Key Points

QuestionÌý What was the prevalence of glaucoma and vision-affecting glaucoma in the US in 2022?

FindingsÌý This meta-analysis estimated that 4.22 million people in the US (1.62% of adults) had glaucoma and 1.49 million people (0.57% of adults) had vision-affecting glaucoma in 2022. Substantial demographic and geographic disparities in glaucoma burden were also found.

MeaningÌý These estimates may help in the development and prioritization of public health strategies, monitoring of epidemiologic trends, and evaluation of programs tailored for populations at highest risk of glaucoma.

ImportanceÌý Glaucoma is the leading cause of irreversible blindness worldwide and, in the US, disproportionately affects people from racial and ethnic minority groups. Glaucoma prevalence has not been estimated for the US in more than a decade, and state- and county-level estimates are not available.

ObjectiveÌý To estimate glaucoma and vision-affecting glaucoma prevalence by demographic factors and US state and county for the Centers for Disease Control and Prevention’s Vision and Eye Health Surveillance System (VEHSS).

Data SourcesÌý This meta-analysis used data from the National Health and Nutrition Examination Survey (2005-2008), Medicare fee-for-service claims (2019), IBM MarketScan commercial insurance claims (2016), population-based studies of eye disease (1985-2003), and 2022 population estimates from the US Census Bureau.

Study SelectionÌý PubMed was searched for population-based studies of glaucoma prevalence published between 1991 and 2016.

Data Extraction and SynthesisÌý Bayesian meta-regression methods were used to estimate the prevalence of glaucoma and vision-affecting glaucoma stratified by age, undifferentiated sex/gender (a measure that captures an unclear mix of aspects of sex and or gender), race and ethnicity, and US county and state.

Main Outcomes and MeasuresÌý Prevalence of any type of glaucoma (open or closed angle) among people 18 years or older and vision-affecting glaucoma, defined as glaucoma and a visual field abnormality.

ResultsÌý For 2022, an estimated 4.22 million people (95% uncertainty interval [UI], 3.46 million to 5.23 million) in the US were living with glaucoma, with a prevalence of 1.62% (UI, 1.33%-2.00%) among people 18 years or older and 2.56% (UI, 2.10%-3.16%) among people 40 years or older. An estimated 1.49 million people (UI, 1.17 million to 1.90 million) were living with vision-affecting glaucoma, with a prevalence of 0.57% (UI, 0.45%-0.73%) among people 18 years or older and 0.91% (UI, 0.71%-1.16%) among people 40 years or older. Prevalence of glaucoma among people 18 years or older ranged from 1.11% (UI, 0.89%-1.40%) in Utah to 1.95% (UI, 1.57%-2.39%) in Mississippi. Black adults had a prevalence of 3.15% (UI, 2.32%-4.09%) compared with 1.42% (UI, 1.10%-1.85%) among White adults; adults in the Hispanic and all other racial and ethnic categories combined had a prevalence of 1.56% (UI, 1.13%-2.06%).

Conclusions and RelevanceÌý This meta-analysis found that an estimated 2.56% of people 40 years or older have glaucoma, slightly more than estimated by previous studies. Black individuals are disproportionately affected. Prevalence estimates at the state and county level can help guide public health planning.

Glaucoma is the leading cause of irreversible blindness globally,¹ and vision loss due to glaucoma is associated with disability.² In the US, glaucoma disproportionately affects Black individuals.^3-5 Recent estimates of the prevalence of glaucoma and vision-affecting glaucoma in the US and estimates at the state and county levels are lacking.

Due to the slowly progressive nature of glaucomatous vision loss, the disease often goes undetected for many years.⁶ More than half of glaucoma cases in the US are undiagnosed,⁷ preventing treatment and monitoring of the disease, and the populations at the greatest risk for glaucoma have less access to eye care.⁸ Population-level estimates of glaucoma prevalence can help guide the development, prioritization, and evaluation of public health strategies to address glaucoma, particularly in the groups most affected.⁹

The US Centers for Disease Control and Prevention’s (CDC’s) Vision and Eye Health Surveillance System (VEHSS) includes several single-source estimates of the national glaucoma prevalence. One is the National Health and Nutrition Examination Survey (NHANES), a nationally representative survey to assess the health and nutritional status of community-dwelling people in the US, which included eye examinations during 2005 to 2008.^10,11 In these data, the prevalence of glaucoma was 2.1% (95% CI, 1.7%-2.6%) among adults 40 years or older.¹² The VEHSS also includes diagnosed glaucoma documented in administrative claims for commercial insurance and Medicare and Medicaid fee-for-service insurance. Data on glaucoma from NHANES are limited to adults 40 years or older, and claims report only diagnosed glaucoma.

We sought to develop new multisource composite prevalence estimates of of all types of glaucoma and vision-affecting glaucoma in the US. Such estimates can provide insights into the magnitude of the condition, enable forecasting of future trends, and support appropriate allocation of health care resources. State- and county-level estimates that highlight variation in the prevalence of both glaucoma and vision-affecting glaucoma also allow for more tailored public health interventions.

In this meta-analysis, we used Bayesian meta-regression methods developed by the Global Burden of Disease Study,¹³ specifically DisMod-MR, version 1.1.1,¹³ which uses the Python PyMC 2 package.¹⁴ We separately estimate national prevalence of glaucoma and vision-affecting glaucoma. The statistical models are described in eAppendix 2 in Supplement 1. The study adhered to the Meta-analysis of Observational Studies in Epidemiology () reporting guidelines.

Because glaucoma is an adult-onset condition, we set prevalence to 0 in those younger than 18 years. We estimated prevalence by age group (ages 18-19 years and then 5-year age groups between 20 and 100 years), an undifferentiated indicator of sex/gender, and a composite indicator of race and ethnicity. The undifferentiated sex/gender indicator is a composite of disparate sex- and gender-related variables in our data sources; this term from the literature¹⁵ describes a measure that captures an unclear mix of aspects of sex and/or gender, rather than implying only sex or gender (eAppendix 1 in Supplement 1). Each data source contained at most 1 sex- or gender-related variable, and each such variable included only binary values for male and female; the results may not reflect prevalence in gender minority populations, who are known to have worse health outcomes and face barriers in the health care system.^16-18 The race and ethnicity indicator has 4 categories: Hispanic, non-Hispanic Black, non-Hispanic White, and all other non-Hispanic races and ethnicities, which included individuals who are Alaska Native, American Indian, Asian, Pacific Islander, multiracial, and other groups not otherwise classified. We could not report more detailed categories due to a lack of sufficient data (eAppendix 1 in Supplement 1). Race and ethnicity are social constructs, and disparities in prevalence between groups are likely influenced by social factors, such as systemic racism.⁸ However, race and ethnicity are also correlated with genetic ancestry, which may affect glaucoma prevalence,¹⁹ and we were unable to differentiate these in our analysis. Most race and ethnicity information was self-reported; race and ethnicity in the Salisbury Eye Evaluation (SEE) Glaucoma Study was based on administrative data,²⁰ and race and ethnicity in several of the studies analyzed by the Eye Diseases Prevalence Research Group (EDPRG) was assumed based on location.²¹

The meta-regression incorporates published results of population-based studies,^3,21,22 NHANES data from 2005 to 2008,^10,11 Medicare Part B fee-for-service claims in 2019,²³ and IBM MarketScan Commercial Research Database of medical claims to commercial insurers in 2016.²⁴ We extracted each source’s sample size and glaucoma and vision-affecting glaucoma prevalence by age group, sex/gender, and race and ethnicity, where available (eAppendix 1 in Supplement 1). The model assumes that the stratified prevalence has not changed substantially since it was measured in our data sources.

Glaucoma in NHANES was defined by fundus photograph grading. We transformed the polytomous grade into a probability, setting possible to 10% and probable to 90%. We defined vision-affecting glaucoma as glaucoma and a visual field abnormality. We imputed missing glaucoma and vision-affecting glaucoma status information (23.0% and 37.4% of participants aged ≥40 years, respectively). We used NHANES examination weights to calculate glaucoma prevalence by demographics. We performed 2 sensitivity analyses: dichotomizing the grading instead of using probabilities (equivalent to setting possible to 0% and probable to 100%), following Gupta et al,¹² and using a definition as similar as possible to the International Society for Geographical and Epidemiological Ophthalmology (ISGEO) definition of glaucoma.²⁵ The ISGEO definition, like our definition of vision-affecting glaucoma, requires both a structural and a functional abnormality. We imputed missingness for these sensitivity analyses using models adapted to the definitions (eAppendix 3 in Supplement 1); the rest of the methods were the same as in the primary analyses.

We included population-based studies published after 1991 that were representative of the target population and presented primary results or meta-analysis of primary data. We identified 3 such studies: the Los Angeles Latino Eye Study (LALES), conducted from 2000 to 2003²²; a meta-analysis by the EDPRG, which analyzed studies conducted from 1985 to 2000²¹; and the SEE study, conducted from 2001 to 2003.³

Due to a lack of population-based study data on individuals in the other races and ethnicities category and small sample sizes in NHANES for this category, we could not estimate an independent prevalence effect for this subgroup. Therefore, the meta-regression combined the Hispanic and other races and ethnicities groups; prevalence estimates differ between these 2 groups only due to differences in demographic composition.

We used 2022 population estimates stratified by age, gender/sex, race and ethnicity, and county from the US Census Bureau’s Population Estimates Program (PEP).²⁶ These estimates used newly adopted county equivalent units in the state of Connecticut, so we mapped the population of Connecticut to counties using the 2021 PEP county proportions within each demographic group.²⁷ The PEP estimates also top-coded age at 85 years and did not contain an age group cutoff at 18 years; we used county- and race- and ethnicity-specific age distributions from the 2010 US Census²⁸ to disaggregate the population age groups.

We derived county-level estimates by combining our national estimates with county-level demographic compositions and then aggregated county-level prevalence estimates to create state-level estimates. We estimated standardized prevalence of glaucoma and vision-affecting glaucoma by sex/gender and race and ethnicity as the expected prevalence for each stratum assuming the national distribution for age and sex/gender or race and ethnicity.

We estimated that 4.22 million people (95% uncertainty interval [UI], 3.46 million to 5.23 million) were living with glaucoma in the US in 2022, with a prevalence of 1.62% (UI, 1.33%-2.00%) among all adults (aged ≥18 years), 2.56% (UI, 2.10%-3.16%) among those 40 years or older (Table), and 5.20% (UI, 4.12%-6.49%) among those 65 years or older. We estimated 1.94 million individuals (UI, 1.53 million to 2.44 million) with male sex/gender and 2.29 million individuals (UI, 1.77 million to 2.87 million) with female sex/gender living with glaucoma; male adults had an age- and race-standardized prevalence of 1.65% (UI, 1.30%-2.08%) compared with a prevalence of 1.59% (UI, 1.24%-2.00%) for female adults. The age- and sex/gender-standardized prevalence of glaucoma was 1.73 percentage points (pp) (UI, 0.89-2.69 pp) higher among non-Hispanic Black adults, who had a prevalence of 3.15% (UI, 2.32%-4.09%) compared with a prevalence of 1.42% (UI, 1.10%-1.85%) among non-Hispanic White adults. Adults in the Hispanic and other races and ethnicities categories combined had a standardized glaucoma prevalence of 1.56% (UI, 1.13%-2.06%), slightly higher than among non-Hispanic White adults; however, this difference was highly uncertain (0.14 pp; UI, −0.36 to 0.67 pp).

The prevalence of glaucoma increased as a function of age (Figure 1); crude prevalence was 0.08% (UI, 0.06%-0.12%) among those aged 18 to 39 years, 1.09% (UI, 0.84%-1.40%) among those aged 40 to 64 years, 4.44% (UI, 3.41%-5.69%) among those aged 65 to 79 years, and 7.75% (UI, 5.77%-10.35%) among those 80 years or older. State-specific prevalence of glaucoma among adults ranged from 1.11% (UI, 0.89%-1.40%) in Utah to 1.95% (UI, 1.57%-2.39%) in Mississippi (eTable in Supplement 1; Figure 2). Detailed national, state, and county-level data are available on the CDC VEHSS website.²⁹

We estimated that 1.49 million people (UI, 1.17 million to 1.90 million) were living with vision-affecting glaucoma in the US in 2022, with a prevalence of 0.57% (UI, 0.45%-0.73%) among the adult population, 0.91% (UI, 0.71%-1.16%) among those 40 years or older (Table), and 1.83% (UI, 1.41%-2.38%) among those 65 years or older. We estimated that 0.68 million male individuals (UI, 0.51 million to 0.90 million) and 0.80 million female individuals (UI, 0.61 million to 1.05 million) were living with vision-affecting glaucoma; male adults had an age- and race-standardized prevalence of 0.59% (UI, 0.44%-0.77%) compared with a prevalence of 0.55% (UI, 0.42%-0.73%) for female adults. The age- and sex/gender-standardized prevalence of vision-affecting glaucoma was 0.72 pp (UI, 0.41-1.08 pp) higher among non-Hispanic Black adults, who had a prevalence of 1.20% (UI, 0.87%-1.61%) compared with a prevalence of 0.48% (UI, 0.36%-0.65%) among non-Hispanic White adults. Adults in the Hispanic and other races and ethnicities categories combined had a standardized glaucoma prevalence of 0.57% (UI, 0.40%-0.80%), slightly higher than among non-Hispanic White adults; however, this difference was highly uncertain (0.09 pp; UI, −0.11 to 0.31 pp).

The prevalence of vision-affecting glaucoma increased as a function of age (Figure 3); crude prevalence was 0.02% (UI, 0.00%-0.08%) among those aged 18 to 39 years, 0.39% (UI, 0.28%-0.54%) among those aged 40 to 64 years, 1.45% (UI, 1.05%-1.94%) among those aged 65 to 79 years, and 3.13% (UI, 2.28%-4.30%) among those 80 years or older. Prevalence of vision-affecting glaucoma among adults ranged from 0.38% (UI, 0.29%-0.50%) in Utah to 0.70% (UI, 0.54%-0.88%) in Mississippi (Figure 4; eTable in Supplement 1).

Using a dichotomized definition of glaucoma in NHANES, we found 4.00 million people (UI, 3.23 million to 4.94 million) with glaucoma for a prevalence of 1.53% (UI, 1.24%-1.89%) among adults and 1.64 million people (UI, 1.25-2.06 million people) with vision-affecting glaucoma for a prevalence of 0.63% (UI, 0.48%-0.79%) among adults. Using an ISGEO-approximating definition of vision-affecting glaucoma in NHANES, we found 1.19 million people (UI, 0.88 million to 1.58 million) with vision-affecting glaucoma, for a prevalence of 0.45% (UI, 0.34%-0.60%) among adults and 0.72% (UI, 0.54%-0.96%) among those 40 years or older.

We estimate that more than 4 million people in the US have glaucoma and approximately 35% (1.5 million) of them have vision loss from this condition. Non-Hispanic Black adults are approximately twice as likely as non-Hispanic White adults to have glaucoma and nearly 3 times as likely to have vision-affecting glaucoma after adjusting for age and sex/gender. The prevalence of glaucoma and vision-affecting glaucoma increases substantially with age. We provide both state- and county-level estimates of the prevalence of glaucoma and vision-affecting glaucoma, which can help guide public health strategies to support vision health at the local level.

The current estimates of the number of people in the US with glaucoma are higher than previous reports. For example, the EDPRG estimated that 2.22 million individuals had glaucoma in 2004 and projected that 3.36 million individuals would have glaucoma in 2020, nearly a million less than our estimates.²¹ Our findings advance the field by leveraging NHANES data on eye examinations among a representative sample of the US population, which were not available at the time of the EDPRG’s work. Moreover, although glaucoma definitions in the studies analyzed by the EDPRG vary and some require a visual field abnormality, our definition does not. Using NHANES data, Gupta et al¹² estimated a glaucoma prevalence of 2.1% (95% CI, 1.7%-2.6%) among adults 40 years and older in 2005 to 2008. We report a somewhat higher prevalence of 2.56% (UI, 2.10%-3.16%) for this age group, primarily due to imputation of missing glaucoma data and the considerably older demographic composition in 2022 (eAppendix 5 in Supplement 1).

Following previous studies,^3-5,22,30 we report differences by race and ethnicity, with a higher prevalence of glaucoma and vision-affecting glaucoma among non-Hispanic Black individuals compared with non-Hispanic White individuals. In the Baltimore Eye Survey, the prevalence of primary open-angle glaucoma (the most common form of glaucoma) was 3 to 4 times higher among Black compared with White adults 40 years or older (5.6% vs 1.7%).⁵ In the SEE study, the prevalence of open-angle glaucoma among Black and White adults aged 73 to 74 years was 5.7% and 3.4%, respectively. Among those 75 years or older, prevalence increased to 9.4% for White individuals and 23.2% for Black individuals.³ Vision impairment and blindness due to glaucoma was also significantly more common among Black compared with White adults aged 65 to 84 years (0.9% vs 0.1%).⁴ On the basis of these previous studies^3-5,22,30 and NHANES,^10,11 our study reports a higher crude prevalence of glaucoma among non-Hispanic Black individuals compared with non-Hispanic White individuals 40 years or older.

Some prior studies have also reported a high prevalence of glaucoma among Hispanic adults. In LALES, the prevalence of open-angle glaucoma among Hispanic individuals 40 years or older was 4.8%.²² In contrast, Proyecto Vision and Eye Research, a representative study of Hispanic adults in southern Arizona, reported a prevalence of open-angle glaucoma of 2.0% in those 40 years or older.³⁰ However, that study also reported a more rapid increase in prevalence with age than had been previously observed among non-Hispanic White or Black individuals. We estimated that the crude prevalence of glaucoma was 0.96% among all Hispanic adults and 1.81% among those 40 years or older. This apparent discrepancy with LALES, which was conducted only in Los Angeles, may be due to the national scope of the current study. Hispanic individuals represent a diverse ethnic group, and the prevalence of glaucoma may vary considerably by geography within this demographic.

Glaucoma is difficult to objectively define, and although diagnostic criteria varied among data sources, we considered all definitions equivalent. The first sensitivity analysis demonstrates that the use of probabilities instead of dichotomizing the ordinal glaucoma grading scale in NHANES did not have a substantial effect on our results. However, the second sensitivity analysis demonstrates that an ISGEO-approximating definition of glaucoma is more stringent than our primary definition of vision-affecting glaucoma, underscoring the difficulty of comparing prevalence estimates that are based on different definitions.

We used nationally representative NHANES data as the basis for our estimates. These data were collected during 2005 to 2008, and the included population-based study data were even older. Although this study accounted for changes in the demographic composition of the US population over the last 15 years, updated NHANES eye health data could help us to better understand US trends in glaucoma. Our use of missing data imputation yielded an estimate that was 19.2% higher for glaucoma and 31.3% higher for vision-affecting glaucoma compared with complete-case analysis, which reflects the older age and increased risk characteristics of participants with missing data (eAppendix 5 in Supplement 1). Lacking studies on the prevalence of glaucoma and vision-affecting glaucoma among older adults living in institutional settings, such as assisted living or nursing homes, we extrapolated data from a noninstitutionalized population, likely underestimating the true prevalence.

The NHANES sample sizes were not large enough to model a race effect for non-Hispanic people who are not White or Black. Therefore, we used a single parameter for the Hispanic and other races and ethnicities groups, making our standardized prevalence estimates the same between these groups. Future vision data collection in NHANES could allow more granular analysis of race and ethnicity, providing critical insights for planning public health strategies to address disparities in vision health.

The measurement of sex/gender and race and ethnicity in our data sources was inconsistent and often not described in detail. Future data collection efforts could be more transparent about how these concepts were defined and provide more granular response categories, for example, for gender minority populations. Our descriptive results are consistent with social drivers of disparities, such as systemic racism, and should not be interpreted as evidence of biological differences; more research would be needed to understand the mechanisms that drive disparities in glaucoma.

In contrast to models for previous conditions,^31-33 we did not include county-level prevalence data in models for glaucoma and vision-affecting glaucoma. County-level insurance claims data are available, but these data only reflect diagnosed glaucoma and vary geographically in ways that are unlikely to be driven by true glaucoma prevalence.³⁴ This may be due to geographic differences in health care use and diagnostic practices; glaucoma is more likely to be undiagnosed and has more subjective diagnostic criteria than other common eye conditions.

Approximately 4.22 million people in the US (1.62% of adults) are estimated to have glaucoma, and 1.49 million people (0.57% of adults) are estimated to have vision-affecting glaucoma, with substantial variation in prevalence across demographic subgroups and US states and counties. These estimates may help provide information for the development and prioritization of public health strategies and interventions, the monitoring of epidemiologic trends, and evaluation of programs tailored for communities and populations at highest risk of glaucoma. Future population-based data collection may allow for updated prevalence estimates and estimation of trends, which could be particularly relevant to an aging and increasingly diverse US population.

Accepted for Publication: July 30, 2024.

Published Online: October 17, 2024. doi:10.1001/jamaophthalmol.2024.3884

Corresponding Author: Joshua R. Ehrlich, MD, MPH, Department of Ophthalmology and Visual Sciences, University of Michigan, 1000 Wall St, Ann Arbor, MI 48105 (joshre@umich.edu).

Author Contributions: Dr Flaxman had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Ehrlich, Burke-Conte, Wittenborn, Saaddine, Omura, Flaxman, Rein.

Acquisition, analysis, or interpretation of data: Ehrlich, Burke-Conte, Wittenborn, Omura, Friedman, Flaxman, Rein.

Drafting of the manuscript: Ehrlich, Burke-Conte, Flaxman.

Critical review of the manuscript for important intellectual content: Burke-Conte, Wittenborn, Saaddine, Omura, Friedman, Flaxman, Rein.

Statistical analysis: Burke-Conte, Wittenborn, Flaxman.

Obtained funding: Wittenborn, Rein.

Administrative, technical, or material support: Ehrlich, Omura, Flaxman, Rein.

Supervision: Ehrlich, Saaddine, Flaxman, Rein.

Conflict of Interest Disclosures: Dr Ehrlich’s effort was funded through an Intergovernment Personnel Act agreement with the Centers for Disease Control and Prevention Vision Health Initiative. Dr Friedman reported receiving consulting fees from Abbvie and Life Biosciences outside the submitted work, receiving speaking fees from Thea Pharmaceuticals outside the submitted work, and serving as a board member of Orbis International and the Glaucoma Research Foundation. Dr Flaxman reported receiving fees from Janssen, SwissRe, Evaluation Merck for Mothers, and Sanofi for assistance in analysis and interpretation of licensed data previously produced by the Institute for Health Metrics and Evaluation and advising fees from Agathos outside the submitted work. No other disclosures were reported.

Funding/Support: This work was supported by cooperative agreement NU58DP007190 from the Center for Disease Control and Prevention Vision Health Initiative.

Role of the Funder/Sponsor: The Centers for Disease Control and Prevention Vision Health Initiative participated in the study design, research, analysis, data collection, interpretation of data, reviewing, and approval of the publication.

Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

Data Sharing Statement: See Supplement 2.