Percentage change in the number of people in the United States with diabetic retinopathy, 2005-2050.
Percentage change in the number of people in the United States with diabetes who have glaucoma, 2005-2050.
Percentage change in the number of people in the United States who have cataracts, 2005-2050.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Saaddine JB, Honeycutt AA, Narayan KMV, Zhang X, Klein R, Boyle JP. Projection of Diabetic Retinopathy and Other Major Eye Diseases Among People With Diabetes MellitusUnited States, 2005-2050. Arch Ophthalmol. 2008;126(12):1740–1747. doi:10.1001/archopht.126.12.1740
Copyright 2008 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2008
To estimate the number of people with diabetic retinopathy (DR), vision-threatening DR (VTDR), glaucoma, and cataracts among Americans 40 years or older with diagnosed diabetes mellitus for the years 2005-2050.
Using published prevalence data of DR, VTDR, glaucoma, and cataracts and data from the National Health Interview Survey and the US Census Bureau, we projected the number of Americans with diabetes with these eye conditions.
The number of Americans 40 years or older with DR and VTDR will triple in 2050, from 5.5 million in 2005 to 16.0 million for DR and from 1.2 million in 2005 to 3.4 million for VTDR. Increases among those 65 years or older will be more pronounced (2.5 million to 9.9 million for DR and 0.5 million to 1.9 million for VTDR). The number of cataract cases among whites and blacks 40 years or older with diabetes will likely increase 235% by 2050, and the number of glaucoma cases among Hispanics with diabetes 65 years or older will increase 12-fold.
Future increases in the number of Americans with diabetes will likely lead to significant increases in the number with DR, glaucoma, and cataracts. Our projections may help policy makers anticipate future demands for health care resources and possibly guide the development of targeted interventions.
Efforts to prevent diabetes and to optimally manage diabetes and its complications are needed.
Diabetic retinopathy (DR) is a common and serious condition. It is the leading cause of blindness among working-age adults in the United States.1In 2004, blindness from DR accounted for approximately $500 million in direct medical costs among Americans 40 years or older.2However, interventions to detect and treat DR to prevent vision loss are cost-effective and may result in cost savings.3,4People with diabetes mellitus also have a higher prevalence of other eye diseases, such as cataracts and glaucoma, than the general population.5-10Vision loss related to eye disease among people with diabetes is an important disability that threatens independence and can lead to depression, reduced mobility, and reduced quality of life.11The Eye Diseases Prevalence Research Group (EDPRG) classified DR into 2 major composite outcomes: (1) any DR as any DR consisting of mild, moderate, or severe DR, diabetic macular edema, or any combination and (2) vision-threatening DR (VTDR) as DR likely to result in vision loss in the absence of treatment, consisting of proliferative DR, diabetic macular edema, or both. Among people 40 years and older with known diabetes in the year 2000, an estimated 4.1 million persons had DR and approximately 0.9 million had signs of VTDR.12
Several factors will likely combine to increase the number of Americans with diabetes, including a steady increase in the total US population, an increase in the average age of the population, and disproportionate growth in the number of Hispanics and blacks, among whom the prevalence of diabetes is higher than among whites.
Previous studies have projected the number of people with DR and VTDR using forecasts of the number of people with diabetes as reported in 2001 by Boyle et al13of the Centers for Disease Control and Prevention (CDC). These forecasts were based on the estimated 29 million people in the United States who will have diabetes by 2050. However, a more recent CDC study14projects that 48 million people will have diabetes in 2050. This analysis accounts for the recent increase in the incidence of diagnosed diabetes between 2000 and 2004, the decrease in death rates among people with diabetes, and expected changes in the age and racial/ethnic composition of the US population through 2050. As the number of people with diagnosed diabetes increases, the number of people with eye complications leading to severe vision loss is also likely to increase, unless glycemic and blood pressure control among people with diabetes improves.
In this article, we provide age-, sex-, and race/ethnicity-specific estimates of the number of people with DR, VTDR, glaucoma, and cataracts among those 40 years and older with diagnosed diabetes in the United States for the years 2005-2050. These forecasts may help policy makers prepare for increased future demand for vision-related health care and develop interventions to prevent or delay vision loss among people with diabetes.
To project the numbers of Americans with diabetes who will have moderate to severe visual disorders through 2050, we combined eye disease prevalence rate estimates from the EDPRG12,15,16with projections of the numbers of Americans with diagnosed diabetes. We first briefly describe the model used to project the US burden of diabetes. We then describe our methods to project the number of people with diabetes and eye disease. We quantified uncertainty in our projections of the number of people with eye disease using 95% Bayesian confidence intervals (CIs; ie, the 2.5% and 97.5% quantiles of the posterior distributions). We reported these CIs for the totals.
To project the number of Americans with diabetes from 2005 to 2050, we used an updated discrete-time (1-year intervals), incidence-based Markov model with 3 states (no diagnosed diabetes, diagnosed diabetes, and death) to project the number of people with diabetes in each year from 2005 to 2050. A complete description of the model structure, data sources, and parameter estimation can be found in the studies by Honeycutt et al17and Narayan et al.14In each cycle of the model, we developed projections for 808 population subgroups defined by age, sex, and race/ethnicity (age [in years]: 0, 1, . . . , 99, ≥100; sex: male, female; race/ethnicity: non-Hispanic white, non-Hispanic black, Hispanic, other).
We used US National Health Interview Survey data to estimate age-, sex-, and race/ethnicity-specific diabetes prevalence and incidence rates of diabetes for 2004. Future incidence rates were assumed constant at 2004 levels. We generated all these estimates using Bayesian logistic regression with improper flat priors. The age-, sex-, and race/ethnicity-specific 2004 prevalence rate estimates were combined with US Census Bureau population estimates for 2004 to determine the initial distribution of individuals across the nondiabetes and diabetes states. US Census Bureau projections of US live births, mortality rates, and net migration were also required because our diabetes projection model was constrained to yield the total population projections implied by these quantities. Finally, we required estimates of the relative risk of mortality for people with diabetes. These estimates were obtained from a published study18that applied Bayesian methods and logistic regression models to North Dakota death certificate data because of the lack of national data that would support direct estimates of the relative risk of mortality from diabetes. These relative risks and the US Census mortality rates determined projected nondiabetes and diabetes mortality rates.
We used Markov chain Monte Carlo simulation to generate 5000 draws from the posterior distributions of all model parameters (age-, sex-, race/ethnicity-specific prevalence, incidence, and nondiabetes and diabetes mortality rates). The Markov model used these posterior distributions as inputs to generate posterior distributions of the projected number of people with diabetes. The actual projections were the means of these posterior distributions.
Estimates of the prevalence of DR, VTDR, glaucoma, and cataracts in the United States have been reported previously by age, sex, and race/ethnicity.12,15,16Our projections of the number of Americans who will have these eye diseases were based on the assumption that the prevalence of these conditions will remain constant through 2050. We projected the number of people with eye disease by multiplying the appropriate age-, sex-, and race/ethnicity-specific prevalence rate in Table 1by the age-, sex-, race/ethnicity-specific projection of the number of people with diagnosed diabetes. Thus, for example, to estimate the number of diabetic non-Hispanic white men aged 40 to 49 years with DR in any year, we multiplied the DR prevalence rate by the projected number of white men with diabetes in this age range. To project the number of Americans with diabetes who will have glaucoma and cataracts, we used the prevalence estimates for the total US population that are not specific to people with diabetes. The EDPRG generated prevalence estimates for DR and glaucoma for whites, blacks, and Hispanics; however, prevalence estimates for cataracts were only for whites and blacks.
Finally, we used prevalence estimates from previous reports and their 95% CIs to generate 5000 independent normal distributions in logit space for each estimate. Logistic transforms of these draws yielded 5000 draws from the posterior distribution of the prevalence estimates. Draws from these distributions were combined with draws from the posterior distributions of the number of people with diabetes to obtain posterior draws of the number with eye disease. The 0.025 and 0.975 posterior quantiles determined 95% CIs for the projected number of people with diabetes and eye disease. These CIs are reported only for totals in Tables 2, 3, and 4.
The US Census projections used in the projections of diagnosed diabetes mellitus underlying the eye disease projections in this study forecast a total US resident population of 402 million in 2050: 213 million non-Hispanic whites, 53 million blacks, 98 million Hispanics, and 38 million people of other races.
Our forecasts suggest that the number of people with DR and VTDR will almost triple during the next 45 years (Figure 1). The number of people with DR is expected to increase from 5.5 million in 2005 to 16.0 million in 2050, and the number with VTDR is expected to increase from 1.2 million in 2005 to 3.4 million in 2050.
Our projections indicate an even larger growth in the number of blacks and Hispanics with DR and VTDR, especially among those 65 years or older (Table 3). We estimated that 5.9 million whites, 1.0 million blacks, and 2.9 million Hispanics 65 years or older will have DR in 2050. The projected increases for VTDR are considerably smaller but also show expected increases to 1.0 million for whites, 0.4 million for blacks, and 0.5 million for Hispanics 65 years or older by 2050.
Increases in the number of people with DR and VTDR are also expected for younger age groups. The forecasted growth is higher for Hispanics than for whites and blacks across all age groups (Table 3), mirroring population growth and its expected impact on the number of people with diabetes.
For some age, sex, and racial/ethnic groups, the number of DR and VTDR cases is expected to peak before 2050. For example, among white women aged 65 to 74 years, we projected that the number of cases will peak in 2030; this peak reflects the changes in the number of women with diabetes in 2030. Our projections of eye disease among people with diabetes were based on unchanging prevalences based on data from published studies. Thus, changes in the numbers of diabetic people with eye disease are expected to follow changes in the numbers of people with diabetes.
Among Americans with diabetes, we projected that the number of cases of glaucoma will increase between 2005 and 2050 for all demographic groups but particularly among blacks 50 years or older and among Hispanics across all age groups (Table 4). We projected an 11- to 12-fold increase in the number of Hispanics 65 years or older with glaucoma (Figure 2). By 2050 our forecasts suggest that the number of people 65 years and older with diabetes who have glaucoma will be approximately 0.23 million for white men, 0.40 million for white women, 0.21 million for black men, 0.24 million for black women, 0.14 million for Hispanic men, and 0.23 million for Hispanic women.
The number of whites and blacks with diagnosed diabetes who have cataracts is expected to increase for all age and sex groups from 2005 through 2050 (Table 4). In 2050, among people with diabetes, almost 2.7 million white men, 4.6 million white women, 0.5 million black men, and 1.2 million black women with diabetes 65 years or older will either have cataract or have had cataract surgery. The number of cataract cases among people with diabetes 75 years or older is expected to increase 637% for black women and 677% for black men between 2005 and 2050 (Figure 3).
Although a previous study13reported estimates of the number of people with DR and VTDR in the United States, the authors of that study assumed that the prevalence of diagnosed diabetes would remain constant at 1999 levels or increase only slightly. However, a recent study by Narayan et al14demonstrated that diabetes prevalence in the United States is likely to increase dramatically through 2050, given recent increases in the incidence of diagnosed diabetes, decreases in diabetes-related mortality, and expected changes in the age and racial/ethnic composition of the US population. Accounting for these changes implies a large and growing burden of eye diseases associated with diabetes, in particular DR and VTDR.
We forecast that the number of people in the United States with any DR will increase approximately 3-fold from 5.5 million in 2005 to 16 million by 2050. Our forecasts also suggest that the number of people with diabetes who have glaucoma will increase to 1.4 million (>300% increase vs 2005 levels) and the number with diabetes who have cataracts will reach approximately 10 million by 2050 (>200% increase vs 2005 levels).
The EDPRG estimated that 6.1 million people with diabetes will have any DR by 2020, but the EDPRG forecast assumed constant diabetes prevalence over time.19When the EDPRG accounted for increasing diabetes prevalence rates, they estimated that 7.2 million people will have DR in 2020. In contrast, our estimate that 9.7 million will have DR by 2020 is 35% higher than the EDPRG estimate. We also estimated that 16 million Americans with diabetes (95% CI, 14.8 million to 17.4 million) will have any DR in 2050 and that 3.4 million of those (95% CI, 3.1 million to 4.1 million) will have VTDR. Even if the prevalence of DR among people with diabetes remains constant over time, in the absence of normalization of glycemia and blood pressure through intensive control, the number of people with DR is likely to increase rapidly because of expected growth in the numbers and changes in the race/ethnicity composition of people with diagnosed diabetes between 2005 and 2050.
Our results also suggest that the older age groups—those 65 to 74 years and 75 years and older—will experience the largest increases in the number of people with diabetes-related eye disease. Hispanics of all ages with diabetes are forecast to have particularly large increases in the number of people with DR and other eye diseases associated with diabetes. For example, growth in the number of Hispanics between 40 and 49 years with DR and VTDR is comparable to expected growth in DR and VTDR for whites 75 years and older.
Although we assumed that the prevalence of DR and VTDR will remain constant at 2000 published levels through 2050, results of recent national and international population-based studies among people with type 1 diabetes have shown a lower prevalence of retinopathy than previously reported,20,21possibly because of changes in diabetes management since 1980. Additional studies22have demonstrated improvement in intermediate health outcomes, such as levels of hemoglobin A1cand low-density lipoprotein, between 1988 and 2002. These intermediate outcomes are major risk factors for the progression of DR to severe stages. In our forecasts, we did not account for possible decreases in the prevalence of DR and VTDR that could occur if improvements in intermediate diabetes outcomes continue, which might be expected given the Healthy People 201023emphasis on the prevention and improved management and control of diabetes.
On the other hand, improvements in diabetes care could also result in improved survival.24,25Extending the duration of diabetes, which is a major risk factor for the development and progression of DR may, in part, offset the effects of improved diabetes management on lowering the prevalence of DR and VTDR.
In this study, we focused on generating estimates of DR, VTDR, glaucoma, and cataracts among people with diabetes from 2005 through 2050. We chose to focus on these diseases in our analysis because the availability of age-, sex-, and race/ethnicity-specific prevalence data for these conditions allowed us to explore the impact of changes in the age and racial/ethnic composition of people with diabetes on the number of people with diabetes who also have moderate to severe eye disease. The relationship between diabetes and glaucoma is established in several studies.5,8,10Projecting the number of people with diabetes and glaucoma is important for understanding the impact that preventing and managing this condition will have on our medical care system. Henceforth, there might be a need to enhance or modify our screening program and health care provision planning to tackle this issue appropriately.
Although we did not model the number of people with diabetes who have correctable visual impairment due to refractive error, by using age-, sex-, race/ethnicity-, and diabetes-specific prevalence data, we combined previous data reported in the CDC's Morbidity and Mortality Weekly Report26with forecasts of the number of people with diabetes through 2050 from Narayan et al14and estimated that the number of people with diabetes who have correctable visual impairment due to refractive error will likely increase from 1.2 million in 2005 to 3.5 million in 2050. These estimates suggest a role for interventions to correct refractive error to improve quality of life among people with diabetes.
Our study had several limitations. First, we used a constant prevalence of DR, VTDR, glaucoma, and cataracts. This assumption ignores the impact of factors such as better management of diabetes and hypertension and improved screening for and treatment of eye diseases in people with diabetes, which could lead to decreasing prevalence of these conditions through 2050. Consequently, using constant prevalence may have caused us to overestimate the number of adults with diabetes affected by DR, VTDR, glaucoma, and cataracts. Second, we used general population prevalence estimates for glaucoma and cataracts rather than prevalence estimates specific to people with diabetes. Previous studies27,28have shown that cataracts are more common among people with diabetes than among people without diabetes, and diabetes is associated with cataracts and cataract surgery. Other studies5also demonstrated a higher rate of open-angle glaucoma among people with diabetes. Therefore, using general population prevalence estimates for cataracts and glaucoma may have caused us to underestimate the number of adults with such conditions. Nevertheless, our forecasts reflect the realistic scope of these problems given population aging and a rapidly increasing diabetes epidemic.
Our findings show that the estimated number of people with DR will increase unless approaches are developed to prevent diabetes. Knowledge of this information is important for planning allocation of resources involved in prevention such as education programs and training persons needed to manage people with diabetes at risk of developing these complications. Recent studies23have shown that the percentage of Americans with diabetes who receive recommended eye care, including screening, is below that recommended in Healthy People 2010(75%). One study29showed some increase in the percentage of Medicare recipients who received dilated eye examinations, but only to approximately 60%. Studies based on National Health Interview Survey30,31and Behavioral Risk Factor Surveillance System data32have shown that only approximately 50% to 70% of Americans with diabetes receive annual dilated eye examinations. Simply increasing this cost-effective intervention might reduce the rate of vision loss and blindness associated with diabetes. Continued improvement in the approaches used to diagnose and treat eye diseases in people with diabetes will also help reduce the present and future burden of visual disability and improve quality of life among people with diabetes.
In summary, our projections have shown higher numbers than previously estimated for DR, VTDR, cataracts, and glaucoma among Americans with diabetes. These projections should be helpful to public health policy makers preparing to meet the future demand for vision-related health care among Americans with diabetes and attempting to develop interventions to prevent or delay vision loss in this population.
Correspondence:Jinan B. Saaddine, MD, MPH, Division of Diabetes Translation/Vision Health Initiative, Centers for Disease Control and Prevention, 4770 Buford Hwy, Atlanta, GA 30341 (Mail Stop, K10) (firstname.lastname@example.org).
Submitted for Publication:June 16, 2008; final revision received June 17, 2008; accepted June 25, 2008.
Financial Disclosure:None reported.
Additional Contributions:Sarah Lesesne, BS, research analyst at RTI, provided continuous research assistance with the manuscript.