Klein BEK, Knudtson MD, Brazy P, Lee KE, Klein R. Cystatin C, Other Markers of Kidney Disease, and Incidence of Age-Related Cataract. Arch Ophthalmol. 2008;126(12):1724-1730. doi:10.1001/archophthalmol.2008.502
Copyright 2008 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2008
To investigate the 15-year incidence of 3 specific types of age-related cataract as related to cystatin C and other measures of kidney function.
Examinations of a population-based cohort (n = 4926) occurred at 5-year intervals for 15 years. Assessment of medical history, examination, and photographs of the lens after pupil dilation were performed at each examination. Protocols for photography and grading were used. Laboratory measures were from specimens collected at baseline.
In multivariable analyses, a 1-SD increase in the logarithm of cystatin C was associated with 15-year incidence of cortical (odds ratio [OR], 1.24; 95% confidence interval [CI], 1.09-1.41) and posterior subcapsular (OR, 1.24; 95% CI, 1.02-1.50) cataracts. One SD increase in the logarithm of blood urea nitrogen and creatinine were associated with 15-year incidence of posterior subcapsular cataract (OR, 1.22; 95% CI, 1.04-1.42 and OR, 1.26; 95% CI, 1.03-1.54, respectively).
Increased levels of cystatin C are associated with increased risk of specific types of age-related cataract. Whether the associations are due to the metabolic changes associated with decreased renal function, common genes, or both awaits further research.
Cataracts, the most common cause of visual impairment and blindness worldwide, are concomitants of aging.1- 3Age-related cataracts (ARC) are associated with many risk factors including a variety of environmental and personal exposures such as lifestyle habits, diseases, and metabolic characteristics.4- 18Those with severe kidney disease appear to be at increased risk of ARC; mild kidney dysfunction may also enhance the risk of cataract.19- 23Serum creatinine and blood urea nitrogen (BUN) are used as markers of renal function or glomerular filtration rate (GFR) in standard clinical practice. These markers are imperfect because their levels in the blood are affected by factors other than GFR such as dietary protein, state of hydration, and renal tubular reabsorption or secretion. Researchers have identified another endogenous substance that appears to be a better noninvasive marker of GFR, cystatin C. This substance is a low–molecular weight protein that is a member of the cystatin superfamily of cysteine protease inhibitors. It is filtered by the kidney and then metabolized by the tubules so it cannot be collected in the urine and does not reappear in the blood. It is produced by all nucleated cells in the body. Its production rate is not affected by the subject's diet, but its levels are affected by either hyperthyroidism or hypothyroidism and they fluctuate with other markers of inflammation such as C-reactive protein.24,25Serum cystatin C concentration appears to correlate more closely with GFR than serum creatinine and is more sensitive in identifying subjects with mild renal insufficiency.26
Cystatin C has also been linked to other systemic diseases27,28but, to our knowledge, has not been examined regarding the development of ARC. Our purpose is to determine whether measures of cystatin C and other measures of kidney function are associated with incidence of ARC during a 15-year interval in the population of adults who participated in the Beaver Dam Eye Study.
The population and recruitment methods for the full cohort have been described in previous articles.29- 37In brief, a private census of the population of Beaver Dam, Wisconsin, was performed from September 15, 1987, to May 4, 1988, to identify all residents in the city or township of Beaver Dam aged 43 to 84 years. A total of 5924 eligible individuals were identified and invited for a baseline examination between March 1, 1988, and September 14, 1990.29The Tenets of the Declaration of Helsinki were followed, institutional review board approval was granted, and informed consent was obtained from each subject. Examinations were completed for 4926 persons. The most common reason for nonparticipation after the baseline examination was death. Ninety-nine percent of the population was white, as classified by the examiner. Comparisons between participants and nonparticipants at the baseline have been presented elsewhere.29In brief, nonparticipants (dead or alive) were older, had fewer years of education, were less likely to be currently employed, had poorer visual acuity, were more likely to have cardiovascular disease, had diabetes, smoked more, and had higher systolic blood pressures.
Visits occurred at 5-year intervals from the baseline examination for 3 follow-up evaluations. Of the 5924 enumerated persons aged 43 to 84 years, 4926 participated in baseline examinations from 1988 to 1990. Of these, 3684 (81.1%) participated in 5-year follow-up examinations from 1993 to 1995. Of the 3334 surviving participants in the baseline and second examination, 2764 (82.9%) participated in the 10-year follow-up. Of the 2480 surviving participants who were examined at the baseline, 5-, and 10-year follow-ups, 2119 (85.4%) participated in the 15-year follow-up. Comparison of participants and nonparticipants at each examination phase have been detailed elsewhere.29- 31
For the current incidence analyses, participants had to have attended the baseline examination, not had the cataract type of interest at baseline, provided a blood specimen, participated at the first 5-year follow-up, and may have participated in 1 or more subsequent examinations. Gradable lens photographs had to be available for each relevant visit. A total of 3097 persons contributed to at least 1 analysis.
The same protocols, with few additions or deletions, were used at each examination phase. A brief medical history, including use of medication, was obtained. Photographs of the lens were taken after pharmacologic dilation. Slitlamp photographs were taken to grade the degree of nuclear sclerosis. Retroillumination photographs were taken to grade the presence and severity of cortical and posterior subcapsular cataracts. The protocols for the photography and grading procedures were based on detailed codified rules.38Graders were masked as to subject identity. Scores for nuclear sclerosis were based on comparison with standard photographs, which resulted in a 5-step severity scale based on the opacity of the nucleus. Severities greater than standard (3) were considered nuclear cataracts. Scores for cortical and posterior subcapsular cataracts were based on the estimated amount of involvement based on the total involvement of segments of a grading grid placed under the film image. Cortical opacities involving more than a weighted average of 5% of the total lens were considered cortical cataracts. Posterior subcapsular opacities involving more than 5% of any of the 9 individual grid segments were considered posterior subcapsular cataracts.38Analyses were based on cataract incidence of the first eye to develop a lesion. Persons without gradable photographs in either eye were excluded.
Casual blood specimens were obtained at the time of the baseline examination. An aliquot of serum was used immediately for determination of total cholesterol,39blood glucose,40and BUN. Whole-blood glycosylated hemoglobin was determined using affinity chromatography (Isolab Inc, Akron, Ohio) from casual blood samples. Cystatin C, creatinine, and high-sensitivity C-reactive protein (hsCRP) were measured in 2007 from serum specimens that had been collected at baseline and frozen at −80°C since that time. Laboratory methodologies relating to measurement of these markers are provided below.
Serum BUN levels were measured using the colorimetric method (the Berthelot reaction) on a Technicon RA-1000 AutoAnalyzer (Technicon Instruments, Tarrytown, New York).
Creatinine levels were measured in serum by rate reflectance spectrophotometry using thin-film adaptation of the creatine amidinohydrolase method on the Vitros analyzer (Johnson & Johnson Clinical Diagnostics Inc, Rochester, New York) at the Collaborative Studies Clinical Laboratory at Fairview-University Medical Center (Minneapolis, Minnesota). The reference range in adult women was 0.4 to 1.1 mg/dL (to convert to micromoles per liter, multiply by 76.25) and in adult men was 0.5 to 1.2 mg/dL. The laboratory coefficient of variability (CV) was 2.2%.
Cystatin C levels were determined using the Dade Behring BN100 nephelometer (Deerfield, Illinois) as follows: a solution of polystyrene particles coated with antibodies specific to human cystatin C was incubated with diluted specimen. A reaction occurred between the bound antibody and the cystatin C in the specimen, resulting in particle aggregation and an increase in light absorbance. The cystatin C concentration of the test specimen was determined by comparing its absorbance change to that of a calibration curve. The interassay precision was determined at 2 control levels: 1.72 mg/L (CV, 6.4%) and 0.78 mg/L (CV, 5.2%).
The level of hsCRP was measured in serum using a latex-particle enhanced immunoturbidimetric assay kit (Kamiya Biomedical Company, Seattle, Washington) and was read on the Roche/Hitachi 911 (Roche Diagnostics, Indianapolis, Indiana). The reference range was 0 to 0.5 mg/L (to convert to nanomoles per liter, multiply by 9.524). The interassay CV range in our laboratory was 4.5%. Proteinuria was measured with a dipstick on a casual urine specimen obtained at the baseline examination.
In analyses using a categorical cutpoint, the following were considered abnormally high: serum cystatin C higher than 0.95 mg/L; serum creatinine higher than 1.1 mg/dL for women and 1.2 mg/dL for men; serum BUN of 20 mg/dL or higher (to convert to millimoles per liter, multiply by 0.357); protein levels of 30 mg/dL or higher in urine were considered proteinuria. Smoking and diabetes histories were obtained as part of a medical questionnaire. Persons who did not have diabetes but whose glycosylated hemoglobin or casual blood glucose met age- and sex-specific criteria were included as cases of diabetes.41,42
We examined the relationships between serum cystatin C, creatinine, BUN, and the presence of gross proteinuria to the incidence of 3 types of ARC over 15 years. The SAS version 9 (SAS Institute, Cary, North Carolina) was used for analyzing the data. Multivariate odds ratios (ORs) and 95 percent confidence intervals were calculated from discrete linear logistic hazard models.43These analytical approaches allowed persons who were right-censored (censored at the 10- or 15-year examination due to death, nonparticipation, or cataract surgery) to contribute information to the estimates.
The distributions of serum cystatin C, creatinine, and BUN levels were highly skewed, so we include them in our models after log-transforming them. We present ORs per increase in standard deviation in the risk factor for continuous measures. We also tested for a threshold effect of cystatin C, creatinine, and BUN using predefined values described above. Analyses first controlled for age in 4 categories of 10-year bands and sex. We considered smoking history, glycosylated hemoglobin, serum hsCRP, serum total cholesterol, heavy drinking history, and history of use of oral steroids as additional confounders. Models were also further stratified by baseline diabetes status, but because no interactions were detected, we include only the results for the unstratified analyses in the Tables.
Characteristics of participants are seen in Table 1. Baseline serum cystatin C was significantly higher in those not participating in follow-up examinations. There were no other significant differences in the kidney-related variables between participants and nonparticipants among the characteristics in Table 1.
We examined whether each measure of kidney function was associated with the individual cataract types, controlling first for only age and sex (Table 2). Increased levels of serum cystatin C and decreased levels of BUN were associated with incident nuclear cataract. Serum cystatin C was associated with incidence of cortical cataract. Increased levels of serum cystatin C and BUN were associated with incident posterior subcapsular cataract. After adjusting for smoking, glycosylated hemoglobin, hsCRP, serum total cholesterol, history of use of systemic steroids, and heavy drinking, higher levels of serum creatinine were also significantly related to incident posterior subcapsular cataract. Because the continuous measures of kidney function were relatively highly correlated (Table 3), we did not include them in the same models. Limiting the analyses to include only those without diabetes did not alter the specific variables associated with cataracts; the ORs were similar for those with and without diabetes (data not shown). Alternative analyses in which cut points for high (abnormal) values for the kidney variables were used and yielded similar associations for cystatin C for both cortical and posterior subcapsular cataract and for BUN and posterior subcapsular cataract compared with the continuous models (Table 4). Proteinuria was not significantly associated with any cataract type.
We find that higher levels of serum cystatin C, adjusted for age and sex, is associated with the odds of incident posterior subcapsular and cortical cataract over a 15-year interval and that the ORs remained significant after further controlling for smoking, serum cholesterol, glycosylated hemoglobin, hsCRP, heavy drinking, and use of systemic steroids. This is true even when analyses are restricted to persons without diabetes. Higher levels of serum BUN and serum creatinine are also associated with posterior subcapsular cataract in continuous models. These findings are unique in 2 ways: (1) the relationships are specific for cataract type and (2) a relationship with serum cystatin C is found. Clayton and colleagues21studied 931 patients with cataracts and 325 patients without. Patients with cataracts had higher mean blood urea levels than the comparison group. The relative rarity of posterior subcapsular cataract may explain, in part, why that specific cataract type was not reported. Cortical cataract, while more common, is less likely to result in decreased vision and may therefore be overlooked. Many studies do not have specific grading for the presence of each particular type of ARC, so use of the more generic term cataractmay have obscured the specificity of cataract type in previous reports.
We performed multivariable analyses using cataract surgery as an endpoint. We found similar relationships of the kidney variables to this outcome as we found for posterior subcapsular cataract (data not shown). This is consistent with the finding that the relative risk for cataract surgery is greatest for persons with posterior subcapsular cataract.23
An association between kidney disease and cataract or cataract surgery in persons with diabetes has been reported.23,44In this long-term incidence study, we found that the relationship of kidney function to cataract was not markedly different between those with and without diabetes, suggesting that it is kidney function itself rather than diabetic kidney disease specifically that puts the lens at risk. We have evaluated the association of markers of kidney function and the specific cataract types in data from past examinations. While we found a relationship between posterior subcapsular cataract and BUN in the past, the relationship was not statistically significant and therefore not reported. During the longitudinal follow up, the number of cases of cataract, particularly posterior subcapsular cataract, has increased. This likely accounts for the significance we now find of BUN and creatinine.
There are plausible explanations underlying the biological mechanisms that may account for our findings. It is possible that systemic metabolic acidosis itself is a risk factor for cataract. Many persons with kidney dysfunction are chemically acidotic and it may be that even low levels of acidosis affect the lens. In addition, a specific kidney functional abnormality, renal tubular acidosis, is associated with lens epithelial abnormalities that have been postulated to lead to loss of lens transparency.45We did not have specific measures of systemic acidosis or information to estimate anion gap. We had a single blood pH measure at baseline, and this was not associated with any type of cataract (data not shown).
Cystatin C levels are elevated with mild degrees of renal insufficiency, whereas serum creatinine levels are not.26The metabolic abnormalities present in the early stages of kidney dysfunction are not the result of retained waste products or the failure to maintain the homeostasis of body fluids. These abnormalities are the result of the adaptations made to maintain the homeostasis. The circulating levels of parathyroid hormone are elevated in response to a normal intake of calcium and phosphorus. Sodium regulatory hormones like atrial natriuretic peptide are elevated in persons with unrestricted sodium intake. The capacity to buffer and excrete acid loads is reduced, and excess acid is buffered in tissues like bone. The clinical expression of these adaptive and abnormal processes is often delayed until a later stage of kidney dysfunction. However, in this study, we may be identifying complications (ie, cortical and posterior subcapsular cataracts) that occur in response to some metabolic factor during an early stage of kidney disease.
Oxidative stress may, in part, explain the association of the markers of kidney disease with cataracts. Patients with kidney failure are in a state of oxidative stress.46- 49Oxidative stress may lead to carbamylation of lens proteins, a process that has been suggested to cause cataract.50Thus, markers of kidney function are likely to be indicators of oxidative stress, and there may be a relationship of the severity of kidney disease and oxidative stress across the range of function reflected in the values of the quantitative markers.
Cystatins, naturally occurring inhibitors of cysteine proteinases,51are found in various cell types and body fluids52,53and are considered a component of serum proteins. Serum cystatin C is used currently in other studies as a marker of kidney dysfunction and has been associated with microvascular and macrovascular disease,27,28,54but its usefulness as a predictor of important cardiovascular events is not certain.55While some suggest that cataract and cardiovascular disease share risk factors,56- 58the findings are not universal,59so there may be limited usefulness in comparing new biomarkers for these events in different systems and hoping for parallel findings.
We have been considering cystatin C levels as a marker of kidney function. However, cystatin C, being nearly ubiquitous, is found in tissue and fluid in the eye (aqueous humor, cornea, retina, sclera, and lens epithelium).51Cystatin C has been found to be inhibited by reactive aldehydes that may lead to the abnormal accumulation of proteins, and it has been postulated that this process may lead to cataracts.60
Limitations of our study include the lack of other, perhaps more sensitive, measures of renal function such as inulin clearance or iothalamate excretion. Such testing was beyond the resources of this study. Medical care, specifically cataract surgery for nuclear cataract, selectively eliminated persons (eyes) for incidence follow-up. It is also possible that systemic conditions that we were unaware of or could not detect may have confounded the relationships we studied. We do not know the extent to which medications and other environmental exposures influenced both markers of kidney function and cataract, although we have included the factors that we think would most likely have influenced our findings.
In summary, we have found that markers of kidney function are related to incidence of posterior subcapsular cataract, a lesion that is associated with significant impairment of visual acuity, and to incident cortical cataract. We also found that the severity of kidney dysfunction, as reflected in higher serum levels of cystatin C, BUN, and creatinine, is associated with progressively increased risk. The mechanisms of oxidative stress, other metabolic changes related to kidney function, common genetic etiologies for kidney dysfunction and lens pathology, or all of these are possible and require further study.
Correspondence:Barbara E. K. Klein, MD, MPH, Department of Ophthalmology and Visual Sciences, University of Wisconsin Madison, 610 N Walnut St, Fourth Floor Wisconsin Alumni Research Foundation (WARF), Madison, WI 53726 (firstname.lastname@example.org).
Submitted for Publication:March 31, 2008; final revision received June 3, 2008; accepted June 25, 2008.
Author Contributions:Dr B. E. K. Klein 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.
Financial Disclosure:None reported.
Funding/Support:This study was supported by grant EY06594 from the National Eye Institute; Senior Scientific Awards (Drs B. E. K. Klein and R. Klein) from Research to Prevent Blindness provided further additional support for data analyses.