Prevalence of Cataract and Pseudophakia/Aphakia Among Adults in theUnited States FREE

Cataract is the leading cause of blindness in the world today.¹ It is also the leading cause of vision loss in theUnited States,² responsible for some 60% ofall Medicare costs related to vision.³ Theeffect of age-related cataract can be expected to grow as the US populationcontinues to age. Despite this, few, if any, precise estimates have been madeof the prevalence of cataract in the United States on a national basis.

Measuring cataract prevalence for a truly representative national samplewould likely be very costly and difficult. However, many scientifically designed,population-based studies have recently provided age-specific estimates ofcataract prevalence among population groups relevant to the United States.The current article has attempted to standardize definitions and reportingformat between studies to allow the pooling of prevalence figures for cataractand prior cataract surgery. Age-, race/ethnicity-, and gender-specific prevalencerates derived in this fashion have been applied to US Census data from 2000,⁴ to estimate the prevalence of lens opacity and pseudophakia/aphakiain the US population as a whole. Estimates for 2020, based on US Census projectionsof the population,⁴ are also presented. Thesefigures represent the first estimates of cataract prevalence in the UnitedStates to consider the large number of population-based surveys of eye diseasecarried out over the last decade or more.

In an initiative sponsored jointly by Prevent Blindness America, Schaumburg,Ill, and the National Eye Institute, Bethesda, Md, a meeting of principalinvestigators of large studies of eye disease among populations of white,black, and Hispanic persons was convened in Fort Lauderdale, Fla, in May 2001.It was determined by consensus that the morbidity associated with age-relatedcataract in the United States was best measured by 4 prevalence figures: cataract,prior cataract surgery (eg, pseudophakia/aphakia), and blindness, and lowvision associated with cataract. The current article presents the estimatedprevalence of cataract and of pseudophakia/aphakia in the US population 40years and older in 2000, and the projected prevalence in 2020. Estimates ofthe prevalence of cataract-associated low vision and blindness in the UnitedStates are reported in a companion article in this issue.⁵

An attempt was made to include all scientifically valid, population-basedstudies of cataract relating to white, black, or Hispanic persons publishedin English after 1990 (Table 1).^{6 - 10} Few,if any, population-based studies published before this date measured lensopacity according to predetermined photographic standards. Many earlier studiesalso defined cataract with reference to the visual acuity of the subject,which is difficult to interpret because of the inability to adjust for competingcauses of vision loss. The cutoff date was further chosen to minimize potentialinaccuracies due to changing rates of cataract extraction and other cohorteffects. While studies from Europe and Australia were included in estimatesfor white persons, potentially relevant studies from Africa¹¹ wereexcluded from estimates for black persons owing to concerns over the potentialeffect of rates of cataract surgery significantly different from the UnitedStates.

Investigators from studies listed in Table 1 provided data tables listing the number of persons havingcataract and pseudophakia/aphakia in either eye by 5-year age interval, gender,and (where relevant) race/ethnicity. The number of persons at risk in eachstratum was also provided. Cataract was defined as the presence of 1 or moreof the following in either eye:

Estimates for prevalence of cataract and pseudophakia/aphakia amongblack persons in the current article were based on studies conducted in Salisbury,Md, ⁹ and Barbados, West Indies.⁸ TheSalisbury Eye Evaluation (SEE) Project only examined subjects 65 years andolder. Because of the differences between the Barbados and Salisbury studiesin the cutoff used to define cataract, it was impossible to pool the prevalenceestimates from these studies. Age- and gender-specific prevalence data fromthe SEE Project were used to estimate prevalence for black persons 65 yearsand older. To estimate age- and gender-specific prevalence for individualsaged 40 through 64 years, we applied gender-specific "correction factors"to the Barbados Eye Study 5-year prevalence rates in this age range. The correctionfactors were derived by dividing the reported prevalence for Barbados EyeStudy subjects aged 65 years and older for each gender- and age-specific stratumby the comparable figure for the SEE Project. These fractions for all age-stratawere averaged separately for the 2 genders. The correction factors (0.32 formales and 0.42 for females) were then applied to the Barbados Eye Study datafor each age interval in the range 40 through 64 years to produce age- andgender-specific prevalence estimates of cataract and pseudophakia/aphakiaamong black persons in this age range. In essence, this method imputes ratesfor the SEE Project in the younger-aged groups by adjusting the Barbados EyeStudy rates based on the differences between the 2 studies found in the older-agedgroups.

To derive age- and gender-specific estimates of the prevalence of cataractamong Hispanics and other races/ethnicities (East Asian, Native American,and others), a nonweighted average of the values for white and black personsin each age and gender cell was used. Such values are not useful for estimatingthe prevalence of cataract in these groups but were judged to be the bestavailable approximation for use in generating overall US population estimatesby age and gender. Unpublished data for the prevalence of pseudophakia/aphakiawere available for Hispanic persons (S. West, PhD, communication via e-mail,December 1, 2002) and were used in our estimates for this outcome. Age- andgender-specific prevalence of pseudophakia/aphakia among other races/ethnicitieswas estimated using a nonweighted average of the values for white, black,and Hispanic persons in each age- and gender-specific stratum.

The age-specific prevalence estimates for cataract and pseudophakia/aphakiafor white persons were derived in 2 steps. First, pooled prevalence proportionswere estimated for each gender- and age-specific stratum using minimum variancelinear estimation. Stratum-specific proportions from each study were transformedusing a logarithm odds transformation. Proportion variances were estimatedbased on the binomial distribution. The Cochran test for homogeneity was usedto evaluate the between-study variation for the pooled proportions. Second,logistic regression models were fit to the pooled prevalence proportions usingthe midpoint of each age interval as the independent variable. Models werefit separately for males and females. For black persons, logistic regressionmodels were fit to the age- and gender-specific estimates derived from theSEE Project and the Barbados Eye Study as described in the "StandardizationAmong Studies" subsection.

The number of cases of cataract and of pseudophakia/aphakia in the UnitedStates in each race/ethnicity, gender, and age category was estimated by applyingthe modeled prevalence rate for each year of age to the 2000 US Census populationand summing over the age range for each 5-year age category. Projected estimateswere derived applying the modeled rates for 2000 to the US Census middle-seriesprojections for 2020. Constant age- and gender-specific rates were assumedover this period for both cataract and cataract surgery. Stratum-specificUS prevalence rates were computed by dividing the total number of estimatedcases for each stratum by the stratum-specific US population.

The overall fit for each logistic regression model was evaluated usingthe F test for analysis of variance and the r² measure for proportion of explained variation. Age and race/ethnicityeffects were tested using the model Wald χ² test statistics.Odds ratios (ORs) for race/ethnicity were derived from logistic regressioncoefficients for the appropriate racial comparisons. Tests for gender differenceswere based on the observed age-, race/ethnicity-, and gender-specific ratesfrom each study. Separate Mantel-Haenszel χ² tests were doneby race/ethnicity controlling for both age and study effects.

The pooled age-specific prevalence figures for cataract increased withage for both black and white persons (P <.001for both, χ² test) (Table2). Women had a higher prevalence of cataract among both blacks(OR = 1.75; 95% confidence interval [CI], 1.18-2.56) and whites (OR = 1.35;95% CI, 1.23-1.49). The age-adjusted prevalence of cataract did not differbetween blacks and whites for women (OR = 1.03; 95% CI, 0.97-1.09) but amongmen was significantly higher for whites (OR = 1.09; 95% CI, 1.02-1.16) thanblacks.

The prevalence of pseudophakia/aphakia also increased with age for black,white, and Hispanic persons of both genders (P <.001, χ² test) (Table 3). Pseudophakia/aphakiawas significantly more common among Hispanics (data from the Proyecto VER[Visual Evaluation Research], Nogales and Tucson, Ariz, only) compared withwhites of both genders (OR = 1.52; 95% CI, 1.37-1.68) and blacks of both genders(OR = 3.04; 95% CI, 2.52-3.56), and among white compared with black males(OR = 1.83; 95% CI, 1.17-2.86).

Applying these age-, race/ethnicity-, and gender-specific prevalencefigures to the 2000 US Census data, there were an estimated 20.5 million persons(95% CI, 20.0-20.9) 40 years and older with cataract in the United States,a prevalence of 17.2% (95% CI, 16.8-17.5%) (Table 4). An estimated 6.1 million (95% CI, 5.7-6.5) Americans olderthan 40 years had pseudophakia/aphakia in 2000, a prevalence of 5.1% (95%CI, 4.8-5.5%) (Table 5).

According to our projections, based on US Census estimates for the populationin 2020, the number of persons with cataract will rise to 30.1 million by2020, an increase of 50%. Americans with pseudophakia/aphakia were estimatedto increase in number by almost 60% to 9.5 million by 2020. The number ofHispanic persons with pseudophakia/aphakia was predicted to almost tripleto 1.6 million.

When age- and gender-specific prevalence rates for all included studieswere examined together, the prevalence of cataract (Figure 1) and pseudophakia/aphakia (Figure 2) in the US studies did not seem to differ systematicallyfrom those conducted in Australia and Europe. Derived gender- and age-specificprevalence rates from this study were also applied to the 2000 populationsof Australia¹⁵ and Western Europe.¹⁵ An estimated 1.4 million persons (prevalence 17.2%)were estimated to be affected by cataract in Australia and 36.0 million persons(prevalence 19.3%) in Western Europe. The corresponding figures for pseudophakia/aphakiawere 380 000 (4.7%) and 9.8 million (5.3%), respectively.

Our estimates indicate that cataract prevalence will increase dramaticallyin the coming decades. The large increase in cataract surgical procedurespredicted for the US population as a whole is also of significant health policyimportance. Treatment for cataract already accounts for some 60% of vision-relatedMedicare expenditures.³ Further growth in thisarea will have a substantial effect on health care spending and, potentially,the fiscal stability of the Medicare system.

In determining which studies to include in this article, the decisionwas made to use only data that had been collected on a population basis since1990, using a system of cataract grading with predetermined standards. Thisled to the omission of the Framingham Eye Study,¹⁶ animportant early study that did report on the prevalence of lens opacities,but did not use a grading system with photographic standards, and assignedlens opacity grades based in part on visual acuity. This approach could notbe reconciled with modern grading systems based entirely on the photographicor slitlamp appearance of lens opacities.

Few population-based studies in the United States have reported on cataractprevalence and, thus, the inclusion of data from the Rotterdam Study, Rotterdam,the Netherlands; Blue Mountains Eye Study, Sydney, New South Wales, Australia;and the Melbourne Visual Impairment Project, Melbourne, Victoria, Australia,increased the power of our report, thus, allowing narrower confidence limitsin our estimates, especially for the small, but important, population of thevery old. Estimates for white Americans aged 40 through 64 years would otherwisehave been derived from a single report (Beaver Dam Eye Study, Beaver Dam,Wis). Most Australians immigrated originally from the same European countriesfrom which white Americans came (notably England, Ireland, Scotland, Germany,Italy, and Greece). Still, there are various factors that might lead to differencesin cataract prevalence between countries. These include latitude (as a surrogatefor exposure to cataractogenic UV-B light),¹⁷ differentialrates of cataract surgery, and possible cultural differences with regard todiet, tobacco smoking, and alcohol use. However, our study did not find systematicdifferences between European, Australian, and US studies for the prevalenceof cataract or of pseudophakia/aphakia. The generally similar rates acrossdiverse studies of white persons indicate that pooling is appropriate andsuggest that the estimates are likely to be reliable.

The cataract prevalence data available from population-based studiesfor persons of African descent are sparse; our estimates depended on statisticalmanipulation of data from the Barbados Eye Study to obtain any estimates forblack Americans younger than 65 years. Prevalence of cataract surgery appearssomewhat lower in Barbados than for black Americans (data not shown). Othercultural factors as outlined earlier might be expected to differ between the2 countries. Nevertheless, the only alternative would have been to assumethat the prevalence of cataract and cataract surgery among black and whiteAmericans is the same, an assumption that seems unlikely to be correct.⁹ Some studies of cataract prevalence in Africa exist¹¹ that might have increased the power of our estimates.However, there are significant differences between the United States and Africain availability of cataract surgery and in behaviors potentially affectinglens clarity, rendering such data less useful for our purposes.

Other modern prevalence studies exist that might have provided datarelevant to this article. We have chosen not to attempt to incorporate datafor Chinese populations in Singapore¹⁸ andelsewhere as US Census data identify individuals as being of East Asian originonly, without providing the exact country of origin. It is unclear whetherestimates derived from Chinese living in Singapore would improve the accuracyof our data for persons originating from Japan, Korea, Vietnam, the Philippines,and others, now residing in the United States. Other population-based studiesof cataract prevalence among European-derived persons are also available,¹⁹ but it was impossible for us to obtain data fromthese studies in a format that allowed these data to be incorporated intothe current article.

One difficulty of combining cataract prevalence data from differentstudies lies in attempting to divide an inherently continuous variable suchas lens opacity into discrete units and in reconciling the differing gradingsystems that have been used to do so. This study followed the consensus opinionof a group of experts, the principal investigators of the studies cited herein,in setting an arbitrary cutoff for cataract. As cortical cataract is generallymeasured by the area opacified on retroillumination photographs or at theslitlamp, it was simple to pick a cutoff area or proportion of the lens involvedand then choose the grade in each system that came closest to approximatingthat area. For PSC, it is likely that the Wilmer Grading System,¹² whichonly required that PSC be present, might have included some opacities thatwould not have reached the standard for PSC in LOCS II¹³ (>3%of the visible lens area involved) or the Wisconsin Cataract Grading System¹⁴ (≥ 0.625% of the visible lens area).

Nuclear cataract is graded in all of the systems in current use withreference to standard photographs depicting different degrees of opalescence(brunescence as measured in the LOCS II¹³ andIII²⁰ systems was not considered in our studydefinition of cataract). In choosing the penultimate nuclear category in eachsystem as our cutoff, we have attempted to identify a degree of opalescencethat is approximately the same in each of the studies cited. However, smalldifferences in the cutoffs used in the various grading systems, and the impossibilityof arriving at any definite equality of visual significance between differentcataract types, will to some extent limit the accuracy of our conclusionswith regard to cataract prevalence. The widespread use of more objective anduniversal systems to quantify lens opacity may improve accuracy in this areain the future, but it seems unlikely that any objective equivalence of gradesof the different opacity types can ever be determined. Providing separateprevalence estimates for the different cataract types might have avoided thisproblem, but it was felt that policy makers required a summary prevalencefigure for cataract.

The increasing prevalence of cataract and pseudophakia/aphakia withage, and among women, has previously been reported in many studies.²¹ The higher prevalence of pseudophakia/aphakia amongHispanic persons when compared with white and black persons has not, to thebest of our knowledge, been documented previously. However, this finding isbased on a single study and might possibly be influenced by practice patternsspecific to the 2 locales (ie, Nogales and Tucson, Ariz) included in the ProyectoVER sample. Further, this study of Mexican Americans is not necessarily representativeof the full range of Hispanic persons in the United States, which includesCuban Americans, Puerto Ricans, and persons from elsewhere in Latin America.The ongoing Los Angeles Latino Eye Study (LALES)²² willprovide an opportunity to further study rates of cataract extraction amongHispanic Americans. If, in fact, Hispanic persons undergo cataract extractionat a significantly higher rate than other racial groups, this will be of increasingimportance owing to the rapid growth of this segment of the US population.

Application of this study's findings cannot be made without a clearawareness of its weaknesses. As mentioned earlier, our estimates for cataractprevalence in the United States rely in part on data from Western Europe,Australia, and Barbados, areas that may differ from the United States in cataractsurgical rates and many other cultural factors likely to influence the prevalenceof lens opacity. For some groups, such as Hispanics and blacks, our prevalenceestimates rely on the results of a single study, and are, thus, likely tobe affected by local variations in surgical practices and by dietary, sunexposure, tobacco smoking, and genetic profiles that are highly specific tothe population reported. There are many important groups that are unlikelyto be represented by any of the study populations cited by us, including theurban poor and those living in the rural southeastern part of the United States.

Our projections of the prevalence of cataract and pseudophakia/aphakiain 2020 are based on assumptions of constant incidence. Such assumptions maynot be accurate, particularly for future rates of cataract surgery, whichare known to fluctuate with levels of reimbursement²³ amongother factors. Finally, as discussed in detail earlier, the accuracy of ourestimates of cataract prevalence must be limited to some extent by the necessityof combining results using different grading systems.

Nevertheless, these estimates are the first to combine the results ofseveral population-based studies of cataract prevalence with newly completed2000 US Census data and population projections. As such, they are likely toprovide the most complete information available on the most important causeof visual disability in our country. Our projections of a greatly increasedcataract burden and need for surgical services, despite their limitations,almost certainly reflect the realistic scope of this problem in a rapidlyaging population. Without strategies to prevent or delay the onset of lensopacity, the health care system will be challenged with an unprecedented demandfor cataract care.

In addition to underscoring the need for further research into cataractprevention strategies, this study also highlights the complete lack of dataon the prevalence of eye disease among important population groups such asAsian Americans. There is also a clear need, if cataract prevalence data areto be of practical use to health policy planners at a national or internationallevel, to develop methods of comparing existing cataract grading systems orto agree on a single system for universal use.²⁴

Corresponding author: Nathan Congdon, MD, MPH, Wilmer Eye Institute,Wilmer 120, 600 N Wolfe St, Baltimore, MD 21287 (e-mail: ncongdon@jhmi.edu).

Submitted for publication April 3, 2003; final revision received January23, 2004; accepted January 23, 2004.

From the Dana Center for Preventive Ophthalmology, Wilmer Eye Institute,The Johns Hopkins University (Drs Congdon, West, Friedman, and Kempen); Departmentof Ophthalmology, Erasmus Medical Center, Rotterdam, the Netherlands (Dr Vingerling);Department of Ophthalmology, University of Wisconsin, Madison (Dr Klein);Macro International, Inc, Calverton, Md (Ms O'Colmain); Department of PreventiveMedicine, School of Medicine, Stony Brook State University of New York atStony Brook (Ms Wu); Centre for Eye Research Australia, University of Melbourne,East Melbourne, Victoria, Australia (Dr Taylor); Department of Ophthalmology,University of Sydney, Centre for Vision Research, Westmead, New South Wales,Australia (Dr Wang); Department of Epidemiology and Biostatistics, Schoolof Public Health and Health Services, George Washington University MedicalCenter, Washington, DC (Ms O'Colmain).

This study was supported by Prevent Blindness America, Schaumburg, Ill;and the National Eye Institute, Bethesda, Md.

The members of the Eye Diseases Prevalence Research group are as follows: Baltimore Eye Survey, Baltimore, Md: James M. Tielsch;Alfred Sommer; Joanne Katz; Harry A. Quigley. Barbados EyeStudies, Barbados, West Indies: M. Cristina Leske; Suh-Yuh Wu; BarbaraNemesure; Anselm Hennis; Leslie Hyman; Andrew Schachat. Beaver Dam Eye Study, Beaver Dam, Wis: Barbara E. K. Klein; RonaldKlein; Kristine E. Lee; Scot E. Moss; Sandra C. Tomany. Blue Mountains Eye Study, Sydney, New South Wales, Australia: PaulMitchell; Jie Jin Wang; Elena Rochtchina; Wayne Smith; Robert G. Cumming;Karin Attebo; Jai Panchapakesan; Suriya Foran. MelbourneVisual Impairment Project, Melbourne, Victoria, Australia: Hugh R.Taylor; Cathy McCarty; Bickol Mukesh; LeAnn M. Weih; Patricia M. Livingston;Mylan Van Newkirk;, Cara L. Fu; Peter Dimitrov; Matthew Wensor; Yury Stanislavsky. Proyecto VER (Vision Evaluation Research), Nogales and Tucson,Ariz: Sheila West; Jorge Rodriguez (deceased); Aimee Broman; BeatrizMuñoz; Robert Snyder. Rotterdam Study, Rotterdam,the Netherlands: Paulus T. V. M. de Jong; Johannes R. Vingerling; RogerC. W. Wolfs; Caroline C. W. Klaver; Albert Hofman; Redmer van Leeuwen;M. KamranIkram; Simone de Voogd. Salisbury Eye Evaluation Project,Salisbury, Md: Sheila West;, Gary Rubin; Karen Bandeen Roche; BeatrizMuñoz; Kathy Turano;, Oliver Schein; Donald Duncan; Susan Bressler.

Resource Centers

Coordinating Center: David S. Friedman; Nathan G.Congdon; John H. Kempen; Benita J. O'Colmain. National Eye Institute,Bethesda, Md: Frederick L. Ferris III.