from a detailed interview of participants’ use of dietary supplements, as well as graded fundus photographs of participants 40 years and older. This extensive, nationally representative sample allows study of an association between calcium intake through self-reported dietary supplementation and a diagnosis of AMD based on fundus photography.
We used publicly available data from the 2007-2008 administrations of NHANES, a cross-sectional series of interviews and examinations of the civilian, noninstitutionalized population of the United States, to assess the association between AMD assessed by fundus photography and self-reported calcium supplement intake.19 This research used publicly available deidentified data and as a result is not considered human subjects research. Therefore, it was exempted from review by the Committee on Human Research at the University of California, San Francisco. NHANES is administered by the CDC for the purpose of providing US health statistics, and it includes interviews and examinations of approximately 5000 persons per year. NHANES uses a stratified multistage sampling design that requires a weighting scheme to most accurately estimate disease prevalence in the US population. These weighted data were used throughout our data analysis, which included 3191 participants in the 2007-2008 NHANES who were 40 years and older and underwent the interview and examination portions of the study.
The primary predictor was self-reported calcium intake from dietary supplements and antacids. NHANES included an interview about prescription and nonprescription dietary supplements and antacids during the 30-day period before the interview. NHANES aggregated the self-reported calcium intake from all reported antacids and dietary supplements, both prescription and nonprescription, and calculated a mean daily intake for each nutrient. For this study, participants were classified into 1 of 6 categories, including the 5 quintiles of self-reported calcium intake or no intake. We safely assumed that those who had a missing value for calcium intake had zero calcium intake because practically no values were missing for questions detailing the self-reported number of supplements or antacids each individual took.
The main outcome was the presence of AMD diagnosed by NHANES on fundus photography. NHANES investigators performed fundus photography on participants 40 years and older.19 These images were graded at the University of Wisconsin, Madison, by 9 experienced graders: a preliminary grading coordinator, 2 preliminary graders, and 6 detail graders. The graders viewed each fundus image with a high-resolution monitor, using the EyeQLite image processing software and database, and referenced the written protocol and the digital photographic standards to evaluate retinal abnormalities. All images were graded by at least 2 individuals: a preliminary grader and a detail grader. On systematic grading, if the first 2 graders did not agree on a diagnosis, a third individual graded the eye images. If 2 of these 3 graders disagreed, the image was evaluated by an adjudicator to make a final decision. Inspection, calibration, and maintenance of the equipment and supplies were performed on a regular basis.
Potential confounders included age, sex, ethnicity, annual household income, and educational level; health-related behaviors, such as smoking (current, past, or never), alcohol use (number of alcoholic drinks per day in the last year), and exercise (total number of metabolic equivalent tasks [METs] in minutes per week); comorbid medical conditions, such as self-reported history of osteoporosis, hyperlipidemia, hypertension, coronary heart disease, and obesity determined by body mass index (calculated as weight in kilograms divided by height in meters squared); comorbid eye conditions, such as self-reported history of cataract extraction, diabetic retinopathy, and glaucoma; and self-reported general health condition (self-rated as excellent or very good, fair, poor, or very poor).
We compared the distribution of possible confounding variables between participants with and without AMD using design-adjusted Rao-Scott Pearson χ2 and Wald tests for categorical and continuous variables, respectively. We also compared the distribution of possible confounding variables among participants in each self-reported quintile of supplemental calcium intake using design-adjusted Rao-Scott Pearson χ2 and Wald tests for categorical and continuous variables, respectively. A multivariable logistic regression model was created to determine the odds of an AMD diagnosis among participants consuming each quintile of self-reported calcium vs participants not consuming calcium after adjusting for confounders. Potential confounding comorbidities with P ≥ .10 in univariable models were excluded from the final model, including sex, highest level of education, history of diabetic retinopathy, exercise (total number of METs in minutes per week), and hyperlipidemia. However, given that hyperlipidemia20 and female sex21 are associated with AMD in previous studies, we chose to include these covariates in our final model. To calculate CIs around estimates for the US national population, all data analyses were performed using STATA statistical software, version 13.0 (Stata Corp), and NHANES weighted data. The SEs of population estimates were calculated using Taylor linearization methods. To provide an indication of the relative explanatory power of the variables in this model, we also calculated the McFadden pseudo-R2 values for the individual variables included in the model and for the final model that includes all variables.
For a better understanding of the influence of age in the association between self-reported calcium supplementation and the presence of AMD, we also examined subgroups of the population sample, comparing younger (40-67 years) and older individuals (≥68 years), taking into account that the mean age of AMD participants was 67 years. We created multivariable logistic regression models to determine the odds of AMD between participants in each quintile of self-reported calcium vs participants who did not self-report calcium consumption for 2 subgroups of the population sample determined by age. We hypothesized that the odds of having AMD would be more pronounced in the older group because of a number of reasons, including possible longer durations of exposure to calcium supplements or a decreased ability of cells to maintain calcium homeostasis with increasing age.22
In the 2007-2008 NHANES data set, 3191 participants who were 40 years and older participated in the interview and examination portions of the study and had graded fundus photographs to determine whether they had AMD. Of the participants who received a graded fundus photographic examination, 248 (7.8%) were diagnosed as having AMD. Of the participants who received an AMD diagnosis, 220 (6.9%) were diagnosed as having early AMD, and 28 (0.9%) were diagnosed as having late AMD.
Table 1 and Table 2 provide information on the demographic characteristics, health-related behaviors, comorbidities, general health condition, comorbid eye diseases, body mass index, and self-reported calcium intake among those participants with and without a diagnosis of AMD by fundus photography. The mean age of participants with and without AMD was 67.2 and 55.8 years, respectively. eTable 1 in the Supplement provides the distribution of demographics, comorbidities, and health-related behaviors by quintiles of self-reported calcium supplement consumption.
The odds of being diagnosed as having AMD was higher among participants in the top quintile of self-reported calcium intake compared with those who did not consume calcium after adjusting for the following confounders: age, sex, ethnicity, annual household income level, smoking status, alcohol intake, obesity, history of osteoporosis, cataract operation, glaucoma, hypertension, stroke, coronary heart disease, and hyperlipidemia (odds ratio, 1.85; 95% CI, 1.25-2.75; Table 3). To provide an indication of the relative explanatory power of the variables in this model, eTable 2 in the Supplement reports the McFadden pseudo-R2 values for the individual variables and for the final model that includes all variables.
Given that the mean age of participants with an AMD diagnosis was 67 years, we separately examined younger (40–67 years) and older individuals (≥68 years) with regard to AMD prevalence based on self-reported calcium consumption. We found that among individuals 68 years or older, the odds of an AMD diagnosis were higher among participants in the top quintile of calcium intake compared with those who did not consume calcium after adjusting for confounders (odds ratio, 2.63; 95% CI, 1.52-4.54) (Table 4). No similar association was found between an AMD diagnosis and self-reported calcium consumption among those aged 40 through 67 years (Table 5).
Our study of a US national population-based sample of adults 40 years and older found that participants who self-reported consuming more than 800 mg/d of supplementary calcium had higher odds of having a diagnosis of AMD by fundus photography, after adjustment for a variety of confounders. Subgroup analysis revealed that this association was specific to individuals who were older than the mean age of participants with an AMD diagnosis.
To our knowledge, this is the first study looking at the association between self-reported supplemental calcium intake and AMD. We did not identify a dose-response association between the quintiles of self-reported calcium intake and AMD. The strong association between the highest quintile of intake and AMD prevalence in elderly participants suggests a threshold above which susceptible individuals are most vulnerable to the effects of calcium consumption. It is noteworthy that the 800-mg cutoff point for the highest quintile of self-reported calcium intake is below the recommended total daily intake of calcium for men and women in the United States.23 There are several possible explanations for why participants 68 years or older had the most pronounced association between self-reported calcium intake and AMD. Age is an important risk factor for AMD, and older individuals may be more susceptible to potentially harmful ingestants. Aging is also related to the dysregulation of calcium homeostasis,11 so older individuals may be more sensitive to fluctuations in calcium levels. In addition, older individuals are likely to have taken calcium supplements for a longer duration with potentially cumulative adverse effects.
Calcium plays an important role in neuronal cell regulation and function. Not surprisingly, changes in normal calcium homeostasis can lead to permanent neurologic damage, and excessive levels of calcium trigger caspase-dependent cell death.12 It is believed that the age-related dysfunction of cell membranes and intracellular organelle membranes, which are normally involved in maintaining calcium homeostasis, contributes to this neuronal dysfunction and increased susceptibility to further cell damage.11 Many neurodegenerative diseases are thought to be at least partially mediated by neuronal calcium dyshomeostasis, including glaucoma, Alzheimer disease, Parkinson disease, Huntington disease, autosomal dominant spinocerebellar ataxias, amyotrophic lateral sclerosis, epilepsy, schizophrenia, traumatic brain injury, brain stroke, and human immunodeficiency virus dementia.11 In patients with Alzheimer disease, age-related calcium dyshomeostasis is compounded by environmental and genetic stressors, which causes increased production of toxic β-amyloid, leading to neuronal cell death. Given previously described pathologic similarities between AMD and Alzheimer disease, including amyloidogenesis, plaque formation, inflammation, oxidative stress, and changes in glial cell function,24- 27 it is reasonable to hypothesize that similar calcium-dependent pathologic mechanisms occur in these 2 neurodegenerative diseases and are particularly harmful with increasing age.
We acknowledge that our study has several limitations. Assessment of calcium intake is based on an interview requiring participants to recall the supplements they ingested within the past 30 days and does not distinguish between those who took supplemental calcium for 30 days and those who may have been taking calcium supplements for many years. In addition, self-reported calcium supplementation may not be accurately reflected by serum calcium levels. Previous studies28- 31 have found that oral calcium administration is associated with increases in plasma ionized calcium and urinary calcium and decreases in plasma parathyroid hormone. However, there is a linear decrease in intestinal calcium absorption efficiency with increasing age, particularly after 40 years of age.32 In addition, calcium absorption is regulated to meet the body’s needs, with greater efficiency of absorption when dietary calcium levels are low.32 Prior work33 has found discrepancies between self-reported calcium supplementation and calcium levels and reported that the accuracy of self-reported calcium supplementation is partially dependent on the type of questionnaire that is administered. Self-reported calcium supplementation may be inaccurate because of the ease with which consumption can be started and stopped. In one study,34 34% of those who reported taking daily supplements in NHANES I were no longer daily users at the time of NHANES I follow-up, and 25% of those who reported not taking daily supplements in this study were taking daily supplements at follow-up. To summarize, limitations of our study include inaccuracy in reporting actual calcium supplementation and the lack of a strong correlation between such supplementation and serum calcium levels. Another limitation of the study is that the self-reported supplemental calcium intake does not reflect a participant’s total calcium intake because dietary calcium is excluded. Finally, we found a potential association between self-reported supplemental calcium intake and AMD prevalence, but, as with any cross-sectional study, we cannot make any claims regarding causation, which would require longitudinal analysis to determine incident cases of AMD among those consuming varying amounts of calcium. It is possible that the association we have found between self-reported calcium supplement consumption and AMD could be explained by some other underlying condition that is associated with both self-reported calcium consumption and AMD. It is also possible that the association is a chance finding. Until further prospective studies are conducted to assess this association, we would not recommend that patients discontinue taking their calcium supplements in an attempt to reduce their risk of developing AMD, particularly given the known benefits of calcium supplementation for other medical conditions.
We found that there is an increased odds of an AMD diagnosis among participants who self-report consuming more than 800 mg/d of calcium compared with participants who self-report not consuming supplemental calcium after controlling for a number of confounders. This result was most pronounced for individuals older than the 67-year mean age of the AMD population, suggesting that there may be increased risk of developing AMD in individuals who ingest calcium for a longer duration or, alternatively, a greater propensity for calcium to cause harm in terms of AMD risk in the elderly population. There was no dose-dependent response to increasing quintiles of self-reported calcium intake, which leads us to hypothesize that there is possibly a threshold above which the consumption of calcium may be a risk factor for the development for AMD and that this threshold may be below the recommended total daily intake of calcium in the United States.
Submitted for Publication: November 21, 2014; final revision received January 21, 2015; accepted February 1, 2015.
Corresponding Author: Shan C. Lin, MD, Department of Ophthalmology, University of California, San Francisco, 10 Koret Way, Room K301, San Francisco, CA 94143-0730 (LinS@vision.ucsf.edu).
Published Online: April 9, 2015. doi:10.1001/jamaophthalmol.2015.0514.
Author Contributions: Dr 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: Kakigi, Singh, Lin.
Acquisition, analysis, or interpretation of data: Kakigi, Wang, Enanoria, Lin.
Drafting of the manuscript: Kakigi.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Kakigi, Singh, Wang, Enanoria.
Obtained funding: Kakigi, Lin.
Administrative, technical, or material support: Lin.
Study supervision: Singh, Lin.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was funded in part by grant TL1 TR000144 from the Clinical and Translational Research Fellowship Program, a program of the University of California, San Francisco, Clinical and Translational Science Institute that is sponsored in part by the National Center for Advancing Translational Sciences, National Institutes of Health, and the Doris Duke Charitable Foundation. The design and conduct of the study, analysis, interpretation of the data, and preparation of the manuscript were funded by the Clinical and Translational Research Fellowship Program. Analysis was funded by core grant EY002162 from the National Eye Institute, That Man May See Inc, and Research to Prevent Blindness.
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 the decision to submit the manuscript for publication.
Disclaimer: The contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health, University of California, San Francisco, or the Doris Duke Charitable Foundation.